February 12, 2025 • 27 mins
#Holotype #Programming #GameDesign #Python #BoardGames #Science
SummaryA follow-up to our last episode, in this episode we interview Brett Harrison, one of the co-creators of Holotype. Specifically, Brett is the one who programmed a computer to play Holotype against itself 10,000 times per minute in order to precisely balance the point values of different parts of the game. In this interview we'll cover a bit of the background behind Holotype, why he did that sort of optimization
Timestamps- 00:00 - Introductions
- 02:13 - Design of Holotype
- 03:54 - Playtesting and computer simulation
- 11:37 - Designing the right AIs
- 14:50 - What parts got tweaked
- 18:35 - Game length, Bone Wars, and an app
- 21:20 - Lost mechanics, accuracy, and personal favorites
- 25:15 - Closing remarks
Find our socials at https://www.gamingwithscience.net
LinksBrexwerx Games: https://www.brexwerxgames.com/
This episode of Gaming with Science™ was produced with the help of the University of Georgia and is distributed under a Creative Commons Attribution-Noncommercial (CC BY-NC 4.0) license.
Full Transcript(Some platforms truncate the transcript due to length restrictions. If so, you can always find the full transcript on https://www.gamingwithscience.net/ )
Brian 0:06 hello and welcome to the gaming with science podcast where we talk about the science behind some of your favorite games.
Jason 0:12 Today, we'll be interviewing Brett Harrison from Brexwerx Games. All right, welcome back to gaming with science. This is Jason.
Brian 0:22 This is Brian,
Jason 0:23 and today we have another special guest. This is Brett Harrison from Brexwerx Games. Brett, can you introduce yourself?
Brett 0:29 Hi, yeah. Brett Harrison, Brexwerx Games, one of the designers on holotype, which is our board game about paleontology,
Jason 0:35 yes. And when this episode drops, about two weeks before we will just put out our episode on holotype itself, we we're talking with some paleontologists, but this was because I wanted to ask some more about what you did for the game. So our listeners are already familiar with the game itself, how it plays the science in it, but I wanted to really follow up with an aspect I learned about when I was doing research for that episode about you created a computer simulation to basically play test the game, and we'll get to that in a moment, but I think we need to learn a little bit more about you first. So can you give us, like, what's your background? Like, how did you get into making board games, paleontology, that sort of thing?
Brett 1:13 Yeah. So I've always been into paleontology since I was a little boy. My grandfather brought me a book from the Los Angeles library, which I still have, so it's incredibly overdue, but it was all about dinosaurs, so he would read it to me when I was like six, and that's how I kind of got hooked on it. And then I eventually went to college for computer science and needing to get electives and stuff, I always chose paleontology or geology and stuff like that. So that's where I get that background always been a gamer since really young, started out with like, axis and allies, and went all up to all the Euros and everything else that's out there. So really big into board gaming. I've designed games, computer games and board games in the past, but nothing of this nature until COVID hit, and me and my buddy were like, Let's make a board game. So that's when we started working on holotype. And the theme was dinosaurs from the beginning, because I was so into paleontology, but we also want to make it super scientifically, you know, as accurate as possible within something that's supposed to be a fun game mechanic
Brian 2:11
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Brett (00:01):
Brian,
Brian (00:06):
hello and welcome to the gaming with science podcast
where we talk about the sciencebehind some of your favorite
games.
Jason (00:12):
Today, we'll be interviewing Brett Harrison from
Brexwerx Games. All right,welcome back to gaming with
science. This is Jason.
Brian (00:22):
This is Brian,
Jason (00:23):
and today we have another special guest. This is Brett
Harrison from Brexwerx Games.Brett, can you introduce
yourself?
Brett (00:29):
Hi, yeah. Brett Harrison, Brexwerx Games, one of the
designers on holotype, which isour board game about
paleontology,
Jason (00:35):
yes. And when this episode drops, about two weeks
before we will just put out ourepisode on holotype itself, we
we're talking with somepaleontologists, but this was
because I wanted to ask somemore about what you did for the
game. So our listeners arealready familiar with the game
itself, how it plays the sciencein it, but I wanted to really
follow up with an aspect Ilearned about when I was doing
(00:56):
research for that episode aboutyou created a computer
simulation to basically playtest the game, and we'll get to
that in a moment, but I think weneed to learn a little bit more
about you first. So can you giveus, like, what's your
background? Like, how did youget into making board games,
paleontology, that sort ofthing?
Brett (01:13):
Yeah. So I've always been into paleontology since I was a
little boy. My grandfatherbrought me a book from the Los
Angeles library, which I stillhave, so it's incredibly
overdue, but it was all aboutdinosaurs, so he would read it
to me when I was like six, andthat's how I kind of got hooked
on it. And then I eventuallywent to college for computer
science and needing to getelectives and stuff, I always
(01:36):
chose paleontology or geologyand stuff like that. So that's
where I get that backgroundalways been a gamer since really
young, started out with like,axis and allies, and went all up
to all the Euros and everythingelse that's out there. So really
big into board gaming. I'vedesigned games, computer games
and board games in the past, butnothing of this nature until
COVID hit, and me and my buddywere like, Let's make a board
(01:57):
game. So that's when we startedworking on holotype. And the
theme was dinosaurs from thebeginning, because I was so into
paleontology, but we also wantto make it super scientifically,
you know, as accurate aspossible within something that's
supposed to be a fun gamemechanic
Brian (02:11):
That was really obvious. So we played with a couple of
paleontologists for our episodeon holotype. They really
appreciated the attention todetail, and so do I, because I
think it's really important fora science game that, like,
accuracy, at least, is you cannever have it fully accurate,
right? You always have to decidehow much you're going to
simulate and like, where is thefun point? Where are you going
to lose fun for the sake ofaccuracy, that you guys did a
(02:33):
really good job, maybe if you doanother game for the Patreon,
you can have a stretch goal topay those overdue library fees.
Brett (02:40):
All right, yeah, I just, I'm never gonna ask you if it's
overdue.
Jason (02:44):
Well, I think you succeeded as far as, like, the
fun and the accuracy, becauseour listeners will have already
heard at the end of ourepisodes, we basically give a
grade for both the fun of thegame and the, like, the science
in the game. And you got A's onboth of them, like, right up
there with some of ourfavorites, like wingspan and
cytosis. So y'all did very well,
Brett (03:03):
awesome. Yeah, and those kind of, there's so many
dinosaur games out there thatare, like, you know, the theme
park and dinosaurs taking overand all that kind of stuff. It
was like, we wanted to makesomething that was, we called
our love letter to paleontology.And we did work with
paleontologists to make sure itwas accurate. We worked with the
Southwest paleontological groupto, you know, kind of get those
resources of having a PhD, goover and check and there's stuff
(03:25):
that got changed in the gamebecause it was not accurate,
because ornithophods, like, Iguess, stegosaurs and
ankylosaurs are not technicallyornithopods. So we had to go
with a much more inclusive cladethat included them and the
ceratopsians, which is whythere's genasaurs. Those used to
be called ornithopods, and thatwas not correct.
Brian (03:44):
So yeah, I know I didn't know the term either for sure,
that was a surprise, and weactually talked about that a
little bit as like, What theheck is a genasaur? But they
appreciated the use of the term.They said this is a better term.
Jason (03:54):
So I now want to drill down to the whole reason why I
wanted to get you on thispodcast is because while I was
looking up information on thisgame, I ran across some of your
other promos, interviews, andyou mentioned the play testing
of this. And can you walk usthrough the play testing
process? Like I'm reallyinterested in this computer
part, but how did you play testthe game in all its aspects? And
(04:14):
then we'll focus on thecomputer.
Brian (04:16):
Let me give some context for why Jason is excited about
this. He is a programmer. I amnot, so I'm gonna sit here and
listen politely.
