Improve IvanHoe?

[benstoker]

BB--

Having deconstructed Ippo, and being apparently knowledgeable of the state of the art in computer chess, can you suggest how Ippo/Ivanhoe might be improved?

[kingliveson]

BB--

Having deconstructed Ippo, and being apparently knowledgeable of the state of the art in computer chess, can you suggest how Ippo/Ivanhoe might be improved?

I don't think it's nice to call out people like that. You could just phrase a general question and let people contribute. Especially when he is not the author. You should have asked Sentinel instead.

[BB+]

Having deconstructed Ippo, and being apparently knowledgeable of the state of the art in computer chess, can you suggest how Ippo/Ivanhoe might be improved?

Twiddling with parameters and a whole lot of testing. Seriously, though, I do think that the use of various "modern" testing methods (for instance, play fast games and extrapolate the results to longer time controls) is one of the main things that Rajlich has added to computer chess. [Letouzey also seems to have had a "scientific" view of testing, as his Fruit readme.txt had: Computer testing time is especially welcome, but be warned that I am quite demanding. "I can include test versions in my Fritz-GUI swiss tournament." -> forget it, as well as my email address please, thanks a lot!]

[BB+]

Recently on TalkChess:

I don't think so. BB (or BB+) said (in open forum) that fast tc cluster testing was Rajlich's (only?) contribution to comp chess eval testing ..are you saying that you [Bob Hyatt] invented it instead?? Could be that BB was wrong of course.

Here is the quotation:

Seriously, though, I do think that the use of various "modern" testing methods (for instance, play fast games and extrapolate the results to longer time controls) is one of the main things that Rajlich has added to computer chess.

My comment was concerned mostly with (hyper)-fast time controls, not merely cluster-testing (which Bob might have well done first, and I'm not sure exactly what a "cluster" has to do with it). In this regard, Nelson Hernandez nominated Marc Lacrosse for "important new analytic insight of 2006" when he noted that (say) testing at 1m/move down to 1s/move gave largely the same results.

I also think Rajlich's contribution is not merely limited to this, but further includes: championing/improving Kaufman's material imbalance tables, investigation of "statistical" and hard pruning in search, and hiring and/or hooking up with the right people (Noomen and Kaufman for example). Other things he has done, such as starting his own forum, introducing the cluster program, etc., are at least novel.

[Richard Vida]

BB--

Having deconstructed Ippo, and being apparently knowledgeable of the state of the art in computer chess, can you suggest how Ippo/Ivanhoe might be improved?

Is this question addressed purely to Mark (BB+)? Or is others contribution welcomed too?

[BB+]

Is this question addressed purely to Mark (BB+)? Or is others contribution welcomed too?

This question is more than a year old! I just bumped it because there was a bit of TalkChess discussion over what I meant regarding Rajlich's contribution. Feel free to express any opinions for improvement (including banal ones like: "change the name").

[Richard Vida]

Is this question addressed purely to Mark (BB+)? Or is others contribution welcomed too?

This question is more than a year old! I just bumped it because there was a bit of TalkChess discussion over what I meant regarding Rajlich's contribution. Feel free to express any opinions for improvement (including banal ones like: "change the name").

Sorry, I did not notice the date of the first post in this thread. Anyway... If any of the Robbo/Ivanhoe developers read this, there is a an easy one:

(I do not have the actual sources at hand, so I will use my own terminology)

In IPPOLIT sources in 'low_depth' search there was a possibility to return immediately with SCOUT-1 when the EVAL is much bellow SCOUT (> value of queen). This test was done *before* hash table lookup. I noticed that in recent Ivanhoes this was removed - and for a good reason (I have tried this in Critter and the result was something like -6 elo).

This cut-off condition in itself is OK, but it should be moved *after* the hash lookup. Just think about it a little and the reason becomes obvious.

And if you really want to play it safe, this can be further enhanced. In my experiments (with Critter, IPPO* behavior may be different - YMMV) I tried to restrict this cut-off with several other conditions:
a) no pawns at 7th rank
b) no queen capture is possible
c) no mate threat (in IPPO framework this information is not available, I think)
d) king safety of the side to move did not worsen relative to the previous ply
.

I won't tell which combination is the best - test yourself

[veritas]

Richard

Congratulations on your engine success


Would you please consider enabling your engine to support robobases . the OLD one's , new ones are buggy and for size see no real improvement over old ones
many find the old robobases far better than others as well as a very manageable size , i'm sure both peter pan and KLO would assist should you wish
It would be HUGELY appreciated

[ThinkingALot]

Recently I ran several tests of Stockfish's history pruning (futility_move_count). It appears that this feature makes the engine perform better at fixed depth controls. However when the prunings are turned off the engine reaches the same depth a bit faster. There are also some reasons to believe (without sufficient statistical basis yet) that the ELO boost provided by history prunings is greater at greater depths. Therefore it may be useful to use history prunings in Ivanhoe, especially in low and cut nodes.

P.S. The easiest way of improving Ivanhoe: disassemble Houdini (relatively easy) and copy its ideas .

[Peterpan]

hi ThinkingAlot

Yes one could do that,well not anyone,but people familiar with this kind of thing... and for them i assume it would be relatively easy to do so.
But this will make us no better than them (the ones accused of stealing segments of code and ideas from for example Fruit and Crafty.And then others stealing from and Ivanhoe & Robbolito and calling it their own)
I am not a chess programmer,just a baby,but if/when i write a chess engine or modify Ivanhoe i will not use this way you have described in your post.
I will use ideas available openly and create new ideas,not walk in the shadow of others and this way keep my integrity.

Although one may say then what if the ideas were stolen by others by reverse engineering and posted on the web as open now for anyone to use...?? this is a whole new debate and then this thing gets complicated

computer chess is getting too complicated now