| View previous topic :: View next topic |
| Author |
Message |
Ron Langeveld
Joined: 05 Jan 2010 Posts: 47
|
Post subject: Re: Houdini 3 Posted: Sat Mar 10, 2012 8:54 am |
|
|
| lkaufman wrote: |
| M ANSARI wrote: |
| lkaufman wrote: |
| M ANSARI wrote: |
| I this entire debate about scaling of Komodo vs. Houdini is pointless, rather than handicap all engines to play on one core so that they can compete with Komodo, it makes a lot of sense to make Komodo MP to compete with other engines. Time is a constant in computers, and this constant gets exponentially improved with each new generation of hardware. But I will be first to admit that some engines perform dramatically better at very short time controls than others, and yes the Ippolit derivatives are extremely strong at very fast blitz times. I don't know why that is so, but I always had the belief that in the case of Rybka, it has a lot of overhead in initializing start of processes (as it is a process based engine). These overheads are in mili seconds, but they can quickly become prominent when time runs down and games are very fast 1 0 or 2 0. |
Start-up times may be a factor for Rybka at super-fast speeds, but for Komodo and all of the other top programs these are not a significant factor unless you are talking about something like game in one second. They do not play a role in explaining why Komodo and Stockfish are weaker than say Ivanhoe at bullet (1 min) chess but stronger at blitz (5 min) and much stronger at longer time limits. We will have our MP soon, but because so many people insist on judging engines by very fast games, we really need to figure out precisely why we cannot compete at these bullet speeds, in order to fix the problem if possible. It is almost surely some detail about the search that Ippo uses and that SF and Komodo do not use, but which detail is responsible I cannot guess, as we have tried almost every Ippo idea, mostly without success. |
I am sure there are other reasons, the fact is that there are some evaluation values that simply don't work at fast time control. A good example in Houdini 1.5 would be the overly high evaluation of a queen vs pieces. Just like in human chess, strong tactical players like to have a queen against weaker players because there are many more dynamic reasons to go wrong. But when you have plenty of time to see the tactical pitfalls, a queen can quickly get over extended and the pieces can get coordinated to quickly snuff out the advantage of the queen and thus the "real" evaluation wins. I remember seeing this when beta testing R4 against H 1.5 and I saw many games that confirmed this. |
I believe that what you say is indeed a contributing factor, but in my opinion the observed disparity in relative strength between bullet and blitz is way too large to be explained by such eval factors. It must be something in the search that is responsible. By this I mean something that Ippo (and Houdini) does which Komodo and Stockfish do not do. But as we've tried pretty much every search idea in Ippo and nothing helps the problem, it remains a mystery. |
Larry, maybe the reason for this discrepancy in strength is much simpler than you think. I cannot deduct that my idea is right but I have a feeling that it is just related to the shape of the search tree in relation to time or reflection time. I like to presume that a shallow/thin tree is best for elo provided this tree is 'correct' from a PV point of view. I also presume that no engine does what I would call 'permanent pruning', but that temporary pruned moves can resurface at a later iteration due to longer reflection time. I belief that engines with a shallow/thin search tree could spot such temporary pruned (better) moves and shift the tree (in time) without actually changing its 'form' because it stays thin. This could explain that the better move is eventually spotted given enough time. This better move normally is not the first (and only) move of the PV, but can also be pinpointed due to a second better move down the line that was pruned initially. Spotting this type of dependency (coherence) between moves can be a strength or weakness of an engine but is also influenced by the general tree shape. If it is a strength and the tree is thin and deep then you have a very strong engine. Now please compare Komodo with Houdini for example and you will notice that on average the single thread Komodo after prolonged time reaches the same depth as Houdini running three threads in that same time. This is a clear indication that Komodo's search tree is thin compared to Houdini's and that the latter suffers less from initially pruned moves. This should be more prominent when reflection time is fast.
Another more subjective way of formulating this principle would be to look at the 'quality' of the given PV line. If just the first few moves are 'correct' the PV line is 'flawed' even though the first move is the correct one. It's a matter of time though before you run into examples where the second best move is chosen due to inferior moves in the PV. Coherence in the chain is crucial, especially in long chains.
