blitz/bullet rating is the only thing that matters in internet chess

Well , one thing is that the beasts never blunder like humans do , they dont hang pieces , overlook simple mates , etc . Humans , even strong humans do blunder sometimes , also time usage is one way to tell , if you are playing an engine it uses time very differently than a human does . 

It depends on how sneaky the cheater is. Some will make a bunch of engine moves in a row and analyzing the game with an engine will reveal them. The smart cheats only make enough engine moves to win, like in critical situations. Then it's almost impossible to tell, did they see that tactic or did they have help????? You'll never know.

Shygirl6985 wrote:

Don't mean to be irrelevant here but how can one accurately tell for sure whether or not someone is playing using a program? I a just a novice player and I can't really tell whether certain moves are "program" moves versus "human" moves.

Erik, that sort of smart cheat is not likely to play blitz/bullet because the time is so limited. 

Reb, I don't know whether the 5-minute pool on ICC is full of cheats. I rarely play in such long time controls! But I do know that even GMs are paranoid and will accuse non-GMs of cheating if they get beat. The real poison in chess is paranoia, which is what happened at the European Women's Championships last month.

Still, I rarely play increments online for the same reason that +2600 rated GM gave up on the 5-minute pool. 

We used a D wave Quantum computer at my university to play bullet chess against.

This is a pretty common experiment at many good universities across the world.

It can move more than 100 billion positions in a second, but anyone with about a 2000+ Elo in bullet can beat it 9 out of 10 times.

you can not really cheat at bullet, its like trying to number every atom in the universe and articulating position.

of course its likely not impossible, any more than light speed travel. but its so far beyond our reach in 2015 that it can't even be imagined by an educated theorist of mathematics and computer engineering

and even if you could, it would be so painfully obvious that it is a cheat, ..making the point moot

You're wrong. Bullet is not chess.

You're wrong. Try learning the rules before accusing people of violating them.

If someone is good enough to know when to use an engine and then disciplined enough to only rarely use it, then you are correct, it is almost impossible tell. How many people, that are willing to cheat, are that good to fit both of those?

Having a string of moves match an engine isn't that big of a red-flag; I've had it happen in OTB casual and tourney games. Some positons are easier to play. It is when they statstically play moves that correlate well with engines over a large number of moves, over multiple games that will really  point out a cheater (there are probably other ways too).

Computers actually do better than humans in fast time control, I'm not sure where you're getting your information from. Even if computers don't play absolutely perfect move, they don't fall for any mates or tactics and every move is a pre-move.

Here is a stockfish vs. stockfish game with one minute time control, I think it shows stockfish's high level of chess play gets preserved no matter how short the time controls are.

(Sorry, I don't know what all that extra text is.)

Stockfish even admits in post analysis that some of the moves were not great after it thinks about it for a second. But remember that most moves were made in under half a second. I think this shows that while computers suffer in tight time controls, they suffer much less than humans. Even GMs will start making blunders in bullet chess, and remember that if a GM even falls for a 5 - 8 move combination tactic, stockfish will see it fast enough to premove it (what I mean is that if stockfish answers in less than 1/10 of a second, it was so fast that it might as well have been a premove by a mere human).

Remember that this game is 0 blunders, 0 mistakes, 0 inaccuracies (at least as far as I know, I may be wrong). Many times, human bullet ratings are inflated because we face up against other humans who never bring these kinds of numbers to the table. It is not easy for a human to reliably play such a good game, so reliably, on such short time controls.

Blitz benefit the stronger player, his or her chess knowledge, experience and intiution gives him or her the advantage in blitz. The problem with correspondence chess the tools available to a user, opening reference, endgame books, games date base and moving pieces around, all sudden this player play like a grandmaster in the opening and early middlegame, some player even think they are at a level they are not; I give you an example, this a really joke correspondence grandmaster with a online rating of 2500 - 2600, sound strong but if you actually look at the quality of their game it is at best 2100 elo to 2200 elo level. This not so with blitz, my friend who is a very strong master has otb rating 2400 uscf and his blitz and bullet rating is 2400 level, he can't never go to 2500 blitz and stay there, as soon he get a 2500 blitz rating he drops. What I like about blitz is you know you are playing your opponent with his current knowledge and not some opening manuel, endgame book or games data base, and that is real chess.

What's your opinion on standard time control, such as 30|0 or 45|45? Do you think it accurately represents chess skill?

no, it isn't (title)

indeed. Correspondence Chess allows you to think, play the best move you can. It's truer than championships in some ways

Yeah, it's not like you use the same board, the same pieces and the same rules.

Statistically, the OP is right. Blitz ratings here has a much higher correlation to OTB ratings than correspondence ratings do. 

His hypothesis that this is due to cheating has been stated by many others, but there's not enough evidence to support such a claim. 

What other reasons are there for the disparity in ratings?

xmon720 wrote: What's your opinion on standard time control, such as 30|0 or 45|45? Do you think it accurately represents chess skill? 

Yes, I think it is more accurate than correspondence, at least you playing your opponent and not the sources that could be available in correspondence chess. I prefer to play the player with current knowledge and not what sources they can use against me ( that is not real chess game).

Let's just think simply.
Most people either play one or the other.
A player that plays hours of blitz while juggling with dozens of correspondence is pretty rare.
Obviously, thus, there'll be a difference between the ratings. 

I have decent correspondencerating, and blitz 500 below. There are reasons for it. In correspondence i have the explorersupport, which secures that I dont loose in the start, but the main factor is time. Give me 90 minutes +30 sec pr move and I can both avoid a lot of blunders and do some nice thinking. Give me ten minutes, and I will not have enough time to think well and examine the board good enough.

Its different with the sharp young players. They are much faster, and can also play ok quite fast.

I have relatively little experience, a very low degree of automation, so I really needs a lot of time to figure out things. New problems pops up everywhere, which requires a lot of thinking.

IronedSandwich wrote: indeed. Correspondence Chess allows you to think, play the best move you can. It's truer than championships in some ways 

I agree, that is the beauty of correspondence chess, I played many beautiful and complicated endgames and if they were blitz or bullet game I would ruin some these endgames. So, I agree if a players are honest correspondence can be a big benefit to a player. But is it real chess when your opponent is playing like a grandmaster in the opening and his online rating is a mere 1500 and sudden he play like grandmaster in the endgame because of endgame manual; I want to know I playing my opponent true strength not what chess sources is available to him or her.

I'm about the same age as Djonni. I've won my city's blitz and standard championships several times. Blitz is far closer to standard chess than correspondence is. 

 I give you another example in correspondence let say 200 to 300 rating points more to your otb rating, so if player is 1500 uscf his online rating can be 1600, 1700 and 1800. I know a player from my chess club who is a 1500 uscf and his highest rating was 1650 uscf but his correspondence rating is 2450; there is something wrong here. In this same chess site I brought four experts, one the strongest is rated 2160 uscf and his online rating is a high 2300 and now chess.com he was at 2400 and drop to a mid 2300. How can a very strong expert be outrated by a mere 1500 uscf in correspondence, think about; the conclusion is this 1500 uscf is cheating. The beginning of this year this 1500 uscf play in our chess club and in six round he only score two points and lost to players lower than his otb rating.