Jason (04:22):
Yeah, so I'm a bioinformatician, so I study, I
use computer programming toanswer biology questions. I have
this long running goal of havinga little like, a few minisodes
about teaching computers to playgames, which is kind of what you
did. So I'm really curious aboutthis.
Brett (04:37):
So play testing, I mean, it started very early on in the
process of like, okay, these arethe mechanics that we want to
do, and this is all done with meprototyping ridiculously bad
computer programmer graphics onpaper and cutting them out and
gluing them on cards and allthat kind of stuff. And granted,
this all took place during thepandemic, so our normal gaming
(04:57):
group that would normally gettogether that wasn't as.
Possible, but we'd have a setgroup of about four people that
started play testing it early onin that form, and that's where
you learn stuff that's fun,that's not fun, that is
completely, you know, seemedlike a great idea, but once you
get it in practice, it makes thegame 20 minutes longer, that
kind of stuff. So all that kindof play testing took place
(05:18):
within a very small group tobreak out those, you know, core
mechanics and then we went tousing a program called Tabletop
Simulator on steam to do ourvirtual play testing, because
that way we could get peoplefrom all over the country
playing it. We didn't have toget people together, because we
couldn't with COVID, that kindof thing. So that was basically,
there's no real automation inthat one. It's just replicate
(05:40):
all the components so people canplay in a virtual table. And so
with that, we again fine tunedmore of the mechanics, and then
once all the mechanics were kindof solidified and we had our
this is fun. It plays fast. Theloop's, right. That's when I
went into doing the simulationin Python that would actually
play the game, so that I coulddo all the balancing that was
(06:01):
necessary to make sure it wasreally tight for victory points,
and there wasn't one strategy orcard that threw the balance out.
Jason (06:08):
Okay? So this sounds like this was the late stage
polishing of the game.Basically, you you got the big
things done with playing withother people, either in person,
online, but then this is whenit's like, okay, let's fiddle
all the little knobs and tweaksto try to just get it nice and
smooth. Is that basically,right?
Brett (06:23):
Yeah. And the reason behind that is, like, the
simulation can't tell you ifit's fun. So that would be just
like, too early in the process,because there were things it's
like, I would have programmedthat we just threw out, because
it was like, yeah, that's not afun mechanic or, you know, so it
had to be at that stage for itto be useful, and then it
becomes incredibly useful forbalancing and determining, even
(06:47):
to the point of, like,manufacturing the amount of
little wooden cubes that are inthe thing that increases weight.
Each one of those costs a penny.So how many of those do I
include of each type? And that'sbasically I can calculate during
a play test. But I want thatover, you know, 10,000 play
tests to figure out what's themaximum that's ever in use for
this thing, because then I cango, okay, that's what we
(07:07):
manufacture, and no one willever run out and we don't over
produce.
Jason (07:10):
I hadn't even thought about that. I was thinking pure
game mechanics, not even gamecomponents, and how many cubes
do we need?
Brett (07:16):
And that every penny adds up to five cents, which adds up
to weight and shipping it overfrom the manufacturer and all
that kind of stuff. So...
Jason (07:23):
How hard was it to code the game into a computer? I've
tried this once or twice, andquickly realized that even a
relatively simple game actuallyhas a lot of moving parts when
you get down to computer code,like, how hard was that to
actually instantiate it as adigital thing?
Brett (07:37):
Let's see. I did it in Python. So coding wise, wasn't
too difficult, just becausethat's fairly easy to code in.
The game mechanics beingfinalized made it a lot easier
to replicate what was actuallyhappening in the game. So the
main thing was just getting thegame states, having all the
cards, their values, isbasically building that database
of what are all the gamecomponents, what are their
values within the game? And thenI could move to actually making
(07:59):
the player AI that would playthat game, which is basically
just manipulate the game state.The other thing is, because it's
a simulation, I don't have toworry about player input and all
that kind of stuff, you know. Sothat takes a lot of the UI out
of making something like that.
Jason (08:12):
Yeah. And for listeners not familiar with the term, UI
is user interface. It'sbasically how you interface with
the computer. It's the thingthat looks pretty and that is
meant for you to do. And if thecomputer is just running with
itself, it doesn't need that,because it's just talking to
itself with code.
Brett (08:27):
There was a UI in the sense of, I needed to have a
graphical interface that wouldshow what's happening so I could
just see how the players werebehaving. Was it actually
playing correctly? Or had Icoded something, you know,
wrong, instead of just lookingat the data, I wanted to see the
visual. You know what was goingon, but that's very just crude
graphics. And really, you know,nothing you'd want people
(08:47):
seeing. But it was basically asanity check of like, is this
actually playing the gamecorrectly?
Brian (08:51):
We were joking. It seems like Holotype might have been
played more times than any othergame just because of the
simulation time.
Brett (08:58):
That's a good that's and interesting point because I ran
my basic run of the simulation.Is once everything was working
and doing what it was supposedto do, I would turn the graphic
part of it off, and it wouldplay basically 50,000
simulations of the game in abouta minute. So that would spit out
a it took me way longer to dealwith getting that data into
(09:18):
Excel and making reports that,you know, gave me some
interesting info than it did toactually run the simulation once
everything was said and done,and any time that something
would be shown that was like,Oh, I that should be, you know,
modified a little bit. I couldjust go in, go into the JSON
file, change the value of acard, or change the, you know,
point value of something, andrun that simulation again right
(09:38):
away to get another output tobe, "Did that change it? Did
that go in the direction wewanted?" that kind of stuff. So
yeah, to say it was run 50,000times. It was run 50,000 x 100
appear.
Jason (09:50):
So like, now that I've heard this, this sounds like an
obvious thing to do for tryingto balance and play test the
game, but I can't say I've everheard of another game actually
doing this. I hear of openbetas, I hear of people doing
stuff online, digital versionsof the game for people to play,
but I haven't heard of someonejust having the computer play
itself a whole bunch. Was thisan original idea? Did you hear
about this from somewhere else?
Brett (10:11):
I didn't hear about it from someone else. I just
because of my programmingbackground. And it wasn't, I
don't think it was something weset out to do from the
beginning. It was just as thedesign was coming together, and
I was like, I want to make surethis is balanced. I went, Oh, I
can easily do that myself andplay test it way more than
asking a bunch of people to playtest it to get those, you know,
real true averages over a crazyamount of time. And I think it
(10:33):
would be useful for other boardgames and stuff like that. It's
because I was the programmer andone of the designers, I knew the
game so well that it wasprobably easier for me to
implement that once I explainedit, I had other designers that
were like, Hey, can you do thatfor my game? And I was like, uh,
you know, I could, but do I wantto? So I think it would be
useful, because that's one of myfrustrations sometimes when
(10:56):
playing games, is you'll playit, and then someone will find a
strategy just like, Okay, that'sall you ever do, and you'll just
win, because it's justunbalanced. Through that one
thing I've had, I've had gameswhere I was like, Oh, I figured
it out, and then it was like,there was no fun for the other
people, because someone was justblowing other people away
because they played a certainway. And that's something that
we really wanted to haveholotype be super balanced. And
(11:17):
I think a lot of the feedbackwe've gotten has been that,
yeah, it's incredibly balanced,especially point wise at the
end. There's been plenty oftimes where people have, you
know, had to go to the tiebreaker because they were right
in the, you know, same area,point wise and stuff.
Brian (11:29):
So we talk about a game being solved right, where
there's always an obvious choiceto make. It sounds like you were
working to keep the game frombeing solved.
Brett (11:37):
Yeah. And there's, there's an interesting point
there that goes to designing theAIs. When I made the AI to play
the game, it had to play it theway a human would play it,
because sometimes gamers aren'toptimal. I didn't want to make
the AI figure out like, this isthe best way to play this game
and optimize it for somethingplayers never do. So it was, it
was definitely more of like,okay, this is the way the
(11:59):
players play. It's weighted totry different strategies and
stuff, but it was very muchinformed by how people would
play the board game. If thatmakes sense.
