1400 bots: mateo-bot, david-bot, ali-bot

mateo-bot:

this game felt wild. i went for an aggressive opening and stayed on the offense. apparently we were playing queens gambit accepted, tho i dont know how to recognise that one yet.

there were pawns everywhere. i started checking early and often. and the game became a cramped fight over my promotable pawn while there were still most of the pieces on the board.

someday i will remember that en passant exists, but today was not that day. i gave away a pawn with this move:

luckily i was already ahead enough that it didnt stop me from getting the win.

the engine was so proud of us.

i do wish i woulda noticed that i coulda absolutely trapped this rook, and had one less thing to worry about during cleanup:

but despite my mistakes, there was never a moment where i worried i might lose.

david-bot:

this was a messy one, with a messy rook endgame.

i made more mistakes and inaccuracies than it felt like i was making. here’s one where i simply didnt consider that if i go to attack a knight with a knight, the knight can attack back:

luckily i could at least take back, but i spent like 15 minutes calculating other moves from this position and thats what i came up with??

i suppose its not entirely accidental that my position was coordinated enough to be able to soak the unintentional trade, but still. i could do better.

i havent studied rook endgames, so i had to really think a long time to not mess it up too bad once things got technical in the endgame. i was pretty sure there was a way to win, and also ways to lose with one wrong move, and in the end i brought my king over to defend a promotion and lure in the other rook. 

i thought i could win with just a king and rook, tho it was nice of david-bot to let me promote the other pawn anyway, to speed things up. but in retrospect i think i woulda rather done the suggested move to guarantee it.

not my best work.

ali-bot:

this was a straightforward game, i think. there were no blunders, and tho i had some misses, my middlegame harmony was sufficient to punish ali-bot’s mistakes and bring all my plans to fruition. i felt like i was playing the best i possibly could given my current knowledge and experience.

ali-bot started with the english opening, which i havent studied. i didnt feel confident about playing in the middle, so i went for a kingside fianchetto to threaten a potentially future vulnerable rook on a1, and played “guess the book move”. we made it 4 turns of book.

as always, i have trouble figuring out when to pawn trade and when to build pressure. in this case it looked like there was no need to rush and i thought i could come out on top if i took the time to develop more. apparently this was the wrong choice, even tho it worked out for me this time.

i have found that i enjoy long-lasting xray staredowns! but perhaps i should leave long staredowns to the bishops and queens?

when i did finally take the pawn, apparently that was wrong too. i missed my chance. but my intention was to keep the pressure on the pawn stuck between the queens, and maintain the possibility of a nice queenside pawn chain.

after the trade, white’s d pawn moved forward, which i wasnt expecting. i sometimes forget to consider pawns can just move up instead of trade. but it immediately felt fun, because i realized the pawn and horse were not actually threats, and i could try to draw them in rather than counterattack. my strategy became to limit white’s possibilities for good moves, while improving my harmony and maneuvering to move up my a and b pawns.

in the following sequence of moves through the middlegame, my good moves felt like good moves, and my opponents mistakes felt like mistakes, and thats always gratifying. blunders are usually immediately apparent but im much less consistently able to say “wait, thats a mistake, right?”

ali-bot seemed unwilling to back off of the unsuccessful attack, and shuffled pieces around in a way i was able to detect as uncoordinated and unhelpful. i was able to take advantage and get myself into a winning position through a few decisive moves. a lot of my games are full of ups and downs and chaos, but the flow of this felt better. it felt like i was winning because id lured my opponent into making mistakes, rather than because they were programmed to let me win.

the endgame wasnt quite optimal and i had a couple misses, but it didnt need to be perfect, and i was ready to not take so long on each move. 

and that’s all the intermediate level bots! its been a few levels since i beat all 3 on the first try, so thats a good sign that im still improving faster than the bots are. or maybe im just playing slower and getting more stubborn. but im still hopelessly terrible at 10 min games so i take my bot wins with plenty of salt.