Topic: Game theory: Superrational thinking vs rational in BitCoin (Read 815 times)

I found and shared a link a couple of days ago: https://bitcointalksearch.org/topic/m.23084247

Incidentally, the most recent Nobel Prize for economics was awarded to a guy who has been trying to tell people for 40 years that the financial markets keep getting people caught out simply because people are completely incapable of perfect rationality. Even banks during the mortgage crisis 8-10 years ago failed to act rationally, realising they were getting into worse debt by accepting bad deals. Logically, they could only head down into disaster but went down that path anyway until it was simply too big to hide.

I've seen it also in crypto: Mt Gox, other exchanges, even dogetipbot (https://gizmodo.com/reddit-users-lose-real-money-after-meme-currency-bot-di-1795125165). At the first signs of trouble, they just seem to dig deeper and deeper holes until they go completely bust.

Rational models are deeply flawed for as long as they are not applied to perfect computers =)

P.S. Bitcoin is not a fad and is no longer in danger of becoming one, I agree. I know more and more people who buy it in a fad-like frenzy but people do that with gold too, and gold's not a fad.

You are 100% correct about Segwit2x being a greed driven product.  I would say you are wrong about Game Theory not playing a part.  Because of it being a Nash Equalibrium (all players do not have to change their strategy based on the other persons decision to recieve the reward.)  When one of the 2 chains becomes compromised then logically one will have to change their strategy to keep the equalibrium.  So the 2 chains logically must settle down at some time so their is no problem at all for supporters and non-supporters in way of economic gain but the point was raised interestingly that greed is the strategy behind segwit2x.  If over %50 percents of the mining power leaves the original chain then that chain should follow suit immediately and dump the original coins into seg2x.  I mean all said and done it is just a way to get instant cash. There is to much money into it now for bitcoin itself to go anywhere. Hard forks like BCCash are just alt coins that don't require the work of starting at the bottom

Segwit2x is an attempt to steal the "Bitcoin" brand, it's as simple as that. This sums it up and everyone can understand it without having a doctorate in game theory.

The rabbit goes indeed really deep with bitcoin's governance model. Those that really care about Bitcoin and want to keep it sovereign will never align themselves with corporate takeover attempts such as segwit2x, but you always depend on a certain hashrate to keep the legacy chain alive. I think this is the case once again, the so called 95% support for 2x is fake and will be proven fake soon.

As 2x chain dies and 2x tokes are dumped, the legacy Bitcoin supporters will end up with more BTC, which means more power to attack further forks. We already saw this BCash. People that dumped BCash now hold more BTC, which means more power against forks.

The problem would be if miners actually left the legacy chain and there would be no way out but a change of PoW, which is a hardfork within itself, so im not sure if we could call any chain legacy anymore.

Dude, relax... I specifically used 'bitcoin as a fad' as a hypothetical scenario to depict a game-theory limitation. I am not saying it will be a fad. In fact, if you read my posts, you will understand that I am a big believer in its potential to be the "king" and only survivor to any serious damage to the crypto-world.

John Nash, the father of game theory, said himself that "game theory may serve no actual purpose other than satisfying some intellectual exercise [paraphrased]." In other interviews later on in life, after he had really thought about his contribution, he equated it to "intellectual masturbation [his direct words]."

So, next time, before you go and attack others and make assumptions about what their beliefs are and say stuff like "kinda tells me you don't know nearly enough about it", go and remember that game theory has some serious limitations in understanding aspects of the world and even economics. It is very narrow and any serious investor will tell you that. Game theory gives you very limited advantage on the stock market, or more seriously established investment environments, where you are playing against big and professional players for the most part and which look at fundamentals vs. the perception of fundamentals - which game theory fails to address as it is always stuck in the 'perception of fundamentals'. As an example, are you rational? You took my hypothetical example so seriously that you assumed what my true beliefs were.

Bitcoin will win primarily because of its branding, what the technology has managed to create (by it being implemented more as a protocol than as a pure service) and by having come into existence at the right time in history. Don't let your game theory constructs pull you away from harder, real analysis of the world.

I started this topic about Game Theory and the correlation to the new crypto economic world.  It is my firm belief BitCoin will NOT be a fad for any time in the near future in fact you saying it is a possibility kinda tells me you don't know nearly enough about it. But lets entertain that idea:

The rational meantioned in Game Theory is not the definition you are giving it.

Rationality is have a preference order that is complete and transitive. That is the definition of rationality in a game-theoretical sense.

In economics and game theory, it is sometimes assumed that agents have perfect rationality: that is, they always act in a way that maximizes their utility, and are capable of arbitrarily complex deductions towards that end.  -

So anyway.  Im still learning a lot about game theory. I don't have perfect rationality but I would call the thought of calling BitCoin a fad happening in my lifetime irrational.

http://gametheory101.com/courses/game-theory-101/rationality/

Another interesting aspect to consider is the meaning of "rationality" itself. Often times, you don't know if a behavior is truly rational until enough time has passed to prove it as such. Many things can be rationalized, which is where the trap lies.

