How SCAM Penalizes Bad Behavior with Matt Stephenson & Kevin Olsen

Transcript

foreign [Music] ation on the Pod today we have Matthew Stevenson who is head of crypto economics at Pantera and Kevin Olson who is engineering lead at git coin so basically why I wanted to have them on the podcast is that we're talking about behavioral Game Theory today and its applications in the regen web3 ecosystem and public goods funding and how we can shift the crypto economics of web 3 to more regenerative systems so behavioral game theory is a way of modeling the irrationality of the actors of people in economic system and scam stands for sub

game incredible anti-mev so basically a game theoretic way of solving for Mev on the podcast today we talk a little bit about applying behavioral Game Theory and scam to solving some of the biggest civil resistance problems inclusion resistance problems that get coin and quadratic funding have so I think a really fun episode because it dives into public Goods funding behavioral Game Theory and crypto economics in a big way so I think you're going to enjoy this episode Athena Dao is addressing the underfunded areas of women's health such as menopause and endometriosis Athena is a Community Driven doubt giving women a say in decision making and funding the scientific breakthroughs that impact their lives with cutting-edge technology like molecules IP nft framework they

invest in research IP that ensures exclusive licensing rights Athena Dao is Raising awareness to women's health research working alongside existing organizations so join Athena Dao today to help change the landscape of women's health research what's up Kevin what's up Matthew what's going on man how you doing Kevin pretty good excited to talk about a scam but before we get into that Matt what is behavioral Game Theory and how does it help us think about public goods yeah awesome so but behavioral game theory is basically a kind of modern extension of game theory that lets it be a little more respective of real human behavior it's incorporated into mainstream economics now but it's not really fully into crypto economics formally even though we use it all the time so a classic example that we use in

in the sort of Behavioral uh Game Theory world is say charitable giving Which relevant to bitcoin obviously right you when you're donating to the matching cool part of that behavior is charitable why do you do it well there's maybe a warm glow effect right you may enjoy seeing these projects get built you may enjoy the feeling of giving to them you know maybe there's a sort of status and reputation effect uh where I care what other people think of me and it's nice to sort of be associated with this cool project and get my name on the letterhead all those things are fundamentally behavioral game theory in the sense that they're like me caring about these non-material payoffs right around uh you know benefiting others or things like that which you know is not to say that we don't think material payoffs matter we certainly do so you

know there may be totally self-interested reasons to give to the matching pool I think in a lot of cases there probably are for for marketing reasons but it's kind of about saying both and ultimately you just see this stuff all over crypto right like we talk about noise trading when we're modeling uh D5 with that's irrational Behavior we talk about social identity with Dao's I mean nft Behavior obviously and then you know vitality's excellent post about social legitimacy sort of like uh underlying a lot of the security models and things that we do so I think I think behavioral game theory is kind of everywhere but uh it's nice to get a little more formal about it so I'm really excited to talk what I'm hearing at least the way this might be too reductionist but the way it kind of places in my brain is like game theory

is what would perfectly rational agents do in a very contained system but behavioral Game Theory sort of recognizes the diverse tapestry of uh things that shape our Behavior social preferences social utility psychological factors like behavioral Game Theory feels like it's more applied but also just recognizes the complexity of how people are kind of predictably irrational is that true reductionist or is there anything you'd you'd alter about that I think that's great I think that's great yeah it's just kind of funny how the the ecosystem has evolved I feel like in 2017 when I got into the ecosystem everyone was writing these icos and these white papers and the the kind of thought was like oh if you take a human and you put in a token you will get out in action and it turns out that all these projects raised billions of

dollars of capital and most of them never got past the white paper stage of deploying things and you know there was a lot of things wrong with the Ico level but taking again into account uh that humans are just more complex than just like oh I get token I produce Behavior I think was a big thing for me absolutely yeah and again so I we talked more about the like preference side of things like social preferences another part of hate real game theory is you know cognitive limitation beliefs things like that right so that's the kind of the other side of it which I think is certainly relevant to the Ico bubble from what I saw um where it seemed like in some cases people would just see these complex looking white papers and go I don't understand it and some people think it's cool therefore I'm gonna eat uh and so