Best wishes,
Ron |
|
| Back to top |
|
 |
|
| Subject |
Author |
Date/Time |
Houdini 3 |
Robert Houdart |
Thu Mar 01, 2012 4:37 pm |
Re: Houdini 3 |
Bram Mourik |
Thu Mar 01, 2012 5:07 pm |
Re: Houdini 3 |
Robert Houdart |
Thu Mar 01, 2012 5:17 pm |
Re: Houdini 3 |
Alfonso Turri |
Thu Mar 01, 2012 7:43 pm |
Re: Houdini 3 |
Dr.Wael Deeb |
Thu Mar 01, 2012 7:55 pm |
Re: Houdini 3 |
Adam Hair |
Thu Mar 01, 2012 7:55 pm |
Re: Houdini 3 |
Robert Houdart |
Thu Mar 01, 2012 8:00 pm |
Re: Houdini 3 |
Ignacio Garcia |
Thu Mar 01, 2012 8:37 pm |
Re: Houdini 3 |
Robert Houdart |
Thu Mar 01, 2012 9:18 pm |
Re: Houdini 3 |
Robert Flesher |
Sat Mar 03, 2012 12:23 am |
Re: Houdini 3 |
Robert Houdart |
Sat Mar 03, 2012 1:25 am |
Re: Houdini 3 |
Mel Cooper |
Thu Mar 01, 2012 8:39 pm |
Re: Houdini 3 |
Robert Houdart |
Thu Mar 01, 2012 9:41 pm |
Re: Houdini 3 |
Jimmy Huggins |
Fri Mar 02, 2012 6:57 am |
Re: Houdini 3 |
Robert Houdart |
Fri Mar 02, 2012 10:24 am |
Re: Houdini 3 |
Larry Kaufman |
Fri Mar 02, 2012 11:02 pm |
Re: Houdini 3 |
Richard Vida |
Sat Mar 03, 2012 2:47 am |
Re: Houdini 3 |
Dr.Wael Deeb |
Sat Mar 03, 2012 9:05 am |
Re: Houdini 3 |
George Speight |
Sat Mar 03, 2012 9:45 am |
Re: Houdini 3 |
Larry Kaufman |
Sat Mar 03, 2012 6:05 pm |
Re: Houdini 3 |
Maurizio Maglio |
Sat Mar 03, 2012 6:12 pm |
Re: Houdini 3 |
Larry Kaufman |
Sat Mar 03, 2012 7:09 pm |
Re: Houdini 3 |
Thorsten Czub |
Sat Mar 03, 2012 7:58 pm |
Re: Houdini 3 |
Thorsten Czub |
Sat Mar 03, 2012 7:40 pm |
Re: Houdini 3 |
Larry Kaufman |
Sat Mar 03, 2012 7:56 pm |
Re: Houdini 3 |
Eelco de Groot |
Sat Mar 03, 2012 8:06 pm |
Re: Houdini 3 |
Carl Bicknell |
Sat Mar 03, 2012 8:25 pm |
Re: Houdini 3 |
Eelco de Groot |
Sat Mar 03, 2012 8:43 pm |
Re: Houdini 3 |
Thorsten Czub |
Sat Mar 03, 2012 8:30 pm |
Re: Houdini 3 |
Maurizio Maglio |
Sat Mar 03, 2012 8:50 pm |
Re: Houdini 3 |
Uri Blass |
Sat Mar 03, 2012 9:05 pm |
Re: Houdini 3 |
Mike Scheidl |
Sat Mar 03, 2012 9:10 pm |
Re: Houdini 3 |
Uri Blass |
Sat Mar 03, 2012 9:24 pm |
Re: Houdini 3 |
Mike Scheidl |
Sat Mar 03, 2012 9:33 pm |
Re: Houdini 3 |
Thorsten Czub |
Sat Mar 03, 2012 10:13 pm |
Re: Houdini 3 |
Uri Blass |
Sat Mar 03, 2012 9:38 pm |
Re: Houdini 3 |
Thorsten Czub |
Sat Mar 03, 2012 10:05 pm |
Re: Houdini 3 |
Mike Scheidl |
Sat Mar 03, 2012 10:44 pm |
Re: Houdini 3 |
Larry Kaufman |
Sun Mar 04, 2012 4:21 am |
Re: Houdini 3 |
Maurizio Maglio |
Sun Mar 04, 2012 10:43 am |
Re: Houdini 3 |
Larry Kaufman |
Tue Mar 06, 2012 4:11 am |
Re: Houdini 3 |
Bigler |
Tue Mar 06, 2012 6:23 am |
Re: Houdini 3 |
Larry Kaufman |
Tue Mar 06, 2012 3:37 pm |
Re: Houdini 3 |
Bram Mourik |
Tue Mar 06, 2012 6:05 pm |
