?

Log in

No account? Create an account

Previous Entry | Next Entry

nothing ventured

to follow up on my last post about the chess, one thing that really stands out for me is how many times both players, in the analysis of computer-assisted observers, have failed to make midgame or lategame moves that would have set them up for a near certain victory. i would not be surprised if it is the case that players who've reached this level are fairly risk-averse overall and all too willing to accept a draw rather than lose (and perhaps look foolish) and although i don't know this to be true it is quite possible those moves were unorthodox and risky. still, it says a lot to me that our greatest human players can fail to glimpse all of the possibilities once they've been substantially narrowed, even as they clearly see so many of them; that even they are quite fallible. i wonder how much of the admixture each contributes to these failings, between risk-averse play, failing of ability, and over-reliance on known lines of play causing unorthodox unprepared-for lines of play to be dismissed out of hand (out of mind...). have our best players perhaps become more 'robotic' than actual (with figurative license) robots?

Comments

nightspore
Nov. 21st, 2018 10:20 pm (UTC)
Endgames are much more about calculations than openings or middle games. A computer can almost always outthink a human in an end game. The maxims of strategy (as internalized by grandmasters) are far thinner in the end game than before. So it's much more a question of calculation, and calculating all the possibilities down 20 moves on a nearly empty board where pieces can move almost anywhere is really daunting unless you're a calculating machine.
jones_casey
Nov. 26th, 2018 03:31 am (UTC)
as i said i'm far too unskilled to judge and so i take your point without hesitation. yet i'm pretty surprised that this is true. i would have thought most endgames would offer fairly obvious (to the experienced) best moves simply due to the dearth of pieces and positions, especially since expert players can very quickly concur on a draw without playing it out to try to generate the slightest chance at an advantage. on another note, i don't know the fuzzy bright line between when endgame starts and midgame ends, but it seemed to me that the computers still have the advantage as some point in the midgame.
nightspore
Nov. 26th, 2018 04:30 am (UTC)
There's no intuitive way to see subtle differences in an endgame. For example computers showed -- to everyone's surprise -- that two bishops against a king could force a mate after about 90 moves. Until then two bishops was regarded as a certain draw between grandmasters. None of the bishop moves were intuitive. They went all over the board in apparent indifference to the other king. But the vice slowly closed. Go seems to be like that too now, when the best programs play it. Humans have no idea what the programs are doing but the programs win. Whereas in middle games you can see what computers are trying to do -- break out, attack, defend. The resistance of so many opposing pieces means that the computer's pieces don't have free run of the board as in endgames. The near term goals are a lot more important in middle than in end games. Humans can intuit over longer stretches in middle games than in endgames.

Edited at 2018-11-26 04:32 am (UTC)
jones_casey
Nov. 26th, 2018 08:16 am (UTC)
i appreciate your insights. i seem to vaguely recall reading about that 2 bishops thing -- i did at one point read a book about endgames when i was trying to see if polishing my gane would reveal any underlying value but i didn’t retain much and it was probably 10-14 years ago.

Profile

blue legacy
jones_casey
cleaning up so well

Latest Month

June 2019
S M T W T F S
      1
2345678
9101112131415
16171819202122
23242526272829
30      

Tags

Page Summary

Powered by LiveJournal.com
Designed by Witold Riedel