Jason (12:08):
What does that look like? Because that was the thing I was
really curious about, is, howdid you program an AI to
essentially fill the role of ahuman? Because when I'm playing
this game like I'm looking at mycards, I'm looking at my
personal goal. I'm looking atthe global objectives. I'm
looking at other people's stuff.Obviously, a computer can keep
track of that pretty easily. Butthen making, like, integrating
that and making the decision of,what should I do next? That even
(12:29):
I don't know how I make thatdecision sometimes. So how did
you represent that for an AI tomake that decision.
Brett (12:34):
It's represented in basically the way you would as a
player. So you're looking atyour cards. You have, I have a,
you know, theropod from theTriassic. I have this from the
Jurassic. I have this manyresources of that, and I have a
personal objective that I'mgoing for. So those actions are
all given weights based onwhat's in your hand and what
resources you have, even to thepoint of like, okay, I have, I
(12:55):
have these two Jurassic cards,but this one's worth more, but
it also costs less in resources.So all those things are just
weighted, and then it'sbasically making a decision on
which one of those actions hasthe highest value, which
eventually really just relatesto which of these actions is
going to get me more the mostpoints at the end of the game.
And then, if I want to test adifferent strategy, basically
you're just changing thoseweights like this, AI is going
(13:16):
to want to publish globals morethan the other ones do. That
kind of thing. And then at theend, which AI wins more of the
time, which actions, that kindof thing. And then that leads to
just more the balancing.
Brian (13:27):
Okay, so your your AIs have some biases in them, built
in some preferences.
Brett (13:32):
Preferences, and you can change those to, you know,
investigate other assumptions.Like, one of the things as we
were playing was like, certainplayers were like, I never go
for globals, or I always go forglobal and then so by doing AIs
that did different preferences,you could see like, is that a
better strategy? Is thisstrategy not even viable if
someone just does this? So thatallows you to test that over
(13:53):
1000s and 1000s of plays, whichyou just can't quite get in the
plane with other humans, becausepeople just have their own
preferences on how they play. Wewant to make sure that if they
did have those preferences, didthat mean they always lost or
always won? They'll balance it.
Brian (14:05):
So, you've got the AI Jason who always wants to play
optimally, and the AI Brian whowants to collect the cool
dinosaurs,
Brett (14:11):
yeah, and that, and that was one of the one of the AIS is
just like, whatever cards thefirst in my list is what I do. I
call it my eight year old AI,she's just like, what's the
first one? I that's what I'mgoing for.
Jason (14:25):
Did you try to make it so that that AI could occasionally
win? Or was that sort of, like,the control AI of you have to
beat this one to be considered adecent AI.
Brett (14:32):
There's definitely the control of just like, this is
the most simple thing. Does thatone ever win out of 50,000 play
throughs, maybe a couple times.But you didn't want that one to
have, you know, the same odds ofwinning as someone who's
actually, like, paying attentionto what's happening, going and
looking for certain things, fortheir personal objective and
that kind of thing.
Jason (14:50):
And then, as you were using this to polish everything,
what aspects of the game did youtry to tweak? I mean, I was
looking over the box and, like,I can think of dozens of things
you could try to do. Places onthe board, the number of things
on each face of the dice, thehow long it takes to get a
graduate student, the pointvalues on the cards. Obviously,
you already mentioned those. Didyou make a list of like, here's
(15:10):
everything in the game that canbe tweaked. I'm not going to
systematically test all ofthese. Or were there some that
he's like, Okay, I need to tweakthese. These are not so
important. Like, how did youdecide what to check?
Brett (15:20):
So there's certain parts of the game design that are just
like, once you've decided it,that's what influences
everything else, and that thatcomes down to, like, what is an
action worth in an actioneconomy game, and what is a
resource worth? So there wasbasically before any of this
simulation in Python, all thatstuff was in Excel, so that it
was like, Okay, if I have, ifthis is how many pips I have on
(15:42):
a, you know, purple die, what'sthe distribution of that and
probabilities that all thesethings happen so that I know
when you roll this many dice,this is the average pips you'll
get of this type of resource,those things kind of had to be
like baseline, and you don'tmess with those once you get
into the simulation, becausethat was just core mechanics.
And really all that determiningis, what are those values on the
(16:03):
cards? How does that turn into avictory point? And that's where
the simulation so it was all thevalues on the cards, which comes
from the point values on thespecimen cards, and the point
spreads on the the personalobjectives and the global
objectives. So those were themain things. It was like balance
the point values for all thatstuff using these simulations.
So that was the big one. Andthen some of the stuff that got
(16:24):
changed is personal objectives.Definitely got changed, because
there would be certain personalobjectives based on the
distribution of those specimensin the in the deck. How often
would this person win with thispersonal objective? How much did
that skew? Same thing withglobal objectives. How often did
the game end early, becausethese global objective cards got
filled. So those values were theones that got the major changes
(16:45):
from the simulation, as well asthe like. Now we know you only
need 35 Triassic cubes out of50,000 playthroughs. Those only
ran out 50 times that kind ofthing.
Jason (16:54):
All right. and so looking at what you just said, so the
global objectives, personalobjectives, specimen cards,
those are the ones where playershave a choice. Basically the
like the excavation cards, thedice rolls, those things are
outside of player control. So itsounds like you set those as
these are just aspects of thegame you have to deal with, and
it's the ones where players canchoose where you were fiddling
with things. Is that right?
Brett (17:15):
Yeah, that's correct. Now, some of the like generation
cards, like the fieldexpeditions, those values could
be tweaked too to test that kindof stuff. But those things were
typically set pretty early on,like the amount of cards of each
type in the deck. That kind ofstuff was kind of hard coded.
That it was going to be the 10Triassic, 20 Jurassic, 30
Cretaceous that went back toearly design, because it was
just like there weren't manydinosaur type in the Triassic.
(17:38):
Then they started, you know,being more in the Jurassic, and
then even more in theCretaceous, so that that became
a constraint of, like,constraint of like, we can't
have more than 10 Triassicdinosaur cards because there
weren't enough specimens thatwould be easy for us to import
into the game.
Brian (17:49):
Now I regret that I overlooked that. I really wish
I'd noticed that specificpattern and pointed it out, but
I guess for our listeners, theyfound it up now, so no worries.
Brett (17:57):
Yeah, and if you notice on the on the pips on the dice,
purple Triassic dice only havesix the blue Jurassic dice only
have eight, and the Cretaceousdice have the green ones have
10.
Brian (18:08):
The lure of the Triassic was just too strong for me. I
knew it was more valuable, and Ikept chasing it and would lose.
So I was one of those bad AIs,
Brett (18:16):
and sometimes it worked out. You never know,
Jason (18:18):
yeah, but it was funny watching him get so frustrated
as the paleontologists we'replaying with would just keep
tossing these Triassic fossils,Triassic specimens to go after
the cool dinosaurs they reallylike.
Brett (18:28):
There's the personal preference AI, the one that only
wants to publish Triceratops.Yeah, we had one of those.
Jason (18:35):
Also, there's a question with all of your hundreds ot
thousands or millions ofplaythroughs, you might be able
to ask Brian and I, every timewe play this game, we have the
question, how long does a gamelast? Like, how many actions
does each player get before thegame ends? Because we're kind of
thinking the metaphor of like,okay, this is the
paleontologist's career. Is eachone of these actions a month, a
(18:56):
semester, a year? It meansnothing for the game. But we're
curious. So how long does a gameusually last? How many actions
does each person get?
Brian (19:05):
Good question
Brett (19:05):
that's interesting. Now, roughly about 40 during a game,
and it's going to changedepending on there's different
variants, right? You can playthe longer, shorter game by
using the different playertracks. So there's been times
where, you know, I'll play a twoplayer game with the five player
board, just because I want topublish a ton. But the average
amount of publishes for eachplayer is around seven to eight
by the time you're done with thethings, I'm sure a
(19:26):
paleontologist could answer thisbest, but, but I expect that,
you know, maybe that's a decadeor more research if you're
publishing, you know, eighttimes. Because when you're going
out in the field and discoveringstuff, it doesn't definitely
takes a while. But then if yougo back, you know, way back to
the bone War era Marsh, andcope, they were pumping out new
ones like you know,
Brian (19:41):
that makes me think that the best analogy to overlay on a
paleontologist's career for 40rounds is 40 years, because I
think there are plenty ofpaleontologists who are still
working into their 70s.
Jason (19:51):
yeah, but you hope that they published more than eight
times in those forty years
Brian (19:54):
you would again. No simulation is perfect, right?
Jason (19:57):
Yeah, although you mentioned the Bone Wars. That
was the number one request fromour paleontologist friends, is
that they want the Bone Warsexpansion to the game.
Brian (20:05):
Yeah, I want the Bone Wars expansion. I want the
traitor mechanics.
Brett (20:08):
Yeah, that's definitely been asked for, for sure, where
that's like, I'm going to stealyour fossils, or that, or I'm
going to, you know, denounceyour spec, your publication,
because that's wrong, and nowmine becomes the better one.
Brian (20:20):
Yeah, you need a fourth type of meeple, the, I don't
know,
Jason (20:23):
the saboteur? the infiltrator?
Brian (20:25):
The saboteur is probably the best way to put it.
Brett (20:27):
I like it
Jason (20:28):
someone with like the cloak a fedora and like little
spy mask,
Brian (20:32):
since they have to be they have to be indicated by the
type of hat.
Jason (20:36):
So Brian, do you have any last questions you want to ask?
Brett,
Brian (20:39):
I did so one thing I was curious about, so a lot of board
games now, wingspan evolutionand a lot of other ones root to
have these digital components tothem. There's an app you can
play wingspan on. It's arepresentation of the game. Lots
of games have made their way tosteam. It seems like you've done
all the work on the back end todo that for holotype. Is there
any interest in pushing it to anapp?
Brett (20:58):
Yeah, I guess so. That just seems like more programming
work for me, which I enjoydoing, but I don't have a lot of
experience with App Stores andgetting and doing stuff like
that. Mine's more on just theback end, spitting out all the
stuff. But I know there's, youknow, companies out there that
will take those games and makethem look beautiful and nice and
playable on a on an app store.So we'd be more than interested
(21:20):
in that.
Brian (21:20):
I had one other question, and you don't have to answer,
but I'm curious, if you want togive a look behind the curtain,
what was the mechanic you guystossed?
Brett (21:27):
Oh, this is a great one. It was a horrible one. That's
why I got tossed. So all themilestones that you get during
the game. So adding, you know,extra research or another, every
time you hit one of those greenstars, you add a milestone to
your board to get a littlebetter at something. Those used
to be in the game. Now, everyplayer has access to the same
set of those in the originalgame. They didn't
Brian (21:49):
there was zero sum?
Brett (21:50):
yeah. So there was the number of players minus one
available. Oh, and and ourbrilliant catch up mechanic was
okay, when you hit those greenstars, everybody adds up their
current points. Whoever's inlast place gets to choose first.
Brian (22:06):
So that's a little bit too close to real life.
Brett (22:09):
Yeah. And, and the problem is, while that was cool
in the thought of like, Oh, it'sa catch up mechanic, someone
could snag something thatsomeone else, the leader would
get last pick, and usually getsomething, you know, less good.
It made the games take aridiculous amount of time
longer, because that wholescoring process you would have
to do four times throughout thegame. Then you'd have the
picking and someone waiting tobe like, do I get this? Do I get
(22:30):
this? That indecision? And wewanted a really fast, snappy
playing game so that that onehad to get thrown out. And I
think it got thrown out for thebetter, because now everybody
having their own set, it's easy.You know, when's you know, you
can make your choice on yourtime, and it makes a little more
fair. Everybody's got the samething. You can't complain that
someone else snagged this beforeyou.
Brian (22:47):
I'd be really sad if I didn't get the storage closet
for the extra fossils.
Jason (22:51):
All of us always chose the storage closet first because
we're hoarders.
Brett (22:56):
That's interesting. Yeah, that's that's super interesting
because that's one of the ones Iit's very subjective to me on
whether I choose that one ornot, like, how many fossils are
currently flooding my storage?
Jason (23:04):
Was that digitally play tested? Did you change the AI's
preference for which one ofthose it would go for first
Brett (23:10):
in a couple states? Yeah, because there's only six in the
base game, that was a fairlyeasy thing. I'm going to get
four of them throughout thegame. I'm just going to randomly
pick them. Or this one's pickedalways first. Does that, you
know, skew it too much. So thatwasn't too hard to test. And
it's funny trying to make those,you know, game things into like,
how does this, you know, relateto the field of paleontology?
Because, like, I spent a lot oftime on making sure, like, all
(23:31):
the global objectives hadmeaningful names within
paleontology. And people thatplay in the game just don't
care. Paleontologists, they're,like, systematic what is that? I
don't care
we have, like, systematics,cladistics, and phylogeny, and
one of the paleontologists brokedown the differences between
them, and at a certain point Ijust, I like, yeah, okay, um,
sure.
We wanted to be true to it, sothat someone that was really
(23:53):
into it was like, Oh, I reallyappreciate same thing with,
like, the colors of the dicebeing related to the
international stratigraphychart, that kind of stuff. It's
like, that's an unnecessarystep, but people sure appreciate
it when they learn about it orknow about it.
Brian (24:05):
We were trying to figure out why only one person can
publish per turn.
Brett (24:09):
That's purely game mechanics. We wanted the one
spot that was going to force alot of, you know, bumping
mechanic going on.
Brian (24:15):
We retconned that there was a departmental copy editor
or the press release person.They only had so much time.
Brett (24:20):
I like it. Yeah, there's, there's there's certain things
where it's like when we playedwith the paleontologist, like we
realize research is not as easyas just going to the University
library. Please, please takethis as a you know, conceit to
game mechanics.
Jason (24:33):
One last question. This is the million dollar question,
what is your favorite dinosaurin the game?
Brett (24:40):
Oh my gosh. So Well, my favorite dinosaur would be
Stegosaurus. My favoritecreature from the mesozoic is a
marine reptile, which is theMosasaur,
Brian (24:50):
anyone in particular?
Brett (24:51):
oh my gosh, probably a tylosaurus.
Brian (24:53):
I don't know if I have a favorite. I was a sucker for
ichthyosaurs personally. But myfavorite dinosaur is
parasarolophus.
Brett (24:59):
Those. Are pretty awesome. I'll have to send you
one of our Parasaurolophus pins.
Brian (25:03):
That would be, yes, you can do that. I will allow it.
Jason (25:08):
Oh, yeah, just twist his arm there.
Brian (25:10):
Yeah, we'll have to send you stuff too, of course. Thank
you awesome. We'll do anexchange.
Jason (25:15):
All right, that's a good place to wrap it up. Brett,
thanks so much for coming on.
Brett (25:19):
My pleasure.
Jason (25:20):
We've already told people a lot about holotype, but if
people want to look you, to lookyou up or other than advertising
your game, is there anythingelse you want to talk about,
other stuff that may be in theworks from brexwerxs, or other
things you want to mention?
Brett (25:30):
Yeah, we've got some stuff in the works, but you can
always check us out on brexwerxsgames.com we just released our
8-legged Peacock , which is agame about the Maratus peacock
spiders from Australia and theirmating dance.
Brian (25:41):
we saw that. I'm excited about that. That might be what
something we put into one of ourlight games,
Brett (25:47):
light science game.
