How to make Gothmog 0.4.7 50 points stronger

[Tord Romstad]

Geschrieben von:/Posted by: Tord Romstad at 20 March 2004 10:48:09:

Change the "selectivity" setting from the default value of 4 to 6. If you run Gothmog in
Winboard mode, this is done by modifying the gothmog.ini file. In UCI mode, changing
the line in the init file has no effect. The selectivity setting must be set from the GUI
instead.
I have not made enough tests to be entirely sure that selectivity=6 is better, but my
first results are very promising. I have played two 100-game blitz matches against
Pepito 1.59 and Yace Paderborn. Here are the results:
Selectivity 4:
Gothmog 0.4.7-Pepito 1.59: 42.5-57.5
Gothmog 0.4.7-Yace Paderborn: 40-60
Selectivity 6:
Gothmog 0.4.7-Pepito 1.59: 47.5-52.5
Gothmog 0.4.7-Yace Paderborn: 51.5-48.5
The second match aganst Yace, by the way, showed the importance of playing a big
number of games. Gothmog lost the first 6 games, did not win a game before the
13th, and had only 3 points after 16 rounds.
Until I tell you otherwise, I would like everybody who starts tournaments with Gothmog
0.4.7 among the participants to use selectivity=6. Those who want to experiment with
various selectivity settings are more than welcome to do so. It is possible that other
values are even better than 6 (I am not even 100% sure that 6 is better than 4 yet).
Tord

[Kurt Utzinger]

Geschrieben von:/Posted by: Kurt Utzinger at 20 March 2004 11:28:01:
Als Antwort auf:/In reply to: How to make Gothmog 0.4.7 50 points stronger geschrieben von:/posted by: Tord Romstad at 20 March 2004 10:48:09:

Change the "selectivity" setting from the default value of 4 to 6. If you run Gothmog in
Winboard mode, this is done by modifying the gothmog.ini file. In UCI mode, changing
the line in the init file has no effect. The selectivity setting must be set from the GUI
instead.
I have not made enough tests to be entirely sure that selectivity=6 is better, but my
first results are very promising. I have played two 100-game blitz matches against
Pepito 1.59 and Yace Paderborn. Here are the results:
Selectivity 4:
Gothmog 0.4.7-Pepito 1.59: 42.5-57.5
Gothmog 0.4.7-Yace Paderborn: 40-60
Selectivity 6:
Gothmog 0.4.7-Pepito 1.59: 47.5-52.5
Gothmog 0.4.7-Yace Paderborn: 51.5-48.5
The second match aganst Yace, by the way, showed the importance of playing a big
number of games. Gothmog lost the first 6 games, did not win a game before the
13th, and had only 3 points after 16 rounds.
Until I tell you otherwise, I would like everybody who starts tournaments with Gothmog
0.4.7 among the participants to use selectivity=6. Those who want to experiment with
various selectivity settings are more than welcome to do so. It is possible that other
values are even better than 6 (I am not even 100% sure that 6 is better than 4 yet).
Tord

Hi Tord
Most interesting news. If selectivity 6 should be stronger at blitz,
I would suppose that this effect brings even more at higher time
controls.
Kurt

[Tord Romstad]

Geschrieben von:/Posted by: Tord Romstad at 20 March 2004 12:36:13:
Als Antwort auf:/In reply to: Re: How to make Gothmog 0.4.7 50 points stronger geschrieben von:/posted by: Kurt Utzinger at 20 March 2004 11:28:01:

Change the "selectivity" setting from the default value of 4 to 6. If you run Gothmog in
Winboard mode, this is done by modifying the gothmog.ini file. In UCI mode, changing
the line in the init file has no effect. The selectivity setting must be set from the GUI
instead.
I have not made enough tests to be entirely sure that selectivity=6 is better, but my
first results are very promising. I have played two 100-game blitz matches against
Pepito 1.59 and Yace Paderborn. Here are the results:
Selectivity 4:
Gothmog 0.4.7-Pepito 1.59: 42.5-57.5
Gothmog 0.4.7-Yace Paderborn: 40-60
Selectivity 6:
Gothmog 0.4.7-Pepito 1.59: 47.5-52.5
Gothmog 0.4.7-Yace Paderborn: 51.5-48.5
The second match aganst Yace, by the way, showed the importance of playing a big
number of games. Gothmog lost the first 6 games, did not win a game before the
13th, and had only 3 points after 16 rounds.
Until I tell you otherwise, I would like everybody who starts tournaments with Gothmog
0.4.7 among the participants to use selectivity=6. Those who want to experiment with
various selectivity settings are more than welcome to do so. It is possible that other
values are even better than 6 (I am not even 100% sure that 6 is better than 4 yet).
Tord

Hi Tord
Most interesting news. If selectivity 6 should be stronger at blitz,
I would suppose that this effect brings even more at higher time
controls.

Very hard to guess. It's possible that you are right, but it is impossible to guess
without lots of testing. It could also be that high selectivity values only work well
in fast games.
I'm running a third blitz match now, against AnMon 5.21. After 16 games, Gothmog
is leading by 12-4 (+11,-3,=2). Selectivity=6 still looks very good ...
Tord