Because they have longer to think.

I never made a claim that a gap characterizes engine use. In fact, I correlated (loosely) the gap to a difference in the number of games played. Further, I think the ratings don't reflect anything because basing it on a pseudorandom pool of other players of arbitrary selection whose ratings are also based on a pseudorandom pool of players of arbitrary selection (ad infinitum) is mashed potato math. I could push my rating down if I only play players much higher than me, a little bit at a time, after I play the set number of games where the +/- rating points to be gained are set (unlike in the first few games, where the +/- is much wider). Etc., etc. So, by F all, I mean that the ratings have much less accuracy than what they are being used for - to screen for players of a certain skill level in the first place when looking for games. Hence, my suggestion of letter categories. Finally, I don't use an engine, but other than looking at how many games and tactics I lost and failed at, you'd have to just take my word for it, wouldn't you?

I think that the people who defend online ratings the most vociferously are just beholden to them. Frankly, I could care less. I set my search range for games for 50 below to 150 higher than what my rating has been floating at recently, because I would like to play people who are slightly to fair dimsum better than me. But even if I screen for even higher rated players, I get schlepper wankers who don't deserve their rating and are wasting my time.

And frankly, based on how people are ranked or rated in other online games, the system would do better if you started out with no rating at all for having played less than some number of games against rated players from each successive class. But whatever.

This topic was create with the overt intention of discussing cheating.  As everyone knows - at least should know, discussion of cheating isn't permitted in the main forums. I'm moving this topic to the Off Topic Forum with a warning to the OP (a second warning - the first being language.)

I never made a claim that a gap characterizes engine use.

You're right on this; I mixed you up with Dodger111.  Sorry.

No problem, accepted.

Yes, I agree with the above and everything else you have written in this thread.  But consider the likelyhood that many players who try out correspondence chess here (the too-vaguely named "online chess"), continue to play as if they were playing casual fast-chess.  They use only a few minutes per move (out of the multiple days available!), lacking the time or patience to look deeply.  And they probably mostly ignore chess books, databases, and videos.  ("If I'm going to work that hard, I want to get paid!")

These less-serious correspondence players are essentially the equivalent of free rating points added to the ratings of the more serious correspondence players, as chrka already pointed out.  This bottom-feeding of course occurs in fast-chess, but probably to a much smaller degree since the average difference between (the thinking-time spent by) casual and serious players is much smaller for fast-chess.

Cheating with engines probably occurs at all levels.  The assumption that such cheating is hard to arrange for fast time controls is questionable.

IM Pfren's ratings show the same divergence between his Blitz and CC ratings.

It's perfectly NORMAL for CC ratings to be higher than Blitz ratings.

End of Story.  Get off your hobby horse, PLEASE.

Lastly, why aren't the number of moves per game used to weight the +/-? That data is clearly available but sits disregarded. If I beat someone better than me in <20 moves, that should count for more. Ditto for material point differential. I hate that a loss in <20 moves counts the same as a loss on time in >50 but with a 12+ point advantage. In the first, I got spanked, in the second, I was clearly winning. The whole idea of giving a four digit number based on a simplistic statistical distribution is anathema to me not because it uses a statistical distribution but because what it's based on is so hollow.

"A model is a representation of something else. Whether the model gives an accurate enough picture of the real entity depends entirely on which features are important to you." --Gary Chartrand, Introductory Graph Theory

Na, that's rubbish.

I often simplify even if I probably have a mating attack.

Winning an endgame a piece up isn't 'worse' than winning quicker with a mating attack.

Sure, it's rubbish, because that's not what I was comparing. I was comparing losing an endgame up multiple pieces just on time (i.e., when I was winning) to losing an early game outright (i.e., because I made a horrible mistake and couldn't recover). I'm going to stop because I think you're biased now on anything I'm saying about it and it shows.

My current correspondence rating is 1947 and my blitz rating is 1022 (off six games but it is accurate enough).

I am 64 and no longer very quick thinking. But give me time and I can penetrate a position well enough.

The skills required for the two games are different.

If i got this right, you're saying a person that wins earlier deserves more points? So a beginner that pulls off a Scholar's Mate deserves more points than a master that grinds out a tough win? The rating system would favour tactical over positional play when it shouldn't be decided by either.

Material count shouldn't matter. Do you think players that sacrifice pieces for the win deserve less points?

Both cases seem to punish good play.