Jason (25:48):
Well thank you so much. This has been great, and it was
really nice to have thesequestions answered that I was
telling Brett beforehand, like Iwas thinking, Man, I wish I
could just know the reasoningbehind, like, wait a second, I
run a podcast? I can just askhim and get him on the air
that's content.
Brian (26:06):
Thanks for being our first creator willing to come on
and talk to us about their game.
Brett (26:09):
no problem anytime.
Jason (26:11):
All right. Thank you so much. Brett, good luck with
everything. And I guess youknow, if your current job
doesn't work out, you can alwaysstart up a new job as a a board
game designer, play tester,because you apparently have a
skill set in short supply.
Brian (26:24):
And if there's anything new from brexwerxs, send us an
email. We'll make sure that wepromote it for you.
Brett (26:29):
Sounds good. Thank you so much, guys. And for our
listeners,
Jason (26:31):
thank you for tuning in. We hope you enjoyed this first
interview we've done, and untilnext time, have a good week and
happy gaming. Have
Brian (26:37):
fun playing dice with the universe. See ya.
Jason (26:41):
This has been the Gaming with Science Podcast copyright
2025 listeners are free to reusethis recording for any non
commercial purpose, as long ascredit is given to Gaming with
science. This podcast isproduced with support from the
University of Georgia. Allopinions are those of the hosts,
and do not imply endorsement bythe sponsors. If you wish to
purchase any of the games wetalked about, we encourage you
to do so through your friendlylocal game store. Thank you and
have fun playing dice with theuniverse.
Brian,
Brian (00:06):
hello and welcome to the gaming with science podcast
where we talk about the sciencebehind some of your favorite
games.
Jason (00:12):
Today, we'll be interviewing Brett Harrison from
Brexwerx Games. All right,welcome back to gaming with
science. This is Jason.
Brian (00:22):
This is Brian,
Jason (00:23):
and today we have another special guest. This is Brett
Harrison from Brexwerx Games.Brett, can you introduce
yourself?
Brett (00:29):
Hi, yeah. Brett Harrison, Brexwerx Games, one of the
designers on holotype, which isour board game about
paleontology,
Jason (00:35):
yes. And when this episode drops, about two weeks
before we will just put out ourepisode on holotype itself, we
we're talking with somepaleontologists, but this was
because I wanted to ask somemore about what you did for the
game. So our listeners arealready familiar with the game
itself, how it plays the sciencein it, but I wanted to really
follow up with an aspect Ilearned about when I was doing
(00:56):
research for that episode aboutyou created a computer
simulation to basically playtest the game, and we'll get to
that in a moment, but I think weneed to learn a little bit more
about you first. So can you giveus, like, what's your
background? Like, how did youget into making board games,
paleontology, that sort ofthing?
Brett (01:13):
Yeah. So I've always been into paleontology since I was a
little boy. My grandfatherbrought me a book from the Los
Angeles library, which I stillhave, so it's incredibly
overdue, but it was all aboutdinosaurs, so he would read it
to me when I was like six, andthat's how I kind of got hooked
on it. And then I eventuallywent to college for computer
science and needing to getelectives and stuff, I always
(01:36):
chose paleontology or geologyand stuff like that. So that's
where I get that backgroundalways been a gamer since really
young, started out with like,axis and allies, and went all up
to all the Euros and everythingelse that's out there. So really
big into board gaming. I'vedesigned games, computer games
and board games in the past, butnothing of this nature until
COVID hit, and me and my buddywere like, Let's make a board
(01:57):
game. So that's when we startedworking on holotype. And the
theme was dinosaurs from thebeginning, because I was so into
paleontology, but we also wantto make it super scientifically,
you know, as accurate aspossible within something that's
supposed to be a fun gamemechanic
Brian (02:11):
That was really obvious. So we played with a couple of
paleontologists for our episodeon holotype. They really
appreciated the attention todetail, and so do I, because I
think it's really important fora science game that, like,
accuracy, at least, is you cannever have it fully accurate,
right? You always have to decidehow much you're going to
simulate and like, where is thefun point? Where are you going
to lose fun for the sake ofaccuracy, that you guys did a
(02:33):
really good job, maybe if you doanother game for the Patreon,
you can have a stretch goal topay those overdue library fees.
Brett (02:40):
All right, yeah, I just, I'm never gonna ask you if it's
overdue.
Jason (02:44):
Well, I think you succeeded as far as, like, the
fun and the accuracy, becauseour listeners will have already
heard at the end of ourepisodes, we basically give a
grade for both the fun of thegame and the, like, the science
in the game. And you got A's onboth of them, like, right up
there with some of ourfavorites, like wingspan and
cytosis. So y'all did very well,
Brett (03:03):
awesome. Yeah, and those kind of, there's so many
dinosaur games out there thatare, like, you know, the theme
park and dinosaurs taking overand all that kind of stuff. It
was like, we wanted to makesomething that was, we called
our love letter to paleontology.And we did work with
paleontologists to make sure itwas accurate. We worked with the
Southwest paleontological groupto, you know, kind of get those
resources of having a PhD, goover and check and there's stuff
(03:25):
that got changed in the gamebecause it was not accurate,
because ornithophods, like, Iguess, stegosaurs and
ankylosaurs are not technicallyornithopods. So we had to go
with a much more inclusive cladethat included them and the
ceratopsians, which is whythere's genasaurs. Those used to
be called ornithopods, and thatwas not correct.
Brian (03:44):
So yeah, I know I didn't know the term either for sure,
that was a surprise, and weactually talked about that a
little bit as like, What theheck is a genasaur? But they
appreciated the use of the term.They said this is a better term.
Jason (03:54):
So I now want to drill down to the whole reason why I
wanted to get you on thispodcast is because while I was
looking up information on thisgame, I ran across some of your
other promos, interviews, andyou mentioned the play testing
of this. And can you walk usthrough the play testing
process? Like I'm reallyinterested in this computer
part, but how did you play testthe game in all its aspects? And
(04:14):
then we'll focus on thecomputer.
Brian (04:16):
Let me give some context for why Jason is excited about
this. He is a programmer. I amnot, so I'm gonna sit here and
listen politely.
Jason (04:22):
Yeah, so I'm a bioinformatician, so I study, I
use computer programming toanswer biology questions. I have
this long running goal of havinga little like, a few minisodes
about teaching computers to playgames, which is kind of what you
did. So I'm really curious aboutthis.
Brett (04:37):
So play testing, I mean, it started very early on in the
process of like, okay, these arethe mechanics that we want to
do, and this is all done with meprototyping ridiculously bad
computer programmer graphics onpaper and cutting them out and
gluing them on cards and allthat kind of stuff. And granted,
this all took place during thepandemic, so our normal gaming
(04:57):
group that would normally gettogether that wasn't as.
Possible, but we'd have a setgroup of about four people that
started play testing it early onin that form, and that's where
you learn stuff that's fun,that's not fun, that is
completely, you know, seemedlike a great idea, but once you
get it in practice, it makes thegame 20 minutes longer, that
kind of stuff. So all that kindof play testing took place
(05:18):
within a very small group tobreak out those, you know, core
mechanics and then we went tousing a program called Tabletop
Simulator on steam to do ourvirtual play testing, because
that way we could get peoplefrom all over the country
playing it. We didn't have toget people together, because we
couldn't with COVID, that kindof thing. So that was basically,
there's no real automation inthat one. It's just replicate
(05:40):
all the components so people canplay in a virtual table. And so
with that, we again fine tunedmore of the mechanics, and then
once all the mechanics were kindof solidified and we had our
this is fun. It plays fast. Theloop's, right. That's when I
went into doing the simulationin Python that would actually
play the game, so that I coulddo all the balancing that was
(06:01):
necessary to make sure it wasreally tight for victory points,
and there wasn't one strategy orcard that threw the balance out.