For instance, let's say bitcoin ends up being a total fad for some reason in the future, was any participation in bitcoin rational? What would count as rational participation in bitcoin and irrational?

People forget the age-old 'it has to stand the test of time' and I think that applies in a lot of these digital economy cases. Usually, in everyday life, it's easier to judge rationality, as we all have 'life experience' and can make a good, educated judgement as to the outcome people will have if they undertake certain decisions. But for bitcoin, the most we can do is draw parallels to the current fiat economies and gold and whatnot and compare, however it will not be exact as it is a new 'game' on its own.

Time will tell, but I would caution against using 'rational' to describe a lot of the players (even those who vote for the main chain may not be voting out of rational motives).

Interesting topic!

So when all is said and done BitCoin core will prevail as it set the initial rules of crypto as far as we know by introducing the blockchain.  As long as any crypto uses any blockchain it just means it has different rules. Core is the best word to call BitCoin and until there is a %51 percent consensus all hard forks are just alts fueled by greed until another fork brings the alt back into the original chain or the alt dies... Or at least will never be worth as much. The name BitCoin remains forever in the cryptocurrency market.

I believe all the abstracts you guys mentioned can be added to the economic calculations which is why game theory is so interesting when applied to the current digital economy.  For instance the coolest variable mentioned is knowing whether or not players are rational or not.  Superrational thinking cannot always go for group thinking and cannot cooperate every round as the individualized more greedy rational thinking will take advantage. Although the number of players is finite the exact number of individuals is unknown.  So both rationalities will battle. Completely irrational players lose quickly.  Superrational thinkers being inherently less greedy will try to use group effort to maintaining the highest odds of winning although having less returns then game-theoretic thinking.  This is why one cannot be proven to be superior in multiple round games with in unknown number of rounds due to the unknown variables of the other players. Example game Superrational thinking is player A and game-theoretic thinking player B

Round one- A chooses cooperate for highest probability and B chooses defect for highest returns. Although both have a degree of selfishness B is statistically more selfish since he is at greater risk the more rounds are played.  
A -1 B+2

Round 2-  A now knows that B logically has a greater percentage of choosing defect albeit an unknown amount.  B vise-versa.  If A chooses to still cooperate then he is at slightly greater risk now knowing B is no longer at 50% probability for defect 50.1% maybe?) B now is the opposite.  B won but has to factor in the endless loop of an endless defect/defect response.  This would be a draw.  In humans B to remain selfish has to change at sompoint.  So does A.  These factors increase exponentially every round.  In a game of chess both players of equal rating theoretically should make the same amount of mistakes or blunders.  The more you know about any game the more you don't have to worry about human thinking. That is why a superior player does not need to know all the emotional nuances of the other player. Only his own. He will win as long as every move is strictly logical.  The only human interference is an external force (for example someone coming over and knocking the board down or in BitCoin government and business interference).  So all said and done you hit the nail on the head with happiness factors. Chess is a non-cooperative game so it cannot use the same logic as cooperative games like these economic games that have team play. However, cooperative games can use logic applied in uncooperative games. If someone is going for highest monetary returns STRICTLY then quality of life theoretically goes down (as aging and death decrease the value of money overtime and overall life value is determined by other factors like family and friends and experiences)  Otherwise to keep the game of money going and guarantee at least some reward every round cooperating is best in my opinion.  The game of life which it seems you were hinting at when you mentioned happiness is the greatest problem we are all trying to solve. Imperfection and irrational thinking throw the math out the window with the economic games.  For instance poker.  One irrational thinker can ruin any chance a rational thinkers (be it Superrational or game-theoretic) to calculate returns and odds each round as he is not employing strategy but just pure luck of the unknown. He could go all in on an 7/2 off suit pre-flop for no reason but the flop is 777 giving him a %100 chance of winning the round.  The random numbers in computer games are the unknown variable.  In human economics then both types of thinkers benefit from the irrational players or self-sacrificing players.  One type being more greedy does not get too concerned about the other players welfare if they are winning each round.  In humans many only "team" up with friends and family and when they fare badly they switch to cooperate.  Some... Few...  team up with the human family and battle the defectors constantly as they take risks protecting the irrational.  That is a whole other topic but in BitCoin I was thinking forks of BitCoin at all are inherently greedy because  supporters (who think they should be the %51 or hate that miners have power) do not want to lose BitCoin core completely because of already held value and if they essentially double the coins their is no immediate risk it seems completely logical.  It does very negatively impact the irrational self-sacraficers which ones without care and concern for group cooperating will see as disposable. That goes for all businesses. The other Superrational thinkers are not affected as it cannot be proven which one is better as the number of rounds is unknown.  That is where human morals, ethics and spirituality plays a part as the true value of life is not determined by physical wealth. Those things cannot be computed but do not make a person irrational which is why I think Superrational thinking will prevail always in life's value but not in tangible physical economic value. That is why we have the saying "absolute power corrupts absolutely" to have this power one must remove certain moral and spiritual ideologies. But I digress...

EDIT: Poker analogy off. Odds are not 100% as anyone with a poker pair of 8s or higher getting a 4 of a kind on turn and river is possible.

