Dylan Sharp Vs. Harvey Williamson (G4)

Discussion of computer chess matches and engine tournaments.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
User avatar
Ovyron
Posts: 2746
Joined: Tue Jul 03, 2007 2:30 am

Re: Dylan Sharp Vs. Harvey Williamson (G4)

Post by Ovyron » Thu Oct 31, 2019 6:58 am

carldaman wrote:
Thu Oct 31, 2019 5:05 am
It was only a matter of time before this contest imploded. :|
It seems like they always do.

(they = forum games)
Huh? This is the very first time a game had to become private because of audience disruption.
I've been playing correspondence chess games in forums since 2008 (as Vytron or Uly), and it took me 11 years to find someone like Zullil (which doesn't have respect for one of the players and would refuse to provide evidence that he's actually doing what he claims after the game is over.) I don't think Zullil deserves to see my moves as they happen, so nobody else gets to see them.

maxdeg
Posts: 191
Joined: Thu Mar 10, 2016 5:17 pm
Location: Canada

Re: Dylan Sharp Vs. Harvey Williamson (G4)

Post by maxdeg » Thu Oct 31, 2019 8:45 am

Ovyron wrote:
Wed Oct 30, 2019 8:29 pm
The game will continue by private message, we'll come back with the full PGN of the rest of the moves after the game ends.

Thanks everyone for watching.
Thats NOT nice .. :D :mrgreen:
The mob got to you right ?
An update every few moves would be nice for your fans .. :D :)

zullil
Posts: 5670
Joined: Mon Jan 08, 2007 11:31 pm
Location: PA USA
Full name: Louis Zulli

Re: Dylan Sharp Vs. Harvey Williamson (G4)

Post by zullil » Thu Oct 31, 2019 8:50 am

Ovyron wrote:
Thu Oct 31, 2019 6:58 am
carldaman wrote:
Thu Oct 31, 2019 5:05 am
It was only a matter of time before this contest imploded. :|
It seems like they always do.

(they = forum games)
Huh? This is the very first time a game had to become private because of audience disruption.
I've been playing correspondence chess games in forums since 2008 (as Vytron or Uly), and it took me 11 years to find someone like Zullil (which doesn't have respect for one of the players and would refuse to provide evidence that he's actually doing what he claims after the game is over.) I don't think Zullil deserves to see my moves as they happen, so nobody else gets to see them.
I'm happy to stop kibitzing and posting evals, if that has distressed you.

About what I've posted and claimed, why on earth would I be fabricating anything? :shock: And why do you feel that I later need to provide you with Stockfish's deep analysis? You've repeatedly implied that my 20-core machine is an expensive waste, and that your hardware and methods are sufficient.

In any case, I'm sorry that my posts have bothered you. Had I realized that, I would have stopped upon request.

I actually left the search running, and was a bit surprised to see that depth 78 is still not resolved. Once that happens, I'm happy to send a screenshot, including the current PV, to a trusted third-party, as evidence that I've done exactly what I've claimed. Maybe Dann would be interested in it?

User avatar
Ovyron
Posts: 2746
Joined: Tue Jul 03, 2007 2:30 am

Re: Dylan Sharp Vs. Harvey Williamson (G4)

Post by Ovyron » Thu Oct 31, 2019 10:37 am

(repeated post)
Last edited by Ovyron on Thu Oct 31, 2019 10:41 am, edited 1 time in total.

User avatar
Ovyron
Posts: 2746
Joined: Tue Jul 03, 2007 2:30 am

Re: Dylan Sharp Vs. Harvey Williamson (G4)

Post by Ovyron » Thu Oct 31, 2019 10:37 am

zullil wrote:
Thu Oct 31, 2019 8:50 am
In any case, I'm sorry that my posts have bothered you. Had I realized that, I would have stopped upon request.
Really? Have you forgotten my post where I requested you to stop? It's here:
Ovyron wrote:
Sat Oct 26, 2019 12:04 pm
zullil wrote:
Sat Oct 26, 2019 10:15 am
By the way, I don't understand your interest in tails of PVs or "precise" evaluations. Whatever Stockfish is looking at 80-120 plies from the root, it's not changing the first dozen moves of the PV. Stockfish is a chess engine. It's designed to play a "best" move in the current root position. I'm content to just let it do its thing.
I'm now going to claim that I don't believe you :)

It's very easy to make up scores about some supposed Stockfish dev high depth and claim it has found all the moves we've been playing unassisted. Storing the PVs and showing them after the game ends or whatever would work as proof that you have actually been doing something, but they have been conveniently discarded by you, who supposedly runs the engine all that time and only nods that the move matches with whatever we play every time, and that our moves have been in the PV all along?? Come on!