Jason (06:08):
Okay? So this sounds like this was the late stage
polishing of the game.Basically, you you got the big
things done with playing withother people, either in person,
online, but then this is whenit's like, okay, let's fiddle
all the little knobs and tweaksto try to just get it nice and
smooth. Is that basically,right?
Brett (06:23):
Yeah. And the reason behind that is, like, the
simulation can't tell you ifit's fun. So that would be just
like, too early in the process,because there were things it's
like, I would have programmedthat we just threw out, because
it was like, yeah, that's not afun mechanic or, you know, so it
had to be at that stage for itto be useful, and then it
becomes incredibly useful forbalancing and determining, even
(06:47):
to the point of, like,manufacturing the amount of
little wooden cubes that are inthe thing that increases weight.
Each one of those costs a penny.So how many of those do I
include of each type? And that'sbasically I can calculate during
a play test. But I want thatover, you know, 10,000 play
tests to figure out what's themaximum that's ever in use for
this thing, because then I cango, okay, that's what we
(07:07):
manufacture, and no one willever run out and we don't over
produce.
Jason (07:10):
I hadn't even thought about that. I was thinking pure
game mechanics, not even gamecomponents, and how many cubes
do we need?
Brett (07:16):
And that every penny adds up to five cents, which adds up
to weight and shipping it overfrom the manufacturer and all
that kind of stuff. So...
Jason (07:23):
How hard was it to code the game into a computer? I've
tried this once or twice, andquickly realized that even a
relatively simple game actuallyhas a lot of moving parts when
you get down to computer code,like, how hard was that to
actually instantiate it as adigital thing?
Brett (07:37):
Let's see. I did it in Python. So coding wise, wasn't
too difficult, just becausethat's fairly easy to code in.
The game mechanics beingfinalized made it a lot easier
to replicate what was actuallyhappening in the game. So the
main thing was just getting thegame states, having all the
cards, their values, isbasically building that database
of what are all the gamecomponents, what are their
values within the game? And thenI could move to actually making
(07:59):
the player AI that would playthat game, which is basically
just manipulate the game state.The other thing is, because it's
a simulation, I don't have toworry about player input and all
that kind of stuff, you know. Sothat takes a lot of the UI out
of making something like that.
Jason (08:12):
Yeah. And for listeners not familiar with the term, UI
is user interface. It'sbasically how you interface with
the computer. It's the thingthat looks pretty and that is
meant for you to do. And if thecomputer is just running with
itself, it doesn't need that,because it's just talking to
itself with code.
Brett (08:27):
There was a UI in the sense of, I needed to have a
graphical interface that wouldshow what's happening so I could
just see how the players werebehaving. Was it actually
playing correctly? Or had Icoded something, you know,
wrong, instead of just lookingat the data, I wanted to see the
visual. You know what was goingon, but that's very just crude
graphics. And really, you know,nothing you'd want people
(08:47):
seeing. But it was basically asanity check of like, is this
actually playing the gamecorrectly?
Brian (08:51):
We were joking. It seems like Holotype might have been
played more times than any othergame just because of the
simulation time.
Brett (08:58):
That's a good that's and interesting point because I ran
my basic run of the simulation.Is once everything was working
and doing what it was supposedto do, I would turn the graphic
part of it off, and it wouldplay basically 50,000
simulations of the game in abouta minute. So that would spit out
a it took me way longer to dealwith getting that data into
(09:18):
Excel and making reports that,you know, gave me some
interesting info than it did toactually run the simulation once
everything was said and done,and any time that something
would be shown that was like,Oh, I that should be, you know,
modified a little bit. I couldjust go in, go into the JSON
file, change the value of acard, or change the, you know,
point value of something, andrun that simulation again right
(09:38):
away to get another output tobe, "Did that change it? Did
that go in the direction wewanted?" that kind of stuff. So
yeah, to say it was run 50,000times. It was run 50,000 x 100
appear.
Jason (09:50):
So like, now that I've heard this, this sounds like an
obvious thing to do for tryingto balance and play test the
game, but I can't say I've everheard of another game actually
doing this. I hear of openbetas, I hear of people doing
stuff online, digital versionsof the game for people to play,
but I haven't heard of someonejust having the computer play
itself a whole bunch. Was thisan original idea? Did you hear
about this from somewhere else?
Brett (10:11):
I didn't hear about it from someone else. I just
because of my programmingbackground. And it wasn't, I
don't think it was something weset out to do from the
beginning. It was just as thedesign was coming together, and
I was like, I want to make surethis is balanced. I went, Oh, I
can easily do that myself andplay test it way more than
asking a bunch of people to playtest it to get those, you know,
real true averages over a crazyamount of time. And I think it
(10:33):
would be useful for other boardgames and stuff like that. It's
because I was the programmer andone of the designers, I knew the
game so well that it wasprobably easier for me to
implement that once I explainedit, I had other designers that
were like, Hey, can you do thatfor my game? And I was like, uh,
you know, I could, but do I wantto? So I think it would be
useful, because that's one of myfrustrations sometimes when
(10:56):
playing games, is you'll playit, and then someone will find a
strategy just like, Okay, that'sall you ever do, and you'll just
win, because it's justunbalanced. Through that one
thing I've had, I've had gameswhere I was like, Oh, I figured
it out, and then it was like,there was no fun for the other
people, because someone was justblowing other people away
because they played a certainway. And that's something that
we really wanted to haveholotype be super balanced. And
(11:17):
I think a lot of the feedbackwe've gotten has been that,
yeah, it's incredibly balanced,especially point wise at the
end. There's been plenty oftimes where people have, you
know, had to go to the tiebreaker because they were right
in the, you know, same area,point wise and stuff.
Brian (11:29):
So we talk about a game being solved right, where
there's always an obvious choiceto make. It sounds like you were
working to keep the game frombeing solved.
Brett (11:37):
Yeah. And there's, there's an interesting point
there that goes to designing theAIs. When I made the AI to play
the game, it had to play it theway a human would play it,
because sometimes gamers aren'toptimal. I didn't want to make
the AI figure out like, this isthe best way to play this game
and optimize it for somethingplayers never do. So it was, it
was definitely more of like,okay, this is the way the
(11:59):
players play. It's weighted totry different strategies and
stuff, but it was very muchinformed by how people would
play the board game. If thatmakes sense.
Jason (12:08):
What does that look like? Because that was the thing I was
really curious about, is, howdid you program an AI to
essentially fill the role of ahuman? Because when I'm playing
this game like I'm looking at mycards, I'm looking at my
personal goal. I'm looking atthe global objectives. I'm
looking at other people's stuff.Obviously, a computer can keep
track of that pretty easily. Butthen making, like, integrating
that and making the decision of,what should I do next? That even
(12:29):
I don't know how I make thatdecision sometimes. So how did
you represent that for an AI tomake that decision.
Brett (12:34):
It's represented in basically the way you would as a
player. So you're looking atyour cards. You have, I have a,
you know, theropod from theTriassic. I have this from the
Jurassic. I have this manyresources of that, and I have a
personal objective that I'mgoing for. So those actions are
all given weights based onwhat's in your hand and what
resources you have, even to thepoint of like, okay, I have, I
(12:55):
have these two Jurassic cards,but this one's worth more, but
it also costs less in resources.So all those things are just
weighted, and then it'sbasically making a decision on
which one of those actions hasthe highest value, which
eventually really just relatesto which of these actions is
going to get me more the mostpoints at the end of the game.
And then, if I want to test adifferent strategy, basically
you're just changing thoseweights like this, AI is going
(13:16):
to want to publish globals morethan the other ones do. That
kind of thing. And then at theend, which AI wins more of the
time, which actions, that kindof thing. And then that leads to
just more the balancing.
