May 16, 2025 • 57 mins
Baptiste Morel-Prieur is a author on a recent paper titled Submaximal force-velocity relationships during mountain ultramarathon: Data from the field
Research from the 2023 UTMB offers ground-breaking insights into trail running performance by analyzing how runners handle different gradients throughout a race. Batiste Moral shares his data-driven approach that identifies specific areas where athletes can improve from year to year.
• Analyzing Strava data from 600 UTMB participants reveals distinct "force" (uphill ability) and "velocity" (flat ground speed) components of performance
• These components have no correlation with each other, meaning they represent different physiological abilities requiring targeted training
• Case study of Jim Walmsley shows 76% of his time difference behind Kilian Jornet in 2022 occurred on gradients steeper than 20%
• After focused training in the Alps, Walmsley specifically improved his steep climbing ability before winning UTMB 2023
• Durability analysis suggests flat terrain speed deteriorates more than uphill capability during ultras, indicating flat-ground training after fatigue might be more valuable
• Western States vs. UTMB comparison shows dramatically different terrain profiles despite both being 100-mile races<br>• Simple heart rate-based test can help identify your personal gradient-specific weaknesses
Papers discussed-Decoding Ultramarathon: Muscle Damage as the Main Impediment to Performance
Sign up for CTS Coaching-https://trainright.com/coaching/ultrarunning/
Subscribe to Research Essentials for Ultrarunning-https://www.jasonkoop.com/research-essentials-for-ultrarunning
Information on coaching-
https://www.trainright.com
Koop’s Social Media
Twitter/Instagram- @jasonkoop
Buy Training Essentials for Ultrarunning:
Amazon-https://www.amazon.com/dp/B09MYVR8P6
Audible-https://www.amazon.com/dp/B09MYVR8P6
#ultrarunning #trailrunning #running #sports #sportsperformance
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:08):
Trail and ultra runners.
What is going on?
Welcome to another episode ofthe coop cast.
As always, I am your host,coach Jason coop, and on this
podcast we oftentimes gleaninsight from the lab and then we
use it in the field ofcompetition.
Other times, we look at what isactually going on in the field
of competition and we gleaninsights from that on how to
(00:30):
train and race smarter.
And today's guest was an authoron a paper that did just that.
Welcome to the podcast today forthe first time, batiste Moral,
who took publicly availableStrava data from the 2023 UTMB
to determine what pacingstrategies worked better, and he
also used those same techniquesto determine how people in, or
(00:51):
how the elite field specifically, evolve over time.
And the results of all of thisanalysis is something that I
feel we can take to training andracing, and it's something that
I'm taking into practice withmany of the athletes that I work
with currently.
I'm going to let Batisteexplain exactly what he did and
what the results of his researchactually are and how we can
(01:12):
apply it.
So with that out of the way,I'm getting right out of the way
.
Here's my conversation withBatiste Moral.
Batiste, welcome to the podcasttoday.
I appreciate you coming on.
I'm really excited to talkabout some specific pieces of
research that you've kind ofrecently published.
(01:33):
But before we get into it toomuch, can you give the listeners
a little bit more of abackground of who you are, how
you got involved in trailrunning and looking at this from
a scientific perspective in thefirst place?
Speaker 2 (01:43):
and looking at this from a scientific perspective in
the first place.
Yeah, thank you for theinvitation.
First, I'm a researcher rightnow, so it's been 10 years.
I defended my PhD and I startedworking on performance sports
performance, but more on rugbyplayers, but we will see later.
There is some comparison thatwe can do with trail runners.
So I work on physiology andbiomechanics fields and I do
(02:06):
research on sport performance,but also on health and even on
animals in ecology.
That's another question, butyou will see there is some
crossroads between all of thesetopics.
So that's my background.
Speaker 1 (02:18):
Perfect.
So, the part of your researchthat really piqued my interest
really gets to the core and thesoul of what athletes go through
every year.
But most ultramarathon athletes, they're experienced athletes.
They're not novel or new to thescene.
They have marathon PRs andthey've done ultramarathon races
before and things like that.
(02:38):
And the coaches who arelistening to the podcast can
resonate with this run of show.
At the end of every year and atthe beginning of the next season
they're going through theirRolodex of athletes, they're
doing their consults and thequestion that comes up how am I
going to get better?
Where am I going to get better?
What am I going to do to takemy performance from X to Y based
(03:01):
on the previous year?
It's something that's kind oflike top of mind for everybody
and really, when you think aboutit, it's kind of the root of
all interventions whether it'sinterval intervention or a
nutrition intervention, like asupplement or anything that you
can kind of like throw at anathlete is all kind of aimed to
improve them specifically fromone year to the next.
And a lot of your and a lot ofyour research has looked at that
(03:23):
specifically, like where haveathletes actually improved?
And so I want you to kind oflike take it from there with
this.
Maybe we can look at howcertain athletes have improved
from year to year and how we canactually take this really broad
brushstroke of looking at arace and trying to dissect it in
(03:44):
terms of where can thoseathletes actually improve?
That's a big question.
Speaker 2 (03:50):
that's a big question because it's not easy with so
much performance factors,especially in trail running and
ultra trail running even moreprobably, and there is multiple
way to to approach and toaddress this question, and I
tried to do some new way toanalyze the athlete performance
(04:12):
based on the data that we.
You know, there is a lot of bigdata coming from the sports
performance and this issomething that we can go deeper
on it to see how were theperformance on one year and how
it can evolve, and so you canhave like objective metrics like
that help you to choose, tolook at if the intervention and
(04:32):
the training, the thing that youchange, maybe the nutrition and
all we mentioned how iteffectively and change my
profile and so making me moreperformant.
Speaker 1 (04:43):
And the way that you kind of like looked at this was
looking at race profiles, right,and specifically UTMB race
profiles, and we can take acouple of high level athletes,
but let's take it at like kindof like a population level first
.
Right, you wanted to look atUTMB from an entire race
perspective.
What is the entire you know,the entirety of the race
actually doing?
What did you glean from that?
(05:04):
First off, what are youspecifically looking at?
We're going to have to get intoforce velocity relationships
and some of the thingsspecifically that you're looking
at and define those for theusers, but take it from the
highest level possible.
How are you trying to dissectthis problem of where can
athletes actually get better?
Speaker 2 (05:19):
yeah, actually the first thing I did was in 2023,
at the end of August, justbefore the UTMB this year.
And that was for fun, becausesome of my colleagues just asked
me who will won the UTMB thisyear and I have no idea.
But I say, okay, we'll take alook at the data and see if
there is some piece ofinformation in there.
So I look at the race, Ianalyze the race, coming from
(05:43):
Kilian John that won the 2022edition and I look at one of the
guys who will probably win thisyear that was Jim Wemsley in
2023.
So I look at the race in 2022and say, okay, let's dissect the
race and we will cut it intosegments, into portion, every
(06:04):
100 meters of this race.
And I wanted to compare theirprofile on each 100 meters of
the race to see who is the beston which segment.
And there was a result reallyinteresting because you know,
there is flat portion, there isuphill with more or less steep
gradient, really steep gradient.
(06:24):
And the thing that was reallyinteresting, that was that if
you look at the race and all thesegments when it was flat or
low gradient, or downwind oruphill, there is not that much
differences between the race ofkian and jim.
But when you compare them, onlylooking at the steep gradient
let's say above 20% gradient,there was a big difference
(06:49):
between Kylian and Jim this year.
Actually, jim lost.
Almost all of the difference onthe finish line was because of
the steep gradient.
So that was the first approachto say, okay, he seems to have
kind of a weakness on thisspecific part of the race, so
that's only one way to approachit because the determinants of
(07:11):
the performance are complex.
But there was like a first step.
Okay, there is somethingdifferent between these two top
elite athletes on this specificportion and to get into that
specifically 76%.
Speaker 1 (07:26):
So three quarters of the time difference was lost on
sections above a 20% grade.
That's what your analysis foundcorrect, yeah yeah, that's
correct.
Speaker 2 (07:38):
So more than one hour was only on those portions, but
they only represent less than10% of the whole race, so most
of the difference was done onvery specific portion of the
race.
Speaker 1 (07:51):
Okay.
So that's kind of like we'regoing to take this like Jim
versus Killian and now we'regoing to move it into like Jim
versus Jim, because I do thinkthat this sets this up for the
population level thing.
A lot of people listening willsay, okay, yeah, big deal,
killian's like pushing it harderon the steepest climbs, because
that story is very good.
But there's another analysisthat you looked at which is gym
(08:12):
versus gym.
So gym pre-winning, utmb, gymtraining in the Alps for a
couple of years very famousstory here in the US.
Some you know an American movesover here specifically to train
for here.
I'm now located in Chimney,france.
So for the listeners out therewho are wondering why I'm using
that vocabulary, it's because ofthat Comes over to the Alps
(08:34):
specifically to train and winUTMB and ends up doing it.
Do you analyze Jim'sperformances before that
transition and then after thattransition?
What did that actually reveal?
Speaker 2 (08:50):
Yeah, actually that was the purpose of this analysis
, because everybody knew that hecame to the ops probably trying
to win the utmb on the nextyear and actually he did a 100
race in april this year.
It was in croatia, if Iremember well.
So so I we found the data on.
We found the data of the raceand analyze it to see if there
is a changing in the profile ofgym, especially in this specific
(09:11):
steep slopes.
So we did exactly the sameanalysis, comparing the profile
of gym in the utmb 2022 and theprofile of the istria, the 100
istria race in croatia in 2023.
And actually we saw a bigimprovement on the ability to
maintain a good speed in thesteep slopes, in the steepest
(09:32):
slopes.
That's why we thought thatprobably the training he did
this year in the Alps helped himto improve in this specific
part, the part that was aweakness for the last, the year
previous, the previous year.
So he was better on thisspecific slopes and he was as
(09:53):
good as before on all the othersegment of the race.
So that's why we, at this time,I thought that he was probably
able to be one of the topathletes and probably winning
the race because of thisanalysis.
So it's always a verychallenging race and with a lot
of random aspects that we can'tcontrol.
(10:14):
But they're training hello himto be better on this specific
part of his profile, so probablybetter at the utmb next year.
Speaker 1 (10:24):
Well, okay, so now we're going to get it into the
general population right Withthat as a little bit of a back,
with that as a little bit of abackdrop.
That happens specifically, mostlikely because of an athlete's
environment changes.
They change their environment.
They train more in steepertypes of train.
It makes all the sense of theworld with that.
That's where they are going toimprove in the most.
(10:44):
But let's back this, let'sbroaden this out a little bit,
because we're talking about twovery good athletes that are
training for a very specificpurpose and, as you mentioned,
they have narrow margins toimprove upon.
And usually those narrowmargins get even narrower
because it's a very specificelement that they're improving
with.
I know I go through this with myelite athletes every single
(11:06):
year when are you going toimprove?
And it's usually not a generalfitness proposition.
We try to maintain that as muchas possible.
Usually we're trying to improvea specific component of their
fitness, maybe a specificcomponent of the entire of the
preparation arc where we screwedup one phase or something like
that.
Maybe it's something withstrength training.
Maybe it's something withstrength training, maybe it's
(11:31):
something with nutrition, but itusually is because they are so
good it's sliced out and it'sfocused on in a very narrow
perspective.
You contrast that with yourgeneral pop athlete, somebody
who's only been training forthree or four years.
They kind of improve across theboard with just generalize
training.
So let's take that biggerpicture approach.
You were part of a study thatlooked at 600 athletes.
Use the Strava data and it cangeneralize athletes into a few
different like profiles.
(11:52):
So first off go over like whatthe profile is that you're
actually looking at and thenwhat the actual research found,
and then we'll get into somepractical takeaways from that.
Speaker 2 (12:01):
Yeah, actually, at the beginning it was an analysis
of just two top elite athletesfor fun, with my team and we did
a meeting at the end of Augustand say, oh, that's very
interesting, we probably need togo to see all of the other
participants to see what we canidentify in terms of profile.
So, to improve, so I bring myteam at home and we open all of
(12:25):
our computers so to bring backall the data and say, okay, now
we have the database that we cananalyze to see how we can
describe the profile of all theathletes during this specific
race of the under edition of2023 of the UTMB.
So the idea was to do the exactsame thing that for the NMGM
(12:48):
look at every 100 meters segmentof the race and to see at which
speed they were able to run fordepending on the gradients.
