Topalov at 3/4. BJJ looking like a genius!

Another amazing crazy win by Levon and a hilarious interview with Maurice (starting @3:30:25 in the official stream recording)!
What stops you from closing the chat box by some other window?

So what happened to Caruana? He was on top of the chess world a year ago.

his live rating is like 2800. What do you mean what happened to him? He had one super outlier tourney and shot up above his real rating. Do you think Topalov is 2825 strength now and if a year from now he is "only" 2800 something must have happened? This is natural variance

Right, chess observers somehow tend to underestimate the extent of variance that's signficant even at the highest level. That variance comes of course not from the RNG (which only determines colours) but from the fluctuations of shape of players and the fact that the time to act is finite (it's hard to calculate deep, especially just before the 40th move when there's no increment, so some positions become a gamble because the players have to use heuristics instead of more or less exhaustive calculation). It's bigger than if single games were played with decent intervals between them - when someone is on a slump or on fire, it lasts for at least several rounds (i.e. their results are highly positively correlated, not independent).

this sounds like a question for BJJ! Is there really a hot hand effect? Are chess players really very streaky, or do we just fool ourselves into seeing patterns in randomness? I think there is enough high quality data to measure, say, if a player's results one day apart are more highly correlated than his results 3 months apart.

It makes plenty of sense that a player's mind is clear one week and less so the other. The hard part is predicting inflection points. Take So the week his mom was stalking him - likely had a huge effect on his level of play.

I'm actually in the process of developing a methodology by which to study the hot hand phenomenon! I started building it earlier today. My core question is: In my simulations is it more predictive to use live ratings or just use the same pre-tournament ratings the whole way through?

I currently use live ratings, figuring they capture "form" to some extent, but that's only better if late round results actually are correlated to early round results. If there's no detectable correlation then I might as well use pre-event ratings the whole way through.

I think I've determined how I'll attack it, and have started building a spreadsheet to track and analyze the question, but data collection will be slow as it stands right now. I'll definitely publish an article on the topic when I have answers!

Another complicating factor that I'm worried about is that anecdotally I sense there's a tendency for players who are both out of contention in the late rounds of tournaments to play quick draws more frequently. Haven't confirmed, but if draw rates do noticeably increase in later rounds for this reason it could look like regression to the mean and possibly make it look as if there is no hot hand phenomenon, when it's actually an unrelated causal influence (tournament position). I have not yet figured out a way to control for this, so I'll probably ignore it in my initial study unless something comes to mind, and just mention it as a footnote "topics for further study" type of thing.

Is it that easy to play a "quick draw" though? I mean isn't the ability to secure a draw when you want one a skill?

Depends on context. In later rounds if one guy needs a draw to secure first place and their opponent is having a bad tournament and just wants to be done, a draw is probably very easy as both players may willingly go into a knwon line that is very drawish.

Well, when BOTH players are fine with a draw (last round, both out of contention, for instance) it's super easy. That's the scenario I'm thinking of. Not sure if it will actually have a detectable effect in the data though. We'll see, I guess.

Why aren't they playing with increments? I thought that was standard these days.

Yeah I wouldn't worry about "agreed upon" draws noisying up your signal and representing a regression to the mean for the otherwise hot player. Even in these cases where both players want a draw, if one slips up the other wins. So it's still an "earned" half point really. It's not like the player forfeits, earning a goose egg and truly making the data less "natural".

If an exotbitant % of these types of draws happened as white, against inferior opponents then that would screw the data up a bit but this should smooth out over time.

Would it be reasonable to exclude the last round and see if that changes your results, or is that too much extra work?

Those confession box thingies that they use - are they available for viewing anywhere?

I think it's clear that live ratings carry more information and should be better predictors than pre-tournament ratings. And we also know that the fide k-factor is conservative and a higher k would give live ratings more predictive power (this is not necessarily bad because ratings are not used only for predicting future results).

If there's a hot hand effect, though, you should overweight the recent results, move the rating used for the sim by more than it moved for the live rating - even more than it would move with a k-factor optimised for long term predictive power.

That depends on how my data gathering goes, and more to the point how the data ends up being formatted. If the formatting allows for it to be checked easily then I would do so.

I think that's clear too, intuitively, but would feel better if I could prove it. Updating everyone's rating to reflect the most recent round before starting the next round is tedious enough that if it turns out intuition is wrong here, and if data says it doesn't help, then it would be nice to stop doing it

Plus it feels like the kind of study that once I have the data, could also be useful for other things too. After the all-decisive first round I looked a little at those odds, and then it occurred to me to wonder about the odds of a round with NO decisive results (all five games drawn) and from there I started thinking about whether there might be any kind of round by round grouping effects in draw rates.

In theory, if every game is truly independent, and if draws happen at a consistent rate of 56%, then the odds of all five games in a round being decisive are 1/61 and the odds of all five games in a round being drawn are 1/18. This means that in any given 10-player round robin (nine rounds, five games per round) there should be a 60% chance that EVERY round has at least one decisive game. Seeing all five games in a round be drawn should be rare.

Norway saw all draws in round 7 (event draw rate was 56%). Looking again at how it "feels" to be (rather than actual data) this doesn't seem that uncommon to me. Not as uncommon as it "should" be statistically, unless there's a phenomenon that "groups" draws into certain rounds (maybe later rounds?), which would have important implications for my simulation methods.

Another not statistically significant data points: Tbilisi Grand Prix had 11 rounds (6 games per round), so a round with all draws should have been even rarer if the draw rate was the same. First nine rounds: at least one decisive game every round, total draw rate 58%. Final two rounds: 12 draws. Counterpoint: there was at least one decisive game in all 12 rounds at Khanty-Mansiysk. It's frustratingly hard to know what "patterns" we think we see are real, and which ones are just our overactive imaginations. This is why I love data

Long story short, if there is an effect that increases the draw rate late in tournaments, building that into my model (along with the corresponding DECREASED draw rate in early rounds, since that's how averages work), it would be useful for me to identify and quantify. The data set for that is probably the same as the data set for my live ratings study. Plus who knows what other unrelated and unexpected discovery might come from the research. Or maybe I'm totally overthinking it all, and nothing useful will come from it at all, I am prone to that

Lagrave is now my favourite player. That confessional was the nuts, lol.

(It's @55:34 in the stream recording, thanks for sharing! )
Meanwhile, the Topalov vs Caruana game is going berserk - White is down a rook for 3 pawns, but has 4(!) connected passers rolling at the king (Well, it's as of move 41; it might have been temporary because Black can return the bishop for two pawns - 41... Be4 42. f3 Bxf3 43. Qxf3 Qxe6. Instead, Caruana played 41... Qh7, allowing the 4-pawn march, oops.)

I don't think Grischuk is going to be invited to the next GCT... he chronically gets into time pressure and thus is at risk of making suboptimal moves 30-40, and now he's agreed for a draw at move 30 (the earliest when it's allowed), meh...

Oh, and Wesley has just collapsed under time pressure (37... Qb5?), it looks as though the champion is luckboxing back into the top.

caruana is playing at 2800 strength right now but clearly not at 2810 strength. imo.

Why is the rest day Friday, not Thursday, when it would be more logical? Those spectators who have real jobs could come earlier on Friday (though 3 AM is still too early for the start time).

That Carlsen win over So is impressive, at least from my view as an H-rated player. He was down and his position didn't look good and he's just like this is winable ldo.

I think at no point was Carlsen actually losing most likely, he had compensation at all points at the least. It was also quite impressive. It seems he missed a few ways of forcing the win earlier but this definitely seems like his style of getting a winning advantage and then not letting his opponent wriggle free. It definitely seems like Carlsen is "back" or at least partially back anyway.

Trying to figure out how strong So is seems very tough. What are his results vs. top 20 or so in the world? It seems to me he's done fine with them but overall has gotten battered around a bit. Has he really shown any instances yet of being able to contend?

It's kind like with Giri where I feel he seems super solid but I don't ever see games from him that make it appear like he can just go out there and win against the very best without them making significant mistakes. Maybe he outprepares people though and wins primarily by that method though.

E.g. Caruana seems like for sure capable of contending since he can go out and beat the very best even when they are on their "b" game (I'm going to just assume the best cannot be beaten on their "a" game). The problem is he seems to be on his "b" game himself too often.

but im gonna lose my roll if toppy doesnt win!