Komodo 4536 v Critter 1.6a- 40/40, 50-Game Match Begins!

[geots]

We have what I call the "Big 3" right under Houdini 3 and they are poised to make a run at him. I am not saying they have caught Houdini version 3, in fact I haven't broached that subject. Everything in due time- right now I am more concerned with the order these 3 engines sit in at the moment. Understanding that k-4536 is definitely a beta, so look for improvement. And it is not written in stone that Strelka 5.6 is about to be released by Yuri, either. These are interesting times- with some extremely strong engines.

At any rate, this match has just begun. Throw out the record books for this one................



Rest and sleep are important-

george

[geots]

We have what I call the "Big 3" right under Houdini 3 and they are poised to make a run at him. I am not saying they have caught Houdini version 3, in fact I haven't broached that subject. Everything in due time- right now I am more concerned with the order these 3 engines sit in at the moment. Understanding that k-4536 is definitely a beta, so look for improvement. And it is not written in stone that Strelka 5.6 is about to be released by Yuri, either. These are interesting times- with some extremely strong engines.

At any rate, this match has just begun. Throw out the record books for this one................



Rest and sleep are important-

george

Well, 12 games have been completed and the best thing I can say is that it is off to a good start. Because it is a bit early for adding anything further.






XPS 8500 Intel i5-4 True Cores
Fritz 11 gui
1CPU/64-bit
128MB hash
Bases=NONE
Ponder_Learning=OFF
Perfect 2012b.ctg w/12-move limit
40/19 Repeating (Benched to adapt to 40/40)
Match=50 games

Code: Select all

XPS-i5, 19m/40+19m/40+19m/40  3/5/2013  

                           
Komodo 4536.00 64-bit    +2/-1/=9   54.17   6.5/12
Critter 1.6a 64-bit      +1/-2/=9   45.83   5.5/12

Applying any elo value at this point in time is useless- so go with this info. It is plenty.




Until another 20 or so hours-

george

[geots]

Maybe I should have mentioned that I am considering another 40/40 match exactly like this one, except Houdini 2 will be playing Komodo 4536. First I have to decide the point and significance of such a match.


Been up an hour or so and I am worn out-

george

[geots]

Now 35 of the 50 games have been completed. Everyone has their own opinions about how far Komodo has gone in distancing himself from Critter. One thing for sure- good things usually come to people who work hard and tend to their own business, as Don and Larry do.




XPS 8500 Intel i5-4 True Cores
Fritz 11 gui
1CPU/64-bit
128MB hash
Bases=NONE
Ponder_Learning=OFF
Perfect 2012b.ctg w/12-move limit
40/19 Repeating (Benched to adapt to 40/40)
Match=50 games

Code: Select all

XPS-i5, 19m/40+19m/40+19m/40  3/5/2013   

                           
Komodo 4536.00 64-bit    +9/-6/=20   54.30%   19.0/35       
Critter 1.6a 64-bit      +6/-9/=20   45.70%   16.0/35  

I believe the next update will be the final- but it is possible there could be one before that.




zzzzzzzzzzzzz,

george

[geots]

Well, this k-4536 v Critter 1.6a match has concluded. And it ended up being a lot closer than I actually expected. I have hundreds upon hundreds of matches at all different controls, and as I go back and look, I find matches that Critter 1.4 or 1.6 lost. But so far none where Critter was not competitive- and none yet where he was just destroyed. Even all the matches with Houdini 2.0c at different controls were all close.

My first thought was Critter 1.6a and Houdini 3. I haven't checked yet to see if that match has been played on my end- but if it hasn't it likely won't be at present. As I am beta testing Komodo 4536 and Strelka 5.6.

