Strelka 5.5 vs Houdini & Critter

[LudiBuda]

Strelka 5.5 seems to be the strongest 1CPU engine in the world.

TC is 40 moves in 30sec repeating:

Strelka 5.5 vs Houdini 1.5 : 73 - 55 with 112 draws (240 games)

Strelka 5.5 vs Critter 1.4 : 48 - 36 with 115 draws (199 games)


Very strong in opening and middlegame. Has room for improvements in the endgame.

[Jouni]

In my short test under Fritz 12 GUI (no ponder) Strelka lost most games against Komodo 3 by bad time control - the old Houdini/Ivanhoe bug again

[LudiBuda]

I used CuteChess to run the matches.

Houdini lost 2 games on time in the first match.
Strelka lost 2 games on time (in a drawn position) in the second match.

[lkaufman]

Strelka 5.5 seems to be the strongest 1CPU engine in the world.

TC is 40 moves in 30sec repeating:

Strelka 5.5 vs Houdini 1.5 : 73 - 55 with 112 draws (240 games)

Strelka 5.5 vs Critter 1.4 : 48 - 36 with 115 draws (199 games)


Very strong in opening and middlegame. Has room for improvements in the endgame.

The time limit you cite is far too fast to justify any conclusion about which is the strongest 1 cpu unit in the world. Also, at the time limit you are using, you need to run Houdini 2.0 rather than 1.5, because it is well established that Houdini 2.0 is much stronger than 1.5 at bullet chess, just not at normal chess. If Strelka 5.5 can convincingly defeat Houdini 2.0 at any time limit, it will at least demonstrate that it is not basically just a Houdini clone. Also, you don't mention this, but presumably these are all just 32 bit versions, so until Strelka 64 bit is released your conclusion is wrong for yet another reason.

[Houdini]

The time limit you cite is far too fast to justify any conclusion about which is the strongest 1 cpu unit in the world. Also, at the time limit you are using, you need to run Houdini 2.0 rather than 1.5, because it is well established that Houdini 2.0 is much stronger than 1.5 at bullet chess, just not at normal chess. If Strelka 5.5 can convincingly defeat Houdini 2.0 at any time limit, it will at least demonstrate that it is not basically just a Houdini clone. Also, you don't mention this, but presumably these are all just 32 bit versions, so until Strelka 64 bit is released your conclusion is wrong for yet another reason.

Larry, below some facts to enlighten your confusion:

1) At ultra-fast TC the difference in strength between Houdini 2 and Houdini 1.5 is exactly 24 Elo +/- 2 Elo.

2) Strelka 5.5 does not "convincingly defeat Houdini 2.0", in fact it is 10 to 20 Elo weaker than Houdini 1.5a. See the other thread about Strelka 5.5 in the "General Topics" Section.

3) Strelka 5 eval and search are 99% identical to Houdini 1.5a, but all the surrounding code (Time Management, UCI input/output handling etc) appears to be Ivanhoe-inspired or original.

4) The whole Strelka 5.5 business, with a cheat table containing 30,000 test positions, appears to be a late April Fools joke.

Robert

[lkaufman]

The time limit you cite is far too fast to justify any conclusion about which is the strongest 1 cpu unit in the world. Also, at the time limit you are using, you need to run Houdini 2.0 rather than 1.5, because it is well established that Houdini 2.0 is much stronger than 1.5 at bullet chess, just not at normal chess. If Strelka 5.5 can convincingly defeat Houdini 2.0 at any time limit, it will at least demonstrate that it is not basically just a Houdini clone. Also, you don't mention this, but presumably these are all just 32 bit versions, so until Strelka 64 bit is released your conclusion is wrong for yet another reason.

Larry, below some facts to enlighten your confusion:

1) At ultra-fast TC the difference in strength between Houdini 2 and Houdini 1.5 is exactly 24 Elo +/- 2 Elo.

2) Strelka 5.5 does not "convincingly defeat Houdini 2.0", in fact it is 10 to 20 Elo weaker than Houdini 1.5a. See the other thread about Strelka 5.5 in the "General Topics" Section.