Brian (13:27):
Okay, so your your AIs have some biases in them, built
in some preferences.
Brett (13:32):
Preferences, and you can change those to, you know,
investigate other assumptions.Like, one of the things as we
were playing was like, certainplayers were like, I never go
for globals, or I always go forglobal and then so by doing AIs
that did different preferences,you could see like, is that a
better strategy? Is thisstrategy not even viable if
someone just does this? So thatallows you to test that over
(13:53):
1000s and 1000s of plays, whichyou just can't quite get in the
plane with other humans, becausepeople just have their own
preferences on how they play. Wewant to make sure that if they
did have those preferences, didthat mean they always lost or
always won? They'll balance it.
Brian (14:05):
So, you've got the AI Jason who always wants to play
optimally, and the AI Brian whowants to collect the cool
dinosaurs,
Brett (14:11):
yeah, and that, and that was one of the one of the AIS is
just like, whatever cards thefirst in my list is what I do. I
call it my eight year old AI,she's just like, what's the
first one? I that's what I'mgoing for.
Jason (14:25):
Did you try to make it so that that AI could occasionally
win? Or was that sort of, like,the control AI of you have to
beat this one to be considered adecent AI.
Brett (14:32):
There's definitely the control of just like, this is
the most simple thing. Does thatone ever win out of 50,000 play
throughs, maybe a couple times.But you didn't want that one to
have, you know, the same odds ofwinning as someone who's
actually, like, paying attentionto what's happening, going and
looking for certain things, fortheir personal objective and
that kind of thing.
Jason (14:50):
And then, as you were using this to polish everything,
what aspects of the game did youtry to tweak? I mean, I was
looking over the box and, like,I can think of dozens of things
you could try to do. Places onthe board, the number of things
on each face of the dice, thehow long it takes to get a
graduate student, the pointvalues on the cards. Obviously,
you already mentioned those. Didyou make a list of like, here's
(15:10):
everything in the game that canbe tweaked. I'm not going to
systematically test all ofthese. Or were there some that
he's like, Okay, I need to tweakthese. These are not so
important. Like, how did youdecide what to check?
Brett (15:20):
So there's certain parts of the game design that are just
like, once you've decided it,that's what influences
everything else, and that thatcomes down to, like, what is an
action worth in an actioneconomy game, and what is a
resource worth? So there wasbasically before any of this
simulation in Python, all thatstuff was in Excel, so that it
was like, Okay, if I have, ifthis is how many pips I have on
(15:42):
a, you know, purple die, what'sthe distribution of that and
probabilities that all thesethings happen so that I know
when you roll this many dice,this is the average pips you'll
get of this type of resource,those things kind of had to be
like baseline, and you don'tmess with those once you get
into the simulation, becausethat was just core mechanics.
And really all that determiningis, what are those values on the
(16:03):
cards? How does that turn into avictory point? And that's where
the simulation so it was all thevalues on the cards, which comes
from the point values on thespecimen cards, and the point
spreads on the the personalobjectives and the global
objectives. So those were themain things. It was like balance
the point values for all thatstuff using these simulations.
So that was the big one. Andthen some of the stuff that got
(16:24):
changed is personal objectives.Definitely got changed, because
there would be certain personalobjectives based on the
distribution of those specimensin the in the deck. How often
would this person win with thispersonal objective? How much did
that skew? Same thing withglobal objectives. How often did
the game end early, becausethese global objective cards got
filled. So those values were theones that got the major changes
(16:45):
from the simulation, as well asthe like. Now we know you only
need 35 Triassic cubes out of50,000 playthroughs. Those only
ran out 50 times that kind ofthing.
Jason (16:54):
All right. and so looking at what you just said, so the
global objectives, personalobjectives, specimen cards,
those are the ones where playershave a choice. Basically the
like the excavation cards, thedice rolls, those things are
outside of player control. So itsounds like you set those as
these are just aspects of thegame you have to deal with, and
it's the ones where players canchoose where you were fiddling
with things. Is that right?
Brett (17:15):
Yeah, that's correct. Now, some of the like generation
cards, like the fieldexpeditions, those values could
be tweaked too to test that kindof stuff. But those things were
typically set pretty early on,like the amount of cards of each
type in the deck. That kind ofstuff was kind of hard coded.
That it was going to be the 10Triassic, 20 Jurassic, 30
Cretaceous that went back toearly design, because it was
just like there weren't manydinosaur type in the Triassic.
(17:38):
Then they started, you know,being more in the Jurassic, and
then even more in theCretaceous, so that that became
a constraint of, like,constraint of like, we can't
have more than 10 Triassicdinosaur cards because there
weren't enough specimens thatwould be easy for us to import
into the game.
Brian (17:49):
Now I regret that I overlooked that. I really wish
I'd noticed that specificpattern and pointed it out, but
I guess for our listeners, theyfound it up now, so no worries.
Brett (17:57):
Yeah, and if you notice on the on the pips on the dice,
purple Triassic dice only havesix the blue Jurassic dice only
have eight, and the Cretaceousdice have the green ones have
10.
Brian (18:08):
The lure of the Triassic was just too strong for me. I
knew it was more valuable, and Ikept chasing it and would lose.
So I was one of those bad AIs,
Brett (18:16):
and sometimes it worked out. You never know,
Jason (18:18):
yeah, but it was funny watching him get so frustrated
as the paleontologists we'replaying with would just keep
tossing these Triassic fossils,Triassic specimens to go after
the cool dinosaurs they reallylike.
Brett (18:28):
There's the personal preference AI, the one that only
wants to publish Triceratops.Yeah, we had one of those.
Jason (18:35):
Also, there's a question with all of your hundreds ot
thousands or millions ofplaythroughs, you might be able
to ask Brian and I, every timewe play this game, we have the
question, how long does a gamelast? Like, how many actions
does each player get before thegame ends? Because we're kind of
thinking the metaphor of like,okay, this is the
paleontologist's career. Is eachone of these actions a month, a
(18:56):
semester, a year? It meansnothing for the game. But we're
curious. So how long does a gameusually last? How many actions
does each person get?
Brian (19:05):
Good question
Brett (19:05):
that's interesting. Now, roughly about 40 during a game,
and it's going to changedepending on there's different
variants, right? You can playthe longer, shorter game by
using the different playertracks. So there's been times
where, you know, I'll play a twoplayer game with the five player
board, just because I want topublish a ton. But the average
amount of publishes for eachplayer is around seven to eight
by the time you're done with thethings, I'm sure a
(19:26):
paleontologist could answer thisbest, but, but I expect that,
you know, maybe that's a decadeor more research if you're
publishing, you know, eighttimes. Because when you're going
out in the field and discoveringstuff, it doesn't definitely
takes a while. But then if yougo back, you know, way back to
the bone War era Marsh, andcope, they were pumping out new
ones like you know,
Brian (19:41):
that makes me think that the best analogy to overlay on a
paleontologist's career for 40rounds is 40 years, because I
think there are plenty ofpaleontologists who are still
working into their 70s.
Jason (19:51):
yeah, but you hope that they published more than eight
times in those forty years
Brian (19:54):
you would again. No simulation is perfect, right?
Jason (19:57):
Yeah, although you mentioned the Bone Wars. That
was the number one request fromour paleontologist friends, is
that they want the Bone Warsexpansion to the game.
Brian (20:05):
Yeah, I want the Bone Wars expansion. I want the
traitor mechanics.
Brett (20:08):
Yeah, that's definitely been asked for, for sure, where
that's like, I'm going to stealyour fossils, or that, or I'm
going to, you know, denounceyour spec, your publication,
because that's wrong, and nowmine becomes the better one.
Brian (20:20):
Yeah, you need a fourth type of meeple, the, I don't
know,
Jason (20:23):
the saboteur? the infiltrator?
Brian (20:25):
The saboteur is probably the best way to put it.
