horrid positional play in a solid tactical searcher

Discussion of chess software programming and technical issues.

Moderators: hgm, Rebel, chrisw

Ras
Posts: 2487
Joined: Tue Aug 30, 2016 8:19 pm
Full name: Rasmus Althoff

Re: horrid positional play in a solid tactical searcher

Post by Ras »

smcracraft wrote:Can you quantify in millipawns?
Between 5 and 40 centipawns. It depends also on what you want to emphasise to what extent. Don't attribute too much or the engine will start to sacrifice material. You also need some equilibrium between these factors.

But all of that depends also on what other factors you value (space, mobility, king threats), and what style the engine prefers.
tttony
Posts: 268
Joined: Sun Apr 24, 2011 12:33 am

Re: horrid positional play in a solid tactical searcher

Post by tttony »

Ras wrote:
smcracraft wrote:Can you quantify in millipawns?
Between 5 and 40 centipawns. It depends also on what you want to emphasise to what extent. Don't attribute too much or the engine will start to sacrifice material. You also need some equilibrium between these factors.

But all of that depends also on what other factors you value (space, mobility, king threats), and what style the engine prefers.
My engine just use pst for eval, it's super basic, I'm trying to add one feature at time, I marked what I have doubts, is not the same: mobility and space?
Ras
Posts: 2487
Joined: Tue Aug 30, 2016 8:19 pm
Full name: Rasmus Althoff

Re: horrid positional play in a solid tactical searcher

Post by Ras »

Henk wrote:Skipper doesn't have that problem anymore:
If 1. ... h5 doesn't qualify as horrid, I don't want to imagine what might do?!

Btw, is that engine downloadable somewhere?
Ras
Posts: 2487
Joined: Tue Aug 30, 2016 8:19 pm
Full name: Rasmus Althoff

Re: horrid positional play in a solid tactical searcher

Post by Ras »

tttony wrote:is not the same: mobility and space?
Grabbing space is about pushing pawns (with mobility as side-effect) while mobility is about piece play (with occasional pawn pushes as side-effect).

A typical mobility-centred approach lets the opponent push the pawns, then undermines the centre and destroys it. Suddenly, the advanced pawns become a liability. That's the idea behind openings like Alekhine's Defence. Of course, that can also blow back and you'll be overrun.
Henk
Posts: 7216
Joined: Mon May 27, 2013 10:31 am

Re: horrid positional play in a solid tactical searcher

Post by Henk »

Ras wrote:
Henk wrote:Skipper doesn't have that problem anymore:
If 1. ... h5 doesn't qualify as horrid, I don't want to imagine what might do?!

Btw, is that engine downloadable somewhere?
Was just a joke. Engine is kept private to not inflict psychic damage to others.

It still has problems with beating Fairy max. So progress is -30 ELO points a year.
Ras
Posts: 2487
Joined: Tue Aug 30, 2016 8:19 pm
Full name: Rasmus Althoff

Re: horrid positional play in a solid tactical searcher

Post by Ras »

Henk wrote:Was just a joke. Engine is kept private to not inflict psychic damage to others.
Too bad, I would have liked to do a game with Skipper on a PC at 3.7 GHz (single-thread) against the CT800 on a Cortex-M4 microcontroller at 168 MHz. :D
smcracraft
Posts: 737
Joined: Wed Mar 08, 2006 8:08 pm
Location: Orange County California
Full name: Stuart Cracraft

Re: horrid positional play in a solid tactical searcher

Post by smcracraft »

I stripped out ALL evaluation other than material and pc/sq.

Its first game against someone was:

d2d4 g8f6
e2e3 d7d5
g2g3 b8c6
f1h3 e7e5
h3c8 a8c8
d4e5 c6e5
d1d4 e5c6
d4a4 f8c5
b2b3 d5d4
g1e2 d4e3
a4c4 e3f2
e1f1 d8d1
f1g2 d1h1
g2h1 f2f1Q
e2g1 f1g1

I'll add in mobility and do some tests between the above and the above+mobility.

All the weird, bizarre moves it used to make.... looks refreshed!

Thanks,
--Stuart--
smcracraft
Posts: 737
Joined: Wed Mar 08, 2006 8:08 pm
Location: Orange County California
Full name: Stuart Cracraft

Re: horrid positional play in a solid tactical searcher

Post by smcracraft »

It played this in its second game:

e2e4 g8f6
e4e5 f6d5
g1f3 d7d6
e5d6 e7d6
f1c4 d5b6
c4b5 c8d7
b1c3 g7g6
e1g1 b8c6
d2d4 f8g7
b5c6 d7c6
d4d5 c6d7
c1g5 g7f6
f1e1 e8f8
g5f6 d8f6
f3d2 f6g5
d2e4 g5f5
d1d4 f5e5
d4e5 d6e5
e4c5 d7f5
e1e5 f5c2
g2g4 f7f6
e5e2 c2a4
c5a4 b6a4
c3a4 f8f7
a1c1 b7b5
c1c7 f7f8
a4c5 a8d8
e2e7

I tested it without mobility and it scored about 2536 or so on a tactical test.
smcracraft
Posts: 737
Joined: Wed Mar 08, 2006 8:08 pm
Location: Orange County California
Full name: Stuart Cracraft

Re: horrid positional play in a solid tactical searcher

Post by smcracraft »

Mobility dropped the score on the test by about 40 points.

Probably plays more positionally tho.