Author: Uri Blass
Date: 15:23:05 12/04/03
Go up one level in this thread
On December 04, 2003 at 17:35:47, Kurt Utzinger wrote: >On December 04, 2003 at 10:51:22, Sally Weltrop wrote: > >>On December 04, 2003 at 08:56:52, Brian Katz wrote: >> >>>On December 04, 2003 at 06:03:41, Kurt Utzinger wrote: >>> >>>>On December 04, 2003 at 00:36:53, Brian Katz wrote: >>>> >>>>>Is there anyway to use Fritz 8.0.0.8 vs 8.0.0.23 in the same Engine folder and >>>>>know which engine is which? >>>>>You can place both engines in the same folder, but they both show up as just >>>>>Fritz 8. At least with the old Fritz 6 upgrades, the new engine would become >>>>>Fritz 6 and the old engine would stay in the engine folder with the name >>>>>converted to Fritz 6 Old. >>>>> >>>>>At least you could then play both engines against one another in the same PC. >>>>> >>>>>Any info would be welcome. >>>>> >>>>>Thanks, >>>>>Brian >>>> >>>> It can't be done and by the way it does also >>>> not make much sense to play old vs new version. >>>> Kurt >>> >>>Thanks >>>Actually it does make a lot of sense to play old versus new. Just as much sense >>>as playing any engine vs engine. >>>There are always questions regarding whether or not an upgrade has made an >>>engine weaker or stronger. Such is the case (and has been for quite some time >>>now) with Fritz 8.0.0.8 and Fritz 8.0.0.23. >>>You can play endless games with both Fritz 8 engine versions vs other engines >>>and still not know.............Or you can play a number of games between Fritz 8 >>>versions and get a pretty good idea whether or not the upgrade has increased or >>>decreased the performance. >>> >>>I think testing it this way would be much a much more efficient way of running a >>>test. And Quicker!!! >>> >>>And above all, what does it matter, if one enjoys testing chess engines in this >>>way or not? To say it does not make much sense is just a matter of personal >>>opinion. What ever one enjoys and makes you happy is all that really matters. >>>This is only a hobby for most of us anyway,is it not?? >> >>I think it makes sense as from what I have seen from the past years. Many >>versions are released and sometimes they are weaker. >>>Thanks, >>>Brian > > My experience with testing CM9-settings and as a beta-tester > for a specific (here not mentioned) program shows the contrary. > "Engine X default" gets a 75 % score vs "engine X version a" > but "engine X version a" obtains much better results vs other > engines like Fritz and Shredder. Testing vs other engines is > in my opinion usually more efficient than against another > version of the same program. And if you get better results > vs other engines, it is absolutely of no interest that > "engine X default" plays better vs "engine X version a". > For these reasons I have stopped testing an engine with > different versions against each other. > Kurt You may be right about changing weights in the evaluation but as a programmer I think that the main gain that I can get is not by changing weights and a good change should make the program stronger against everything. At this point I think that improving the search is the most important thing. Improving the evaluation is also important but I believe that the main gain to get is by adding knowledge and not by changing numbers. Uri
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.