As i had announced I left the -by now boring- thread about solving chess;
yet as someone suggested people like me may 'have problems to solve' I
will comment/react to this person here:
Chesskobra
i'm not offended, but i was annoyed, for reasons (eg .the hosting here, and the on average
rather low level of discussion) i already explained in a previous posting in this awesome thread .
Some postings there (in the solving chess thread) confirm/show again that also math/game theory is a sociological
process. It sometimes takes time and a changed mindset ( i dont like the word paradigm change) before a new truth
is understood . It's not about finding one winning line against SF, it's about finding a winning strategy for White; which is
impossible; let's say you find one winning line with 1.h4 after 1.d5 (probably due to some Black mistakes but
anyway); then you simply change 1.d5 to 1.e5 in the opening book and again, it's a draw.
So it's not about beating SF in Game, it's about finding a winning line in eg. the
Chinese database (instead of -probably- losing line like 1.g4). Mission impossible.
So i may have a problem to solve, namely, writing a clear and unambigous article
about my findings over last decades and showing chess is a darw.
Chesskobra it's Speelman, not Spielman(n) the latter was someone else; and i don't
care what eg. an Elkies or Terence Tao or the Pope would (currently) say about it.
An other example besides those i already mentioned is Arno Nickel, organizer of the
-now defunct- site infinitychess for advance chess; they found out chess is a draw
already years, ago, therefore he addressed the draw problem in 2015
https://en.chessbase.com/post/correspon ... aw-problem
https://en.chessbase.com/post/how-many- ... for-a-draw
For the rest it appears only onetrickpony there (in that horrible other thread) seems to get it, and i agree with
his suggestion of 'weakly solved in practice'
Dann Corbit suggested this may be true but we only lack a formal proof. Indeed for ultraweakly solving you need
a math proof, but for me Mcts plus Zermelo is indicating an ultraweak proof in practice. The Chinese database
indicates a weak solution in practice. And my computer when running Shashchess with 6 thrds Mcts
indicate a weak solution in practice. A few more years and we simply omit the words 'in practice.
Some math purists may then still continue to debate about nomenclature, well i don't care;
the current definitions are on wikipedia, but i couldn't find much about the origins of such definitions
(starting from eg. Neumann, Nash etc. it looks like it's some terms invented by
one specific inner circle of game theorists, eg. originating from Jvd Herik etc. If someone
has clear references besides wikipedia, let us know. Oh and yes you can corroborate
that Hex has been 'ultraweakly' solved by using an Mcts searcher; it has been done you know.
https://webdocs.cs.ualberta.ca/~hayward ... ts-hex.pdf
An ultraweak (horrible word) solution is not so weak btw
By now it's not 99.x pct certain that chess with current rules is a draw, it's 100.0000.. pct certain.
Because i say so (quote GM Ben Finegold with his favorite expression when teaching 1.h4 kiddies on yt).
Yes kiddies, play 1.e4 or 1.d4, or 1.Nf3 (not h4); don't believe
me syzygy, well then run lco nibble for some time and you'll see.