So stop with the lies or I'm going to start reporting your posts to moderation because all you're causing is game disruption, who knows if you're reaching Depth 40 and fudging the score or who knows what. We can't know because you discard the whole point of high depth :mrgreen:
So, no, you didn't stop.
zullil wrote:
Thu Oct 31, 2019 8:50 am
I actually left the search running, and was a bit surprised to see that depth 78 is still not resolved. Once that happens, I'm happy to send a screenshot, including the current PV, to a trusted third-party, as evidence that I've done exactly what I've claimed. Maybe Dann would be interested in it?
Was it so hard to keep screenshots and show them after the game was over? That was my entire problem with what you were doing, that I was allowing you to see my moves for free, and you were going to keep your analysis secret, and worse than that, you were discarding it.

So now you get nothing. And least, not while it matters.

Sure, go ahead and send the screenshot to Dann, everybody else that wanted to see the screenshot will be screwed over. That's what I find bothersome, that the only way for you to show anything (and to a single random person of the audience, at that) was to make the game private, otherwise, you had no reason to prove that you're not just lying, and you wouldn't have done so.

All your posts on this game have been mean-spirited, only to attack Harvey and me, because we are incapable of playing an "original line" because all our moves can be predicted by your Stockfish. That's what has been your entire objective all along, and nothing else.

But if that was true, and you were able to predict all our moves, you'd be able to predict the outcome of the game. Yet, you have NOTHING. Your score and PV at Depth 76 TELLS YOU NOTHING. You have no idea if I hold the draw or Harvey wins. You have a prediction that the game ends at Black's 63rd with an unresolved result, and its tail of that PV is garbage, and nobody would be able to tell at what point that tail begins because you weren't going to save any line.

Making the game private makes it impossible for you to continue discrediting Harvey's and me's analysis methods, you can no longer accuse us of being unoriginal because your Stockfish plays the same move, and for me, that's a small victory.

zullil
Posts: 5670
Joined: Mon Jan 08, 2007 11:31 pm
Location: PA USA
Full name: Louis Zulli

Re: Dylan Sharp Vs. Harvey Williamson (G4)

Post by zullil » Thu Oct 31, 2019 11:01 am

Ovyron wrote:
Thu Oct 31, 2019 10:37 am
zullil wrote:
Thu Oct 31, 2019 8:50 am
In any case, I'm sorry that my posts have bothered you. Had I realized that, I would have stopped upon request.
Really? Have you forgotten my post where I requested you to stop? It's here:
Ovyron wrote:
Sat Oct 26, 2019 12:04 pm
zullil wrote:
Sat Oct 26, 2019 10:15 am
By the way, I don't understand your interest in tails of PVs or "precise" evaluations. Whatever Stockfish is looking at 80-120 plies from the root, it's not changing the first dozen moves of the PV. Stockfish is a chess engine. It's designed to play a "best" move in the current root position. I'm content to just let it do its thing.
I'm now going to claim that I don't believe you :)

It's very easy to make up scores about some supposed Stockfish dev high depth and claim it has found all the moves we've been playing unassisted. Storing the PVs and showing them after the game ends or whatever would work as proof that you have actually been doing something, but they have been conveniently discarded by you, who supposedly runs the engine all that time and only nods that the move matches with whatever we play every time, and that our moves have been in the PV all along?? Come on!

So stop with the lies or I'm going to start reporting your posts to moderation because all you're causing is game disruption, who knows if you're reaching Depth 40 and fudging the score or who knows what. We can't know because you discard the whole point of high depth :mrgreen:
So, no, you didn't stop.
zullil wrote:
Thu Oct 31, 2019 8:50 am
I actually left the search running, and was a bit surprised to see that depth 78 is still not resolved. Once that happens, I'm happy to send a screenshot, including the current PV, to a trusted third-party, as evidence that I've done exactly what I've claimed. Maybe Dann would be interested in it?
Was it so hard to keep screenshots and show them after the game was over? That was my entire problem with what you were doing, that I was allowing you to see my moves for free, and you were going to keep your analysis secret, and worse than that, you were discarding it.

So now you get nothing. And least, not while it matters.

Sure, go ahead and send the screenshot to Dann, everybody else that wanted to see the screenshot will be screwed over. That's what I find bothersome, that the only way for you to show anything (and to a single random person of the audience, at that) was to make the game private, otherwise, you had no reason to prove that you're not just lying, and you wouldn't have done so.

All your posts on this game have been mean-spirited, only to attack Harvey and me, because we are incapable of playing an "original line" because all our moves can be predicted by your Stockfish. That's what has been your entire objective all along, and nothing else.