3) Strelka 5 eval and search are 99% identical to Houdini 1.5a, but all the surrounding code (Time Management, UCI input/output handling etc) appears to be Ivanhoe-inspired or original.

4) The whole Strelka 5.5 business, with a cheat table containing 30,000 test positions, appears to be a late April Fools joke.

Robert

I wasn't confused, in fact for once I was on your side. My point was that Houdini 2 is stronger than 1.5 at game in a minute or faster, just as you say, so his test was invalid for determining the strongest 32 bit single core engine even at bullet chess. I don't doubt your statement that Strelka eval and search are 99% Houdini, but why in your opinion is Strelka 5.1 apparently a bit weaker than Houdini 1.5, just time-control issues?
The existence of free near-clones of top programs hurts both of us, so we have a common interest here. I imagine that Strelka's author felt justified by the fact that you did the same to Ippolit originally, but two wrongs don't make a right.

[LudiBuda]

Ok, so to be perfectly correct, Strelka 5.5 is the strongest engine under my test conditions. Good enough?
I don't see why Strelka has to be the Houdini clone. Houdini 1.5 is just slightly modified Robbolito. I disaassembled its search and found only minor changes compared to the Robbolito 0.085
If Strelka author also started from the Robbolito base, he will end up with the engine very similar to Houdini.

[Dr.Wael Deeb]

The time limit you cite is far too fast to justify any conclusion about which is the strongest 1 cpu unit in the world. Also, at the time limit you are using, you need to run Houdini 2.0 rather than 1.5, because it is well established that Houdini 2.0 is much stronger than 1.5 at bullet chess, just not at normal chess. If Strelka 5.5 can convincingly defeat Houdini 2.0 at any time limit, it will at least demonstrate that it is not basically just a Houdini clone. Also, you don't mention this, but presumably these are all just 32 bit versions, so until Strelka 64 bit is released your conclusion is wrong for yet another reason.

Larry, below some facts to enlighten your confusion:

1) At ultra-fast TC the difference in strength between Houdini 2 and Houdini 1.5 is exactly 24 Elo +/- 2 Elo.

2) Strelka 5.5 does not "convincingly defeat Houdini 2.0", in fact it is 10 to 20 Elo weaker than Houdini 1.5a. See the other thread about Strelka 5.5 in the "General Topics" Section.

3) Strelka 5 eval and search are 99% identical to Houdini 1.5a, but all the surrounding code (Time Management, UCI input/output handling etc) appears to be Ivanhoe-inspired or original.

4) The whole Strelka 5.5 business, with a cheat table containing 30,000 test positions, appears to be a late April Fools joke.

Robert

I wasn't confused, in fact for once I was on your side. My point was that Houdini 2 is stronger than 1.5 at game in a minute or faster, just as you say, so his test was invalid for determining the strongest 32 bit single core engine even at bullet chess. I don't doubt your statement that Strelka eval and search are 99% Houdini, but why in your opinion is Strelka 5.1 apparently a bit weaker than Houdini 1.5, just time-control issues?
The existence of free near-clones of top programs hurts both of us, so we have a common interest here. I imagine that Strelka's author felt justified by the fact that you did the same to Ippolit originally, but two wrongs don't make a right.

I just happen to love computer chess

[LudiBuda]

It's like Spanish soap opera, isn't it

[lkaufman]

Ok, so to be perfectly correct, Strelka 5.5 is the strongest engine under my test conditions. Good enough?
I don't see why Strelka has to be the Houdini clone. Houdini 1.5 is just slightly modified Robbolito. I disaassembled its search and found only minor changes compared to the Robbolito 0.085
If Strelka author also started from the Robbolito base, he will end up with the engine very similar to Houdini.

That's an interesting observation, but Strelka 5.5 seems to be much closer to Houdini 1.5 in strength than either are to Robbolito. Since you have disassembled Houdini search, can you comment on those differences you found that might account for the huge strength increase from Robbolito to Houdini 1.5? Critter author Richard Vida did so about a year ago, but the differences he cited were not enough to account for the large strength increase, in my opinion. Perhaps his list was incomplete, or else the gains came more from eval, I don't know.