Author: Rolf Tueschen
Date: 08:13:20 01/15/06
Go up one level in this thread
On January 15, 2006 at 11:04:26, Albert Silver wrote: >On January 15, 2006 at 10:57:25, Rolf Tueschen wrote: > >>On January 15, 2006 at 10:23:33, Albert Silver wrote: >> >>>On January 15, 2006 at 10:08:23, Rolf Tueschen wrote: >>> >>>>On January 15, 2006 at 09:38:02, Albert Silver wrote: >>>> >>>>>On January 15, 2006 at 09:22:48, Rolf Tueschen wrote: >>>>> >>>>>>On January 15, 2006 at 08:43:13, Albert Silver wrote: >>>>>> >>>>>>>On January 15, 2006 at 02:07:06, Marc Lacrosse wrote: >>>>>>> >>>>>>>>> >>>>>>>>>Lacrosse's analysis showed above all that in the 87 positions he tested, that >>>>>>>>>Shredder 9 and Rybka scored 57% given 10 seconds, and Fruit and Toga and company >>>>>>>>>are much weaker with so little time, and thus much weaker in blitz. >>>>>>>> >>>>>>>>> >>>>>>>>> Albert >>>>>>>> >>>>>>>>Just a little point, Albert. >>>>>>>> >>>>>>>>What my little experience shows is not an argument for telling that engine A is >>>>>>>>better or worse than engine B at faster or slower time control. >>>>>>>> >>>>>>>>What I precisely did is the following : >>>>>>>>let say : >>>>>>>>- engine A solves "x" positions in 180 seconds and >>>>>>>>- engine B solves "y" positions in 18o seconds. >>>>>>>>I recorded: >>>>>>>>- what percentage of "x" engine A had already solved after 10 seconds >>>>>>>>- what percentage of "y" engine B had already solved after 10 seconds >>>>>>>> >>>>>>>>So each engine is compared at 10 seconds with the number of positions that it >>>>>>>>will solve _itself_ at 180 seconds >>>>>>>> >>>>>>>>So when I record that Rybka has a 57% score and Fruit a 39%, this does _not_ say >>>>>>>>that Rybka is "stronger" or "weaker" than Fruit, and we could have a much weaker >>>>>>>>1800 elo engine getting a 80% (or a 15%) score in the same test. >>>>>>>> >>>>>>>>What the little test tends to show is just that rybka has already shown 57% of >>>>>>>>its own analysis capacity at 10 seconds whereas Fruit has a larger margin of >>>>>>>>improvement (compared with itself) when given a larger time control. >>>>>>> >>>>>>>Actually, it doesn't even show what you suggest, that Rybka has already shown >>>>>>>57% of it's capacity in 10 seconds, and as a consequence I'm afraid your >>>>>>>conclusions are incorrect. >>>>>>> >>>>>>>The positions you tested with have definite solutions I presume, thus once that >>>>>>>solution is reached there is no room for improvement. How can you claim that >>>>>>>Rybka cannot improve its analysis when the positions you gave it cannot be >>>>>>>improved upon after the solutions are found? >>>>>> >>>>>>Please read again: "larger margin". Does it mean "cannot improve"? >>>>> >>>>>It can't have a 'larger' margin of improvement if it is being compared to >>>>>something with no possible margin. >>>> >>>>This is another claim but of course you cant say anything at all if the problems >>>>are too easy to solve. >>> >>>It makes no difference whatsoever. >>> >>>>- Let me repeat what i wrote to Marc. His little >>>>experiment and discussion proves that he has done something valuable. >>> >>> >>>Ok, if the value of the experiment is in the debates that it created, then fine. >>>But the numbers and results have no value whatsoever if one wants to draw any >>>conclusions from them. >>> >> >>No way. You cant say that. This WHATSOEVER is wrong and destructive towards >>Marc. > >Not only can I say it, I even showed why. What does Marc have to do with this? >I'm talking about the data and the inferences drawn. > > Albert Ok, if you think that it wouldnt bother him if he had made experiments without any value whatsoever - then you can make such conclusions. I agree.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.