you know cognitive limitations and things like that uh can can apply and they even get there as well well uh I'd love to get into scam and what is it how does it help us fund public goods how does it help us coordinate so let's get into it what did scam yeah yeah so so stem is a sort of Applied behavioral Game Theory mechanism that I was calling uh you know sub game credible anti-nev because it came up I was you know I work at Pantera and flashbots is a is a portfolio company I was talking to them and they were like you should do a weird behavioral Game Theory approach to Ned so it sort of emerged out of that but um but it's broader than that and I mean I think through conversations with with governing yourself to the idea that it could be applied to anti-civil the

essential idea with scam is that whenever you're going to have somebody make a decision and that decision has externalities like this ability to harm others what you can do is you can have them put up some sort of a stake and then you can allow others including you know maybe mostly the people who are possibly harmed to retroactively earn that stake at cost to themselves right so originally with sub game credible anti-meb but we might just call it sub game credible altruism because that gets us the M uh because what that is is it's altruistic sanctioning when you when you burn that money so I can actually give us a little game example if you want so like imagine Olson you you give me 10 bucks and you tell me hey split this 10 bucks somehow between uh between you and

you and I walk right and uh and here's the game so I have this 10 bucks and Milwaukee I'm just gonna split it and then however I split it you get to say yes and then the split goes through or else you get to say no and the money gets burnt we both get nothing right okay so I have 10 I have 10 10 die let's say right okay I'm gonna I'm gonna take nine I'm gonna give you one yes or no no I only get one you jerk yeah exactly so so that's really common right I think any upper less than three gets burned like no one gets anything but now now look at that in a traditional game theory model you just chose zero dollars over one dollar we assume people don't do that but what did you also do you also took a dollar and set it on fire so

you could burn my nine dollars it feels like a game of chicken between us there it's like well I'll burn this whole place down if you don't give me the my mutualistic five exactly but it's credible because for you uh it's not a theme of chicken where you have to lie and say oh I'm gonna do the duh it's credible because you feel the emotion of screw you I'm not taking a dollar for your well you get nine jerk like that is utility maximizing so that's the uh sub game quotable part so it feels like to me like traditional Game Theory would be like oh Kevin will choose to get one dollar instead of zero dollars uh whereas behavioral game theory has that emotive sort of like fiery I'm gonna burn this place down kind of set unless I get five thing and behavioral Game Theory would predict that uh that

I'll shoot for five over five whereas regular game theory would maybe say oh I'll settle for one because it's better than zero yep yep your traditional Game Theory will settle for Epsilon for basically the tiniest traction I could possibly give you you'll always see yeah these sort of mechanisms uh which I just sort of we Illustrated one with that game have been studied and replicated widely across the world and they're they vary according to culture but they're but the fact that people are willing to punish Injustice really does not seem to vary as far as we can tell some sort of unfair offer gets punished no matter where you conduct these sort of games so to translate what the game we just played back into the scam um we could say that like what I just did was I put up nine dollars of stay

right and you got to use a dollar to burn that nine dollars to punish me for my unfair Behavior so that's that's fairly generally applicable and and I mentioned it came up initially in the in the context of Med because if somebody got Med you have a white hat hacker let's say who figured out a way to sort of unlock a 100K that was stuck in a smart contract and then somebody before that transaction gets uh gets you know it's in the mempool somebody simulates it and figures out that before it can go that's worth a hundred thousand dollars on the front run that and I'm gonna take the hundred thousand for myself and screw the white hot Hackberry and the person whose funds they were trying to unlock completely right so that's a frustrating thing and if you instead made the Builder the the Mev or in this

case whoever it is put up a stake then you could use that sense of unfairness to um to sanction the person who's doing the wrong thing right and so the point here is not that we do a lot of sanctioning the point is that everybody knows that sanctioning is available so they behave more fairly right so if we really played that game I would have probably just offered you five or four maybe right uh I think The Optimist stopping for because people usually say yes to that uh just because I know that you'd burn anything less so in the uh when you apply the scam you don't see a lot of punishment you just see a lot of good behavior so um it feels like it's a way of constructing a behavioral Game Theory uh mechanism to stop Mev and could you take us through

