AWSmith61 wrote:
11hub: Daily Chess rating is 'correspondence chess'. You have days to complete each move. I *think* it's OK to use databases but not a computer engine for those. I'd love some official comment or reference to chess.com's rules for daily chess and databases. I never use them in it b/c well... it smells like cheating and I don't see the point of playing chess if one's going to cheat. My rating is mine. And I really play against my own mind.
...
Basically, for Daily, no engines, tablebases, or actual people giving you moves. Books, databases (that don't contain a bunch of numerical computer evaluations and are mostly human vs human games), videos, magazines, etc are allowed to be referred to in ongoing games.
Regarding playing a computer that claims to be 1400. Chess.com doesn't have a lot of data about how a chess.com rating translates to a USCF or FIDE rating. Some studies have been done, but they are old and I don't know how accurate they were. What you can know from your rating is that after you play ~20 games, your rating will be pretty accurate as a predictor of your performance against other players who have played ~20 games. The way the math works out is you have a 50% chance of beating someone at the same rating. I'm not so sure about the odds vs someone 100 points higher but in my own games I win about 45% (?) of the games vs +100 rated players. The ratings system used here is a form of the ELO rating system which is accurate enough for betting agents to actually figure odds for matches and make money the same way a casino does. I'm fascinated with how accurate the rating system is.
With regards to this fragment; you are expected to win 36% of the time against opponents rated exactly 100 points higher than you. The probability of winning is 0.36. Winning about 45% against a pool of opponents would mean, on average, you are rated ~35 elo lower.
If there is a 200 or 300 rating gap the probability of winning slims to 0.24, respectively 0.15.