But if that was true, and you were able to predict all our moves, you'd be able to predict the outcome of the game. Yet, you have NOTHING. Your score and PV at Depth 76 TELLS YOU NOTHING. You have no idea if I hold the draw or Harvey wins. You have a prediction that the game ends at Black's 63rd with an unresolved result, and its tail of that PV is garbage, and nobody would be able to tell at what point that tail begins because you weren't going to save any line.

Making the game private makes it impossible for you to continue discrediting Harvey's and me's analysis methods, you can no longer accuse us of being unoriginal because your Stockfish plays the same move, and for me, that's a small victory.
Don't even know how to respond to this, so I won't try. Sorry that you're so upset.

It certainly seems to me that you really had no idea how strong unassisted Stockfish is on good hardware. That said, I don't think anything played so far would be much of a challenge for any strong engine even on moderate hardware. Maybe I'm wrong.

Good luck with the game. For what it's worth, I think you can still hold the draw.

User avatar
Ovyron
Posts: 2746
Joined: Tue Jul 03, 2007 2:30 am

Re: Dylan Sharp Vs. Harvey Williamson (G4)

Post by Ovyron » Thu Oct 31, 2019 11:30 am

zullil wrote:
Thu Oct 31, 2019 11:01 am
It certainly seems to me that you really had no idea how strong unassisted Stockfish is on good hardware.
I'm surprised at how weak it is. I was able to build lines with Stockfish Depth 24 and it predicted we'd reach our current position, and it didn't take me much time to do that (less than a day.) I'd have expected Depth 60 to improve over that at some point, given how long it'd take to reach it on my machine (what is it, 10 days?) Alas, it provided nothing new, nothing that would make it running worth it.
zullil wrote:
Thu Oct 31, 2019 11:01 am
That said, I don't think anything played so far would be much of a challenge for any strong engine even on moderate hardware. Maybe I'm wrong.
You have no idea what you're talking about, and you will never do until you start playing games and your puny unassisted Stockfish gets defeated. Specially by an opponent that knows it's playing against it, and can predict its next dozen depth 60 moves in less than a day, without ever needing to go beyond depth 24.

jp
Posts: 825
Joined: Mon Apr 23, 2018 5:54 am

Re: Dylan Sharp Vs. Harvey Williamson (G4)

Post by jp » Thu Oct 31, 2019 1:42 pm

Ovyron wrote:
Thu Oct 31, 2019 11:30 am
I'm surprised at how weak it is. I was able to build lines with Stockfish Depth 24 and it predicted we'd reach our current position, and it didn't take me much time to do that (less than a day.) I'd have expected Depth 60 to improve over that at some point, given how long it'd take to reach it on my machine (what is it, 10 days?) Alas, it provided nothing new, nothing that would make it running worth it.
This is making huge assumptions about how much an engine "should" change its best move with increasing depth.

Maybe if you built the lines with Komodo depth 24 the argument wouldn't sound so odd.

Raphexon
Posts: 99
Joined: Sun Mar 17, 2019 11:00 am
Full name: Henk Drost

Re: Dylan Sharp Vs. Harvey Williamson (G4)

Post by Raphexon » Thu Oct 31, 2019 1:46 pm

It's honestly fucking annoying (besides cluttering everything) when somebody constantly reports his eval, and egoizes that both players are following SF lines.

User avatar
Ovyron
Posts: 2746
Joined: Tue Jul 03, 2007 2:30 am

Re: Dylan Sharp Vs. Harvey Williamson (G4)

Post by Ovyron » Thu Oct 31, 2019 2:03 pm

jp wrote:
Thu Oct 31, 2019 1:42 pm
Maybe if you built the lines with Komodo depth 24 the argument wouldn't sound so odd.
For the record, I made the same experiment with Komodo 11.3.1 at MultiPV=4 at Depth 25, and it couldn't hold, it was very easy to refute all its lines and see nothing but lines were black won. The same with Houdini 6 at Depth 21. Among others. And also with Stockfish MultiPV=4 at Depth 23, which is surprisingly weak and again provides nothing of value. I had expected a gradual jump, but Depth 24 remains undefeated to this day (within 24 takebacks).

Worth noting is that Stockfish MultiPV 4 Depth 24 easily crumbled after 2.Bg2, 2.Nf3, 2.h3, 2.e3 and 2.d4; 2.g5 was the only "unbeatable" line, so if Harvey can beat me I could become a believer that 1.g4 is a mate in X position, because alternatives to 2.g5 are much easier to beat.

Post Reply