That was the idea.
We all know that there is agrade adjusted pace.
You can find that on your watch, on your Strava or Garmin
account.
But most of the time thisgrade-adjusted pace is just
(13:10):
computed on the mean of apopulation.
So it gives you an idea of howyou can translate a speed when
you are going up compared to theflat.
Speaker 1 (13:18):
But actually what we observed was that there were
really big differences on how anathlete will run, depending on
his own profile okay, before wego any further, I want to take
some time and explain what ngpand gap actually are, because
(13:38):
both of these terms are kind ofan alphabet soup, so to speak,
and they try to do the samething although they get about it
at different ways.
So NGP or normalized gradedpaste, is a is something
proprietary to training peaksand all it tries to do is it
tries to make the pace that youare actually seeing on a run
(13:59):
during a post hoc analysis whenyou're actually looking at the
file and make an equivalent forthat pace of what it would be on
flat level ground.
Now the two things thattraining peaks is adjusting for,
our first off, thenormalization piece of it.
That's the first part of theterm normalized graded pace and
the second piece of it is thegraded part of it.
(14:20):
Now, most people will befamiliar with the grading part
of it.
It just means running uphill isharder than running on the
flats and running downhill iseasier than running on the flats
and running uphill.
The second part of that, thenormalization part it is a
little bit trickier and honestlydoesn't have that big of an
influence on the number, but itnormalizes the pace to what you
(14:42):
could run if you ran the pacecompletely evenly.
So for example, if you go outand run one mile in six minutes
and you do every single lap ofthat mile at 90 seconds, so the
first quarter is 90 seconds, atthe halfway mark or at 800
meters you hit 130.
At the 1200 meter mark you hit430.
And then at the end of the mileyou hit six minutes, you've run
(15:03):
that mile completely evenly.
Your normalized graded pacewould be six minutes per mile.
Now if you decided to mismatchthat pace, or you decided to run
the first quarter in 60 secondsand the second quarter slower
and the third quarter faster,the normalization component of
that would be faster than sixminute miles, because the
physiological toll is harderthan six minutes per mile.
(15:25):
So all it's trying to do is it'strying to make a physiological
equivalent of both the variationin pacing and the variation in
grades and normalize it to flatlevel terrain.
Strava's equivalent of this,which is gap or grade adjusted
pace, only takes the gradingcomponent of it.
So it looks at how fast you'rerunning and what the grade
(15:45):
you're running on and it triesto make a physiological
equivalent for that for flatlevel terrain.
Now, as Batiste mentioned, bothof these are based off of
research that is done on tonsand tons of people and, as we
know, certain people are moreefficient or effective at
running uphill and downhill andso their particular
physiological cost might not bequite accurately reflected in
(16:07):
these grade adjusted paces andnormalized graded paces.
And if we can look at thatacross a whole host of athletes
we might be able to glean someinsight.
But, needless to say, with bothof these, all of all that they
are trying to do is it's tryingto look at the variables around
the actual pace that you ranuphill, downhill, the variation
in those paces and trainingpeaks case and saying what is
(16:28):
the physiological equivalent, or, more accurately, the oxygen
cost equivalent of if thisathlete actually ran on flat
level terrain, what would thepace actually be?
Speaker 2 (16:43):
I would say you all know people that are really good
on flatland and on levelrunning but when it's going up
it's hard and you have guys thatnot that much fast on flatland
but when you are going up theyare really good.
And that was this idea how wecan model the profile of an
athlete to determine his owngrade, adjusted pace so we know
(17:06):
he is he able to run with a lotof speed.
This is the velocity part ofthe profile, and when you are
going up you have to push more,produce more force so you can
elevate your center of mass in abiomechanical perspective.
So how force and velocityinteract within a profile, can
we characterize an athlete'sspeed on this profile?
(17:27):
And that will probably opensome new information on how we
can train for the next year orfor the next race.
Speaker 1 (17:36):
And so what were the different profiles that you
found throughout all of the UTMBparticipants, and how did those
profiles correlate to their endperformance, so to speak?
Was there a profile that wasmore advantageous for the race,
least advantageous for the race,more advantageous for the
second half of the race versusthe first half of the race?
What did the data miningactually glean out if we were to
(17:57):
build this perfect athlete thatwe wanted to put on the start
line for UTMB?
Speaker 2 (18:02):
For the comparison between the first half and last
half of the race, we will talkabout that later because it will
imply durability aspects, solet's talk to that after.
But for the force, velocity,how you're able to run fast when
it's steep, with steep gradient, or how you're able to run fast
when it's flat, if you look atall the participants, actually
(18:22):
we had the race data coming fromthe winner of this edition and
the race data coming from thelast guy who crossed the finish
line, so that's evident.
But, yes, the best athletes arebetter on each portion of the
race, right?
they run faster on flat andfaster when it, when you have,
when they are going uphill.
But but we ran a secondanalysis looking at people and
(18:45):
runners with quite the sameperformance.
So we cut all the population in10 percentiles.
So let's take the 10 bestathletes of this race, or the
athlete between the 40 and 50percentile, so we can analyze a
comparable performance.
And in this specific analysis,the very interesting thing with
(19:06):
that was that there was notoptimal profile.
You can do the exact same racetime but with a very different
profile, and probably the guyyou would cross the finish line
with him you never saw himbecause when you are behind and
when it's going down you arebelow.
Speaker 1 (19:28):
Everybody's had that experience.
I mean, that's like a reallyrealistic thing.
You can cross the finish line.
You look to the person whofinished in front of you, the
person who finished behind youand two places behind you, and
maybe you saw one of thosepeople for a part of the race.
Aside from that, they're likeyou said they're faster in
different sections and slower ondifferent sections,
particularly races that have alot of different features steep
(19:50):
uphills gradual uphills, steepuphills, gradual uphills, steep
downhills, gradual downhills,flats and things like that.
Speaker 2 (20:00):
Yeah, and I feel like this is something that everyone
who did a trail, who ran atrail race, know that, because
you know that you will be betteron some portion, and then the
guy that let you down you willcome back, et cetera.
But the thing interesting isthat now we have a model so we
can precisely evaluate this.
And since we can evaluate this,we have indices indexed that
can use for testing and foroptimizing training or to open
(20:23):
new idea of how I can train nowso to be better.
For example, if you compareyour profile with the guy who
finished the race at the sametime of you, but if you see that
for every flat portion he wasbetter, you probably need to
train more on these flatsections and it's probably this
(20:45):
training that can help you to bebetter on the next race.
Speaker 1 (20:51):
Let's check in with Coach Adam really quickly on how
we as coaches determine thestrengths and weaknesses with
athletes.
Speaker 3 (21:00):
So short answer.
It's not always easy, but thefirst place I start is actually
just asking the athlete, and youcan kind of get this when you
have a call after a race andthey have some perspectives on
what went well and what didn't.
And I also want to combine thatwith the real, objective data
that we're seeing in the racefiles.
(21:21):
So a lot of times people mightfinish a race and say, man, I
suck at climbing, and I alwayswant to make sure that is that
the sensation of sucking atclimbing, or is it a potentially
a reality?
Because there's a lot of thingsthat are simply hard running in
the heat, things like that.
I mean, it's easy to perceive aweakness when it's really just
everyone kind of performs worsein those conditions.
(21:42):
So we see if, potentially,there's a mismatch with your
normalized graded pace in theclimbs versus the descents in
something that would besignificant or material.
Usually we probably would haveseen that in training already,
but hopefully we can see that issome sort of weakness.
But it might be something thatshows up only late in a race too
(22:02):
.
So that's something we want tolook at.
Another thing that we havethat's a tool at some races is
the kind of live tracking whereit shows your position
throughout the race and that canactually give you some good
ideas of your strengths andweaknesses, based on where are
you passing people and where areyou being passed.
I like to use that quite a bit,actually.
So there's no perfect answer,and I think one of the
(22:24):
weaknesses that we have rightnow is that only the more strong
edge cases stand out, maybe thepeople on the edges of the bell
curve.
But with some more tools whatwas done in this study we might
be able to use a bit more of afine tooth comb.
Speaker 1 (22:41):
So that's what I wanted to kind of get into next
before we get into the firsthalf back half piece, and just
as a little bit of a depthbackdrop.
The listeners who have beenkind of constantly tuned into
this podcast will realize thatthis concept of durability that
is, a relatively newerphenomenon or concept that we've
started to talk about in sportsscience and in performance has
(23:02):
continued to come up in thispodcast via a lot of different
angles and sometimes ones thatwhere I'm not really like
seeking it out, it just happensto be part of the fabric of the
dialogue.
And so this podcast runs trueto form, where we're going to
talk a little bit aboutdurability as well.
But we're going to put it,we're going to put a pin in that
.
For just one second, thesynopsis of how to look at this
(23:24):
data can really be boiled downto how can one person get better
knowing their profile from yearto year?
Not necessarily and I was kindof baiting you with the way that
I phrased the question or likehow to build the perfect athlete
for the race, right?
I don't think that's the valuein knowing this information.
It's taking Jason Koops' datafrom UTMB or from other races or
(23:49):
whatever looking at it and thenextracting something from that
to train for the next year andthe year after that.
How might one actually do that?
I mean, this is something thatyou and your team put together,
but can we put this in morepractical aspects for the
coaches and the athletes thatare out there that do look at
their race files and do look atother people's race files a lot?
(24:11):
How can we actually go aboutdoing that?
Speaker 2 (24:14):
yeah, actually there is something interesting that we
are doing also in my lab.
This is an experiment andresearch that we're doing right
now, but that had reallypractical application.
The idea is that I told you thatwhen you're going up, you are
producing more force, andsomething that we did was to
compare the force that youproduce when you are going up by
(24:35):
running or using sled training,like like in track and track
and training resisted sprintingor resisted pushing right.
Yeah, exactly, and what we sawwas that actually this is almost
the same biomechanicalorganization of your body when
you are pulling a sled or whenyou are going up.
(24:55):
So that's opened newopportunities to train, for
example, if you have a profilethat is not that good in the
force portion.
So when you're going up andlet's say, maybe you are living
in a city with not that muchgradients, so you are not able
to train in this kind ofgradients you probably can find
(25:16):
some alternative with this kindof training.
So that's just an example.
But to say, you can be betterand be more endurance in the
fourth part of your profile, notonly by running uphill, but you
can find other alternativeswith strength training and, in
this specific example, with sledtraining that can mimic a slope
(25:38):
.
Speaker 1 (25:38):
And so what you're trying to do is, like you said,
improve the force part of theequation.
Speaker 2 (25:42):
Ah, because the thing interesting was that there is
no correlation between thevelocity part and the force part
.
So, just to be clear, thevelocity part is your when we
are talking about train, is yourability to run fast on flat
land, and the force part is areyou able to run fast or to walk
fast when you are going up?
So this is how we candifferentiate these two parts,
(26:05):
and the interesting thing isthat there is no correlation
between these two parts of theprofile.
That's what we were talkingabout at the beginning of the
talk.
So if it's not the same ability, you can probably train that
differently.
And when I said you can improvethe force part of your profile,
(26:26):
you probably need to find a wayto train your ability of your
muscle to produce higher levelof force for a longer time and
it probably will improve betteryour profile on this part, but
not necessarily changing theother part being the velocity
part the.
Speaker 1 (26:42):
So the analogy that I've always used to to that
where you're saying the forcepart is not related to the
velocity part, or let me kind ofcolloquialize that a little bit
the uphill performance is notrelated to the flat part, or let
me kind of colloquialize that alittle bit the uphill
performance is not related tothe flat level performance is
that a fair colloquialization ofwhat you're trying to describe
from a, from a biomechanicsstandpoint?
Speaker 2 (27:05):
yeah, totally.
Actually, we all.
What we are doing is definitelynot a revolution.
You don't know that, that kindof aspect, because we saw that
in training and we saw that whenwe're racing.
But the thing interesting isthat we can know model it, we
can know evaluate it.
So our objective informationsso we can stand on to build the
(27:26):
training of your athletes yeah.
Speaker 1 (27:28):
so the analogy that I've always used for that the
uphill component and the flatcomponent and I actually add two
more components to that that wego completely off the rails, we
can talk about those for alittle bit is the downhill
component and the walkingcomponent, and I always liken
those to different sports.
They're sure linked by thecardiopulmonary system and the
(27:50):
musculoskeletal system to somecertain degree and have a little
maybe a little bit of Venndiagram overlap.
Although you're describing theuphill component and the flat
component as not being related,but I still think that from a
training standpoint we do needto think about training all four
of these.
Just like a triathlete trainsall three swim, bike run I think
(28:10):
that analogy kind of playsthrough and the more
statistically we look at raceperformance, the more that
analogy starts to actually come.
It actually comes to thatrealization that it's not just
an analogy, it's actually what'sgoing on.
Speaker 2 (28:24):
Yeah, but you agree that if you have a limited time
to train, a limited volume, youcannot do all the training you
want, because you can enrich,over reaching or over training
or that especially for eliteathletes.
So, to be more precise, and ifyou more tailored the
intervention and the trainingthat you do because you know
that you cannot do infinitetraining probably you should
(28:48):
find the the best part of yourprofile to train.
And just something that Iwanted to say before.
But it also depends on the race, because 100 miles race
training, race can be reallydifferent and the profile can be
really different, and not onlyon the total elevation, because
it's something that is oftenlooked at okay, there is, let's
(29:10):
say, 10,000 meters of elevationgain in this race, but you can
do 10 000 meter elevation gainin very different way correct
and if you're going really withreally steep slopes at more than
40 or 50 percent, or only along portion with 15 percent
slopes, that's not at all thesame in terms of the physical
(29:33):
ability that your trainer needsto have.
Speaker 1 (29:36):
Yeah.
So I get this question actuallya lot because there's this
section of my book where we talkabout how to tailor the uphill,
downhill components of trainingto the race.
And the way that I illustratedin the book was with the
broadest brushstroke, and I'mgoing to do it in.
I'm going to do it in us unitsas opposed to metric units, so
(29:59):
we can create the.
We can create the translationlater.
But if your race has 200 feetof elevation gain per mile,
which would be approximatelywhat the Western States 100 has,
try to average that over thecourse of a week of training.
Now to your point.
You can come up with that ahundred different ways.
You can do all of thatelevation gain in one run up a
(30:20):
40% slope, or you can do it up5% slope, 7% slope, 5%.
There's a lot of different waysto do it.
I guess is what I'm saying, andso I get a lot of questions on
my book related to just that.
How do I come up with that?
So the first step is just hitthe average.
Second step is try to hit thegrades that you're actually
going to experience during therace.
(30:43):
Interestingly enough, I watcheda presentation that you gave.
It was in French, but theEnglish subtitles are apparently
pretty good, according to myFrench speaking counterparts.
And one of the really neatslides that you had during the
course of that presentationcompared UTMB and Western States
in terms of the percentage ofthe distance that the runners
(31:04):
would spend at each grade.
And the point that you'retrying to make is between zero
and 15%.
They're about equal, meaningrunners are going to spend about
the same amount of time runningacross those grades 15 to 30%.
Utmb has over twice the amountof distance as compared to UTMB
(31:25):
at that grade.
And then over 30%, utmb hasseven times the amount of
distance at that grade comparedto Western States, and there's
hardly any part of the WesternStates and there's hardly any
part of the Western Statescourse.
Having been on both of thosecourses, that is at that percent
grade.
But what that reallyillustrated to me is how two
marquee 100-mile races, themarquee 100-mile races in the
world, can be so dramaticallydifferent.
(31:47):
And I look at it through thelens of they're dramatically
different from a demandsperspective, because the
gradient that they're actuallyrunning on not just the total
elevation gain, elevation loss,but the actual gradient that
they're running on can bethought of as a specific demand
of the race yeah, yeah,definitely.
Speaker 2 (32:02):
That's really important to understand that,
because we often look at therace.
Metrics are the mean big thing,but we we definitely need to go
more into detail on how the raceoccurred really and what are
the gradients that you will findin this race, because it's not
the same to do one kilometer at30 percent or two kilometers at
(32:24):
15.
It's just not the same in termsof how your body will react.
And maybe another thinginteresting is that we're a lot
talking about the steep gradientand the flat and the level
running, but actually there isalso a component for
intermediate gradients.
I mean that we found someprofiles of runners that were
able to run at the same speed instrip gradients, the same speed
(32:47):
in flats land, but they werenot able to run at the same
speed for the intermediategradients, especially in the
area of the walk-run transitionand when you have to walk at
high speed or to run at lowspeed, some of the assets are
not good and other ones are good, and that's another component
of the profile that we analyzewith the model we proposed.
Speaker 1 (33:09):
I have a colleague of mine, jackson Brill.
Shout out to Jackson, who wasactually an intern with us a
long time ago.
He was actually trying to workon this in Roger Cromwell's lab
at the University of Coloradoand find out where the right
walk to run transition actuallyexists.
And does it actually exist?
And the answer is no, you'vegot to use your own intuition
for it.
It might be a conversation foranother day, but I'll link that
(33:32):
conversation up that I had withJackson in the show notes for
anybody that's interested inmaybe trying to find it or
trying to find a better way todetermine when that walk to run
transition for you shouldactually be.
I want to spend a little bitmore time on this Western state
versus UTMB component of it.
One of the things that that thatillustration that I just
mentioned the percentage ofdistance that the runners are
(33:53):
spending at each kind of likegradient bin or gradient chunk
being so different between thosetwo races, also highlights to
me how difficult it is to getboth of those races right, even
though they're both a hundredmile races, even though they're
both long and you know duration,and it's very far out to the
right on the power durationcurve which we haven't really
(34:16):
gotten into yet, and they seemlike low intensity activities
and things like that.
You're running at a lowerpercentage of your VO2 max way
lower than a marathon oranything like that.
But because the courses are sodifferent, it's hard to get both
of those right in a short, in areally short timeframe, and
we've seen that play outespecially really short
timeframe, and we've seen thatplay out especially in the elite
field over the years, that it'shard to have an athlete do both
of those.
(34:37):
And I happen to have an athlete, Katie Scheid, who did both of
those successfully one year.
But it is exceedingly rare.
And one of the reasons it'sexceedingly rare is because we
see this kind of marketdifference in the course
profiles, which is kind of ano-duh moment.
But your research is reallybringing to light how different
they can actually or howdifferent the demands actually
(34:58):
are.
I guess is what I'm trying tosay.
Speaker 2 (34:59):
Yeah, yeah, actually we can think that it's the same
race, because just it's 100miles right 160K.
But if I say, if you take a100-meter dash sprinter, you
will never perform on marathonand the opposite, that's evident
for every everybody.
But actually that's the samefor various different race of
(35:20):
training race that do notrequire the same abilities and
physical capabilities yetexactly okay, before we go on to
this durability component, Ikind of want to like rehash some
of this from a practicalstandpoint.
Speaker 1 (35:35):
So the athletes out there and they look at they can
look at this conceptually andsay, ok, you know, I understand
that if I looked at myperformance on this race for a
particular year and how I rackand stack compared to the people
around me, I'm going to noticedifferent strengths and
weaknesses.
And I want to focus on theweaknesses part generally in
(35:57):
order to kind of extract moretime out of it.
But how, aside from theobservation piece, people are
going away from me on the climbsand I'm bringing them back on
the flats, right?
So the climbs are my weaknessesand the flats are my and the
flats are my strengths.
And then, taking that to anintervention standpoint, can an
athlete at home statistically gothrough this?
(36:17):
Are there tools or methods outthere that are similar to what
you actually went through to,where an athlete can like break
down their own stuff, or a coachcan actually break this down
for an athlete?
Because I've done the samething with my athletes and it's,
trust me, it's a painstakingprocess.
We've like we generalize itinto like a five minute summary,
but maybe you can kind of takethe athletes through what are
the listeners, through what youwould actually have to do in
(36:39):
order to extract this from astatistical perspective, as
you've done.
Speaker 2 (36:43):
I'm not sure I understood you.
You mean more how everybody cando that kind of analysis at
home.
Speaker 1 (36:49):
Yeah, exactly Like if I were to just take, you know,
my race file from the.
The Cocodona two fifties isjust happening and I happened to
do that race a couple of yearsago, so it's at the top of my
Cocodona two, 50 or whateverrace.
How can I take that informationand say, okay, statistically
speaking, this is where mystrengths and weaknesses lie.
Speaker 2 (37:08):
Okay, actually, that's something that we try to
develop.
We have a patent on all theprocess that we developed and
now we are currently discussingwith some companies to try to
implement this kind of analysis.
Speaker 1 (37:22):
Somebody's going to charge us for it.
That's what you're telling me.
There's going to be an app thatI'm going to have to download
in order to tell it.
Okay, I get it.
Speaker 2 (37:30):
Yeah, and actually because we talked a lot about
the race analysis.
But actually we can do the samekind of thing than the record
profile instead of the data ofall your training and not only
your race and to look how yourprofile evolved by taking all
your records for all thegradients that you encountered
in your training.
But we also try to developsomething that I think that you
(37:53):
can try at home.
It's a test that we try todesign.
So let's say you go to runoutside and you will try to keep
your heart rate at a givenintensity that interests you for
your training, let's say 100and 60 RPM, because it's the
intensity that interests you.
(38:14):
And then you can go running invarious gradients and look at
the speed that you can maintainwith this specific heart rate.
So you have an idea of what isthe velocity that you're able to
run on this specific gradient.
And you can do that for anothergradient, etc.
And with only a few segments ofvarious gradients you only need
(38:37):
a few minutes to stabilize theheart rate.
So you can have the fullrelationship of what is the
velocity that I can use for thisgiven heart rate and in this
given gradient and then you cancompare to others, but you can
also compare to yourself.
Let's say, I need to train inthe steep gradient.
So you try to train, you startto train for a few weeks and you
(39:02):
want to see okay, I amimproving in this specific part.
You can do again the samesegment, going on this steep
slope, and look at the sameheart rate Are you going faster?
So that's a simple way, a quitesimple way that you can use,
with a simple testing to, to seeif you improve in this specific
ability that you targeted inyour training yeah, what
(39:24):
everybody's searching for intrial running right now is
analogous to the power durationcurve in cycling.
Speaker 1 (39:31):
so you can take any cyclist who has a reasonable
amount of power data and you canproduce a power duration curve
for cycling.
So you can take any cyclist whohas a reasonable amount of
power data and you can produce apower duration curve for them,
and then you could fit thatpower duration curve against a
Tour de France rider, a worldclass time trial athlete,
somebody who runs or somebodywho rides the kilometer on the
track any of those riders.
You can kind of pit themagainst each other, so to speak,
and see where rider A is goingto be better than writer B and
(39:53):
how they can like rack and stackand they even do a lot of like
still some old school like teamselections, kind of based off of
these, based off of theseprofiles.
People are searching for thatin trail running, which you're
you know what you're kind ofproposing, and I think that this
is a good tool to have.
This is essentially a velocitygrade curve curve as opposed to
a power duration curve, tocomplete some part, some
(40:16):
component of that analysis.
Speaker 2 (40:18):
Yeah, actually it's a force velocity duration Right,
because in cycling you canchange the gears so you can just
stick to the best condition,but when you are going up you do
not have gears when you'rerunning, so you have to go up
with your foot and no way other.
So you have to distinguish theforce and the velocity part in
(40:41):
the power.
That's why you need todistinguish force and velocity.
And then how force and velocityevolve with duration is another
component, and force andvelocity not necessarily evolve
the same way in the duration andfor different individuals they
can have a better durability invelocity or better durability in
(41:02):
force.
And that's another thing.
Speaker 1 (41:05):
And another question and another that we can analyze
the statistics nerds out therewill certainly get a lot of a
big kick out of some of thematerial that will be in the
show notes.
Speaker 2 (41:17):
Let's kind of move the conversation forward,
because we've kind of gonethrough this from a practical
standpoint in terms of howathletes can actually analyze
this no-transcript component andhow athletes kind of exhibit
(42:04):
some of the or how some of thethings that we were looking at
earlier change over the courseof a big, long race like utmb so
if we stick to the analogy ofthe force and velocity part or
the flat and happy componentthat that's what I was just
saying that there is notcorrelation between the ability
to continue being good andrunning fast on flat and all
(42:28):
uphill we can start with thefirst picture being the mean
picture of all of the UTMBfinishers that we analyzed.
