Page 1 of 7

Houdini 5 dev in TCEC

Posted: Thu Sep 08, 2016 8:13 am
by Jouni
Robert has really improved engine. In CCRL, CEGT and IPON Houdini 4 was 119, 154 and 137 points below Komodo 10.1. So progress has been around 140 ELO! Also the qualification to superfinal was the greatest surprise in last ten years or ever!!

Re: Houdini 5 dev in TCEC

Posted: Thu Sep 08, 2016 9:27 am
by JJJ
Jouni wrote:Robert has really improved engine. In CCRL, CEGT and IPON Houdini 4 was 119, 154 and 137 points below Komodo 10.1. So progress has been around 140 ELO! Also the qualification to superfinal was the greatest surprise in last ten years or ever!!
Yeah I agree with that :)

Re: Houdini 5 dev in TCEC

Posted: Thu Sep 08, 2016 9:42 am
by Henk
Maybe because Houdini 5 was not in the test set of the other engines.

Re: Houdini 5 dev in TCEC

Posted: Thu Sep 08, 2016 1:00 pm
by shrapnel
Jouni wrote:the greatest surprise in last ten years or ever!!
I bet it was for the Komodo Team too, albeit an unpleasant one ! :lol:
Imagine the sheer ARROGANCE of using Contempt 10 against Houdini in the first game of Stage 3 !
Luckily for them, the Backup Plan worked, and the Game was called off.
Full Marks to Robert Houdart, who had full Faith in his Engine and persevered without complaint.
In the end Truth and Justice Prevailed.
Amen.

Re: Houdini 5 dev in TCEC

Posted: Thu Sep 08, 2016 1:09 pm
by JJJ
shrapnel wrote:
Jouni wrote:the greatest surprise in last ten years or ever!!
I bet it was for the Komodo Team too, albeit an unpleasant one ! :lol:
Imagine the sheer ARROGANCE of using Contempt 10 against Houdini in the first game of Stage 3 !
Luckily for them, the Backup Plan worked, and the Game was called off.
Full Marks to Robert Houdart, who had full Faith in his Engine and persevered without complaint.
In the end Truth and Justice Prevailed.
Amen.
The settings problem wasn't Komodo team fault, and you have to know that.
I m not as expert as Mark or Larry are, but I have the feeling than the contempt 7 was probably too much for this stage, but the impact of it might have been non existent :)

Re: Houdini 5 dev in TCEC

Posted: Thu Sep 08, 2016 4:18 pm
by shrapnel
JJJ wrote:The settings problem wasn't Komodo team fault
If you believe that, you probably also believe in Fairies and Santa Claus, too ! Cho Chweet !
:lol:
You're right about the Contempt, though.
Contempt 7 was ridiculous !
Perhaps Contempt -7 would have been more appropriate. :lol:

Re: Houdini 5 dev in TCEC

Posted: Thu Sep 08, 2016 4:32 pm
by Sean Evans
Jouni wrote:Robert has really improved engine. In CCRL, CEGT and IPON Houdini 4 was 119, 154 and 137 points below Komodo 10.1. So progress has been around 140 ELO! Also the qualification to superfinal was the greatest surprise in last ten years or ever!!
Plagiarizing other authors code and adding it to your program to make it stronger is not doing work; it is called pirating.

Cordially,

Sean

Re: Houdini 5 dev in TCEC

Posted: Thu Sep 08, 2016 4:50 pm
by lkaufman
JJJ wrote:
shrapnel wrote:
Jouni wrote:the greatest surprise in last ten years or ever!!
I bet it was for the Komodo Team too, albeit an unpleasant one ! :lol:
Imagine the sheer ARROGANCE of using Contempt 10 against Houdini in the first game of Stage 3 !
Luckily for them, the Backup Plan worked, and the Game was called off.
Full Marks to Robert Houdart, who had full Faith in his Engine and persevered without complaint.
In the end Truth and Justice Prevailed.
Amen.
The settings problem wasn't Komodo team fault, and you have to know that.
I m not as expert as Mark or Larry are, but I have the feeling than the contempt 7 was probably too much for this stage, but the impact of it might have been non existent :)
Of course contempt 7 was too high for the games with SF and Houdini, but probably too low for the other five opponents. Overall at least close to the right value, maybe 5 or 6 would have been a tad better. But that wasn't the main problem. The main thing is that Komodo just moves too fast in general in these games, and I think we finally figured out why and what to do about it, but too late for this year's TCEC.

Re: Houdini 5 dev in TCEC

Posted: Thu Sep 08, 2016 4:52 pm
by JJJ
lkaufman wrote:
JJJ wrote:
shrapnel wrote:
Jouni wrote:the greatest surprise in last ten years or ever!!
I bet it was for the Komodo Team too, albeit an unpleasant one ! :lol:
Imagine the sheer ARROGANCE of using Contempt 10 against Houdini in the first game of Stage 3 !
Luckily for them, the Backup Plan worked, and the Game was called off.
Full Marks to Robert Houdart, who had full Faith in his Engine and persevered without complaint.
In the end Truth and Justice Prevailed.
Amen.
The settings problem wasn't Komodo team fault, and you have to know that.
I m not as expert as Mark or Larry are, but I have the feeling than the contempt 7 was probably too much for this stage, but the impact of it might have been non existent :)
Of course contempt 7 was too high for the games with SF and Houdini, but probably too low for the other five opponents. Overall at least close to the right value, maybe 5 or 6 would have been a tad better. But that wasn't the main problem. The main thing is that Komodo just moves too fast in general in these games, and I think we finally figured out why and what to do about it, but too late for this year's TCEC.
I hope it will give you some nice elo Jump. Anyways, it was nice to see Komodo in tough competition against Houdini. And now there is again 3 top chess program. I think it's good for chess.

Re: Houdini 5 dev in TCEC

Posted: Thu Sep 08, 2016 5:57 pm
by beram
lkaufman wrote:
JJJ wrote:
shrapnel wrote:
Jouni wrote:the greatest surprise in last ten years or ever!!
I bet it was for the Komodo Team too, albeit an unpleasant one ! :lol:
Imagine the sheer ARROGANCE of using Contempt 10 against Houdini in the first game of Stage 3 !
Luckily for them, the Backup Plan worked, and the Game was called off.
Full Marks to Robert Houdart, who had full Faith in his Engine and persevered without complaint.
In the end Truth and Justice Prevailed.
Amen.
The settings problem wasn't Komodo team fault, and you have to know that.
I m not as expert as Mark or Larry are, but I have the feeling than the contempt 7 was probably too much for this stage, but the impact of it might have been non existent :)
Of course contempt 7 was too high for the games with SF and Houdini, but probably too low for the other five opponents. Overall at least close to the right value, maybe 5 or 6 would have been a tad better. But that wasn't the main problem. The main thing is that Komodo just moves too fast in general in these games, and I think we finally figured out why and what to do about it, but too late for this year's TCEC.
..A tad better... is about 0,4 Elo? :lol:
Well at least Komodo scales better :lol: :lol: