Chess by the Numbers

Chess fan, spreadsheet fan, stats fan. author:

[Chess by the Numbers @ChessNumbers]

@OnTheQueenside So as a rule my benchmark for big gains is that if [new rating] < [performance rating to get there] I'll give it the benefit of the doubt as likely relatively legit, while if it's the other way around I get a lot more skeptical of the new rating.

[Chess by the Numbers @ChessNumbers]

@OnTheQueenside Not enough to *prove* anything, but my point is this isn't one of the cases (that also happen) where someone jumps from like 2200 to 2500 with a performance rating for the month of 2350ish, by squeezing in a ton of games to one month

[Chess by the Numbers @ChessNumbers]

@OnTheQueenside *performance ratings in May

[Chess by the Numbers @ChessNumbers]

@OnTheQueenside Worth noting that both of them had performance ratings in June (2448 and 2380) that were *higher* than their new ratings. So this isn't k-factor shenanigans and overshooting their levels of play, it's just the k-factor working correctly to reflect their strong performances!

[Chess by the Numbers @ChessNumbers]

@EnronCompliance @ChessvisionAI I'll bite, I would love to hear an argument that Keymer had a better 2025 than Pragg

[Chess by the Numbers @ChessNumbers]

@mcelsouza @MehmetMars7 Surprisingly Judit didn't hit top 50 until she was 16 (although she hit #55 at age 12) and when Fischer was that age, the rating system wasn't actually established yet.

[Chess by the Numbers @ChessNumbers]

@OnTheQueenside Just to confirm for the broader audience, your use of the qualifier "non-provisionally" here means outside of the player's first 30 rated games, yes? Because 30 games played is (currently) a requirement for FM title eligibility?

[Chess by the Numbers @ChessNumbers]

This month's Prodigy Watch update is live: https://t.co/kpmmmBETwV

Welcome Charvi Anilkumar not only to the top 5 of the girls list but also to the open top-100 list! (And she's already gained 46 more live rating points on top of the published rating I use here)

[Chess by the Numbers @ChessNumbers]

@undertherainx My data structure is a database, so I have players and their birthdays in one table and published ratings by player and date in another, and can query ratings by age in various ways from that. Don't have a way to just share it all but I could put together a few reports.

[Chess by the Numbers @ChessNumbers]

Over a decade ago I wrote a ton about Wei Yi in the early days of my then-new chess stats blog and early iterations of Prodigy Watch. It's incredibly exciting that I'll finally get to see him in a Candidates tournament!

While I'm a fan of many Candidates, I'll root most for him.

[Chess by the Numbers @ChessNumbers]

@undertherainx Yeah, e.g. Judit Polgar is the highest rated player ever, prior to the (cherry-picked) age of 12.45, but on the "highest before turning 12.44" list she's 46th

Of course that's looking at just before and just after publishing her 190 point rating gain (over 6 months)

[Chess by the Numbers @ChessNumbers]

@undertherainx My data isn't complete either, but does include deceased (where I've added them). I have every 2700+ ever, and most 2650s, but not down to 2600 in all cases, as I've been careful to only add hand-verified data (since FIDE has some errors now and then) and focused on prodigies.

[Chess by the Numbers @ChessNumbers]

@undertherainx I wonder if 2000 might get there. If not Warmerdam, perhaps Theodorou or Puranik could make runs at 2700 over the next decade.

[Chess by the Numbers @ChessNumbers]

@undertherainx Very interesting how from 85-87 we saw 6+ future 2700+ players born each year for three years straight... and it only happened again once in 1990.

Of course those cohorts were age 20-25 when rating inflation peaked so it kinda makes sense.

[Chess by the Numbers @ChessNumbers]

@undertherainx I'm not fully sure exactly what your question is, and it sounds like it probably needs more characters than Twitter offers. I don't specifically have any era-adjustments built into the formula though. Either for inflation/deflation or for back when lists were published less often

[Chess by the Numbers @ChessNumbers]

@undertherainx Prodigy Watch is currently maintained and updated monthly here: https://t.co/kpmmmBETwV

[Chess by the Numbers @ChessNumbers]

@SportForesight @ChessbaseIndia Lmao, all that and I forgot Sharjah had 50+

So yeah, he has to keep his Grand Swiss if I'm right, but still he gets to count with world rapid and world blitz.

[Chess by the Numbers @ChessNumbers]

@SportForesight @ChessbaseIndia Yeah I'm pretty sure I'm right. Considering that the three official rating lists are technically named "standard, rapid, and blitz" (even though we usually refer to the former as classical) and considering that rapid/blitz events are mentioned as eligible separate from standard.

[Chess by the Numbers @ChessNumbers]

@SportForesight @ChessbaseIndia That passage defines "standard tournaments" as ones that are fide rated otb "standard-play". Kind of a tautology and not totally clarifying.

Is there further clarity of that last clause elsewhere? I interpret it as "standard-play, as opposed to rapid or blitz"

[Chess by the Numbers @ChessNumbers]

@SportForesight @ChessbaseIndia I understand "standard" to mean "classical" here. So if he's counting 7 events he must either include his Grand Swiss and World Cup scores or play another big classical event.

Am I wrong about what FIDE means by "standard"? Can world rapid/blitz count towards 4.2.1?