There is something interestingis that on average the runners
are better to.
They conserved more.
They were more close to theirinitial capacity in the fourth
component.
So when it's going up you donot lose that much speed.
(42:52):
When the race is ongoing andbecause of the durability
components, this is an averageanalysis.
But on flats running there is abig difference between the
beginning of the race and theend of the race.
So if we just stop here for afew seconds, I would say on
average you probably need toimprove your ability to run on
(43:15):
flatland for the durability andvery long effort.
That means that probably whenyou prepare this kind of race
you can imagine having apre-fatigue boot or you will run
for, let's say, five, six,eight, ten hours, as you want
for your trail running training,and then go to a flat portion
(43:35):
and try to maintain your highspeed.
For the second part in thisstatic state, that could be a
recommendation based on thiskind of analysis.
Speaker 1 (43:44):
Yeah.
So this intervention, a finishfast run or a progression run or
doing intervals at the end ofthe run as opposed to the
beginning of the run.
Our coaching group has thisdebate kind of like ad nauseum,
and fundamentally you know somepeople in our coaching group,
that's probably why you'relaughing.
Fundamentally there's a lot ofnuance to this, but
fundamentally you're kind oftrading two things right, you do
(44:06):
the hard work at the beginning,you have the best capacity.
Right, you do the hard work atthe beginning, you have the best
capacity right.
So if you have the bestcapacity, theoretically you're
going to do the most kilojoulesof work.
You're operating at the highest, higher percentage of your VO2
max.
You're introducing kind of morestress which theoretically
would lead to more adaptation.
Right, more stress, moreadaptation.
You do the intervals at the end.
You're probably at a reducedcapacity.
(44:27):
Right, you can operate at asmaller percentage of your VO2
max or maybe for not as long ofa period of time, and so the
interval quality is certainlyreduced.
I don't think anybody wouldargue that, right, we might
argue the impact of that reducedquality session, but it
certainly is more specific to anultra marathon where durability
(44:48):
becomes exceedingly important.
Now you're adding in thisanalysis, which, which is once
again, there's a ton ofpractical takeaways to this, and
I think that this is actually abig one.
What your analysis wouldsuggest is that if you are doing
intervals at the end of the run, particularly to accentuate, or
hard work at the end of the run, a progression run, something
(45:10):
like that you're probably bettersuited doing that in a flat
level condition, because that isthe gradient, or lack thereof,
that gets impacted the most whenan athlete is fatigued.
It's not the uphill component.
So a better way to structure adurability type of workout would
(45:31):
be run for two hours and thendo an interval on flat level
terrain, versus run two hoursand do uphill intervals, which
you could still have thatcomponent of it based on the
race that you're actually seeing.
But if you have, if the racehas uphill and flats, the better
piece of it would be to do iton, to do it on the flats.
Is that what the research issuggesting?
Am I reading between the lineswell enough?
Speaker 2 (45:51):
there.
Yeah, actually it's really hardto.
We do not have an easy, rightanswer on the durability
question, because it's a hottopic, it's new even in the
scientific field, so I can't sayyou absolutely must do that.
But based on the analysis thatwe did and since we saw with
(46:11):
evidence that you lose morespeed and you lose more speed at
the end of the run up becauseof the flat running capacity, if
you have to choose what to do,I would probably suggest yes, to
train your durability componenton flats.
Speaker 3 (46:29):
Okay, let's check back in with coach adam on this
really neat concept of when wewould actually want to target
flat ground durability toimprove performance and how we
would actually do it yeah, well,first of all, I do think that
this, when I heard this, it kindof made sense to me, because I
(46:50):
pace people quite often at racesand this is something that I do
see quite a bit where you'rehiking uphill with someone late
in an ultra and they're actuallymoving pretty well, and then
you hit flat ground where youhave to run, and it's a
different story.
So it probably varies person toperson, race to race, but I
actually do think it's a themethat you can just see out there
(47:10):
on the ground as well.
As far as actually training it,I know in the podcast you guys
brought up potentially doingsome sort of faster work or
intervals on flat terrain at theend of a long run.
And another intermediate stepthat I might bring in an athlete
is just, you have your hillylong run and in the last 60, 90
minutes do it on flat terrain.
(47:31):
Maybe not even with the speedwith it, but just placing that
terrain at the end could be onestep along the way.
Speaker 1 (47:39):
Yeah, you know, western States is the classic
course that accentuates thisbecause it gets faster as the
race generally goes on,particularly as you leave
Robinson flat and then Michiganbluff and then forest hill.
It just kind of gets easier andeasier from a surface
standpoint and the athletes thatdo really well, the ones that
can just kind of run past runspecifically, not in quotes
(47:59):
actually run past forest hilland not do a lot of the
prototypical hiking and thingslike that.
But when Batiste actuallymentioned this, I thought of a
paper that a mutual colleague ofmine was a co-author on, and
I'll link that paper up in theshow notes, and it posited the
co-workers, nick Tiller alongwith Guy Mier, who are the two
(48:20):
authors of this paper.
Their primary thesis is thelimiting factor in ultra running
is at the level of the muscle,and particularly muscle damage,
which would be accentuatedduring running more than it
would hiking, because running isa more, let's just say, a
muscularly intensive mode oflocomotion as opposed to hiking.
(48:43):
Uphill, downhill, running wouldprobably be the most muscularly
demanding component of all thethree disciplines, or all the
disciplines that we can likethat we experience in ultra
running.
And so maybe if we took thisresearch step further and
further the speculation that wewould see this.
We would see this componentdiverge even further on the
(49:04):
downhills, where people getdisproportionately people who
are doing or who are slower getdisproportionately worse,
specifically on the downhillsthe most, then on the flats the
second most, and then don'treally deviate that much on the
uphills, but who knows?
But some interesting concepts,so to speak.
And the final caveat that I'lladd before we go back to the
main conversation is rememberthis analysis was done on one
(49:26):
specific race and we have to putthat into context with all the
races that athletes are doingand that may or may not look
like the same course profile.
Speaker 2 (49:37):
But this was an average relationship because we
evidenced, even if they are notthe case, that we saw the most,
but it could be the opposite forsome of the athletes that were
able to maintain quite a goodspeed and flat but was really
slower on the steepest gradients.
So you do, we cannot have justone answer for everybody, but on
(49:58):
average I tell you that smallflatland that you can play, but
specifically individually it canchange well, and if an athlete
wants to apply it to themselves,they can look at.
Speaker 1 (50:10):
This is why post-race notes and post-workout notes
become so incredibly importantDuring the end of the race.
I could do this and I couldn'tdo that.
That's the lens that you canlook through to determine okay,
we can improve this or improvethat.
You can also corroborate thatwith the actual file itself and
this is something that I do withmy athletes is I look at how
(50:31):
each segment deteriorates overtime the uphill segments, the
flat segments and the downhillsegments and the gauge.
I'm kind of like revealing thecoaching process.
There's not like it's likeproprietary to me or anything.
I didn't come up with it.
Somebody taught me how to dothis and then I'm just doing it,
you know 20 years later, butwe're simply using intensity
factor, as the intensity factoris just the ratio of the speed
(50:56):
that they're running to theirthreshold speed.
So if the if they're runningfaster than the threshold speed,
the intensity factor is overone.
If they're running slower thantheir threshold speed, intensity
factors under one.
It's almost always under one inany ultra marathon event, but
we look at how that deterioratesover the course of a race and
do the uphill segmentsdeteriorate more than the flat
segments or do they deteriorateless than the flat segments as
(51:18):
compared to the downhillsegments?
And those are the three bucketsthat I essentially put it in
and that becomes a lens to lookthrough to say, okay, we're
going to improve here from yearto year or based on this race,
and that improvement can comethrough training, it can come
through pacing, it can comethrough nutrition interventions.
There's a whole host of things.
It's not just do progressionruns or anything like that.
(51:39):
That's not the answer toeverything.
But my point with that is wehave tools that we can look at
it from a statisticalperspective as well as a
subjective perspective, how theathletes actually experience
things.
Speaker 2 (51:52):
Yeah, and how the runner and the athletes perceive
something is really important,but I really think that we
should always mix the perceivedthing with the objective
information, because sometimesyou can have a difference
between how the athleteperceives a thing and what
actually the watch and the timesay.
Speaker 3 (52:13):
Exactly.
Speaker 2 (52:15):
Especially at the end of an ultramarathon, where your
mind is, in a way, not in thefresh state that you have at the
beginning, so you can have somediscrepancies between the
perception and what actuallyoccurred.
Speaker 1 (52:28):
Okay, this has been amazing because I like getting
the.
I like getting the curtainpeeled back a little bit with
the people who do these types ofanalyses, and I do think that
this is one where we can come upwith a lot of functional ways
or functional takeaways in termsof how we can start to analyze,
you know, race performance andthings like that.
I know you've got a lot ofthings on your wishlist right
(52:50):
now.
In fact, the people who areviewing the video version of
this will see a whiteboardbehind the Baptiste's shoulders
that has a lot of equations andquadratic formulas and things
like that, written meaningyou're at work doing something.
But I kind of want to know,like, what's on your wishlist to
develop.
Do you now have like, thisframework to say, okay, we can
(53:11):
analyze race performance likethis, how do we like what's on?
What would you like to do withthat?
If you could just likeinstantly wave a magic wand and
put it in the hands of youraverage or elite trail runner in
order to improve theirperformance with, like, your
knowledge base, what would youwant that to look like?
Speaker 2 (53:26):
We have several things that we could do all
seven.
We can go all day with this.
Well, the thing that is, I think, really important is to develop
some tool that we can provideto the community.
So so we can use it and we cannot just look at two elite
runners or not just at thefinisher of the 2023 edition of
(53:47):
the utmb, but but, as you say,it's a descriptive way to see at
the performance, but actuallywe don't know if the durability
of the force or the velocity isrelated to pacing, is related to
the nutrition or related to theheat of the race.
There's a lot of factors that wewant and we want to use this
framework to analyze all of that, all of the nutrition, the heat
(54:10):
, the pacing, etc.
And, for example, I've beensolicited by a team in the UK
that will run an experiment inthe Western States this year in
a few weeks.
They are really interested inthe heat acclimatization aspects
for this race and we will runthis analysis that we did for
the UTMB and the Western Stateand a few participants to see
(54:34):
how the duration, the durabilityprofile on force, on velocity,
on intermediate slope, ondownhill, etc.
How all of that is potentiallyrelated to the heat, to the core
temperature of the athletes.
So that's an example of what wewant to do with this framework
that we developed and that canbe useful tomorrow for analyzing
(54:56):
and reanalyzing again some ofaspects that are related to
performance in trail running,but with a new, potentially more
sensible tool that we developed.
Speaker 1 (55:07):
Amazing.
I hope all of those come tolight.
I'm going to leave links in theshow notes to some of the
research that we developedAmazing, I hope all of those
come to light.
I'm going to leave links in theshow notes to some of the
research that we've eitherdirectly referenced or kind of
tangentially referenced.
I'll get some of those linksfrom you after the fact as well,
because there's probably somethings that I was missing in the
research that our fellowcolleague Fred and I did.
But if people just want to likelook you up and look up the
(55:28):
research and the work that youdo, how are they going to go and
find you?
How they can find me?
Speaker 2 (55:33):
Yeah, Research gate Twitter, Instagram they're going
to direct email address.
Speaker 1 (55:37):
Some people give their cell phone numbers out
every once in a while on thispodcast, which I'm kind of like
a little leery and actually puton the air, but like, how do you
want people to find out moreabout you?
Speaker 2 (55:46):
Yeah, they can find me on all the social media and
twitter, instagram, etc.
I will send you the link sothey can reach me.
And we have also a projectnamed predict trail where we
want to develop this tool thatwe want to spread.
You can find onwwwpredictrailcom.
You can share your link thatthat's another project that we
(56:08):
have and where you can find me.
Speaker 1 (56:09):
Brilliant.
Well, thanks for your time,thanks for your research.
I wish you all the best.
We're going to bring you backon this podcast whenever any of
these probably gated orpaywalled or whatever versions
of these performance analysistools actually come out and you
can explain them.
More power to you.
It's absolutely going to happen.
I'm here for it.
Speaker 2 (56:35):
Any way that we can slice and dice performance is
something that I'm going to be apart of, and we'll nerd out on
how to analyze performanceprobably at another future point
.