the definition like word by word so it's scam which by the way I respect the mean game the meme game calling it scam uh Sub sub game credible anti-mev so is the sub game basically the fact that what is the sub game in this so so when you when you break out these games and you model it in Game Theory you'll you'll often have complex decisions over time and one way you you solve this is you solve it in the sub game which is you take let's say some smaller portion of the larger game and in this case we'll do it as the the final portion right where it ends and you can predict Behavior really cleanly in the last period of that game right because then it reduces to this kind of like One-Shot game which we know

how to model pretty well from there you go backwards via backward induction when you solve again that way so subgrade credibility often determines the equilibrium of these games I know this is super jargony academic nerdy stuff but the essential idea is that when you said this is like a game of chicken and I said that yes in it but it's a sub game credible commitment what I mean is that even if we never interact again you'll still burn that dollar for nine dollars even if it's in some sense the end of the game the sub game you still find a utility maximizing to burn the dollar for the nine and that's what's essential about it because if you if you're relying on these repeated game assumptions and all these other things those are familiar mechanisms that we use all the time right around like

reputation or you know Etc and and that's what's kind of different about this one it almost feels like the divide and conqueror strategy to me so if Mev is this complex social system and uh if I can divide it and Conquer it into a sub game that's solvable then that gives me a foothold into solving me be bubbling up from there I think that's right and and plays nicely with all the solutions people are proposing around Meb which I think is one of the good things about it right like wherever you find a gap that you can't quite fill with a with a with a different mechanism you could apply the scam to sort of say okay well we can't quite figure out censorship in the case of trusted Hardware but you could use the scam on let's say censorship right and even if trusted Hardware is

getting you the rest of the way this can kind of fill in the gaps anyway that's helpful and so I'd be curious I I want to I want to pull in Olson in a sec and we can start talking about how this could be applied to public goods funding uh but I am wondering before we before we pivot into that have you seen scam being used in the anti-mab ecosystem yet or is it is it really in the theoretical State I mean there's talk of implementations in and out locations of it but it was it was just proposed about a month ago so it's still pretty early so I actively be done but I think there's interest there from a couple of people who are actively working at this point I'd love to Pivot the conversation over to you Olson I think that you and I

were sitting in the audience as Matt was presenting skim to to us and I think that I was just reveling in the memetic value of it being called scam and you had some deeper thoughts about how it could be used to solve public goods in in civil so uh can you tell us that Olson for sure man yeah it was uh it was a great talk I don't know if it's going to come out online at some point but you should link that out to folks because it was it just goes that extra kind of level of depth and it's great to see um um and so like I was listening to that and it just felt like I'd like you know kind of struck by lightning right like like this this is like an answer to one of our biggest problems right uh so okay

so I work a Bitcoin uh as everyone here knows but like one of our biggest problems right is uh we talk about civil civil attacks right people you know standing up stock puppet accounts and distorting uh quadratic funding mechanism but one of the issues there is that like um symbols the attacks don't happen in a vacuum right the usually someone has control of a grant and they're using this kind of Bot Farm or bot Army of civil accounts to you know like donate to their own Grant and they're they're you know using their own uh they're cycling them back out we can watch this stuff we always have to catch it in the sort of retroactive sense it's very difficult to gate proactively now we built passport and we try to like prove humanity and that's great up to a point but you don't get the whole picture

right and that's kind of a an infinite game right we've talked about this that like you're going to always be chasing your tail uh trying to prove an individual's Humanity but when you observe these systems retroactively there's this moment where you really want to like you know you find Bad actors and sanction them so is even like the sanctioning term which I'd used internally a few times that's hearing kind of jump out in Matt's presentation like this is exactly what we're looking for glue that on to this other thing we've been talking about get coin a lot is like you know how do we get people to start staking on either their identities or maybe their the grantees need to stake but there's always this like what's the slashing mechanism right and the problem there is the slashing mechanisms are always these like kind of

clunky Court appeals like thing that I was like who's going to do that work right like this just does not feel scalable like you're just kind of pushing this into this like social reputational space it's just going to be fraught right and we've even seen this right like when get coin in the past is like you know sanctioned a civil ring it plays out in the forums and it's awful right no one feels good about it what was really elegant is about this is now it's this mechanism where if a grand key is staking right and you know malfeasance is discovered right someone's observing the donation patterns it says this looks like a scam you basically broadcast the network and there's this mechanism where anybody can burn some of their Capital to then either prevent that person from receiving matching funds at the simplest

