Originally posted by philidor positionI used Hiarcs 12 with 1 GB for hash on my quad.
for the record, my computer doesn't find the mate in 14 in a short time too. I gave up with rybka 3, deep shredder 12 and deep fritz 11 on a quad after 1 minute. Maybe you had your hash uncleared?
Hi PP.
The book in question was a Fred Reinfeld book and 99% of his stuff
was always are check!, defense, check!, capture, bam!: mate.
Hi Varenka
You popped up with your mate in 14 position, something I know you
never got from a Reinfeld book. I stayed within the bounds of the thread,
it is you that stuck in the right angle in 'cos I meantioned the brute force method.
3 good OTB players spotted 1.Qxf8+ wins within seconds looking at 3-4 positions.
"Qxf8 pretty much screams for attention."
The box had to plough through 12600000 positions. I'd say that was pretty close
to brute force. It looks like it tried every legal move before settling on 1.Qxf8+
(if indeed that is the move - I like 1.Rf1)
The postion looks kind of familiar, is it from a game, if so did the human
go for the mate in 14 or did Black resign after 1.Qxf8/1.Rf1.
Originally posted by VarenkaWhile what you say is sound Varenka, my only suggestion for turning on mate search
Why does that matter? I wasn't running an engine with "avoid mate" switched on. 😉 Top engines by default look for mate, even if it's combined with other heuristics.
What engine and settings did you use?
would be to reduce collision.
That said, in the position you've given, collision isn't the issue for the search, and that
large hash field is supplementing the error...Which, without having Mr. Rajlich available,
looks to ME as though its a bad case of the horizon effect.
Here Huygen's principle is applicable, positional reduction would have to follow in
order to reduce the affecting arena of selections. Reduce the affecting party, and
certainly reduce the collidable factors and you might get an engine to work effectively
here.
-GIN
Originally posted by greenpawn34I meantioned the brute force method
And that computers "stumble" upon mates.
The box had to plough through 12600000 positions. I'd say that was pretty close
to brute force.
Did you look at the other figure I quoted as a comparison? The *real* brute force figure? Go on, count the zeros in that figure. 🙂
is it from a game
Yes, Deep Thought versus Gliksman. Black resigned after the computer announced mate starting with Rf1.
Originally posted by philidor positionI can get Deep Shredder 12 to find mate in 58 seconds. Also Zappa Mexico II in under a minute.
that's interesting, maybe hiarcs has something unique that fits into these type of mates. or your computer is a 16 core monster hidden inside a quad-case 🙂. (or, your hash was uncleared 🙂)
Are you using a 64-bit operating system? 64-bit is siginificantly faster for chess engines.
Yo! I knew 1.Rf1 was good - of course I saw the mate in 14 right away.......😉
1.Qf8+ is the move I would have done OTB it's a clean win, easy to calculate,
these days I'm very lazy at the board and have stopped shooting for the brilliancy
prize.
A win is a win.
The lad should have played on and got the box to prove it.
Please post the game including the mate.
Originally posted by VarenkaShredder is the only one among them to spot the mate, but it wasn't mate in 14 in my computer. I do use 64 bit os. Well, monk has quit, so who else is left there to solve this mystery?
I can get Deep Shredder 12 to find mate in 58 seconds. Also Zappa Mexico II in under a minute.
Are you using a 64-bit operating system? 64-bit is siginificantly faster for chess engines.
Originally posted by heinzkatyou hadn't heard of larry david, maybe that's OK, but please don't tell me you haven't heard of monk the defective detective!
Monk, mystery? Ugly game by Gliksman by the way 😕
let's see the bright side of this. if you have survived in this ugly world without monk, now you'll be living in paradise.
http://en.wikipedia.org/wiki/Monk_%28TV_series%29