You aren't looking at the games at all, you just have an engine look at the games.
On another hand I can say for sure that Steve never claimed that he invented an accurate method. He just applied a procedure devised by others and his method only works to detect blatant cheaters, provided that there is a large sample of games.
I may do some of the work, but when I asked for help doing it, the members of the cheating forum did not support the idea.
Show the initiative, then you get the credibility.
This is true.
I suggested for Fezzik to make a showing of work and then offer a proposal based on that work.
I never got a response.
stevecollyer and Polar_Bear have done extensive benchmarking for the top-3 method of engine use detection as well as the analysis of several hundred games each of suspect batches.
Consider them good sources on the subject of T3.
If this society here wants to get rid of progplayers, all we need is the Trophy of Antidoping! When a player gets 3 trophys, all his/her games would been analyzed. That´s simple is that. Unfortunately this procedure would raise costs and membership fee.
First of all, this society needs to get rid of anonymous newcomers, at least for rated tournaments, team matches and team vote games. Then, after eliminating such garbage from the heart of this community and effectively keeping them out until they prove to be trustworthy chessplayers, we can proceed further.
I am anonymous newcomer and I play rated tournaments, TM and VC - I have my reasons to be anonymous on the internet site (I don't want to be googled by any1 who wants to check where I live, facebook ,my job etc.). Why r u calling me garbage???
Do you consider yourself trustworthy?
Cheaters have reasons not to get googled (to avoid shame), but i can't understand the reasons of honest players. It is quite normal that chessplayers play under real names.
Well, if Chess.com wants to move further and become a serious chess site as computer-free counterpart of ICCF-web, then real names must become mandatory.
That sounds like something a witch would say!
and what are we poor mortals supposed to do?
God give me strength!
There are major problems with the top-three method of trying to compare computer moves to human moves.
One basic problem is that there is no documented level of confidence built into the analysis...
waffle, waffle, waffle.
In terms of delivering the goods, aren't you supposed to have started testing your very first benchmark game using just this technique (allbeit with a fixed depth of 20 ply)?
I sugggested using an average error/blundercheck method, but no, you went straight for top 3 match up! 
Months later & you still don't seem to have produced anything.
You don't contribute anything, Dan. You just snipe from the sidelines & feign interest in improving cheat detection methodology. I do wonder what your motives are with regards to your interest in cheat detection & your looming presence in the cheating forum discussions, especially with your 47/17/0 W/D/L turn-based record.
Having analysed some of your games using this method, I think I know the root cause of your cynicism toward it!
I also agree that the penalties for being caught cheating should be harsher. As a matter of fact, I would support a 5 year suspension for being caught cheating the first time and a lifetime ban for a second time. There is one famous club in New York (which I will not name so my post don't get deleted) where I have decided not to play again at because the cheating during tournaments is ridiculous. It doesn't happen through cell phones and is mostly when kids are playing grown ups. I have seen kids giving signals to other kids and worst of all, parents giving signals to their kids. One time I was observing a game in another table when I caught a parent signaling his kid who was playing. When I confronted him he responded by defending his actions because "its not fair that a 9 year old has to play against a 20 something year old". What a great example is that parent teaching his kid!!!
To play Chess 3 matters is needed: 1. Itch 2. Time 3. Opponent #
Board, pieces, pens, papers, internet and computers are only secondary tools. In my opinion there will be always trapping, bluffing, feeling and cheating in games, we humans do play. And that´s what I think.
If this society here wants to get rid of progplayers, all we need is the Trophy of Antidoping! When a player gets 3 trophys, all his/her games would been analyzed. That´s simple is that. Unfortunately this procedure would raise costs and membership fee.
First of all, this society needs to get rid of anonymous newcomers, at least for rated tournaments, team matches and team vote games. Then, after eliminating such garbage from the heart of this community and effectively keeping them out until they prove to be trustworthy chessplayers, we can proceed further.
Sorry, BIG FAT NO!
If a person chooses to join the site using a fake name or cute avatar, that's not for you to decide. Don't forget chess.com's pages are publically accessible and are indexed by google including the forums. Identity thieves, stalkers and other unsavory folks prowl the internet 24/7 looking for victims. Perhaps some of those anonymous players have already been victims of identity theft or they are being prudent about protecting their identity and bank accounts.
Attempting to verify millions of names would be futile anyway. How do you know people are really who they say they are? Haven't we already had players on this site pretending to be a particular IM or GM, when in fact they were not the real person?
On top of which, we've had people on this site who were accused of cheating who appeared to be using their real names. So even if chess.com were to implement your suggestion it wouldn't accomplish anything.
Do you realize that ICCF web, Chessgames.com and FIDE member database work with real names of players? Perfectly searchable online.
Anonymity has one serious disadvantage: it helps cheaters to avoid punishment and sneak back easily if banned. Please, don't argue there were cases of unjustly banned people: there weren't.
There is nothing wrong with creating anonymous account and playing several free blitz games for fun, however things change with team membership and tournament play. I am quite surprised team admins and tournament directors haven't learned yet and still allow untrustworthy nonames to participate.
If this society here wants to get rid of progplayers, all we need is the Trophy of Antidoping! When a player gets 3 trophys, all his/her games would been analyzed. That´s simple is that. Unfortunately this procedure would raise costs and membership fee.
First of all, this society needs to get rid of anonymous newcomers, at least for rated tournaments, team matches and team vote games. Then, after eliminating such garbage from the heart of this community and effectively keeping them out until they prove to be trustworthy chessplayers, we can proceed further.
Sorry, BIG FAT NO!
