zullil wrote: ↑Tue Oct 29, 2019 6:52 pm
Discussion between us seems pointless.
What is interesting is that this experiment further convinces me that Stockfish's most significant weakness is in its static evaluation of near tablebase positions. This is increasingly important, because on decent hardware Stockfish will be reaching such positions from very early in the game, and will make choices that steer the game into (or away from) certain such positions. When a drawn 8-man position is given a static evaluation of 1.5 (for example), Stockfish might well head there, instead of heading for a winning endgame with a static eval of, say, 1.3.
Certainly expert human guidance can help here, and I suspect that many of the (apparently, rare) decisive games in high-level correspondence play result from such human expertise.
So you are learning something. Imagine that!
You are very much right about the endgames influence on early positions in CC. You're also right about SF's eval being an impediment to selecting the right course of action. When it comes to endgame evaluations SF can be a complete nit wit. You will spend half your time trying to convince SF of something YOU already know! But if you want it's help you have to spend the time to get it to realize the “true” value of those positions otherwise it's analysis will be flawed by it's evaluation function.
You are, however, wrong to conclude that wins come from human endgame expertise. It turn's out that many times the players are blinded by the engines single number evaluations. It turns out that this is a less than ideal way to evaluate a chess position.
Regards,
Zenmastur
I'm always interested in learning! And I didn't intend to say that winning at CC comes from human endgame expertise. What I think I wrote (and I just reread it) is that winning comes from expert human guidance (of the engine).
Despite my running Stockfish unassisted during this game, I certainly never claimed that such an approach was optimal in all games. As you yourself suggested, Stockfish unassisted might well suffice in this game, due to the nature of the positions arising from this particular opening.
Ovyron wrote: ↑Tue Oct 29, 2019 8:54 pm
Oops, I missed the word "static", I was talking about the eval that you'd get after searching
OK, no problem. Of course, with a search Stockfish would figure out almost anything. But at the leaves of the tree, when no more search is available, it all comes down to static eval.
I don't like the word "static" here because the opposite of static is dynamic, and an eval at a node includes both static and dynamic features.
I believe it's the standard terminology. But, of course, it's simply "the evaluation". "Static" is simply to emphasize that the evaluation is being produced without any additional looking ahead, strictly from features of the position at the current (quiet) node.
Yes, I guessed it might have been established terminology.
One could claim that "static" is being used in a non-chess sense, but that's also 'bad', because it then sounds like the crazy belief many have that evaluation comes for free.
Raphexon wrote: ↑Tue Oct 29, 2019 8:27 pm
I'm doubtful SF often changes its mind beyond depth 45-50 or so.
And what? We have >65 depths here from Louis since the beginning.
That Ovyron's claim of coming to the same conclusion as depth 70 SF with only 10 minutes of analysis sounds impressive.
But if there's no significant difference between SF at depth 45 and depth 70 then those claims sound a lot less impressive.
Because it's akin to saying you only reached depth 45.
Harvey Williamson wrote: ↑Tue Oct 29, 2019 2:21 pm
[pgn]1. g4 d5 2. g5 e5 3. d4 exd4 4. Nf3 c5 5. Bg2 Ne7 6. c3 dxc3 7. Nxc3 Nbc6 8. O-O d4 9. Ne4 Ng6 10. a3 Bg4 11. Bd2 Bd6 12. Rc1 0-0[/pgn]
I will give you your Qd7 wish:
if 13. Nxc5 Bxc5 14. Rxc5 Qd7
Needless to say, two more Stockfish moves ...
It has 'gotten' boring ....hopefully it picks up...
I guess it's hard to be motivated when you have a lost position so early.
It's only -1.64 at depth 72. No reason to give up hope. Especially if you have a special line that you've already found.
And also -1.64 at depth 73. That's with 15 hours of searching from my entering 12...0-0 (thanks to a large filled hash table). The current PV stretches to Black's 67th move.