level and then you could also burn funds to like slash their Stakes so you're kind of like two ways you could first you could kind of ostracize them right and exclude them from the rewards which they've you know earned through their malicious behavior uh so they just don't get any matching or you could literally like actually slash snake and really really like hurt them right like really punish them and um one of the things that's been sort of stuck in my mind was uh you know ostrom's uh polycentric governance paper one of the big things that jumped out at me is like one of the things you need to have is the ability for these different groups to sanction each other you have to have this possibility for the different groups is the polycentric groups to be able to exclude each other and actually have

some way to manage this common pool resource and that's always been missing for me when people talk about these games is that sanctioning element so what I loved about this it puts the power of sanctioning back into the community the job to be done is really just a broadcast hey we think we found a scam and it's some sort of mechanism level there's ability for people to earn their own money in this like this like very psychologically powerful thing like I burned my own Capital to punish somebody um it also has this Elegance of like there's probably not a great way to like turn this into an economic game of self like a self-interested economic game so all that like came together for me started like right outside and I think it talked to Matt cornered him for a while I was like how do I make this work

uh we've been working together on a on a paper like a like a gov post hit the uh Bitcoin forums here pretty soon um where we try to like spell this out how this could work but uh but for me this is like super elegant solution to you know one of our biggest problems which is like really just securing the allocation of capital in the Bitcoin Network so this for me feels like like a like a like a total like first thing to try um I'm just super psyched about it yeah so let me let me play that back for the audience just to um bring them up to a level of common understanding so uh git coin uh which you're the VP engineering of um is a quadratic funding application primarily quadratic funding Bitcoin grants can do other things but um

quadratic funding has this amazing property where it you take a matching pool and you allocate it to the the results of a crowdfunding campaign but you allocate based off of the number of contributors instead of the amount that they funded so Olson if you have a grant that has a hundred dollars from 100 contributors and I have a grant that has a hundred dollars from one contributor you get 99 of the matching pool which has this amazing Democratic effect of how capital is allocated in that uh it gets allocated to the preferences of the poor and the many over the rich and the few and all also contributors to your Grant can contribute a dollar and they get tens or hundreds of dollars in matching which are two really amazing attributes but it also introduces this Achilles heel of if I can spin up a

bunch of fake accounts then I can drain the matching pool and send it to my grant so that's the Civil resistance problem and basically uh there's a bunch of gating mechanisms on bitcoin passport that help identify civil attackers and and like gate the round there's also a one dollar minimum which means that if you're going to attack the round you need to spend at least a dollar and then there's also retroactive data science where um where similars can be sanctioned and the problem with the sanctioning is that it becomes this whole uh really clunky system where people don't feel really good uh after the sanctioning has happened it plays out in the governance Forum it gets into a debate on Twitter and so is there a more elegant way to sanction the Civil attackers or the

collusion attackers in the system and and that's what you're sort of evaluating scam for did I play that all back in a way that that comports with you Olson totally totally and it actually decentralizes the kind of like this step of our our governance right like it pushes like the decision making that that sanctioning isn't happening from Bitcoin anymore and it can actually happen from any actor in the system uh or any other grantee and act I really like that Dynamic right it actually like removes us as this like you know Central Authority uh governing it and really kind of pushes it where it could actually be you know in the uh mechanism or the protocol itself I'll just add you know on the on the meme front since you

mentioned scam as a meme one of the things in my talks of trying to meme this is as kind of a retro public harm or or retro public bad I think is is sort of uh it wasn't suggested right and that's sort of the commonality I think that helps uh bridge between it might be any V and anti-symbal which as you noted is just something where it's it's a little easier to figure out these things a lot of times after they've happened right and it's very hard to design mechanism around that so you know the scam is it's post slashing because it's anti-slashing is kind of easy if we can tell you exactly what is the wrong thing to do meaning like just before the fact we can easily just slash you for it we could write that into the protocol but for anti-sibil stuff it's a creative act

