mmt Vs. Ovyron (G4 D5 BG2)

[jp]

It's not about me. For a start, could you answer Uri's questions and criticisms? Would you be willing to try to test your claim about predicting SF13's moves now? I don't know what positions we'd choose, but maybe others have ideas.

I have never said anything is "impossible", even things that are actually impossible. Do you actually realize you are misrepresenting what I say? I just asked questions and pointed out some of the many times you have changed arguments, contradicted yourself, made wrong statements, etc. e.g. if you say you have "proved" something when you have not, just pointing it out does not make any claim about whether it would be possible to prove.

[Ovyron]

It is about you, because you're the only person that keeps inquiring about them. Uri didn't make any question, he made statements, like me. He didn't say I was wrong, only that we don't know if I'm right. His point about Stockfish 13 sometimes playing moves weaker than Stockfish 11 wouldn't invalidate that as you find better moves on a position you'd eventually pass S13's moves, you'd just pass them earlier than S11's in those cases.

You'll be able to prove me wrong once S13 is released and you post a position where it finds a move that Stockfish 11 is never able to find no matter how long you leave it to analyze. Good luck with that, but right now a position where Stockfish 11 finds a move that Stockfish 9 never finds would suffice, so go get it.

[jp]

You've gotten into the mode where you just go round in circles, shifting arguments, contradicting yourself, etc. When that happens, unfortunately not much constructive can come out of the discussion.

There are many moves that SF10 does not appear to find that SFdev does. But obviously if we want to make that a serious test, we'd need to decide how long to leave SF10 calculating for. What would make you happy? SF10 at depth 70? Depth 80? 100? I don't think you and I have resources to run SF10 for that sort of depth.

[Ovyron]

Frankly, I don't think anything constructive has come from you. You don't want to learn anything, you just want to be right and me wrong.

[jp]

You take everything so personally. How can questions be trying to show I'm right and you wrong? (Right about what? I haven't even claimed anything.) I just wish you'd care more about rational arguments and evidence.

Claims deserve to be backed up by evidence. That's all. It's the same for everyone's claims. It doesn't mean those claims must be wrong. If we don't have conclusive evidence for or against them, they could be right. (I've repeatedly made that clear.) I also understand that evidence can be hard to provide, but sometimes you don't seem to try; you just act defensive.

[Ovyron]

If you agree everything I'm saying is right then we have nothing else to discuss. If you think there's something wrong then be specific. Evidence should only be provided if the other person does not believe you, but then they have to say what kind of evidence would convince them.

Otherwise all you're saying is "you can't prove your claims", but that's an empty statement. Right now you could claim a lot of true things without evidence or way to prove them and they'd remain true.

[jp]

No, the burden of evidence lies with the claim. It always does. If it's only circumstantial evidence, that's fine. Why would it be so terrible for the person just to say, "Based on this circumstantial evidence, I think this...."? Or "based on no evidence, I believe this..."?

And I'd prefer to focus on things for which we can at least get some evidence in either direction.

[Raphexon]

Also how can you be so sure if an endgame is winning if you're only using a 5 piece TB?

SF has plenty of 6 man positions where it's blind, even with a 5 piece TB. (and even at high node count)
And there are even more 7 man positions where SF is blind with a 6 piece TB.

[zullil]

Also how can you be so sure if an endgame is winning if you're only using a 5 piece TB?

SF has plenty of 6 man positions where it's blind, even with a 5 piece TB. (and even at high node count)
And there are even more 7 man positions where SF is blind with a 6 piece TB.

As someone who has been "infected" by the Ovyron virus---I'm convalescing now---I believe I can answer this for him: "I simply avoid all such positions. Somehow I can recognize all such positions, even those for which tablebases don't yet exist. And, more than that, I claim that avoiding such positions will never cost me anything, because positions I can't assess can never change a game's outcome. Trust me, I'm a centaur!"

[jp]

4rnk1/4rpp1/1npq4/3p1QNp/1P1P3P/4PP2/2N2K2/6RR w - -

Komodo 13.2, depth 28, 29, [30, 31]: -1.34 (Rg2), -1.34, [-1.48, -1.50].

Komodo's eval is starting to move!