At any rate, the final results:





XPS 8500 Intel i5-4 True Cores
Fritz 11 gui
1CPU/64-bit
128MB hash
Bases=NONE
Ponder_Learning=OFF
Perfect 2012b.ctg w/12-move limit
40/19 Repeating (Benched to adapt to 40/40)
Match=50 games

Code: Select all

XPS-i5, 19'/40+19'/40+19'/40  3/6/2013  

                           
Komodo 4536.00 64-bit   +7    +12/-11/=27   51.00%   25.5/50
Critter 1.6a 64-bit     -7    +11/-12/=27   49.00%   24.5/50

Hope you enjoyed-

george

[beram]

Well, this k-4536 v Critter 1.6a match has concluded. And it ended up being a lot closer than I actually expected. I have hundreds upon hundreds of matches at all different controls, and as I go back and look, I find matches that Critter 1.4 or 1.6 lost. But so far none where Critter was not competitive- and none yet where he was just destroyed. Even all the matches with Houdini 2.0c at different controls were all close.

My first thought was Critter 1.6a and Houdini 3. I haven't checked yet to see if that match has been played on my end- but if it hasn't it likely won't be at present. As I am beta testing Komodo 4536 and Strelka 5.6.

At any rate, the final results:





XPS 8500 Intel i5-4 True Cores
Fritz 11 gui
1CPU/64-bit
128MB hash
Bases=NONE
Ponder_Learning=OFF
Perfect 2012b.ctg w/12-move limit
40/19 Repeating (Benched to adapt to 40/40)
Match=50 games

Code: Select all

XPS-i5, 19'/40+19'/40+19'/40  3/6/2013  

                           
Komodo 4536.00 64-bit   +7    +12/-11/=27   51.00%   25.5/50
Critter 1.6a 64-bit     -7    +11/-12/=27   49.00%   24.5/50

Hope you enjoyed-

george

Thx George,
So not yet an indication for improvement when compared to Komodo 5 results against Critter 1.6a in CCRL 40/40, with result 95 − 92 for Komodo 5, +42−39=106

grts Bram

[geots]

Well, this k-4536 v Critter 1.6a match has concluded. And it ended up being a lot closer than I actually expected. I have hundreds upon hundreds of matches at all different controls, and as I go back and look, I find matches that Critter 1.4 or 1.6 lost. But so far none where Critter was not competitive- and none yet where he was just destroyed. Even all the matches with Houdini 2.0c at different controls were all close.

My first thought was Critter 1.6a and Houdini 3. I haven't checked yet to see if that match has been played on my end- but if it hasn't it likely won't be at present. As I am beta testing Komodo 4536 and Strelka 5.6.

At any rate, the final results:





XPS 8500 Intel i5-4 True Cores
Fritz 11 gui
1CPU/64-bit
128MB hash
Bases=NONE
Ponder_Learning=OFF
Perfect 2012b.ctg w/12-move limit
40/19 Repeating (Benched to adapt to 40/40)
Match=50 games

Code: Select all

XPS-i5, 19'/40+19'/40+19'/40  3/6/2013  

                           
Komodo 4536.00 64-bit   +7    +12/-11/=27   51.00%   25.5/50
Critter 1.6a 64-bit     -7    +11/-12/=27   49.00%   24.5/50

Hope you enjoyed-

george

Thx George,
So not yet an indication for improvement when compared to Komodo 5 results against Critter 1.6a in CCRL 40/40, with result 95 − 92 for Komodo 5, +42−39=106

grts Bram

Bram, this is where it gets really tricky. And I have to be extremely careful I say nothing that would look like I was trying to put the most emphasis on "this or that".

I think the fairest way for me to go about this is to say I would NEVER let just 1 50-game match determine my purchase choices- and I know you well enough to know you would agree. I don't know why- but as in the past- at certain times Critter can be almost impossible to handle. Not too long after Houdini 2.0c came out- I ran him in a 50 game match against Critter. Houdini won by 2 games, but he had to win games 49 and 50 to salvage the match.

2 things to please remember: (1) Don and Larry have done some very good work here (2) When you get up into the top 3 or 4 engines, life is just difficult for all involved.




Best regards,

george