Re: Houdini 3 |
Larry Kaufman |
Tue Mar 06, 2012 6:19 pm |
Re: Houdini 3 |
Sedat Canbaz |
Tue Mar 06, 2012 6:48 pm |
Re: Houdini 3 |
Larry Kaufman |
Tue Mar 06, 2012 6:54 pm |
Re: Houdini 3 |
Sedat Canbaz |
Tue Mar 06, 2012 7:05 pm |
Re: Houdini 3 |
Maurizio Maglio |
Tue Mar 06, 2012 10:04 pm |
Re: Houdini 3 |
Larry Kaufman |
Tue Mar 06, 2012 10:42 pm |
Re: Houdini 3 |
Bigler |
Thu Mar 08, 2012 10:36 am |
Re: Houdini 3 |
George Speight |
Tue Mar 06, 2012 8:05 pm |
Re: Houdini 3 |
Uri Blass |
Tue Mar 06, 2012 8:41 pm |
Re: Houdini 3 |
Larry Kaufman |
Tue Mar 06, 2012 9:06 pm |
Re: Houdini 3 |
George Speight |
Tue Mar 06, 2012 11:18 pm |
Re: Houdini 3 |
George Speight |
Wed Mar 07, 2012 12:32 am |
Re: Houdini 3 |
Robert Flesher |
Wed Mar 07, 2012 1:18 am |
Re: Houdini 3 |
Larry Kaufman |
Wed Mar 07, 2012 1:59 am |
Re: Houdini 3 |
Robert Flesher |
Wed Mar 07, 2012 2:09 am |
Re: Houdini 3 |
Robert Flesher |
Wed Mar 07, 2012 12:53 am |
Re: Houdini 3 |
George Speight |
Wed Mar 07, 2012 8:30 am |
Re: Houdini 3 |
Robert Flesher |
Wed Mar 07, 2012 1:45 pm |
Re: Houdini 3 |
George Speight |
Thu Mar 08, 2012 9:17 pm |
Re: Houdini 3 |
Robert Flesher |
Thu Mar 08, 2012 11:34 pm |
Re: Houdini 3 |
Larry Kaufman |
Fri Mar 09, 2012 2:24 am |
Re: Houdini 3 |
Robert Flesher |
Fri Mar 09, 2012 3:14 am |
Re: Houdini 3 |
M ANSARI |
Fri Mar 09, 2012 4:15 am |
Re: Houdini 3 |
Larry Kaufman |
Fri Mar 09, 2012 4:32 am |
Re: Houdini 3 |
Timo Klaustermeyer |
Fri Mar 09, 2012 6:28 am |
Re: Houdini 3 |
Uri Blass |
Fri Mar 09, 2012 8:44 am |
Re: Houdini 3 |
Sedat Canbaz |
Fri Mar 09, 2012 11:04 am |
Re: Houdini 3 |
Timo Klaustermeyer |
Fri Mar 09, 2012 11:28 am |
Re: Houdini 3 |
Sedat Canbaz |
Fri Mar 09, 2012 12:05 pm |
Re: Houdini 3 |
Timo Klaustermeyer |
Fri Mar 09, 2012 2:30 pm |
Re: Houdini 3 |
Sedat Canbaz |
Sat Mar 10, 2012 9:54 am |
Re: Houdini 3 |
M ANSARI |
Fri Mar 30, 2012 6:20 am |
Re: Houdini 3 |
Peter Kasinski |
Fri Mar 30, 2012 11:49 am |
Re: Houdini 3 |
Sedat Canbaz |
Sun Apr 08, 2012 1:52 pm |
Re: Houdini 3 |
M ANSARI |
Sat Mar 10, 2012 4:59 am |
Re: Houdini 3 |
Larry Kaufman |
Sat Mar 10, 2012 6:33 am |
Re: Houdini 3 |
Ron Langeveld |
Sat Mar 10, 2012 8:54 am |
Re: Houdini 3 |
Vincent Diepeveen |
Thu Mar 08, 2012 7:26 am |
Re: Houdini 3 |
Thorsten Czub |
Thu Mar 08, 2012 10:32 am |
Re: Houdini 3 |
Vincent Diepeveen |
Thu Mar 08, 2012 1:06 pm |
Re: Houdini 3 |
Carl Bicknell |
Thu Mar 08, 2012 2:33 pm |
Re: Houdini 3 |
Thorsten Czub |
Thu Mar 08, 2012 2:48 pm |
Re: Houdini 3 |
S.Taylor |
Mon Mar 12, 2012 12:36 am |
Re: Houdini 3 |
Carl Bicknell |
Mon Mar 12, 2012 7:25 am |
Re: Houdini 3 |
Thorsten Czub |
