Bad chess bots (no offense they're kinda stupid at this game)

Hi guys, try going to Play < Computer and try playing a 2200 rated bot, if you're rating is 1200 or higher, then you should be able to beat them. In fact, 1200 is approximately my rating and I beat them once. If chess.com could fix this then I would really appreciate it. It might be a hard task though, so if you can't, I can understand.

Are you playing them in challenge mode? If so, what device are you playing on? They run client side so a slower device won't be as strong

Macbook; No challenge mode.

If you're beating the 2200 bots on Challenge mode that's something different than on Friendly, Assisted, or Custom with some settings (1 minute timed for example)

Thanks .

I can't beat a 2200 bot but I can beat a 2200 elo human

I must really suck at chess then...

Edit:2200s play like an 800 never mind thebots are too easy to beat lol

As a 400 rated player I can defeat 1300 rated bots easily. They play fine in the beginning and then make stupid blunders.

I'm a 1400+ and can play well in the beginning and then make stupid blunders

This game as an example:

https://www.chess.com/analysis/game/computer/218711455?tab=review&move=12

The 2200 Komodo18 (yes 3 crown challenge) simply do not take back the horse, later loses a pawn, then blunders its rook. Analysis said it played 1550, opening best(???), for middlegame no analysis at all is given. I needed 7 minutes with some standard moves, Komodo18 just 2,5 min. in a 30 min. game. If it would need more time to calculate stronger moves, it would have it.

However, ElDeplorableBot 2200 was much harder to beat, took many attempts to have one win.

Sometimes the bots or engines fail to play their level by far. Why?

2000 is about 1600-1800 fida the bore are worse than that... Using fida ratings doesnt make sense when there are ratings on chess.com it would only confuse people

It's really hard to make an engine accurately play like a certain rating. The engine idms configured with an estimated rating and the underlying code changes how the engine evaluates positions based on that. There's always going to be some inaccuracy in that process

In addition, the client capabilities will influence how well the engine plays