You have to also include short-term and long-term effects as parameters. For instance, not everyone will defect if the other chain does not offer the security (or perceived security) that the current chain does. In fact, most players are likely to remain on the main chain and ensure its survival, unless some true contender came up that challenges the core perception of legacy bitcoin.

However, most players will economically be incentivized to vote for the fork (either directly or in their own heads/wishes), because it provides them with free coins that can be worth something or nothing (but there is a chance of them getting some extra reward). It's true that sometimes this extra reward can come at the price of stability on the main chain, but that would require most users to understand that and it is questionable how many users understand this (the current mainstream bitcoin community is often excited about forks as they see them as "dividends" or "free money", without understanding potential consequences attached to this).

If the consequences of a fork would be that the old chain dies out, even if miners and users are still signaling support for it and will use it, then that could tilt the game and people will feel, in a much more direct way, the consequences of promoting all these forks - it wouldn't be very popular in my opinion, as most people understand that you do not risk all this progress for something uncertain. But how you would go about implementing such a thing is difficult.

Game theory is nice to understand things and certain behavior, but it assumes too much that players are rational or always economically-motivated. There are plenty of people who are more "rebellious" and fighting for ideologies, which throws off a lot of the game-theory stuff. I question how much of the B2X split is currently just ideological and not economically-motivated. Someone who thinks "they have nothing to lose" will make decisions in a very unpredictable way, according to rational thinking.

Abstract frameworks like the prisoner dilemma can be difficult to quantify as terms such as reward and rationality can be prone towards failure.

Not to name names but for sake of discussion, Craig Wright might be considered a defector who received the best outcome due to his venture receiving $200 million in funding in exchange for Craig Wright attempting to deceive people into believing he is Satoshi Nakamoto. A typical analysis might conclude Craig Wright is an optimal example of a defector being rewarded for rational behavior. This falls under a materialistic analysis.

But there is a non materialistic school of thought which might suggest having $200 million can less desirable than abstract concepts such as loyalty, integrity and having a clear conscience. There are wealthy celebrities who may have led happier and more content lives if they had never achieved fame or wealth.

With frameworks like the prisoner dilemma we generalize due to an inability to cover rule based exceptions. Topics like reward and rationality can't be quantified as well as we might like them to. Whether reward can be measured solely within materialist bounds such as money or non materialist abstracts tends to be where the prisoner dilemma breaks down.

That said, I like your analysis & think it is a good analogy. I agree there are parallels between the prisoner dilemma and the current state of affairs in crypto that are very interesting.

Not sure I'm very clear on how superrational can be compared against rational. Nice read, nevertheless, here are my comments:

The main flaw of the game theory exists either way: you also have to take into account the realist, which is not necessarily rational. You see, every action, every fork, every chain can be rationalised and, therefore, super-rationalised. Your own post is an example of this.

The realist (I'm not one, by the way) already knows that he does not have to defect or remain, he can choose to stay on all sides, and then still choose later down the road once the benefits/rewards become clearer.

Also, depending on how you stand, you're either overestimating the ability of a centralised entity to change the rules of the game or overestimating the static nature of the existing rules. Because already within the rules are multiple interpretations.

Core being supporting and fork being defectors are also not hard truths. I think more people are realists who think as explained above, rather than rationalists. For me the extrapersonal proof is watching how hash power migrates between BTC and BCH.

I think the prisoners dilemma applies very well to bitcoin forks.  A game-theoretic rational thinker thinks:  If BitCoin players defect the reward is always better for the defector.  Because if:

 If you cooperate and the other defects, the defector gets the best outcome (In BitCoin that would be retaining all of the original chains coins and amassing the forked coins as well) So it is best to defect.

This though is not the best for you individually or for all players.   If all players defect nothing is gained(BitCoin Fails)

This cannot happen to BitCoin easily. Only a controlling force (centralized force) can change the rules so you cannot play the original game. For example government bans.

Superrational thinking assuming the other players thinking is Superrational knows that the worst possibility is when everyone defects. He would automatically cooperate to receive an assured reward although not the best possible.  So hard forks seem completely illogical.  That is ASSUMING Superrational thinking for all players. When the others thinking is unknown then that is where trust plays a part.  Once we convince ourselves this others will cooperate and we make a transaction with bitcoin and all parties receive a reward.  

This explains many things.  The defecting party (fork supporters) is going aftwe the biggest reward. The cooperating (bitcoin core) players believe enough players will stay on their team to be stronger and thus all cooperating players receive a reward.  

Quite amazing how deep the rabbit hole goes.  But I believe that BitCoin started a new economic game and is the controlling source of the game. The games is the blockchain and it is the largest and strongest.  Fear of weaknesses only should play a part in human error.  BitCoin is trusted because computers cannot deviate from the rules.  ALTs are fun mini-games within the game so all BitCoin forks just become alts.

Well that's my rant. The Superrational way to play is in a way less greedy. So although I'm late in the game its still fun I believe the segwit2x community is more selfish at a deep level.

Excuse formatting. On mobile