Wayne_LV, on Dec 23 2006, 05:47 PM, said:
I get a real insight into bridge player's mentality everytime I ever post to one of these threads.
I mentioned a program to flag potential cheaters and some of you go off into how impossible such a program would be to program. Not true... all would be needed would be to periodically check the average scores for players and look for only 3 things. 1. Is the average per board > 1 IMP or MP averages > 58% (pick a number) AND 2. does the player play with the same partner over 80% (again pick a number) of the boards. If the answer to all 3 of those is yes and 3. the player has played over 1000 boards in the past month, I guarantee you that player is cheating in some fashion.
Well, lets test your hypothesis with my name:
1. IMP average is +1.10, and usually it's positive.
2. I've played a lot with "downagain", who is sometimes a f2f partner but mostly online (because we live pretty far away from each other). I'm not going to count the exact percentage
3. 174 boards in the last month. Well guess what, we both had lots of work to do, but I'm sure we could reach 1000 boards if we had enough time.
So suppose with a better time schedule, I would play 1000 boards with "downagain", would probably score +1.10 again, and I would be considered a cheater. I know for sure that I don't cheat, I guess most bridgebase forum people and online friends know (or are pretty sure) that I don't cheat, so these rules don't work. Rules for accusing people should work ALL the time, there's no room for mistake imo.
Ok, so what if we change the rules a bit?
1. IMP average +1.50? +2.0? More? Suppose we play more against GIB's. We average about 30 imps on 15 boards online against GIB's. Perhaps these robots just make assumptions on our bidding, while we give other meanings to the auction. So what, we still win too much don't we?
2. Lets say I'm practicing with a f2f partner for a big tournament and we play exclusively with each other. Whatever percentage you mention, it will not be enough...
3. take 10000 boards? Well, if you can find someone with the time to play 333 boards per day, good luck! 1000 boards already is a lot for most people, so how many cheaters will you catch anyway? All cheaters who play 999 boards a month will not be caught. So should you take a lower limit? 100? Damn then I should be (according to the rules) a cheater.
We can take even more rules into account, but there will always be a percentage of people who will be falsely accused, and you'll limit your search for potential cheaters to a very small percentage of the entire community. You can't make this process automatic because it would fail too many times, and it doesn't even work for 99% of the online players...
"It may be rude to leave to go to the bathroom, but it's downright stupid to sit there and piss yourself" - blackshoe