same with an ed so it's very hard to pre-specify um but yeah it's a mechanism for for punishing retro public bads right for reducing retro public bats certainly civil attacks are are a great example of that and I do think you know as we've talked it through Olson and hopefully you know a lot you agree like that does arouse emotional feelings of unfairness sufficient that like when somebody feels like the matching pool got partially drained um when somebody feels like oh gosh I'm mad at that person for civil attacking and public good that's all you need for the scan to fundamentally be effective and it feels to me like that's there all the ingredients are there right this is time to put them together yeah one of my sort of principles for crypto economic

systems that's heavily informed by listening to a metallic talk about proof of stake is that you need to create an equilibria where it's cheaper to defend the system than it is to attack the system um and so like proof of stake is really elegant with this because I have to stake 32 each which right now is like 64 000 in order to create a validator node and I get incremental rewards for securing the network and if I attack the network my state gets slashed by 16 8 which is like thirty thousand dollars like can happen within 15 seconds and so proof of stake is this really interesting economic equilibria where it's just way cheaper to defend the attack and because of that shelling Point that's what secures all of our

bags on this distributed Network and I I've sort of just been using that as a as like a light in the darkness because Bitcoin grants is this really amazing way of funding public goods with quadratic funding but it just feels like it has this Achilles heel where it's cheaper to attack than defend and so building an equilibria or like a crypto economic system uh that that makes it cheaper to defend than to attack I think is how we scale it from and I'm using we in the Royal we here because of course I've disaffiliated from Bitcoin and it's now up to the Dow and blah blah but um you know the Royal we we just need to make quadratic funding way cheaper to defend than to attack and so you know creating uh on top of the Civil

inclusion problem interesting Innovative new mechanisms that can uh they can just shift that balance a little bit and if each mechanism can can shift the balance a little bit the compounding effect can be that making the entire system cheaper to defendant to attack and so um you know really interested to see how this fits with the Gestalt of other civil mechanisms I know that bitcoin's working on minimum donations and cluster mapping we just had Joel Miller on the Pod and the one dollar minimum and all of that kind of stuff this feels like another tool in the tool belt that like together they can make it cheaper to defend than than to attack I I'd say it's super complementary to something like cluster mapping if I understand that correctly right because that just reduces the cost of no of identifying the the Bad actors

those things are all essential yeah exactly I think that that's like a a really big piece that's missing too is uh you know I mentioned civil attacks in the in the kind of sense that uh here like I have a scam Grant and I'm building a bunch of accounts and I'm draining the matching pool that way but then you have these other actors that are distorting quadratic funding uh as well right you have airdrop farmers and you have people who are just maybe farming for like reputation uh and they're using you know our mechanism and they're kind of clouding the results of it but what's great about cluster mapping is I hope that you know you can Surface which group memberships they belong to and sort of overall you know attenuate the effect that they have which is like one effect sort of like anti-collusion piece but then it also

has this effect too it's just large homogeneous groups you know you can't just rally a bunch of you know look-alikes you know who uh drain a matching pool and actually it has this great kind of like tie into the dsock paper which is really trying to say like you know uh you know a little sort of plurality thing of you know cooperation across uh you know social distance and I'm really excited for like what that would do like what would that amplify in sort of a quadratic funding round if we had something like that you start seeing where there's common support across multiple uh communities and I think that's a very powerful thing to start amplifying not anything to other just totally agree I mean I think that's I you know we we have a the host we've been working on writes some some qualifications into certain ways you

might be sort of like tempted to overstep the the mechanism in certain ways right like one of the things we note is you know if you compensate people who are doing this altruistic sanctioning too much then you can create these kind of feedback effects where there's a symbiotic relationship between the Civil attackers and the people punishing them because the punishers make their living off of the Civil attackers existing right whereas the nice thing about the the kind of CM mechanism is you're perfectly thrilled with unfairness stops right you're not looking to punish people unless they're committing these things so actually you know something we were talking about just before um Olsen is is one of the findings you get I mentioned that if you test this mechanism across the world it works very well and people are willing to uh to

sanction unfairness if you allow people to sanction back against unfairness it becomes a little Messier and a little more culturally a little more culturally buried right so there's just like you have to exercise caution when you're implementing it in a game that might extend further than a sub game right where you will have reputations at stake and so and so on so I think we're we're pretty careful to note that but it's it's some fun behavioral Game Theory findings you can start seeing people getting into like a battle right where they're like you know each round you know they're sanctioning each other and this is where one of the things I like about this is playing with these Ray ratios of what how much Capital you have to burn to sanction and I actually think that this is a great thing to like actually uh you know govern at the

protocol level right so the Dow would that vote on okay so one to ten ratio to burn to prevent uh people from receiving matching funds or maybe to one to one uh to slash them and you can really sort of flatten like the and make it more democratic what the star sanctioning effect is or if we're not getting sanctioning and scamming is happening you could ratch that up to 1 to 100 ratio it's really easy to sanction people uh rather than having to Rally some big group behind you know like excluding some bad actor but you know like these are the things you can play with over time and hopefully keep from getting into you know these sort of like you mentioning those are like you know it was a broad Community right you could you could get these Pockets that get really uh uh you know in this Tit for

Tat kind of battle uh and you'd hate to see that happen right so we need to be able to play with these parameters totally yeah but it's a experimentation will sort of under those sort of Dynamics and and hopefully this will be a worthwhile thing experiment one thing we'll have to figure out and this is by the way my department is what's the meme if it's us if it's not anti-mev anymore are we talking about scas here Subway incredible anti-stibble uh anti-clusion I'm into it you big brains can talk about the mechanism and I'll just figure out the meme while you do but please get any audience members have an idea of what it's called when it's finding similar please let me know we could change it to sub game credible altruism or something instead just because it's all of them are going to include an

altruistic punishment component but scas Scott there we go take me back to the 90s it has a ring to it yeah um well I'd be curious uh if if we've picked all the fruits of of sub game credible anti-civil and anti-conclusion if we could zoom back out and talk about behavioral game theory in and around public goods and in crypto economics um I I'm just really excited about the idea that we can fund public goods with crypto and and you know our bread and butter from my time at Bitcoin um was quadratic funding but there's also conviction voting and retroactive public goods funding and a bunch of different mechanisms and and uh Matt I'm wondering if if you see uh any other

sort of fruitful areas to direct the audience's attention to with respect to behavioral Game Theory and funding public goods solving the tragedy of the commons things of that nature yeah definitely I mean first of all I I big fan of the ones you mentioned right with with like conviction voting is really cool and um you know I'm sure the audience is well aware of the um the laundry list of cool mechanisms you put together I think it's super modular which I think is a fantastic resource um one of the things that uh I've seen recently that I think just got announced maybe last month that I think is really cool on the front on the what you might call kind of a like met up public good you might consider uh this punishment thing a little bit of a metapublic good in a sense right where you're kind of

enforcing the rules of paying for public goods you can think of governance that way too right famously governance itself is a public good and so if you have to if you have to have governance over a public good that's kind of meta governance so there's a there's a company called m0 that's it's it's like a D5 project but they have this really interesting element on their governance where they constantly deflate people who don't vote and it has this it has this fun duel effect and behavioral Game Theory one of them is I'm basically saying I really trust you to vote I want to ensure you vote right so that's implied by the magnets explicit because like you're actively being harmed if you're not voting so it communicates this kind of trust and induces participation that way but the

other thing it does that's kind of fun is you know sometimes people aren't voting because they're worried about regulation because blah blah blah so it starts to also select out the people who are resistant on certain topics and dampens their voice because they're not really kind of putting their money where their mouth is so that's the most recent example of something that feels like a really clean elegant mechanism around governance which is of course you know sort of a nightmare problem to solve so um they're calling it uh the spog all about the uh all about the four letter all caps meme names here but uh big shout out to that one I think that was really cool wait spell that out for me what does Spock stand for that oh my gosh I wish I could remember let me see let me see if Google helps me I'm sure

it's something well you know and just in general I feel like we could do a whole nother part about behavioral game theory in in crypto and in governance and everything like that I mean just thinking about the cognitive limit that we discovered I mean I just remember early on in the get coin days you know you're designing this governance model and you you're or you're thinking like in my mind when we were designing bitcoin's governance back in the day it was like what would a fully informed uh participant that had read the whole all the governance Forum posts like vote and like the reality is that everyone wants to hold governance tokens and no one wants to spend any mental Cycles on anything and so just like the cognitive limits of I didn't even really read the thing but I have to vote on it

