Monday, January 31, 2011

Poll on Engine Users

The votes on what you would do if you knew your opponent was using an engine were pretty evenly divided. Six people would also use an engine just to level the playing field. Six would report them and presumably continue playing. One person voted that they would simply resign in disgust.



I understand resigning. Back in the early 1980’s when Chessmaster was rated about 2000-2100, I was playing in a couple of USCF postal tournaments and became suspicious that some of my opponent’s were using the program. I ran all the games through Chessmaster and discovered that out of 12 opponents at least 8 were using because their moves were identical to those suggested by Chessmaster. After over 20 years of postal chess I was disappointed that that was what it had come down to so I simply resigned all the games and lost over 300 rating points. I resumed CC back in 2004 and found things had not improved any.


For the most part my opponents are over 2000. If I play a lower rated player, I play on my own and actually I prefer players in the 1600-2000 range because they usually are not using engines and are good enough to make the game interesting. Over that and it’s a tough call because somewhere between 2000-2200, maybe they are, maybe they aren’t. If I suspect an opponent is using an engine, I’ll run the game through Fritz and see what moves it’s recommending and if I find a significant number of matchups I will also use one. I didn't always feel that way. There was a time I'd try to see if I could beat the engine, but nowadays that's impossible and I got tired of beating my head against the wall. What usually makes me suspicious is when I start seeing moves that don’t make any sense to me. I figure I should have some idea of what my opponent is trying to do if we are about the same rating but if his moves make no sense, then something’s up.


Recently though it’s becoming harder to detect engine use. Or, rather I should say, which one is being used.  That's because there are so many good ones available and I’ve noticed Fritz rarely chooses the same move as the stronger Houdini. Add to that the fact that sometimes you have to let the engine run a long time get a decent evaluation. However, I’ve noticed most engine users will play whatever Fritz suggests within the first 30 seconds or so. That’s probably because many of them are playing lots of games and don’t have the time to let an engine run overnight. For opponent’s over 2400 there isn’t much to think about…they’re using…trust me. One point I should make here is that there ARE a lot of good players out there. When I started playing chess “good” players were few and far between. It’s a function of numbers; with millions and millions of players a LOT of them are going to be good. That wasn’t the case back when the USCF had 5000 or 6000 members and there were 100 GM’s in the world and you knew all their names!


The problem with reporting engine users is that many sites simply do not have the staff to check hundreds or thousands of players and games. Most sites will require substantial proof and to get the ball rolling, you will probably have to provide some evidence yourself and that means analyzing 15 or 20 of your opponent’s games on your own for submission to the proper authorities. Another reason most sites are loath to kick off engine users is because if they examined the games of their top players most of them would be gone.


One site I’m aware of recently kicked off some people but I think that was probably a token. Another has what they claim is a foolproof method of catching them but I’ve seen 1200’s banned for engine use so I’m a little skeptical. Another major site had a couple of their top players, in a moment of weakness, confess that they used engines, but they promised not to do it anymore so no action was taken.


Final analysis…engine use is here to stay and not much can be done about it. But unless you are playing consistently at around the 2200 level they should not be a problem unless you meet someone on their way up through the ranks. Even then I wouldn’t worry about them because at some point they are going to run into a whole gaggle of other engine users and if they are not really good players who are able to break through the ranks of other engine users they’ll get stuck and go no higher.

4 comments:

  1. I guess you're talking about correspondence chess. That's why I don't like this kind of chess. What is a sense of using computer? Isn't it just wasting time (mine and my opponent's)?

    I think in next years correspondence chess won't be popular at all. Even if we have emails. It's just boring to make moves suggested by computer. What is a role of chess player? Just sending emails?

    ReplyDelete
  2. CC is has become different from what it used to be…a substitute for OTB chess. Nowadays it is a whole different game (unless one is playing at lower levels) and it is not to everybody’s taste.

    Correspondence chess (at least at the master level) has become a challenge of a different kind. Players must be able to correctly evaluate positions that engines don’t play well. I’ve found their evaluations of positions with unbalanced material is often in disagreement with that of GM’s and STRONG CC players. Also to be successful in CC you have to be proficient in endings. Look at the crosstable even in high level CC events and you see some players who win most of their games and some who lose most of them. So clearly there is more to it than just letting the engine select the move. You are correct…for many players it is a waste of time, but for some it is a challenge to try and guide the game into situations in which the engine evaluation is wrong.

    TARTakower + BogolJUBOW = Tartajubow

    ReplyDelete
  3. I'm not experienced in correspondence chess, but can you agree that today is a fewer CC players than for example 25 years ago?

    If so, please tell me why?

    Tartakower+Bogoljubow it's a nice combination. But why not for example Fischer+Kasparov? Why did you take those both first players?

    I'm proud that you've chosen Tartakower, because he was from Poland as I am :).

    Btw, thanks for the note about my blog.

    ReplyDelete
  4. For many cc players engines have ruined the game but for many more it remains a good way to play. It used to be played by postcards and I have played people from around the world that way. It was painfully slow I might add. Still even today some players prefer it. I have played a number of people in prison who don’t have access to the internet so cc is the only way they can play. Postcard users are a dying breed though. CC by server play is pretty popular. For the vast majority of players who are rated below master, engine users are not a real threat so for them it remains a good way to play especially if they do not live near a thriving chess population.
    When I first started playing chess the US had 5-6000 registered players and I played cc games against a few who were highly rated otb masters. That was because there were so few tournaments and they had little opportunity to play otherwise. The USCF cc organization has thousands of players but the official US cc organization, the CCLA, has held pretty steady over the years at just a few hundred members.
    All in all, I think cc remains at about the same level as it did 25 years ago with the main difference being postcards are no longer used because of the internet. The actual number of players is probably higher simply because there are more people playing chess than ever before. Anyway, we cc players were always in the minority and were never consider REAL chess players by the otb crowd.
    Tartakower is one of my favorites …not only was he good, he was imaginative and played some great games. He was also a colorful character. I never thought much of Bogoljubow, but the names seemed to go together well.

    ReplyDelete