Replying to @epicgamer12456 @MehmetKoz1111
Ah yes, these terrible results. Can you believe he didn't beat any 2700s... other than Van Foreest, Erigaisi, Fedoseev, and Aronian?
Replying to @anishgiri
Excellent ask! Yes, here they are!
A graph I can't include as a supplement in a reddit comment so I'm posting it here without context (although it's pretty self explanatory and may be of interest even if you're not on the thread)
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.
Replying to @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?
Replying to @VanillaVICE
Feels silly at this point, as the "breaking down scenarios" part of the analysis is long past, but sure
Last night I set my five year old up with his first lichess account (so he wouldn't try to do MY chess puzzles and tank my rating). Today he got 12/15 puzzles right, mostly mates in one, and coined the fantastic term "immedimate" that definitely needs to catch on!
At this point we all know the stakes. Three spots for four players. It'll be decided over the board, not on a spreadsheet, but just for fun[1] here's my model's take on the World Cup Final Four!
[1]Citation needed
Here is my model's take on the last eight remaining players at the World Cup. Notable is how much of the total winning chance is in the 2v7 quadrant; whoever wins between Wei Yi and Arjun Erigaisi will be a big favorite in the next two rounds.










