Axlerod decided to run a competition and he advertised for experts in game theories to submit strategies. The strategies (14 submitted) were pre-programmed rules for action written in computer language, and then fed into a computer to run the strategies against one another, and once against itself. Axlerod threw in a strategy called Random, which simply played COOPERATE or DEFECT indiscriminately. It served as a non-strategy, so if a strategy could not do better than Random, it was pretty bad. After each strategy had gone through 200 moves of the game, (there were atotal of 225 separate games) the winnings were totaled up and the winner declared. The winner was a strategy called Tit for Tat, submitted by Prof. Anatol Rappaport, a well-known psychologist and game theorist from Toronto. Tit for Tat started out playing COOPERATE on the first move and then replicated the other players actions thereafter. No matter how many times the game is played and which strategies submitted, this is usually the winner. That being said, there are some interesting strategies worth noting. Naïve Prober—like Tit for Tat but it throws in a random DEFECT to cash in occasionally. When confronted with TFT though, Naïve Probers defection invariably starts a string of retaliations. Remorseful Prober—like Naïve Prober except with memory, it remorsefully allows “one free hit” if it has defected previously. Its main problem is that it doesn’t do as well against TFT as TFT does against itself. Tits for Two Tats—lets the opponent two defections before retaliating. This was not played, but Axlerod worked it out that it would have won the initial tournament as it avoided runs of retaliation. The winning characteristics of the first tournament was niceness and forgiveness, and this of course came as a surprise, so Axlerod announced a second tourney, which he received 62 entries, and he also threw in Random again. The game was not fixed at 200 moves, and the programmers had the results from the first tourney. Two camps emerged, one figured niceness and forgiveness were obviously the winning qualities, and the other figured there would be lots of softies, and set out to exploit them mercilessly. Once again TFT won, and in fact the nice strategies did generally better overall winning the top 15 spots, and the bot-tom 15 save one (probably that wuss Tit for Two Tats) were all nasty. Axlerod then took the same 63 strategies and changed the rewards into offspring. The points became replication of the parent, added or subtract-ed to the game. Some strategies became more numerous, and some became extinct, as the proportions changed, so did the ‘climate’ of the game. After about 1000 generations, there were no change in proportions, and stability was reached. The nasty strategies did not survive much past generation 200 and good ol’ TFT ended up winning 5 out of 6 of these tourneys. The climate of the game is interesting because it is certainly feasible that the strategy of ALWAYS DEFECT could easily gain a foothold for an extended period of time, establishing a stability of sorts exploiting any strategies that dare cooperate, until there are no survivors left save the DEFECTORS. Of course, the game quickly ends soon after that because there is nothing left but defectors. Conversely, if TFT starts to dominate the game, eventually a stable infinite environment of cooperation emerges. Whichever stable point comes to dominate the population first will tend to stay dominant. In the mathematical computer simulation of the Prisoner’s Dilemma, this clustering of strategies is left to chance. But how does this happen in ‘real life’? There have to be enough people cooperating on some level in the same area in order to reach a critical mass of reproduction to survive in the midst of defection. Once a critical mass is accomplished, there is a very good chance TFT will out live the always defect strategy. The Always Defect strategy works horribly when clustered together, and eventually gives way to the TFT cluster. Axlerod noted that TFT is ‘nice’ and for-giving, meaning that it has a short memory for past misdeeds. Furthermore, Axlerod noted that TFT is not ‘envious’. To be envious in this case means to strive for more money than the other player, rather than an absolutely large sum from the banker. You are quite happy if the other player wins just as much money as you do. TFT never actually ‘wins’ the game, it just strives to draw and win money from the banker. The word opponent is almost inappropriate with strategies such as TFT. Strangely enough, when psychologists set up games of Iterated Prisoner’s Dilemma between real humans, nearly all players succumb to envy and do relatively poorly in terms of money. Perhaps without even thinking about it, many people would rather go down with the other player than cooperating and doing the banker in. Axlerod’s work shows what a mistake this is. Of course, it is only a mistake in certain kinds of games, the nonzero sum game. Game theorists call a game where one player is the winner and one is the loser a ‘zero sum game’. Prisoner’s Dilemma is a nonzero sum game as it is possible for the two sides to prosper at the expense of the banker. This important fact should not go overlooked. The players will have a perception as to how long the game will last. As it draws to a close, invariably the temptation to DEFECT will be greater, so as to get that last shot at the big money. Dawkins goes on to relay some stories from the First World War that further demonstrate the “Tit for Tat” social strategy (again I’m paraphrasing his writing)... Historian and sociologist Tony Ashworth has written about the non-aggression pacts of German and British troops on the front lines of World War One. During the first two years after 1914 there were many unofficial and unspoken non-aggression pacts that flourished up and down the trenches. Mutual cooperation was bad from the viewpoint of the generals. But for an individual soldier, it was good. Sure, he wanted to win the war, but he also wanted to stay alive. His individual life had no real affect on the outcome of the war, especially the ridiculous trench warfare of World War One. On the other hand, the soldiers on the other side of the battlefield had a lot to do with his fate and future. Every soldier knew he would be in the trenches a long time and so it serves as a great example of Iterated Prisoner’s Dilemma. It is important to note, that for TFT to be a workable strategy, the threat of retaliation must be present. Much like the cowboy shooting out the candle flame in a Spaghetti Western film, the other side must be well aware that they will be dealt with harshly should they dare defect. Of course, TFT is also a forgiving strategy, and this is a very important aspect of its success for it avoids very long runs of recrimination. From the memoir of a British soldier... “I was having tea with A Company when we heard a lot shouting and went to investigate. We found our men and the Germans standing on their respective parapets. Suddenly a salvo arrived but did no damage. Naturally both sides got down and our men started swearing at the Germans, when all at once a brave German got on to his parapet and shouted out “We are very sorry about that; we hope no one was hurt. It is not our fault, it is that damned Prussian artillery.’ In addition, Axlerod emphasizes predictability and ritual in maintaining a stable pattern of mutual trust. A good example is the ‘evening gun’ fired buy the British Artillery with clock-work regularity at a certain part of the line. In the words of one German soldier: ‘At seven it came-so regularly that you could set your watch by it... It always had the same objective, its range was accurate, it never varied laterally or went beyond or fell short of the mark... There were even some inquisitive fellows who crawled out... a little before seven, in order to see it burst.’ The German artillery did the same thing... ‘So regular were they (the Germans) in their choice of targets, times of shooting, and number of rounds fired, that...Col. Jones... knew to a minute where the next shell would fall. His calculations were very accurate and he was able to take what seemed to uninitiated Staff Officers big risks, knowing that the shelling would stop before he reached the place being shelled.’ Axlerod remarks that such ‘rituals of perfunctory and routine firing sent a double message. To the high command they conveyed aggression, but to the “enemy” they conveyed peace.’ I think that the “moral conclusions” that are gleaned from a strategy like Tit-For-Tat are ideally suited to any social cohesion, so much so that it seems embarrassingly obvious to even convey the information. Yet, if we look at our lives and the people around us, it so obviously needs said. Of course, if such an article as this gains it’s adherents, and people really think in such terms as “Tit for Tat”, I can assure you that a not-so-new betrayal technique to be brought out by the “Defectors”. It will be that they will falsely accuse others of “first betrayal”, thereby allowing them to “reciprocate”. This brutally simple technique usually takes the form of unsubstantiated rumors, and I would warn everyone to be wary and on the guard for such tactics. |