If anyone has an idea why the different curves on the last graph look like each other and seem to converge to maia9, even though Leela is a different engine (in particular is there an 'objective' reason moves at - 5% are harder to predict accurately) I would be interested to know :)
Just to report that while playing, the bot left the game (?!) but the clock was still going
Here's the link to the game lichess.org/RdIHbA3LJEEz
is it normal?
Here's the link to the game lichess.org/RdIHbA3LJEEz
is it normal?
I challenged the bots to a match but they refused?
I just tried maia5, and it did feel human-like to me and at the right level for it's rating - especially when it castled into blundering a piece in the opening. Often-times with other programs they'll play perfect and then make a random non-sensible blunder - this felt more natural.
Well done!
Well done!
don't seem able to play from a FEN pasted position - this could be a great feature for practising your openings! Just paste a position in and practice against a human like opponent of the same level.
i did play a game bot1-maia felt stronger than 1100 for a lot of the game and had me on the ropes but eventually blundered and i managed to win (my rating 1925 blitz)
i did play a game bot1-maia felt stronger than 1100 for a lot of the game and had me on the ropes but eventually blundered and i managed to win (my rating 1925 blitz)
So cool! So excited! Can someone tell me the different maia bots?
good job
Maia is defintely human-like as it blundered against me, costing it the game.
👌👌👌👌👌👌👌👌👏👏👏👏👏👏👏👏👏
looks awesome!
This topic has been archived and can no longer be replied to.