Thank you very much.
Speaker 1 (56:42):
All right, folks, there you have it.
There you go.
Much thanks to Batiste andCoach Adam for coming on the
podcast today and, as Imentioned during the intro, I do
think that this is somethingthat we can deploy in our
day-to-day training withathletes and also think about
how we can race smarter.
As always, this podcast isbrought to you sponsorship and
advertisement free, so if youwould like to support this
(57:05):
podcast, the one thing that youcan do is subscribe to Research
Essentials for Ultra Running.
We are actually going to takesome of Batiste's research and
break it down even further inthat newsletter, in an upcoming
version of that newsletter, andso if you really want to geek
out on all the stats, statisticsand find out what we think
about it in a more detailedversion, more detailed manner,
(57:26):
research Essentials for UltraRunning is your ticket.
All right, folks, that is itfor today and, as always, we
will see you out on the trails.
Trail and ultra runners.
What is going on?
Welcome to another episode ofthe coop cast.
As always, I am your host,coach Jason coop, and on this
podcast we oftentimes gleaninsight from the lab and then we
use it in the field ofcompetition.
Other times, we look at what isactually going on in the field
of competition and we gleaninsights from that on how to
(00:30):
train and race smarter.
And today's guest was an authoron a paper that did just that.
Welcome to the podcast today forthe first time, batiste Moral,
who took publicly availableStrava data from the 2023 UTMB
to determine what pacingstrategies worked better, and he
also used those same techniquesto determine how people in, or
(00:51):
how the elite field specifically, evolve over time.
And the results of all of thisanalysis is something that I
feel we can take to training andracing, and it's something that
I'm taking into practice withmany of the athletes that I work
with currently.
I'm going to let Batisteexplain exactly what he did and
what the results of his researchactually are and how we can
(01:12):
apply it.
So with that out of the way,I'm getting right out of the way
.
Here's my conversation withBatiste Moral.
Batiste, welcome to the podcasttoday.
I appreciate you coming on.
I'm really excited to talkabout some specific pieces of
research that you've kind ofrecently published.
(01:33):
But before we get into it toomuch, can you give the listeners
a little bit more of abackground of who you are, how
you got involved in trailrunning and looking at this from
a scientific perspective in thefirst place?
Speaker 2 (01:43):
and looking at this from a scientific perspective in
the first place.
Yeah, thank you for theinvitation.
First, I'm a researcher rightnow, so it's been 10 years.
I defended my PhD and I startedworking on performance sports
performance, but more on rugbyplayers, but we will see later.
There is some comparison thatwe can do with trail runners.
So I work on physiology andbiomechanics fields and I do
(02:06):
research on sport performance,but also on health and even on
animals in ecology.
That's another question, butyou will see there is some
crossroads between all of thesetopics.
So that's my background.
Speaker 1 (02:18):
Perfect.
So, the part of your researchthat really piqued my interest
really gets to the core and thesoul of what athletes go through
every year.
But most ultramarathon athletes, they're experienced athletes.
They're not novel or new to thescene.
They have marathon PRs andthey've done ultramarathon races
before and things like that.
(02:38):
And the coaches who arelistening to the podcast can
resonate with this run of show.
At the end of every year and atthe beginning of the next season
they're going through theirRolodex of athletes, they're
doing their consults and thequestion that comes up how am I
going to get better?
Where am I going to get better?
What am I going to do to takemy performance from X to Y based
(03:01):
on the previous year?
It's something that's kind oflike top of mind for everybody
and really, when you think aboutit, it's kind of the root of
all interventions whether it'sinterval intervention or a
nutrition intervention, like asupplement or anything that you
can kind of like throw at anathlete is all kind of aimed to
improve them specifically fromone year to the next.
And a lot of your and a lot ofyour research has looked at that
(03:23):
specifically, like where haveathletes actually improved?
And so I want you to kind oflike take it from there with
this.
Maybe we can look at howcertain athletes have improved
from year to year and how we canactually take this really broad
brushstroke of looking at arace and trying to dissect it in
(03:44):
terms of where can thoseathletes actually improve?
That's a big question.
Speaker 2 (03:50):
that's a big question because it's not easy with so
much performance factors,especially in trail running and
ultra trail running even moreprobably, and there is multiple
way to to approach and toaddress this question, and I
tried to do some new way toanalyze the athlete performance
(04:12):
based on the data that we.
You know, there is a lot of bigdata coming from the sports
performance and this issomething that we can go deeper
on it to see how were theperformance on one year and how
it can evolve, and so you canhave like objective metrics like
that help you to choose, tolook at if the intervention and
(04:32):
the training, the thing that youchange, maybe the nutrition and
all we mentioned how iteffectively and change my
profile and so making me moreperformant.
Speaker 1 (04:43):
And the way that you kind of like looked at this was
looking at race profiles, right,and specifically UTMB race
profiles, and we can take acouple of high level athletes,
but let's take it at like kindof like a population level first
.
Right, you wanted to look atUTMB from an entire race
perspective.
What is the entire you know,the entirety of the race
actually doing?
What did you glean from that?
(05:04):
First off, what are youspecifically looking at?
We're going to have to get intoforce velocity relationships
and some of the thingsspecifically that you're looking
at and define those for theusers, but take it from the
highest level possible.
How are you trying to dissectthis problem of where can
athletes actually get better?
Speaker 2 (05:19):
yeah, actually the first thing I did was in 2023,
at the end of August, justbefore the UTMB this year.
And that was for fun, becausesome of my colleagues just asked
me who will won the UTMB thisyear and I have no idea.
But I say, okay, we'll take alook at the data and see if
there is some piece ofinformation in there.
So I look at the race, Ianalyze the race, coming from
(05:43):
Kilian John that won the 2022edition and I look at one of the
guys who will probably win thisyear that was Jim Wemsley in
2023.
So I look at the race in 2022and say, okay, let's dissect the
race and we will cut it intosegments, into portion, every
(06:04):
100 meters of this race.
And I wanted to compare theirprofile on each 100 meters of
the race to see who is the beston which segment.
And there was a result reallyinteresting because you know,
there is flat portion, there isuphill with more or less steep
gradient, really steep gradient.
(06:24):
And the thing that was reallyinteresting, that was that if
you look at the race and all thesegments when it was flat or
low gradient, or downwind oruphill, there is not that much
differences between the race ofkian and jim.
But when you compare them, onlylooking at the steep gradient
let's say above 20% gradient,there was a big difference
(06:49):
between Kylian and Jim this year.
Actually, jim lost.
Almost all of the difference onthe finish line was because of
the steep gradient.
So that was the first approachto say, okay, he seems to have
kind of a weakness on thisspecific part of the race, so
that's only one way to approachit because the determinants of
(07:11):
the performance are complex.
But there was like a first step.
Okay, there is somethingdifferent between these two top
elite athletes on this specificportion and to get into that
specifically 76%.
Speaker 1 (07:26):
So three quarters of the time difference was lost on
sections above a 20% grade.
That's what your analysis foundcorrect, yeah yeah, that's
correct.
Speaker 2 (07:38):
So more than one hour was only on those portions, but
they only represent less than10% of the whole race, so most
of the difference was done onvery specific portion of the
race.
Speaker 1 (07:51):
Okay.
So that's kind of like we'regoing to take this like Jim
versus Killian and now we'regoing to move it into like Jim
versus Jim, because I do thinkthat this sets this up for the
population level thing.
A lot of people listening willsay, okay, yeah, big deal,
killian's like pushing it harderon the steepest climbs, because
that story is very good.
But there's another analysisthat you looked at which is gym
(08:12):
versus gym.
So gym pre-winning, utmb, gymtraining in the Alps for a
couple of years very famousstory here in the US.
Some you know an American movesover here specifically to train
for here.
I'm now located in Chimney,france.
So for the listeners out therewho are wondering why I'm using
that vocabulary, it's because ofthat Comes over to the Alps
(08:34):
specifically to train and winUTMB and ends up doing it.
Do you analyze Jim'sperformances before that
transition and then after thattransition?
What did that actually reveal?
Speaker 2 (08:50):
Yeah, actually that was the purpose of this analysis
, because everybody knew that hecame to the ops probably trying
to win the utmb on the nextyear and actually he did a 100
race in april this year.
It was in croatia, if Iremember well.
So so I we found the data on.
We found the data of the raceand analyze it to see if there
is a changing in the profile ofgym, especially in this specific
(09:11):
steep slopes.
So we did exactly the sameanalysis, comparing the profile
of gym in the utmb 2022 and theprofile of the istria, the 100
istria race in croatia in 2023.
And actually we saw a bigimprovement on the ability to
maintain a good speed in thesteep slopes, in the steepest
(09:32):
slopes.
That's why we thought thatprobably the training he did
this year in the Alps helped himto improve in this specific
part, the part that was aweakness for the last, the year
previous, the previous year.
So he was better on thisspecific slopes and he was as
(09:53):
good as before on all the othersegment of the race.
So that's why we, at this time,I thought that he was probably
able to be one of the topathletes and probably winning
the race because of thisanalysis.
So it's always a verychallenging race and with a lot
of random aspects that we can'tcontrol.
(10:14):
But they're training hello himto be better on this specific
part of his profile, so probablybetter at the utmb next year.
Speaker 1 (10:24):
Well, okay, so now we're going to get it into the
general population right Withthat as a little bit of a back,
with that as a little bit of abackdrop.
That happens specifically, mostlikely because of an athlete's
environment changes.
They change their environment.
They train more in steepertypes of train.
It makes all the sense of theworld with that.
That's where they are going toimprove in the most.
(10:44):
But let's back this, let'sbroaden this out a little bit,
because we're talking about twovery good athletes that are
training for a very specificpurpose and, as you mentioned,
they have narrow margins toimprove upon.
And usually those narrowmargins get even narrower
because it's a very specificelement that they're improving
with.
I know I go through this with myelite athletes every single
(11:06):
year when are you going toimprove?
And it's usually not a generalfitness proposition.
We try to maintain that as muchas possible.
Usually we're trying to improvea specific component of their
fitness, maybe a specificcomponent of the entire of the
preparation arc where we screwedup one phase or something like
that.
Maybe it's something withstrength training.
Maybe it's something withstrength training, maybe it's
(11:31):
something with nutrition, but itusually is because they are so
good it's sliced out and it'sfocused on in a very narrow
perspective.
You contrast that with yourgeneral pop athlete, somebody
who's only been training forthree or four years.
They kind of improve across theboard with just generalize
training.
So let's take that biggerpicture approach.
You were part of a study thatlooked at 600 athletes.
Use the Strava data and it cangeneralize athletes into a few
different like profiles.
(11:52):
So first off go over like whatthe profile is that you're
actually looking at and thenwhat the actual research found,
and then we'll get into somepractical takeaways from that.
Speaker 2 (12:01):
Yeah, actually, at the beginning it was an analysis
of just two top elite athletesfor fun, with my team and we did
a meeting at the end of Augustand say, oh, that's very
interesting, we probably need togo to see all of the other
participants to see what we canidentify in terms of profile.
So, to improve, so I bring myteam at home and we open all of
(12:25):
our computers so to bring backall the data and say, okay, now
we have the database that we cananalyze to see how we can
describe the profile of all theathletes during this specific
race of the under edition of2023 of the UTMB.
So the idea was to do the exactsame thing that for the NMGM
(12:48):
look at every 100 meters segmentof the race and to see at which
speed they were able to run fordepending on the gradients.
That was the idea.
We all know that there is agrade adjusted pace.
You can find that on your watch, on your Strava or Garmin
account.
But most of the time thisgrade-adjusted pace is just
(13:10):
computed on the mean of apopulation.
So it gives you an idea of howyou can translate a speed when
you are going up compared to theflat.
Speaker 1 (13:18):
But actually what we observed was that there were
really big differences on how anathlete will run, depending on
his own profile okay, before wego any further, I want to take
some time and explain what ngpand gap actually are, because
(13:38):
both of these terms are kind ofan alphabet soup, so to speak,
and they try to do the samething although they get about it
at different ways.
So NGP or normalized gradedpaste, is a is something
proprietary to training peaksand all it tries to do is it
tries to make the pace that youare actually seeing on a run
(13:59):
during a post hoc analysis whenyou're actually looking at the
file and make an equivalent forthat pace of what it would be on
flat level ground.