Brett (20:27):
I like it
Jason (20:28):
someone with like the cloak a fedora and like little
spy mask,
Brian (20:32):
since they have to be they have to be indicated by the
type of hat.
Jason (20:36):
So Brian, do you have any last questions you want to ask?
Brett,
Brian (20:39):
I did so one thing I was curious about, so a lot of board
games now, wingspan evolutionand a lot of other ones root to
have these digital components tothem. There's an app you can
play wingspan on. It's arepresentation of the game. Lots
of games have made their way tosteam. It seems like you've done
all the work on the back end todo that for holotype. Is there
any interest in pushing it to anapp?
Brett (20:58):
Yeah, I guess so. That just seems like more programming
work for me, which I enjoydoing, but I don't have a lot of
experience with App Stores andgetting and doing stuff like
that. Mine's more on just theback end, spitting out all the
stuff. But I know there's, youknow, companies out there that
will take those games and makethem look beautiful and nice and
playable on a on an app store.So we'd be more than interested
(21:20):
in that.
Brian (21:20):
I had one other question, and you don't have to answer,
but I'm curious, if you want togive a look behind the curtain,
what was the mechanic you guystossed?
Brett (21:27):
Oh, this is a great one. It was a horrible one. That's
why I got tossed. So all themilestones that you get during
the game. So adding, you know,extra research or another, every
time you hit one of those greenstars, you add a milestone to
your board to get a littlebetter at something. Those used
to be in the game. Now, everyplayer has access to the same
set of those in the originalgame. They didn't
Brian (21:49):
there was zero sum?
Brett (21:50):
yeah. So there was the number of players minus one
available. Oh, and and ourbrilliant catch up mechanic was
okay, when you hit those greenstars, everybody adds up their
current points. Whoever's inlast place gets to choose first.
Brian (22:06):
So that's a little bit too close to real life.
Brett (22:09):
Yeah. And, and the problem is, while that was cool
in the thought of like, Oh, it'sa catch up mechanic, someone
could snag something thatsomeone else, the leader would
get last pick, and usually getsomething, you know, less good.
It made the games take aridiculous amount of time
longer, because that wholescoring process you would have
to do four times throughout thegame. Then you'd have the
picking and someone waiting tobe like, do I get this? Do I get
(22:30):
this? That indecision? And wewanted a really fast, snappy
playing game so that that onehad to get thrown out. And I
think it got thrown out for thebetter, because now everybody
having their own set, it's easy.You know, when's you know, you
can make your choice on yourtime, and it makes a little more
fair. Everybody's got the samething. You can't complain that
someone else snagged this beforeyou.
Brian (22:47):
I'd be really sad if I didn't get the storage closet
for the extra fossils.
Jason (22:51):
All of us always chose the storage closet first because
we're hoarders.
Brett (22:56):
That's interesting. Yeah, that's that's super interesting
because that's one of the ones Iit's very subjective to me on
whether I choose that one ornot, like, how many fossils are
currently flooding my storage?
Jason (23:04):
Was that digitally play tested? Did you change the AI's
preference for which one ofthose it would go for first
Brett (23:10):
in a couple states? Yeah, because there's only six in the
base game, that was a fairlyeasy thing. I'm going to get
four of them throughout thegame. I'm just going to randomly
pick them. Or this one's pickedalways first. Does that, you
know, skew it too much. So thatwasn't too hard to test. And
it's funny trying to make those,you know, game things into like,
how does this, you know, relateto the field of paleontology?
Because, like, I spent a lot oftime on making sure, like, all
(23:31):
the global objectives hadmeaningful names within
paleontology. And people thatplay in the game just don't
care. Paleontologists, they're,like, systematic what is that? I
don't care
we have, like, systematics,cladistics, and phylogeny, and
one of the paleontologists brokedown the differences between
them, and at a certain point Ijust, I like, yeah, okay, um,
sure.
We wanted to be true to it, sothat someone that was really
(23:53):
into it was like, Oh, I reallyappreciate same thing with,
like, the colors of the dicebeing related to the
international stratigraphychart, that kind of stuff. It's
like, that's an unnecessarystep, but people sure appreciate
it when they learn about it orknow about it.
Brian (24:05):
We were trying to figure out why only one person can
publish per turn.
Brett (24:09):
That's purely game mechanics. We wanted the one
spot that was going to force alot of, you know, bumping
mechanic going on.
Brian (24:15):
We retconned that there was a departmental copy editor
or the press release person.They only had so much time.
Brett (24:20):
I like it. Yeah, there's, there's there's certain things
where it's like when we playedwith the paleontologist, like we
realize research is not as easyas just going to the University
library. Please, please takethis as a you know, conceit to
game mechanics.
Jason (24:33):
One last question. This is the million dollar question,
what is your favorite dinosaurin the game?
Brett (24:40):
Oh my gosh. So Well, my favorite dinosaur would be
Stegosaurus. My favoritecreature from the mesozoic is a
marine reptile, which is theMosasaur,
Brian (24:50):
anyone in particular?
Brett (24:51):
oh my gosh, probably a tylosaurus.
Brian (24:53):
I don't know if I have a favorite. I was a sucker for
ichthyosaurs personally. But myfavorite dinosaur is
parasarolophus.
Brett (24:59):
Those. Are pretty awesome. I'll have to send you
one of our Parasaurolophus pins.
Brian (25:03):
That would be, yes, you can do that. I will allow it.
Jason (25:08):
Oh, yeah, just twist his arm there.
Brian (25:10):
Yeah, we'll have to send you stuff too, of course. Thank
you awesome. We'll do anexchange.
Jason (25:15):
All right, that's a good place to wrap it up. Brett,
thanks so much for coming on.
Brett (25:19):
My pleasure.
Jason (25:20):
We've already told people a lot about holotype, but if
people want to look you, to lookyou up or other than advertising
your game, is there anythingelse you want to talk about,
other stuff that may be in theworks from brexwerxs, or other
things you want to mention?
Brett (25:30):
Yeah, we've got some stuff in the works, but you can
always check us out on brexwerxsgames.com we just released our
8-legged Peacock , which is agame about the Maratus peacock
spiders from Australia and theirmating dance.
Brian (25:41):
we saw that. I'm excited about that. That might be what
something we put into one of ourlight games,
Brett (25:47):
light science game.
Jason (25:48):
Well thank you so much. This has been great, and it was
really nice to have thesequestions answered that I was
telling Brett beforehand, like Iwas thinking, Man, I wish I
could just know the reasoningbehind, like, wait a second, I
run a podcast? I can just askhim and get him on the air
that's content.
Brian (26:06):
Thanks for being our first creator willing to come on
and talk to us about their game.
Brett (26:09):
no problem anytime.
Jason (26:11):
All right. Thank you so much. Brett, good luck with
everything. And I guess youknow, if your current job
doesn't work out, you can alwaysstart up a new job as a a board
game designer, play tester,because you apparently have a
skill set in short supply.
Brian (26:24):
And if there's anything new from brexwerxs, send us an
email. We'll make sure that wepromote it for you.
Brett (26:29):
Sounds good. Thank you so much, guys. And for our
listeners,
Jason (26:31):
thank you for tuning in. We hope you enjoyed this first
interview we've done, and untilnext time, have a good week and
happy gaming. Have
Brian (26:37):
fun playing dice with the universe. See ya.
Jason (26:41):
This has been the Gaming with Science Podcast copyright
2025 listeners are free to reusethis recording for any non
commercial purpose, as long ascredit is given to Gaming with
science. This podcast isproduced with support from the
University of Georgia. Allopinions are those of the hosts,
and do not imply endorsement bythe sponsors. If you wish to
purchase any of the games wetalked about, we encourage you
to do so through your friendlylocal game store. Thank you and
have fun playing dice with theuniverse.
Popular Podcasts
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
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.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.