noctiferus wrote:I might do it. I would like to try, also, some other less classical methodologies, on these data.
It's my academic area.
However the basic problem, as I said, is to have data about enough engines, in order to have a good predictive model .
Hi Enrico,
That sounds great! I'd like to know the what the total number of engines is that you need their data for? Does 100 sound good? I'd be willing to test 150 or more if that satisfies the necessary requirement to process the set of datas.
I would like to suggest engines for which does exsist a reliable ELO estimation by CCRL and CEGT, otherwise are useless.
Oh, yes I did consider about that factor. I thought there exists enough number of engines in both the lists. I checked it and there were only about 158 engines in CCRL 40/4 and 131 in CEGT 40/4
So, I guess I could only provide datas of about 60 to 70 _single_ Cpu Engines. I'd think that would be enough.
<Swami>
Until you have 8 testsets, I think that 30-35 should be enough (I'd like to do some out-of-sample control).
Let see what happens. We are always in time to add more engines.
Of course, we'll need it when you will increase the number of testsuits.
<Marco>
Yes, of course.
(It's a pity I haven't PLS in my home software)
PS.- just an idea, to be verified: if I could find a student interested in the matter (with the help of an IM that is a colleague and a friend), I think it would be interesting an undergraduate thesis on relations among testsuites(general) and ELO. Don't take it as a promise, however...
noctiferus wrote:<Swami>
Until you have 8 testsets, I think that 30-35 should be enough (I'd like to do some out-of-sample control).
Let see what happens. We are always in time to add more engines.
Of course, we'll need it when you will increase the number of testsuits.
<Marco>
Yes, of course.
(It's a pity I haven't PLS in my home software)
PS.- just an idea, to be verified: if I could find a student interested in the matter (with the help of an IM that is a colleague and a friend), I think it would be interesting an undergraduate thesis on relations among testsuites(general) and ELO. Don't take it as a promise, however...
Sure, I will make the datas available by tomorrow, or day after. I believe I've tested atleast 30 chess engines that are listed in either CCRL or CEGT.
I have few testing to do, I had tested Division 2 and Division 3 for STS 1 to 6. I have to test those engines in STS 7-8.
I hope this suggestion will happen. Looks like a team effort will prevail in creating this. One thought is to leave the "elo" tag out of the title as to avoid confusion with the actual ELO listings. These STS suites are much appreciated.
<Swami>:
Even if you have a limited results table, and it isn't a trouble for you, could you send me it, in order to prepare some templates for the true analyses?
Thanks. Enrico
If it's too much of a trouble for you, forget this mail
noctiferus wrote:<Swami>:
Even if you have a limited results table, and it isn't a trouble for you, could you send me it, in order to prepare some templates for the true analyses?
Thanks. Enrico
If it's too much of a trouble for you, forget this mail
Hi Enrico,
It's almost finished! Will send it to you through mail in few hours. Maybe I can post it here as well. Just send me the pm to let me know of the email address.
I have included WBEC, CCRL and CEGT ratings alongside the STS 1, STS 2.... STS 8 results for about 45 Engines in an Excel File.