Now the two things thattraining peaks is adjusting for,
our first off, thenormalization piece of it.
That's the first part of theterm normalized graded pace and
the second piece of it is thegraded part of it.
(14:20):
Now, most people will befamiliar with the grading part
of it.
It just means running uphill isharder than running on the
flats and running downhill iseasier than running on the flats
and running uphill.
The second part of that, thenormalization part it is a
little bit trickier and honestlydoesn't have that big of an
influence on the number, but itnormalizes the pace to what you
(14:42):
could run if you ran the pacecompletely evenly.
So for example, if you go outand run one mile in six minutes
and you do every single lap ofthat mile at 90 seconds, so the
first quarter is 90 seconds, atthe halfway mark or at 800
meters you hit 130.
At the 1200 meter mark you hit430.
And then at the end of the mileyou hit six minutes, you've run
(15:03):
that mile completely evenly.
Your normalized graded pacewould be six minutes per mile.
Now if you decided to mismatchthat pace, or you decided to run
the first quarter in 60 secondsand the second quarter slower
and the third quarter faster,the normalization component of
that would be faster than sixminute miles, because the
physiological toll is harderthan six minutes per mile.
(15:25):
So all it's trying to do is it'strying to make a physiological
equivalent of both the variationin pacing and the variation in
grades and normalize it to flatlevel terrain.
Strava's equivalent of this,which is gap or grade adjusted
pace, only takes the gradingcomponent of it.
So it looks at how fast you'rerunning and what the grade
(15:45):
you're running on and it triesto make a physiological
equivalent for that for flatlevel terrain.
Now, as Batiste mentioned, bothof these are based off of
research that is done on tonsand tons of people and, as we
know, certain people are moreefficient or effective at
running uphill and downhill andso their particular
physiological cost might not bequite accurately reflected in
(16:07):
these grade adjusted paces andnormalized graded paces.
And if we can look at thatacross a whole host of athletes
we might be able to glean someinsight.
But, needless to say, with bothof these, all of all that they
are trying to do is it's tryingto look at the variables around
the actual pace that you ranuphill, downhill, the variation
in those paces and trainingpeaks case and saying what is
(16:28):
the physiological equivalent, or, more accurately, the oxygen
cost equivalent of if thisathlete actually ran on flat
level terrain, what would thepace actually be?
Speaker 2 (16:43):
I would say you all know people that are really good
on flatland and on levelrunning but when it's going up
it's hard and you have guys thatnot that much fast on flatland
but when you are going up theyare really good.
And that was this idea how wecan model the profile of an
athlete to determine his owngrade, adjusted pace so we know
(17:06):
he is he able to run with a lotof speed.
This is the velocity part ofthe profile, and when you are
going up you have to push more,produce more force so you can
elevate your center of mass in abiomechanical perspective.
So how force and velocityinteract within a profile, can
we characterize an athlete'sspeed on this profile?
(17:27):
And that will probably opensome new information on how we
can train for the next year orfor the next race.
Speaker 1 (17:36):
And so what were the different profiles that you
found throughout all of the UTMBparticipants, and how did those
profiles correlate to their endperformance, so to speak?
Was there a profile that wasmore advantageous for the race,
least advantageous for the race,more advantageous for the
second half of the race versusthe first half of the race?
What did the data miningactually glean out if we were to
(17:57):
build this perfect athlete thatwe wanted to put on the start
line for UTMB?
Speaker 2 (18:02):
For the comparison between the first half and last
half of the race, we will talkabout that later because it will
imply durability aspects, solet's talk to that after.
But for the force, velocity,how you're able to run fast when
it's steep, with steep gradient, or how you're able to run fast
when it's flat, if you look atall the participants, actually
(18:22):
we had the race data coming fromthe winner of this edition and
the race data coming from thelast guy who crossed the finish
line, so that's evident.
But, yes, the best athletes arebetter on each portion of the
race, right?
they run faster on flat andfaster when it, when you have,
when they are going uphill.
But but we ran a secondanalysis looking at people and
(18:45):
runners with quite the sameperformance.
So we cut all the population in10 percentiles.
So let's take the 10 bestathletes of this race, or the
athlete between the 40 and 50percentile, so we can analyze a
comparable performance.
And in this specific analysis,the very interesting thing with
(19:06):
that was that there was notoptimal profile.
You can do the exact same racetime but with a very different
profile, and probably the guyyou would cross the finish line
with him you never saw himbecause when you are behind and
when it's going down you arebelow.
Speaker 1 (19:28):
Everybody's had that experience.
I mean, that's like a reallyrealistic thing.
You can cross the finish line.
You look to the person whofinished in front of you, the
person who finished behind youand two places behind you, and
maybe you saw one of thosepeople for a part of the race.
Aside from that, they're likeyou said they're faster in
different sections and slower ondifferent sections,
particularly races that have alot of different features steep
(19:50):
uphills gradual uphills, steepuphills, gradual uphills, steep
downhills, gradual downhills,flats and things like that.
Speaker 2 (20:00):
Yeah, and I feel like this is something that everyone
who did a trail, who ran atrail race, know that, because
you know that you will be betteron some portion, and then the
guy that let you down you willcome back, et cetera.
But the thing interesting isthat now we have a model so we
can precisely evaluate this.
And since we can evaluate this,we have indices indexed that
can use for testing and foroptimizing training or to open
(20:23):
new idea of how I can train nowso to be better.
For example, if you compareyour profile with the guy who
finished the race at the sametime of you, but if you see that
for every flat portion he wasbetter, you probably need to
train more on these flatsections and it's probably this
(20:45):
training that can help you to bebetter on the next race.
Speaker 1 (20:51):
Let's check in with Coach Adam really quickly on how
we as coaches determine thestrengths and weaknesses with
athletes.
Speaker 3 (21:00):
So short answer.
It's not always easy, but thefirst place I start is actually
just asking the athlete, and youcan kind of get this when you
have a call after a race andthey have some perspectives on
what went well and what didn't.
And I also want to combine thatwith the real, objective data
that we're seeing in the racefiles.
(21:21):
So a lot of times people mightfinish a race and say, man, I
suck at climbing, and I alwayswant to make sure that is that
the sensation of sucking atclimbing, or is it a potentially
a reality?
Because there's a lot of thingsthat are simply hard running in
the heat, things like that.
I mean, it's easy to perceive aweakness when it's really just
everyone kind of performs worsein those conditions.
(21:42):
So we see if, potentially,there's a mismatch with your
normalized graded pace in theclimbs versus the descents in
something that would besignificant or material.
Usually we probably would haveseen that in training already,
but hopefully we can see that issome sort of weakness.
But it might be something thatshows up only late in a race too
(22:02):
.
So that's something we want tolook at.
Another thing that we havethat's a tool at some races is
the kind of live tracking whereit shows your position
throughout the race and that canactually give you some good
ideas of your strengths andweaknesses, based on where are
you passing people and where areyou being passed.
I like to use that quite a bit,actually.
So there's no perfect answer,and I think one of the
(22:24):
weaknesses that we have rightnow is that only the more strong
edge cases stand out, maybe thepeople on the edges of the bell
curve.
But with some more tools whatwas done in this study we might
be able to use a bit more of afine tooth comb.
Speaker 1 (22:41):
So that's what I wanted to kind of get into next
before we get into the firsthalf back half piece, and just
as a little bit of a depthbackdrop.
The listeners who have beenkind of constantly tuned into
this podcast will realize thatthis concept of durability that
is, a relatively newerphenomenon or concept that we've
started to talk about in sportsscience and in performance has
(23:02):
continued to come up in thispodcast via a lot of different
angles and sometimes ones thatwhere I'm not really like
seeking it out, it just happensto be part of the fabric of the
dialogue.
And so this podcast runs trueto form, where we're going to
talk a little bit aboutdurability as well.
But we're going to put it,we're going to put a pin in that
.
For just one second, thesynopsis of how to look at this
(23:24):
data can really be boiled downto how can one person get better
knowing their profile from yearto year?
Not necessarily and I was kindof baiting you with the way that
I phrased the question or likehow to build the perfect athlete
for the race, right?
I don't think that's the valuein knowing this information.
It's taking Jason Koops' datafrom UTMB or from other races or
(23:49):
whatever looking at it and thenextracting something from that
to train for the next year andthe year after that.
How might one actually do that?
I mean, this is something thatyou and your team put together,
but can we put this in morepractical aspects for the
coaches and the athletes thatare out there that do look at
their race files and do look atother people's race files a lot?
(24:11):
How can we actually go aboutdoing that?
Speaker 2 (24:14):
yeah, actually there is something interesting that we
are doing also in my lab.
This is an experiment andresearch that we're doing right
now, but that had reallypractical application.
The idea is that I told you thatwhen you're going up, you are
producing more force, andsomething that we did was to
compare the force that youproduce when you are going up by
(24:35):
running or using sled training,like like in track and track
and training resisted sprintingor resisted pushing right.
Yeah, exactly, and what we sawwas that actually this is almost
the same biomechanicalorganization of your body when
you are pulling a sled or whenyou are going up.
(24:55):
So that's opened newopportunities to train, for
example, if you have a profilethat is not that good in the
force portion.
So when you're going up andlet's say, maybe you are living
in a city with not that muchgradients, so you are not able
to train in this kind ofgradients you probably can find
(25:16):
some alternative with this kindof training.
So that's just an example.
But to say, you can be betterand be more endurance in the
fourth part of your profile, notonly by running uphill, but you
can find other alternativeswith strength training and, in
this specific example, with sledtraining that can mimic a slope
(25:38):
.
Speaker 1 (25:38):
And so what you're trying to do is, like you said,
improve the force part of theequation.
Speaker 2 (25:42):
Ah, because the thing interesting was that there is
no correlation between thevelocity part and the force part
.
So, just to be clear, thevelocity part is your when we
are talking about train, is yourability to run fast on flat
land, and the force part is areyou able to run fast or to walk
fast when you are going up?
So this is how we candifferentiate these two parts,
(26:05):
and the interesting thing isthat there is no correlation
between these two parts of theprofile.
That's what we were talkingabout at the beginning of the
talk.
So if it's not the same ability, you can probably train that
differently.
And when I said you can improvethe force part of your profile,
(26:26):
you probably need to find a wayto train your ability of your
muscle to produce higher levelof force for a longer time and
it probably will improve betteryour profile on this part, but
not necessarily changing theother part being the velocity
part the.
Speaker 1 (26:42):
So the analogy that I've always used to to that
where you're saying the forcepart is not related to the
velocity part, or let me kind ofcolloquialize that a little bit
the uphill performance is notrelated to the flat part, or let
me kind of colloquialize that alittle bit the uphill
performance is not related tothe flat level performance is
that a fair colloquialization ofwhat you're trying to describe
from a, from a biomechanicsstandpoint?
Speaker 2 (27:05):
yeah, totally.
Actually, we all.
What we are doing is definitelynot a revolution.
You don't know that, that kindof aspect, because we saw that
in training and we saw that whenwe're racing.
But the thing interesting isthat we can know model it, we
can know evaluate it.
So our objective informationsso we can stand on to build the
(27:26):
training of your athletes yeah.
Speaker 1 (27:28):
so the analogy that I've always used for that the
uphill component and the flatcomponent and I actually add two
more components to that that wego completely off the rails, we
can talk about those for alittle bit is the downhill
component and the walkingcomponent, and I always liken
those to different sports.
They're sure linked by thecardiopulmonary system and the
(27:50):
musculoskeletal system to somecertain degree and have a little
maybe a little bit of Venndiagram overlap.
Although you're describing theuphill component and the flat
component as not being related,but I still think that from a
training standpoint we do needto think about training all four
of these.
Just like a triathlete trainsall three swim, bike run I think
(28:10):
that analogy kind of playsthrough and the more
statistically we look at raceperformance, the more that
analogy starts to actually come.
It actually comes to thatrealization that it's not just
an analogy, it's actually what'sgoing on.
Speaker 2 (28:24):
Yeah, but you agree that if you have a limited time
to train, a limited volume, youcannot do all the training you
want, because you can enrich,over reaching or over training
or that especially for eliteathletes.
So, to be more precise, and ifyou more tailored the
intervention and the trainingthat you do because you know
that you cannot do infinitetraining probably you should
(28:48):
find the the best part of yourprofile to train.