Mon Mar 12, 2012 6:02 pm |
Re: Houdini 3 |
Thorsten Czub |
Fri Mar 30, 2012 12:28 pm |
Re: Houdini 3 |
Sedat Canbaz |
Thu Mar 08, 2012 7:40 pm |
Re: Houdini 3 |
Ed Schroder |
Fri Mar 30, 2012 12:47 pm |
Re: Houdini 3 |
Sedat Canbaz |
Sun Apr 08, 2012 2:02 pm |
Re: Houdini 3 |
Uri Blass |
Sat Mar 03, 2012 8:30 pm |
Re: Houdini 3 |
Uri Blass |
Sat Mar 03, 2012 8:49 pm |
Re: Houdini 3 |
Sedat Canbaz |
Sun Mar 04, 2012 2:11 am |
Re: Houdini 3 |
Uri Blass |
Sat Mar 03, 2012 10:26 am |
Re: Houdini 3 |
Marijan Stipancic jr |
Sat Mar 03, 2012 10:38 am |
Re: Houdini 3 |
Marek Kwiatkowski |
Sat Mar 03, 2012 11:45 am |
Re: Houdini 3 |
Dr.Wael Deeb |
Sun Mar 04, 2012 8:41 am |
Re: Houdini 3 |
Marijan Stipancic jr |
Sun Mar 04, 2012 10:38 pm |
Re: Houdini 3 |
Dr.Wael Deeb |
Mon Mar 05, 2012 9:27 am |
Re: Houdini 3 |
S.Taylor |
Sun Mar 04, 2012 1:28 am |
Re: Houdini 3 |
Maurizio Maglio |
Sun Mar 04, 2012 11:33 am |
Re: Houdini 3 |
Vael Jean-Paul |
Sun Mar 04, 2012 2:02 pm |
Re: Houdini 3 |
Maurizio Maglio |
Sun Mar 04, 2012 2:16 pm |
Re: Houdini 3 |
Sedat Canbaz |
Sun Mar 04, 2012 5:17 pm |
Re: Houdini 3 |
Maurizio Maglio |
Sun Mar 04, 2012 5:32 pm |
Re: Houdini 3 |
Dr.Wael Deeb |
Sun Mar 04, 2012 5:46 pm |
Re: Houdini 3 |
Sedat Canbaz |
Sun Mar 04, 2012 10:45 pm |
Re: Houdini 3 |
Dr.Wael Deeb |
Mon Mar 05, 2012 9:30 am |
Re: Houdini 3 |
Joel Martus |
Mon Mar 05, 2012 4:12 pm |
Re: Houdini 3 |
Dr.Wael Deeb |
Mon Mar 05, 2012 4:22 pm |
Re: Houdini 3 |
Sedat Canbaz |
Sun Mar 04, 2012 6:23 pm |
Re: Houdini 3 |
Maurizio Maglio |
Sun Mar 04, 2012 6:49 pm |
Re: Houdini 3 |
Sedat Canbaz |
Tue Mar 06, 2012 8:00 am |
Re: Houdini 3 |
Ben-Hur Carlos Langoni |
Thu Mar 01, 2012 9:34 pm |
Re: Houdini 3 |
Robert Houdart |
Thu Mar 01, 2012 9:47 pm |
Re: Houdini 3 |
Teodoriu Catalin |
Thu Mar 01, 2012 9:46 pm |
Re: Houdini 3 |
Robert Houdart |
Thu Mar 01, 2012 9:52 pm |
Re: Houdini 3 |
Nelson Hernandez |
Thu Mar 01, 2012 11:51 pm |
Re: Houdini 3 |
Teodoriu Catalin |
Fri Mar 02, 2012 8:45 am |
Re: Houdini 3 |
Nelson Hernandez |
Fri Mar 02, 2012 2:34 pm |
Re: Houdini 3 |
Teodoriu Catalin |
Fri Mar 02, 2012 5:10 pm |
Re: Houdini 3 |
Marco Costalba |
Sat Mar 03, 2012 9:18 am |
Re: Houdini 3 |
Bram Mourik |
Sat Mar 03, 2012 9:53 am |
Re: Houdini 3 |
Ed Schroder |
Sat Mar 03, 2012 1:04 pm |
Re: Houdini 3 |
Marco Costalba |
Sat Mar 03, 2012 5:18 pm |
Re: Houdini 3 |
Miguel A. Ballicora |
Sat Mar 03, 2012 6:07 pm |
Re: Houdini 3 |
Ben-Hur Carlos Langoni |
Sat Mar 03, 2012 8:28 pm |
Re: Houdini 3 |
George Speight |
Mon Mar 12, 2012 10:02 am |
Re: Houdini 3 |
fernando |
Sun Apr 08, 2012 3:54 pm |
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|