Thursday, January 17, 2013

Detecting cheaters

What happens when cheating is nearly impossible to detect, and has a huge payoff? The popular method seems to be to accuse people years later based on hearsay and gossip, and then punish them as severely as possible as a deterrent to others. That is what is happening to Barry Bonds, Roger Clemens, Lance Armstrong, and others.

It seems much better to me to have objective standards for cheating, and to apply them in a timely manner or not at all. If that does not seem fair, then change the rules until you find rules that people accept as fair.

A recent chess scandal has occurred when a 2200-level Bulgarian played a tourneyment at a 2800 level. People think that he must have cheated, but no one can figure out how. Computer chess programs are now significantly better than humans, so maybe he secretly used a computer somehow. The best chess computer in the world is Rybka, but it has been banned from computer competition because of some obscure cheating allegation.

Another form of cheating is insider trading:
Juries in Federal District Court in Manhattan have convicted all 11 insider-trading defendants who have taken their cases to trial since 2009, the year that prosecutors began bringing charges arising out its multiyear investigation into criminal activity at hedge funds.

Of the 75 people charged with insider trading crimes — a collection of traders, corporate executives, consultants and lawyers — the United States attorney’s office in Manhattan has secured 71 guilty pleas or convictions.
Some of this insider info does not really need to be secret. Maybe if the info is made public in a more open way, the opportunities for insider trading will be lessened.

If insider trading is easy and profitable, then people will do it.

Here is a game theory model of cheating:
The two-player Iterated Prisoner’s Dilemma game is a model for both sentient and evolutionary behaviors, especially including the emergence of cooperation. It is generally assumed that there exists no simple ultimatum strategy whereby one player can enforce a unilateral claim to an unfair share of rewards. Here, we show that such strategies unexpectedly do exist. ... Only a player with a theory of mind about his opponent can do better, in which case Iterated Prisoner’s Dilemma is an Ultimatum Game.
This shows that a mindreading manipulator can outdo a evolutionary best strategy.

Update: Wired magazine writes:
So here’s the thing you need to know: The USADA takedown of Armstrong matters, and it could effect everybody. Because it will enhance the power and reach of a private, non-profit business that has managed to harness the power of the federal government in what’s quickly becoming a brand new war on drugs … with all the same pitfalls brought to you by the first war on drugs.

The USADA is a private outfit. Yet it gets taxpayer money. ...

Nobody cared much about that treaty. And few care much now, really, because it was understood that anti-doping was about testing athletes, and mostly elite ones.

But the Armstrong case isn’t based on testing at all.

No comments: