Gotham claims engines aren’t great at closed positions
Its because engines can calculate tactics 10+ moves ahead. They find open positions “easy” for this reason.
Engines don’t have a preference for open vs closed, post the game if you want and I’ll see if I can explain the engine’s point.
Modern engines are a lot better at closed positions than they used to be, but you can still construct positions that they misunderstand.
There is some Naka game from a long time ago (when I presume engines weren’t as good) where he intentionally creates a closed position vs an engine and then just starts trolling it. Game ends with him having like 6 bishops (from intentional underpromotions).
A thing I’m having trouble with is playing fast enough. I started with 10 minute games and got my ass handed to me early on against rank beginners in a couple of games because the clock was causing me anxiety. I switched to 15|10 games and while I’m not timing out and have won 8 in a row, I still take a bit too long per move. My last game was 25 moves and after checkmate I had a bit over 4 minutes left. If my opponent didn’t blunder into mate in 1, I probably would have reached an endgame and run out of time there.
Crazy variety of openings this WCC from Ding. Was not expecting to see the French get whipped out.
Nepo doesn’t have the ability to handle adversity in-game.
He’s praying that Ding times out
EDIT: JFC does Ding understand the time rules?
Holy shit Ding is completely frozen. Dude has 45 seconds for 8 moves.
Totally blundered at the end. What a choke.
Speechless. Craziest collapse I’ve ever seen in chess.
I played a game against GothamChess for one of his YouTube videos once. I have never played when I knew hundreds of thousands of people would see the game before. Something like this happened to me. Brain would not function.
You can’t say that without linking us to the video!
And while I’ve had moments where I’ve frozen in chess (was just talking about being too slow itt), I do eventually make a move. In Ding’s case, I figured he was planning out the next 10 moves to be done in a blaze instead of focusing on just one move. I was wrong.
Game 4. It’s not very interesting, I just blundered. Like I said, brain would not work.
chess.com profile, my current rapid rating is very bad, switched to Lichess for a while and ate a downswing when I came back, when rating adjustments are high as I haven’t played in a while. Currently 1870 but max was 2043, I think 1950 is about right. Blitz and bullet rating are accurate, though.
At least you weren’t on Guess The Elo.
1 Like
I think I’m a little bit better at Guess The Elo than Levy. I would bet on myself against him.
One big failing I think he has is to overrate opening mistakes in terms of being like “oh they didn’t know what to do on move 4, that means they can’t be over 1500”. It’s been funny to me to watch the current Naroditsky speedrun, the idea of which is to emphasise opening theory, and he started out being like “my Official Recommendation against e4 is the Accelerated Dragon!”. Now he’s played a whole bunch of games against 1. e4 and has gotten the Accelerated Dragon on the board like twice and has otherwise faced an endless array of random anti-Sicilians. He’s discovering that there is no point to learning an opening that you never get on the board anyway, and I think masters way overestimate the ease of learning opening variations, because the fact that they are a master ipso facto means they have an excellent memory for these things.
Which is not to say there’s no point learning openings, I have a > 50% winrate with the Caro-Kann, which is pretty good as Black. But I don’t have that because I am winning out of the opening thanks to my insane theoretical knowledge, it’s because I have good understanding of the positions that arise. And because there’s no such thing as an “anti-Caro”, I am banging out d5 against every second move White plays, we are guaranteed to play in my territory.
To give a practical example, I’ve only just started trying to play the Nimzo-Indian and reached this position today:
I couldn’t remember what to do here, but I do know that the point of the Nimzo is trying to control the e4 square, so I was like fuck it, I’ll just take on c3, that way at least I will for sure achieve that objective. The correct move is actually O-O and Bxc3+ scores poorly in the Masters database and is like half a pawn worse according to the engine than O-O. If this was Guess The Elo, Levy would be like “what an idiot, plays the Nimzo and doesn’t know what to do on move 4”. Cool, but let’s check out how much win% this cost me, in Lichess database, filtering for players of similar strength:
Doesn’t even slightly matter, turns out.
It’s a lot easier to guess the elo of players closer to your rating, that’s probably why most titled players would not be that great at it. Like I could tell the difference between a 1400 and 1800 pretty easily but probably could not tell the difference between a 400 and 800.
Nothing mind-blowing here but I appreciate how unwilling people at the beginner level are to resign games. Gives me a chance to practice my endgame play. I think they’re a bit masochistic but they hang around until the bitter end no matter how crushed they are.
My last opponent was so averse to resigning that they timed out instead of playing on. I mean they might have construed my poor endgame play as something more sadistic and did this out of spite but a win’s a win.
Timing out is tantrum-throwing, generally.
Really interesting world champs match. That QGD in game 3 was the only boring game so far.
I play the nimzo as black and in that position my favorite move is b6 followed by Bb7-Ne4-f5-0-0. That setup in particular is extremely heavy on controlling the e4 square.
1 Like
Has anybody tried using the bots in chess.com for practice? They seem like they could be really useful for that.
I’ve come across some offbeat variations of openings in my rapid games that have been throwing me off and causing me to lose quite embarrassingly. They seem like the type of openings you shouldn’t see at my low rating but they are there.
It’s clear that I need to work on handling them but I’m wondering if I should stick with a strong AI that’ll punish me more for opening mistakes or one that will play at a level consistent with my opponents.