And just something that Iwanted to say before.
But it also depends on the race, because 100 miles race
training, race can be reallydifferent and the profile can be
really different, and not onlyon the total elevation, because
it's something that is oftenlooked at okay, there is, let's
(29:10):
say, 10,000 meters of elevationgain in this race, but you can
do 10 000 meter elevation gainin very different way correct
and if you're going really withreally steep slopes at more than
40 or 50 percent, or only along portion with 15 percent
slopes, that's not at all thesame in terms of the physical
(29:33):
ability that your trainer needsto have.
Speaker 1 (29:36):
Yeah.
So I get this question actuallya lot because there's this
section of my book where we talkabout how to tailor the uphill,
downhill components of trainingto the race.
And the way that I illustratedin the book was with the
broadest brushstroke, and I'mgoing to do it in.
I'm going to do it in us unitsas opposed to metric units, so
(29:59):
we can create the.
We can create the translationlater.
But if your race has 200 feetof elevation gain per mile,
which would be approximatelywhat the Western States 100 has,
try to average that over thecourse of a week of training.
Now to your point.
You can come up with that ahundred different ways.
You can do all of thatelevation gain in one run up a
(30:20):
40% slope, or you can do it up5% slope, 7% slope, 5%.
There's a lot of different waysto do it.
I guess is what I'm saying, andso I get a lot of questions on
my book related to just that.
How do I come up with that?
So the first step is just hitthe average.
Second step is try to hit thegrades that you're actually
going to experience during therace.
(30:43):
Interestingly enough, I watcheda presentation that you gave.
It was in French, but theEnglish subtitles are apparently
pretty good, according to myFrench speaking counterparts.
And one of the really neatslides that you had during the
course of that presentationcompared UTMB and Western States
in terms of the percentage ofthe distance that the runners
(31:04):
would spend at each grade.
And the point that you'retrying to make is between zero
and 15%.
They're about equal, meaningrunners are going to spend about
the same amount of time runningacross those grades 15 to 30%.
Utmb has over twice the amountof distance as compared to UTMB
(31:25):
at that grade.
And then over 30%, utmb hasseven times the amount of
distance at that grade comparedto Western States, and there's
hardly any part of the WesternStates and there's hardly any
part of the Western Statescourse.
Having been on both of thosecourses, that is at that percent
grade.
But what that reallyillustrated to me is how two
marquee 100-mile races, themarquee 100-mile races in the
world, can be so dramaticallydifferent.
(31:47):
And I look at it through thelens of they're dramatically
different from a demandsperspective, because the
gradient that they're actuallyrunning on not just the total
elevation gain, elevation loss,but the actual gradient that
they're running on can bethought of as a specific demand
of the race yeah, yeah,definitely.
Speaker 2 (32:02):
That's really important to understand that,
because we often look at therace.
Metrics are the mean big thing,but we we definitely need to go
more into detail on how the raceoccurred really and what are
the gradients that you will findin this race, because it's not
the same to do one kilometer at30 percent or two kilometers at
(32:24):
15.
It's just not the same in termsof how your body will react.
And maybe another thinginteresting is that we're a lot
talking about the steep gradientand the flat and the level
running, but actually there isalso a component for
intermediate gradients.
I mean that we found someprofiles of runners that were
able to run at the same speed instrip gradients, the same speed
(32:47):
in flats land, but they werenot able to run at the same
speed for the intermediategradients, especially in the
area of the walk-run transitionand when you have to walk at
high speed or to run at lowspeed, some of the assets are
not good and other ones are good, and that's another component
of the profile that we analyzewith the model we proposed.
Speaker 1 (33:09):
I have a colleague of mine, jackson Brill.
Shout out to Jackson, who wasactually an intern with us a
long time ago.
He was actually trying to workon this in Roger Cromwell's lab
at the University of Coloradoand find out where the right
walk to run transition actuallyexists.
And does it actually exist?
And the answer is no, you'vegot to use your own intuition
for it.
It might be a conversation foranother day, but I'll link that
(33:32):
conversation up that I had withJackson in the show notes for
anybody that's interested inmaybe trying to find it or
trying to find a better way todetermine when that walk to run
transition for you shouldactually be.
I want to spend a little bitmore time on this Western state
versus UTMB component of it.
One of the things that that thatillustration that I just
mentioned the percentage ofdistance that the runners are
(33:53):
spending at each kind of likegradient bin or gradient chunk
being so different between thosetwo races, also highlights to
me how difficult it is to getboth of those races right, even
though they're both a hundredmile races, even though they're
both long and you know duration,and it's very far out to the
right on the power durationcurve which we haven't really
(34:16):
gotten into yet, and they seemlike low intensity activities
and things like that.
You're running at a lowerpercentage of your VO2 max way
lower than a marathon oranything like that.
But because the courses are sodifferent, it's hard to get both
of those right in a short, in areally short timeframe, and
we've seen that play outespecially really short
timeframe, and we've seen thatplay out especially in the elite
field over the years, that it'shard to have an athlete do both
of those.
(34:37):
And I happen to have an athlete, Katie Scheid, who did both of
those successfully one year.
But it is exceedingly rare.
And one of the reasons it'sexceedingly rare is because we
see this kind of marketdifference in the course
profiles, which is kind of ano-duh moment.
But your research is reallybringing to light how different
they can actually or howdifferent the demands actually
(34:58):
are.
I guess is what I'm trying tosay.
Speaker 2 (34:59):
Yeah, yeah, actually we can think that it's the same
race, because just it's 100miles right 160K.
But if I say, if you take a100-meter dash sprinter, you
will never perform on marathonand the opposite, that's evident
for every everybody.
But actually that's the samefor various different race of
(35:20):
training race that do notrequire the same abilities and
physical capabilities yetexactly okay, before we go on to
this durability component, Ikind of want to like rehash some
of this from a practicalstandpoint.
Speaker 1 (35:35):
So the athletes out there and they look at they can
look at this conceptually andsay, ok, you know, I understand
that if I looked at myperformance on this race for a
particular year and how I rackand stack compared to the people
around me, I'm going to noticedifferent strengths and
weaknesses.
And I want to focus on theweaknesses part generally in
(35:57):
order to kind of extract moretime out of it.
But how, aside from theobservation piece, people are
going away from me on the climbsand I'm bringing them back on
the flats, right?
So the climbs are my weaknessesand the flats are my and the
flats are my strengths.
And then, taking that to anintervention standpoint, can an
athlete at home statistically gothrough this?
(36:17):
Are there tools or methods outthere that are similar to what
you actually went through to,where an athlete can like break
down their own stuff, or a coachcan actually break this down
for an athlete?
Because I've done the samething with my athletes and it's,
trust me, it's a painstakingprocess.
We've like we generalize itinto like a five minute summary,
but maybe you can kind of takethe athletes through what are
the listeners, through what youwould actually have to do in
(36:39):
order to extract this from astatistical perspective, as
you've done.
Speaker 2 (36:43):
I'm not sure I understood you.
You mean more how everybody cando that kind of analysis at
home.
Speaker 1 (36:49):
Yeah, exactly Like if I were to just take, you know,
my race file from the.
The Cocodona two fifties isjust happening and I happened to
do that race a couple of yearsago, so it's at the top of my
Cocodona two, 50 or whateverrace.
How can I take that informationand say, okay, statistically
speaking, this is where mystrengths and weaknesses lie.
Speaker 2 (37:08):
Okay, actually, that's something that we try to
develop.
We have a patent on all theprocess that we developed and
now we are currently discussingwith some companies to try to
implement this kind of analysis.
Speaker 1 (37:22):
Somebody's going to charge us for it.
That's what you're telling me.
There's going to be an app thatI'm going to have to download
in order to tell it.
Okay, I get it.
Speaker 2 (37:30):
Yeah, and actually because we talked a lot about
the race analysis.
But actually we can do the samekind of thing than the record
profile instead of the data ofall your training and not only
your race and to look how yourprofile evolved by taking all
your records for all thegradients that you encountered
in your training.
But we also try to developsomething that I think that you
(37:53):
can try at home.
It's a test that we try todesign.
So let's say you go to runoutside and you will try to keep
your heart rate at a givenintensity that interests you for
your training, let's say 100and 60 RPM, because it's the
intensity that interests you.
(38:14):
And then you can go running invarious gradients and look at
the speed that you can maintainwith this specific heart rate.
So you have an idea of what isthe velocity that you're able to
run on this specific gradient.
And you can do that for anothergradient, etc.
And with only a few segments ofvarious gradients you only need
(38:37):
a few minutes to stabilize theheart rate.
So you can have the fullrelationship of what is the
velocity that I can use for thisgiven heart rate and in this
given gradient and then you cancompare to others, but you can
also compare to yourself.
Let's say, I need to train inthe steep gradient.
So you try to train, you startto train for a few weeks and you
(39:02):
want to see okay, I amimproving in this specific part.
You can do again the samesegment, going on this steep
slope, and look at the sameheart rate Are you going faster?
So that's a simple way, a quitesimple way that you can use,
with a simple testing to, to seeif you improve in this specific
ability that you targeted inyour training yeah, what
(39:24):
everybody's searching for intrial running right now is
analogous to the power durationcurve in cycling.
Speaker 1 (39:31):
so you can take any cyclist who has a reasonable
amount of power data and you canproduce a power duration curve
for cycling.
So you can take any cyclist whohas a reasonable amount of
power data and you can produce apower duration curve for them,
and then you could fit thatpower duration curve against a
Tour de France rider, a worldclass time trial athlete,
somebody who runs or somebodywho rides the kilometer on the
track any of those riders.
You can kind of pit themagainst each other, so to speak,
and see where rider A is goingto be better than writer B and
(39:53):
how they can like rack and stackand they even do a lot of like
still some old school like teamselections, kind of based off of
these, based off of theseprofiles.
People are searching for thatin trail running, which you're
you know what you're kind ofproposing, and I think that this
is a good tool to have.
This is essentially a velocitygrade curve curve as opposed to
a power duration curve, tocomplete some part, some
(40:16):
component of that analysis.
Speaker 2 (40:18):
Yeah, actually it's a force velocity duration Right,
because in cycling you canchange the gears so you can just
stick to the best condition,but when you are going up you do
not have gears when you'rerunning, so you have to go up
with your foot and no way other.
So you have to distinguish theforce and the velocity part in
(40:41):
the power.
That's why you need todistinguish force and velocity.
And then how force and velocityevolve with duration is another
component, and force andvelocity not necessarily evolve
the same way in the duration andfor different individuals they
can have a better durability invelocity or better durability in
(41:02):
force.
And that's another thing.
Speaker 1 (41:05):
And another question and another that we can analyze
the statistics nerds out therewill certainly get a lot of a
big kick out of some of thematerial that will be in the
show notes.
Speaker 2 (41:17):
Let's kind of move the conversation forward,
because we've kind of gonethrough this from a practical
standpoint in terms of howathletes can actually analyze
this no-transcript component andhow athletes kind of exhibit
(42:04):
some of the or how some of thethings that we were looking at
earlier change over the courseof a big, long race like utmb so
if we stick to the analogy ofthe force and velocity part or
the flat and happy componentthat that's what I was just
saying that there is notcorrelation between the ability
to continue being good andrunning fast on flat and all
(42:28):
uphill we can start with thefirst picture being the mean
picture of all of the UTMBfinishers that we analyzed.
There is something interestingis that on average the runners
are better to.
They conserved more.
They were more close to theirinitial capacity in the fourth
component.
So when it's going up you donot lose that much speed.
(42:52):
When the race is ongoing andbecause of the durability
components, this is an averageanalysis.
But on flats running there is abig difference between the
beginning of the race and theend of the race.
So if we just stop here for afew seconds, I would say on
average you probably need toimprove your ability to run on
(43:15):
flatland for the durability andvery long effort.
That means that probably whenyou prepare this kind of race
you can imagine having apre-fatigue boot or you will run
for, let's say, five, six,eight, ten hours, as you want
for your trail running training,and then go to a flat portion
(43:35):
and try to maintain your highspeed.
For the second part in thisstatic state, that could be a
recommendation based on thiskind of analysis.
Speaker 1 (43:44):
Yeah.
So this intervention, a finishfast run or a progression run or
doing intervals at the end ofthe run as opposed to the
beginning of the run.
Our coaching group has thisdebate kind of like ad nauseum,
and fundamentally you know somepeople in our coaching group,
that's probably why you'relaughing.
Fundamentally there's a lot ofnuance to this, but
fundamentally you're kind oftrading two things right, you do
(44:06):
the hard work at the beginning,you have the best capacity.
Right, you do the hard work atthe beginning, you have the best
capacity right.
So if you have the bestcapacity, theoretically you're
going to do the most kilojoulesof work.
You're operating at the highest, higher percentage of your VO2
max.
You're introducing kind of morestress which theoretically
would lead to more adaptation.
Right, more stress, moreadaptation.
You do the intervals at the end.
You're probably at a reducedcapacity.
(44:27):
Right, you can operate at asmaller percentage of your VO2
max or maybe for not as long ofa period of time, and so the
interval quality is certainlyreduced.
I don't think anybody wouldargue that, right, we might
argue the impact of that reducedquality session, but it
certainly is more specific to anultra marathon where durability
(44:48):
becomes exceedingly important.
Now you're adding in thisanalysis, which, which is once
again, there's a ton ofpractical takeaways to this, and
I think that this is actually abig one.
What your analysis wouldsuggest is that if you are doing
intervals at the end of the run, particularly to accentuate, or
hard work at the end of the run, a progression run, something
(45:10):
like that you're probably bettersuited doing that in a flat
level condition, because that isthe gradient, or lack thereof,
that gets impacted the most whenan athlete is fatigued.
It's not the uphill component.
So a better way to structure adurability type of workout would
(45:31):
be run for two hours and thendo an interval on flat level
terrain, versus run two hoursand do uphill intervals, which
you could still have thatcomponent of it based on the
race that you're actually seeing.
But if you have, if the racehas uphill and flats, the better
piece of it would be to do iton, to do it on the flats.
Is that what the research issuggesting?
Am I reading between the lineswell enough?
Speaker 2 (45:51):
there.
Yeah, actually it's really hardto.
We do not have an easy, rightanswer on the durability
question, because it's a hottopic, it's new even in the
scientific field, so I can't sayyou absolutely must do that.
But based on the analysis thatwe did and since we saw with
(46:11):
evidence that you lose morespeed and you lose more speed at
the end of the run up becauseof the flat running capacity, if
you have to choose what to do,I would probably suggest yes, to
train your durability componenton flats.
Speaker 3 (46:29):
Okay, let's check back in with coach adam on this
really neat concept of when wewould actually want to target
flat ground durability toimprove performance and how we
would actually do it yeah, well,first of all, I do think that
this, when I heard this, it kindof made sense to me, because I
(46:50):
pace people quite often at racesand this is something that I do
see quite a bit where you'rehiking uphill with someone late
in an ultra and they're actuallymoving pretty well, and then
you hit flat ground where youhave to run, and it's a
different story.
So it probably varies person toperson, race to race, but I
actually do think it's a themethat you can just see out there
(47:10):
on the ground as well.
As far as actually training it,I know in the podcast you guys
brought up potentially doingsome sort of faster work or
intervals on flat terrain at theend of a long run.
And another intermediate stepthat I might bring in an athlete
is just, you have your hillylong run and in the last 60, 90
minutes do it on flat terrain.
(47:31):
Maybe not even with the speedwith it, but just placing that
terrain at the end could be onestep along the way.
Speaker 1 (47:39):
Yeah, you know, western States is the classic
course that accentuates thisbecause it gets faster as the
race generally goes on,particularly as you leave
Robinson flat and then Michiganbluff and then forest hill.
It just kind of gets easier andeasier from a surface
standpoint and the athletes thatdo really well, the ones that
can just kind of run past runspecifically, not in quotes
(47:59):
actually run past forest hilland not do a lot of the
prototypical hiking and thingslike that.
But when Batiste actuallymentioned this, I thought of a
paper that a mutual colleague ofmine was a co-author on, and
I'll link that paper up in theshow notes, and it posited the
co-workers, nick Tiller alongwith Guy Mier, who are the two
(48:20):
authors of this paper.
Their primary thesis is thelimiting factor in ultra running
is at the level of the muscle,and particularly muscle damage,
which would be accentuatedduring running more than it
would hiking, because running isa more, let's just say, a
muscularly intensive mode oflocomotion as opposed to hiking.
(48:43):
Uphill, downhill, running wouldprobably be the most muscularly
demanding component of all thethree disciplines, or all the
disciplines that we can likethat we experience in ultra
running.
And so maybe if we took thisresearch step further and
further the speculation that wewould see this.
We would see this componentdiverge even further on the
(49:04):
downhills, where people getdisproportionately people who
are doing or who are slower getdisproportionately worse,
specifically on the downhillsthe most, then on the flats the
second most, and then don'treally deviate that much on the
uphills, but who knows?
But some interesting concepts,so to speak.
And the final caveat that I'lladd before we go back to the
main conversation is rememberthis analysis was done on one
(49:26):
specific race and we have to putthat into context with all the
races that athletes are doingand that may or may not look
like the same course profile.
Speaker 2 (49:37):
But this was an average relationship because we
evidenced, even if they are notthe case, that we saw the most,
but it could be the opposite forsome of the athletes that were
able to maintain quite a goodspeed and flat but was really
slower on the steepest gradients.
So you do, we cannot have justone answer for everybody, but on
(49:58):
average I tell you that smallflatland that you can play, but
specifically individually it canchange well, and if an athlete
wants to apply it to themselves,they can look at.
Speaker 1 (50:10):
This is why post-race notes and post-workout notes
become so incredibly importantDuring the end of the race.
I could do this and I couldn'tdo that.
That's the lens that you canlook through to determine okay,
we can improve this or improvethat.
You can also corroborate thatwith the actual file itself and
this is something that I do withmy athletes is I look at how
(50:31):
each segment deteriorates overtime the uphill segments, the
flat segments and the downhillsegments and the gauge.
I'm kind of like revealing thecoaching process.
There's not like it's likeproprietary to me or anything.
I didn't come up with it.
Somebody taught me how to dothis and then I'm just doing it,
you know 20 years later, butwe're simply using intensity
factor, as the intensity factoris just the ratio of the speed
(50:56):
that they're running to theirthreshold speed.
So if the if they're runningfaster than the threshold speed,
the intensity factor is overone.
If they're running slower thantheir threshold speed, intensity
factors under one.
It's almost always under one inany ultra marathon event, but
we look at how that deterioratesover the course of a race and
do the uphill segmentsdeteriorate more than the flat
segments or do they deteriorateless than the flat segments as
(51:18):
compared to the downhillsegments?
And those are the three bucketsthat I essentially put it in
and that becomes a lens to lookthrough to say, okay, we're
going to improve here from yearto year or based on this race,
and that improvement can comethrough training, it can come
through pacing, it can comethrough nutrition interventions.
There's a whole host of things.
It's not just do progressionruns or anything like that.
(51:39):
That's not the answer toeverything.
But my point with that is wehave tools that we can look at
it from a statisticalperspective as well as a
subjective perspective, how theathletes actually experience
things.
Speaker 2 (51:52):
Yeah, and how the runner and the athletes perceive
something is really important,but I really think that we
should always mix the perceivedthing with the objective
information, because sometimesyou can have a difference
between how the athleteperceives a thing and what
actually the watch and the timesay.
Speaker 3 (52:13):
Exactly.
Speaker 2 (52:15):
Especially at the end of an ultramarathon, where your
mind is, in a way, not in thefresh state that you have at the
beginning, so you can have somediscrepancies between the
perception and what actuallyoccurred.
Speaker 1 (52:28):
Okay, this has been amazing because I like getting
the.
I like getting the curtainpeeled back a little bit with
the people who do these types ofanalyses, and I do think that
this is one where we can come upwith a lot of functional ways
or functional takeaways in termsof how we can start to analyze,
you know, race performance andthings like that.
I know you've got a lot ofthings on your wishlist right
(52:50):
now.
In fact, the people who areviewing the video version of
this will see a whiteboardbehind the Baptiste's shoulders
that has a lot of equations andquadratic formulas and things
like that, written meaningyou're at work doing something.
But I kind of want to know,like, what's on your wishlist to
develop.
Do you now have like, thisframework to say, okay, we can
(53:11):
analyze race performance likethis, how do we like what's on?
What would you like to do withthat?
If you could just likeinstantly wave a magic wand and
put it in the hands of youraverage or elite trail runner in
order to improve theirperformance with, like, your
knowledge base, what would youwant that to look like?
Speaker 2 (53:26):
We have several things that we could do all
seven.
We can go all day with this.
Well, the thing that is, I think, really important is to develop
some tool that we can provideto the community.
So so we can use it and we cannot just look at two elite
runners or not just at thefinisher of the 2023 edition of
(53:47):
the utmb, but but, as you say,it's a descriptive way to see at
the performance, but actuallywe don't know if the durability
of the force or the velocity isrelated to pacing, is related to
the nutrition or related to theheat of the race.
There's a lot of factors that wewant and we want to use this
framework to analyze all of that, all of the nutrition, the heat
(54:10):
, the pacing, etc.
And, for example, I've beensolicited by a team in the UK
that will run an experiment inthe Western States this year in
a few weeks.
They are really interested inthe heat acclimatization aspects
for this race and we will runthis analysis that we did for
the UTMB and the Western Stateand a few participants to see
(54:34):
how the duration, the durabilityprofile on force, on velocity,
on intermediate slope, ondownhill, etc.
How all of that is potentiallyrelated to the heat, to the core
temperature of the athletes.
So that's an example of what wewant to do with this framework
that we developed and that canbe useful tomorrow for analyzing
(54:56):
and reanalyzing again some ofaspects that are related to
performance in trail running,but with a new, potentially more
sensible tool that we developed.
Speaker 1 (55:07):
Amazing.
I hope all of those come tolight.
I'm going to leave links in theshow notes to some of the
research that we developedAmazing, I hope all of those
come to light.
I'm going to leave links in theshow notes to some of the
research that we've eitherdirectly referenced or kind of
tangentially referenced.
I'll get some of those linksfrom you after the fact as well,
because there's probably somethings that I was missing in the
research that our fellowcolleague Fred and I did.
But if people just want to likelook you up and look up the
(55:28):
research and the work that youdo, how are they going to go and
find you?
How they can find me?
Speaker 2 (55:33):
Yeah, Research gate Twitter, Instagram they're going
to direct email address.
Speaker 1 (55:37):
Some people give their cell phone numbers out
every once in a while on thispodcast, which I'm kind of like
a little leery and actually puton the air, but like, how do you
want people to find out moreabout you?
Speaker 2 (55:46):
Yeah, they can find me on all the social media and
twitter, instagram, etc.
I will send you the link sothey can reach me.
And we have also a projectnamed predict trail where we
want to develop this tool thatwe want to spread.
You can find onwwwpredictrailcom.
You can share your link thatthat's another project that we
(56:08):
have and where you can find me.
Speaker 1 (56:09):
Brilliant.
Well, thanks for your time,thanks for your research.
I wish you all the best.
We're going to bring you backon this podcast whenever any of
these probably gated orpaywalled or whatever versions
of these performance analysistools actually come out and you
can explain them.
More power to you.
It's absolutely going to happen.
I'm here for it.
Speaker 2 (56:35):
Any way that we can slice and dice performance is
something that I'm going to be apart of, and we'll nerd out on
how to analyze performanceprobably at another future point
.
Thank you very much.
Speaker 1 (56:42):
All right, folks, there you have it.
There you go.
Much thanks to Batiste andCoach Adam for coming on the
podcast today and, as Imentioned during the intro, I do
think that this is somethingthat we can deploy in our
day-to-day training withathletes and also think about
how we can race smarter.
As always, this podcast isbrought to you sponsorship and
advertisement free, so if youwould like to support this
(57:05):
podcast, the one thing that youcan do is subscribe to Research
Essentials for Ultra Running.
We are actually going to takesome of Batiste's research and
break it down even further inthat newsletter, in an upcoming
version of that newsletter, andso if you really want to geek
out on all the stats, statisticsand find out what we think
about it in a more detailedversion, more detailed manner,
(57:26):
research Essentials for UltraRunning is your ticket.
All right, folks, that is itfor today and, as always, we
will see you out on the trails.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.