michiguel wrote:The purpose of TCEC was not about testing engines and statistic accuracy, it was about chess entertainment and competition. Games have been played one at a time so the spectators can follow, comment, chat etc. It had been a lot of fun.
it seemed to me to become like a forbidden thing to accidentally show that you were getting any pleasure from the incidental testing that was the mechanism of the entertainment, or even to give estimations which engine seemed stronger or not due to any of those games.
It felt like one mustn't be racist and that all engines were meant to be imagined to be equal!
You wrote somewhere that this is not for statistical data but for entertainment. Two Arguments for Ponder ON:
Part of the entertainment is to see what the oppoenent is doing, what it is expecting and if it already sees the problem. For entertainment a ply more or less is irrelevant.
No Tourney (computer or human) in the world where people are facing is played ponder off as thinking while it is the opponents move is the normal way to play chess - that is the nature of the game. Ponder off is used to make more games for better statistical results and nothing else.
I really hope you go for Ponder on.
Btw: I cant vote. It show that the poll is not expired and shows a result even if I havent voted (maybe someone hacked my account and voted for me ...)
Graham Banks wrote:
I wish him the best with that, because I've found that not many spectators tend to chat with each other when they've been watching the live broadcasts I've been running. Fingers crossed.
The couple of times I watched on chessbomb there were a lot of spectators chatting including the authors. But I'm not sure about the new TCEC in the short term - back then not that many people had overclocked 6 cores and they were watching very high quality chess at long-ish time control on fast hardware they did not have themselves. Now with 3 cores, most people have that at home and can watch their own games, assuming that they have the engines. It will be very interesting to see the level of interest. Of course Martin says better hardware will be coming.
You wrote somewhere that this is not for statistical data but for entertainment. Two Arguments for Ponder ON:
Part of the entertainment is to see what the oppoenent is doing, what it is expecting and if it already sees the problem. For entertainment a ply more or less is irrelevant.
No Tourney (computer or human) in the world where people are facing is played ponder off as thinking while it is the opponents move is the normal way to play chess - that is the nature of the game. Ponder off is used to make more games for better statistical results and nothing else.
I really hope you go for Ponder on.
Btw: I cant vote. It show that the poll is not expired and shows a result even if I havent voted (maybe someone hacked my account and voted for me ...)
Bye
Ingo
Hi Ingo,
Thanks for your input. I don't know if you ever followed TCEC while it was going on, but the tournaments are only for entertainment.
Personally, I am tempted to do ponder but I do not like the fact that at this moment the only viable option is to do it with only 1 core.
Modern Times wrote:
The couple of times I watched on chessbomb there were a lot of spectators chatting including the authors. But I'm not sure about the new TCEC in the short term - back then not that many people had overclocked 6 cores and they were watching very high quality chess at long-ish time control on fast hardware they did not have themselves. Now with 3 cores, most people have that at home and can watch their own games, assuming that they have the engines. It will be very interesting to see the level of interest. Of course Martin says better hardware will be coming.
Spot on, Ray. Those are many of my sentiments as well.
Even though people have 4 cores (or the more "bogus" 6-8 cores from AMD), one should consider the speed throughput. The latest Intel (Ivy Bridge) is still far superior to anything AMD have, clock for clock.
Why don't you turn hyper-threading on, and set affinity for engine A to cores 1, 3 and 5, and of engine B to cores 2, 4, 6? Then you can run them at 3 cores with ponder on (and still have 1 physical core, (7 and 8) to spare).
hgm wrote:Why don't you turn hyper-threading on, and set affinity for engine A to cores 1, 3 and 5, and of engine B to cores 2, 4, 6? Then you can run them at 3 cores with ponder on (and still have 1 physical core, (7 and 8) to spare).
Well, the obvious problem would be that I can't run a 24/7 broadcast if I have to set affinity for every engine loaded in every game.
Not sure how the speed would be though. I still think HT doesn't give any benefits for computer chess engines. Or have something happened in the last year?
Thanks for your input. I don't know if you ever followed TCEC while it was going on, but the tournaments are only for entertainment.
I did from time to time but do not remember if it was ponder on or off..
But I was on several live tourneys and a good part of beeing there was the "Hehe the opponent still doesn't see it"-effect, something I miss in P-OFF games and therefore I usually do not follow them. (However I check the results)
Martin Thoresen wrote:
Personally, I am tempted to do ponder but I do not like the fact that at this moment the only viable option is to do it with only 1 core.
I assume you fear that with 4 cores the comp is unstable or your transmission is using to much computational power and yes 2 cores P-ON would be of course more interesting than 1 core but again, no real tourney in the world is played Ponder Off ... and another argument is your "equal ground" reasoning. I would not call it unfair as it is a problem of the 1 core engine but it is even more boring to watch a ponder off game where one engine uses 1 core and the other is using 3 (except by chance you have 2 equaly strong engines then, but that is unintensional)
Martin Thoresen wrote:Not sure how the speed would be though. I still think HT doesn't give any benefits for computer chess engines. Or have something happened in the last year?
Well, you just pointed out that your goal is entertainment, not benefits. Pondering is even more detrimental than hyper threading. Thie way I proposed you can award engines with (better) SMP implementation.
And yes, it seems something has happened: the latest reports I have seen here claimed that Houdini on i7 would run about 20% faster nps-wise with HT on, which is just enough to compensate for the search overhead from doubling the number of threads. So it seems that HT is about neutral. But who cares whether you run 10% faster or slower? That should have no impact on entertainment value.
Finally, I thought that Polyglot did support affinity? Or has this been deprecated again? Normally, you should be able to instruct the GUI to define engine priorities and affinities. Not set them by hand...
