Apparently, you can trick computers

As far as I can tell, you in fact can trick computers. How?

Let me put things in perspective: I beat the chess.com personality bots Wally and Li for the same reason; they captured poisoned pawns with their bishops. But why? I have no idea. And the funny thing is, in the analysis with the game with Wally, it claimed that capturing the pawn was the best move. Also, Noam fell for the same thing! I played the rest of the game all in one go and lost, however... but anyway, you can trick bots! And humans usually don't fall for this trick!

Also, here is a game where Noam made a huge blunder by falling for a pawn fork: [Event "Vs. Computer"][Site "Chess.com"]
[Date "2021-06-01"]
[White "MyNameIsNotBuddy"]
[Black "Noam"]
[Result "*"]
1. Nf3 Nf6 2. g3 d5 3. Bg2 c6 4. O-O Bf5 5. d3 e6 6. h4 Na6 7. Bg5 Bd6 8. Re1 O-O 9. e4 Bg4 10. e5 Be7 11. exf6 Bxf6 12. Bxf6 Qxf6 *

(I played the rest of the game all in one go and lost, too. Sad.)

I never rush games against bots though I could let it run for hours or even days, when I get bored I do something else and get back to it. One thing I can say about the bots is that they are programmed to make a certain number of inaccuracies or blunders per game depending on the rating. The placement of the inaccuracies in the games though could be very odd and uncalled for like straight up letting a tactic happen or just giving away a pawn.

Huh. I usually play bots when I am not connected to internet. Otherwise I would probably play real people. I think they might look at a lower depth depending on their rating?