If a person chooses to join the site using a fake name or cute avatar, that's not for you to decide. Don't forget chess.com's pages are publically accessible and are indexed by google including the forums. Identity thieves, stalkers and other unsavory folks prowl the internet 24/7 looking for victims. Perhaps some of those anonymous players have already been victims of identity theft or they are being prudent about protecting their identity and bank accounts.
Attempting to verify millions of names would be futile anyway. How do you know people are really who they say they are? Haven't we already had players on this site pretending to be a particular IM or GM, when in fact they were not the real person?
On top of which, we've had people on this site who were accused of cheating who appeared to be using their real names. So even if chess.com were to implement your suggestion it wouldn't accomplish anything.
Do you realize that ICCF web, Chessgames.com and FIDE member database work with real names of players? Perfectly searchable online.
Anonymity has one serious disadvantage: it helps cheaters to avoid punishment and sneak back easily if banned. Please, don't argue there were cases of unjustly banned people: there weren't.
There is nothing wrong with creating anonymous account and playing several free blitz games for fun, however things change with team membership and tournament play. I am quite surprised team admins and tournament directors haven't learned yet and still allow untrustworthy nonames to participate.
The tournaments here are for fun only, there is no physical prize and the same goes for team matches. So what's the big difference?
Chess.com says it handles cheating pretty well, and that it's less prevalent then forum hysteria suggests (and I'd definitely agree with the latter).
Identifying on-line is scary for most, and chess.com would most certainly lose membership if it required so. Children and youth, for example, are diligently instructed never to provide their real name on-line, and many reasonable people abide by this "safety tip" even in adulthood.
Chess.com says it handles cheating pretty well, and that it's less prevalent then forum hysteria suggests (and I'd definitely agree with the latter).
Identifying on-line is scary for most, and chess.com would most certainly lose membership if it required so. Children and youth, for example, are diligently instructed never to provide their real name on-line, and many reasonable people abide by this "safety tip" even in adulthood.
1) Chess.com doesn't handle cheating "very well", it handles cheating pretty poor, in fact perhaps nearly all top online turn-based "players" are cheaters. The situation would be much better, if anonymous registrations were disabled.
2) I would like to finally change common public misconception, that ICCF chess with computers tolerated is "serious", while online chess with rules against them is "casual". Just opposite is true. One day I would like to see respected XV. Correspondence Chess World Champion, next in line after Tonu Oim, not some mere computer monkeys.
3) Better cheating detection, pre-emptive measures against potential cheaters and avoiding anonymity are necessary parts, of course. It will be quite unpleasant, but it is common in athletics today, isn't it? With very strict measures, prize tournaments can follow even in correspondence chess.
4) Current approach of staff, team admins and tournament directors i don't find helping. They haven't learned yet that honest chess and anonymity can't mix together.
They haven't learned yet that honest chess and anonymity can't mix together.
I replied once to you: you don't want your boss to be able to google your name and see that you played some blitz games here during the work.
Honest chess and anonimity can mix very well together. Just go on another site (I said which so many times that the staff would really ban me for advertising). Admitedly, there are cheaters there, I encountered some, but they are indeed few.
Honest chess and anonimity can mix very well together. Just go on another site (I said which so many times that the staff would really ban me for advertising). Admitedly, there are cheaters there, I encountered some, but they are indeed few.
I replied too. So again. If you play blitz for fun (and troll in forums), it is OK to stay anonymous. With tournament play, it is not the case.
There is another possible solution. Make real name visible only for logged-in users or for other tournament participants.
I'm with BorgQueen, penalties should be extreme. It's not about forgiveness or protected someone's rating who is a proven cheater. It's about the integrity of the game. 98% of players would never cheat even though they have a burning desire to win.
First offense, reset their rating and strip them of any titles.
Second offense, the same and they can never hold an official rating again.
Honest chess and anonimity can mix very well together. Just go on another site (I said which so many times that the staff would really ban me for advertising). Admitedly, there are cheaters there, I encountered some, but they are indeed few.
I replied too. So again. If you play blitz for fun (and troll in forums), it is OK to stay anonymous. With tournament play, it is not the case.
There is another possible solution. Make real name visible only for logged-in users or for other tournament participants.
If I had children, I wouldn't let them play on an internet site which requires their real name. No way!
And there are othercounter-arguments. I told you: playing under the real name did not stop Dembo or Docx from using computer assistance. Move by move, not from time to time.
However Dembo and Docx paid much higher price for their cheating than Ouachita or Yorkman. Their names are pretty discredited among real chessplayers. And it is the right thing to do.
Children and internet - it is quite complex problem. I wouldn't allow my children to play any tournament, online AND otb, without proper supervision.
I also agree that the penalties for being caught cheating should be harsher. As a matter of fact, I would support a 5 year suspension for being caught cheating the first time and a lifetime ban for a second time. There is one famous club in New York (which I will not name so my post don't get deleted) where I have decided not to play again at because the cheating during tournaments is ridiculous. It doesn't happen through cell phones and is mostly when kids are playing grown ups. I have seen kids giving signals to other kids and worst of all, parents giving signals to their kids. One time I was observing a game in another table when I caught a parent signaling his kid who was playing. When I confronted him he responded by defending his actions because "its not fair that a 9 year old has to play against a 20 something year old". What a great example is that parent teaching his kid!!!
I've never played in a club or any kind of tourney, but if I did, and saw some a-hole giving signals to his kid, I would probably go over there, and punch his lights out.
There are major problems with the top-three method of trying to compare computer moves to human moves.
Yes, there are. Do you have a suggestion, a contribution? I read the rest of your post, it's just trolling.