Play all the bots, one at a time, until you can beat the 2000 bot every time. You will probably find you beat most players at that point.
Can you improve by playing bots/computer?
Hi,
I think playing chess in general will improve your game if you're taking time to find the best moves and analyzing the games afterwards to identify your mistakes. That said, I think bots, particularly lower-level ones, make random errors and then otherwise good moves. It can create very odd games that you likely won't encounter against other players. I think there is a good reason to play against other humans and learn to accept losses with dignity and wins with grace, and practice taking games as free lessons and developing healthy relationship to losses against others as not somehow meaning you're stupid, or inferior, or less of the person you were before the game. You can develop a good habit of playing a game and whatever the outcome may be, learning from it afterwards. You also can miss out on potential social experiences such as post game analysis with the other player and developing some friendships and a sense of community. Chess can become very isolating if you want it to be, but it certainly doesn't have to be.
In summary, yes, I think you can improve at chess playing the engines, but I don't think it's the best way to improve.
-Jordan

In his stream Hikaru nakamura recommended playing higher ranked bots.
Can you plz send me the yt video or blog post link of this statement ???



Ive taken a break from live games with the aim to get to 3000 puzzles and beat 2000 rated bots, well that's what I decided yesterday. See how long it lasts but if I can be a challenge to 2000 rated bots pretty sure I'll be able to hold my own against a few players. I just haven't got the time to play 30 min games most the time so being able to play bots should be a good compromise... Though they do weird things, 1700 rated bot hung its queen against me yesterday 🙄

Playing with bots is OK, just make sure you don't over estimate their rating, for example when I was 600 (around 900 now) in rapid I could beat an 1100 rated bot. I would suggest playing bots that are 400-500 points higher than your rating, maybe even more, if you want a good match.
Also, if you want to practice a trap line, or things that you will often see in a real game, don't play a bot. They either play something totally obscure (like h4 or something) or memorize more lines than most humans do. If you want to practice a certain skill, try playing unrated against a friend, not a bot.

I lost 50 brain cells going against Martin.
I purposely did a Botez Gambit and he ignored the fact there is a nuke in his defenses for like 5 moves. 💀
I went from obliterating Sven (1100), to Nelson (1300), and finally to Antonio (1500) without any help. Just got my first win against Isabel (1600 🫡) too.

You might want to check out Noctie, it's a new chess engine that emulates human play at levels from beginner to master + has some unique training features like instant feedback after every move

I say yes.. I play my regular amount of human to human games on lichess, but I have been working through the chess.com bots. Some of them are quite difficult but will fall for easy traps.. others will take many games to beat. I'd say if you plan to play bots for practice, start with the easiest and work your way to the hardest using both white and black. Try out your new openings and aggressive attacks against them. For me overall I think I've grown in the way I approach attacking just by playing bots.


What are the benefits of playing with bots? 🤔 I think that the benefits are the same as playing with cheaters (all bots use the engine 🤫). Am I right or not?

Take help from a stronger player for analysis.
Maybe most of those stronger players used bots to analyze!! LOL..
I mean you can, but it gets pretty boring after getting crushed every game. Playing humans are better for sure.
No, you obviously shouldn't practice against full strength engines, but weakened to an elo around your level. At lower elos, the engine does play funky lines, but once it's at around 2000 elo, it's very, very solid both tactically and positionally.
There was a Chess Life article a few years back where a guy won the very first tournament he ever played in, beating 2000s and even an IM(I think it was), by only playing against weakened engines.
The best way to improve in Chess is by playing long games against players slightly better than you and reviewing the games afterwards-- however, I find that the human variance below 2000 means most games are decided by one side making an obvious blunder. Yeah, you won, but you know it was simply due to a tactical or major strategical oversight.
Imo, this does not push you to improve as quickly as a game that pushes you to your limits of tactical and strategical, positional understanding. One where you need to review the game afterwards and understand how or why you lost and what were the winning ideas. Humans can give you these type of games, obviously, but because we are inconsistent, very often games become completely onesided even if both are similar elos.
Engines, sufficiently strong yet still weakened, are always consistent in their play. The only way to beat them consistently is to consistently play above that level of play.
The era of Engines are just dumb positionally but are tactical gods is well in the past. Engines positional play is god tier nowadays and only improving.
Who are you to say that they do not practice playing against them? I'm not saying all do, but I'd wager a good amount do. I personally have gained 150 Fide in just two months of only playing against Engines in classical time controls. I'm now 2043 Fide. A 2000 elo player can hang a Rook or even their Queen in 2-3 move combinations every now and then. A 2000 elo engine will never do that. It forces you to refute their inaccuracies with strong lines-- this type of thinking will hone your calculation skills to the next level since it's about pruning for minute inaccuracies and calculating lines for small positional advantages that you can continue to pile up over time.
I remember Hikaru Nakamura stating that players should play against lower elo bots as training. Anyways, I personally think it's great training, and actually having tried it and smash through a plateau adds some support to this claim.