um and also I'm busy with like 17 other favorite like tabs open like those feel like behavioral Game Theory and um I think that we can design better governance systems realizing that probably you only have like a tenth of the the governor's brain when they're voting on something as a opposed to like their full attention um designing for for that kind of world is vastly different than than I think that what we imagined hoped it would be but working around those constraints or recognizing those constraints allows us to work around them yeah 100 I think that's exactly right like limited attention cognitive uh cognitive constraints Etc all explain that stuff and uh you know one of the things I definitely say on the behavioral Game Theory front is that most people who've been working in this space like you all

have they know this stuff in their bones right you might not be modeling informally sometimes like it formally helps sometimes not but most of the problems you're solving are actually probably closer to behavioral Game Theory than regular traditional mathie I mean it's all it's actually pretty Matthew behavioral too but you know the rational actor model we don't always we don't always call it that but I think it's it's always it's always pretty much there right so yeah I mentioned the examples of like noise trading and social identity and stuff but yeah when we talk about Dallas and this was coming up around around like nfts in the way that people who just share a sort of aesthetic preference will act the same way when exposed to certain mechanisms they'll have this kind of like strong in-group bias um it's called a minimal group uh like

like a minimal group bias so you can just sort people minimally and they will suddenly act uh as if this is the end group and this is the out group and that's a that's a pretty strong behavioral effect that only requires a little nudge to produce in some cases obviously that was full of that stuff yeah also anything to add here I'm just laughing at the idea of like subgroups within the Dow and with the nudges we have to act the same way uh no I don't really have much to add on that I do I do love that there's finally you know like formal ways to I think you know confirm like you're saying the things that we've observed and things that we like kind of know in our bones so you know definitely love when uh you know uh Academia confirms the things that we

already think that's a good feeling but also what I really love too is that this is experimental so this is I mean like one of the things you said that uh I don't remember they came up already at this but this stuff's been widely repeated you don't have the reproducibility problem of other social science to Sciences so this is for me like super exciting because it could be reproduced end times across n labs and you get very consistent results that are predictable and repeatable and that's what gives me a lot of like confidence this isn't just like one group found this one weird effect these are like oh truth Sam maybe if you will like within the way people play these games yes widely replicated truths and I actually wind up using the term behavioral Game Theory over like behavioral economics just to sort of communicate that even

though to an academic audience they would kind of be the same thing um the fact is that like behavioral Game Theory kind of began as like the way crypto analysis or crypt analysis is to cryptography where it was like a way of basically pushing against and breaking these systems early behavioral game theory was just no you're wrong and we're going to show it to you via a thousand experiments and because those that was such a fraught battle these things got replicated and replicated so actually um you know when we go back to that game we we first talked about the the splitting the ten dollars you know a natural reaction that people gave is ah that'll go away if it becomes a thousand or ten thousand dollars turns out it does not which is kind of astonishing how robust that sense of fairness is it attenuates a little bit

but it does not go away I mean generally speaking even if it's ten thousand to a hundred thousand of the equivalent you're still willing to burn it because of the unfairness which sounds kind of crazy but but uh that's that's pretty well validated it does become a weaker effect but the effect is still there it doesn't just vanish as you as you need to stay tired which is which is I think a really cool thing have you guys proven this no free will yet yeah have you been hanging around that came up a lot well um we only have a few minutes left I guess is there anything I didn't ask that you want to that you want to say Matt appreciate the opportunity to to talk about this stuff and cover it and the excellent questions it's fun to be able to try and apply it in a new domain

and everything too with anti-sibility I'm really looking forward to the experiment a bit we're able to pull it off can't wait to see where it goes uh and I'll keep uh you know carrying my way by trying to figure out the meme uh if it is successful please yeah help us out Wilson is there anything I didn't ask you that you want to say uh no not so much um I'm really excited to start hearing how the community reacts to this you know as people read uh you know the post and hear this uh like what the sort of feedback we get is but um you know my my enthusiasm is is still there but looking forward to seeing where the debate goes as we try to apply this to bitcoin well thank you both for uh for joining me I'll have a link to both of

your twitters in the show notes and then also Matt you have a Twitter thread uh articulating what scam is and how it works so it encourages listeners to check that out um and then if by the time this airs y'all's uh gov post about how this applies to quadratic funding is live I'll also link that in the show notes cool I think we can do that already thanks for joining me peace and love cool thanks [Music] thank you