CEGT/CCRL and the "forbidden" engines.

[George Tsavdaris]

Am i the only one thinking that these 2 GREAT lists(well they are more than great, their efforts are amazing, the job required huge and the result is superb) are becoming obsolete a bit? Why?

Of course because they don't include several extra strong engines, like Houdini of course, like Ivanhoe and Deep Saros for example. Yes i know the clone stamp is all over their face even if nothing is certain and had been proved, but why that should matter anyway?

I mean most people have them, play with them, and enjoy their analysis etc. Pretending they do not exist would not disappear them!

OK there is the issue that if these engines are clones, then by entering them to CEGT/CCRL we have an indirect support and promotion of clones. Understandable. But the way i see it is that it doesn't matter since MOST people use them anyway, not caring about CEGT's/CCRL's honest and romantic policy.

So by not having the -it seems- number one engine Houdini, the super strong Ivanhoe(although as it seems if they decide to use it then they will have to spend about 2-3 weeks deciding which version they should use from the millions), the new directly and honestly said by its author clone, but nevertheless very strong Deep Saros, the older but very strong Firebird(perhaps i'm missing some more), so by not having these top engine in their list, their list is becoming obsolete and it's anyway not complete.

My question and poll is:

Would you prefer CEGT and CCRL to include the "forbidden" engines in their lists?

•Definitely YES.
•YES but i'm not so sure.
•Definitely NO.
•I would rather say NO.
•I don't care.



(PS: Graham forgive me. I'm not in the other wagon, i just see that it doesn't matter anymore.)

[yanquis1972]

i think houdini has proved itself. it can no longer be namecalled a scrambled ippolit with a little bit of endgame knowledge tossed in (i will readily admit this is all i inititally thought it was as well). imo it should definitely be tested. as for ivanhoe etc, again, when/if they start actually improving strength, sure, include them. let's not fool ourselves into thinking robert is the only author who has used (or even made heavy use of) ippolit code. and he admits it, & is the first to actually improve upon it.

also, there is something to be said for character. robert is not anonymous, releases bug fixes ASAP, & is an active member of the computer chess community. i can't see his engine shouldn't be considered official at this point & thanks to ingo bauer for being the first to consider it as such. without his testing we would just have more speculation & arguments about sample size, etc.

[yanquis1972]

also, this is a problem i have with your poll. i'm still not sure what deep saros is, never used it or researched it, only heard the name a couple times. but of course it would be silly for every nobody who takes ippolit & rebrands it to be allowed to be tested. so i will abstain from voting w/ a concrete nod for houdini & iffy on ivanhoe (other problem with ivanhoe -- to many versions/mods/etc & no authority as to which to test).

[gerold]

Am i the only one thinking that these 2 GREAT lists(well they are more than great, their efforts are amazing, the job required huge and the result is superb) are becoming obsolete a bit? Why?

Of course because they don't include several extra strong engines, like Houdini of course, like Ivanhoe and Deep Saros for example. Yes i know the clone stamp is all over their face even if nothing is certain and had been proved, but why that should matter anyway?

I mean most people have them, play with them, and enjoy their analysis etc. Pretending they do not exist would not disappear them!

OK there is the issue that if these engines are clones, then by entering them to CEGT/CCRL we have an indirect support and promotion of clones. Understandable. But the way i see it is that it doesn't matter since MOST people use them anyway, not caring about CEGT's/CCRL's honest and romantic policy.

So by not having the -it seems- number one engine Houdini, the super strong Ivanhoe(although as it seems if they decide to use it then they will have to spend about 2-3 weeks deciding which version they should use from the millions), the new directly and honestly said by its author clone, but nevertheless very strong Deep Saros, the older but very strong Firebird(perhaps i'm missing some more), so by not having these top engine in their list, their list is becoming obsolete and it's anyway not complete.

My question and poll is:

Would you prefer CEGT and CCRL to include the "forbidden" engines in their lists?

•Definitely YES.
•YES but i'm not so sure.
•Definitely NO.
•I would rather say NO.
•I don't care.



(PS: Graham forgive me. I'm not in the other wagon, i just see that it doesn't matter anymore.)

There are a lot more people testing these engines than the testing at CEGT/CCRL. They have missed the boat.

Best,
Gerold.

[yanquis1972]

there are other people testing them, but not with the same thoroughness or oversight. to me it's not so much about legitimacy -- that seems to be subjective these days unless you attach dollars to your product & are in the chessbase stable. it's more about having a safe, accurate idea of what an engine's strength is.

[Engin]

on this way you will give the cloners a motivation to continue this way !

and may its not take short time we will have 1000 clones of ippolit engines in the end of 2011/2012

[Graham Banks]

(PS: Graham forgive me. I'm not in the other wagon, i just see that it doesn't matter anymore.)

Everybody is entitled to have an opinion George.

[Houdini]

My question and poll is:

Would you prefer CEGT and CCRL to include the "forbidden" engines in their lists?

The issue you address is legitimate, but unfortunately you've translated it into a rather poor polling question which doesn't do justice to the complexity of the problem.

Rating lists are bound to use criteria for which engines to include, the real question is what these criteria should be so that the rating lists keep their relevance in the ever faster moving world of chess engines.

Currently IPON appears to be the only public rating list that has found a satisfactory answer to this question.

[Engin]

sure ?

the IPON list include very few of engine, at the moment, it is better he or they will be only testing IPOLLIT clones for the feature only....

if CEGT/CCRL will include those engine too, then i am sure then in a short time we will get for the first 500 engines will be ippolit clones, because this is a motivation for the cloners to continue this wrong way.

other question to you directly, for a beginner it is very strange to wrote an engine in very short time less then 3 month or even 6 month.

your engine Houdini is very strong at the starting rating, and this is very strange for me....

good night....

[Houdini]

if CEGT/CCRL will include those engine too, then i am sure then in a short time we will get for the first 500 engines will be ippolit clones, because this is a motivation for the cloners to continue this wrong way.

Is it really so difficult to envisage a reasonable solution without falling into extremes?
Like I said, apparently IPON is currently the only public list that has found an intelligent answer to the question.

other question to you directly, for a beginner it is very strange to wrote an engine in very short time less then 3 month or even 6 month.
your engine Houdini is very strong at the starting rating, and this is very strange for me....

Indeed for a beginner this is all quite amazing...
I wonder what makes you think I fit that qualifier. I've been involved in chess, programming and computer chess since the mid 1980's. How about you?