Wow, acording to this thread

http://www.talkchess.com/forum/viewtopic.php?t=22342

the antique Rebel 12 has an elo rating of 3621 on a modern Core2 quad core.

Wow....! Why do people bother messing with Rybka then??!

(laugh)


Obiviously it's just a miscalculation based on a simplistic strength
estimator in Rebel.

It probably did a simple speed test, and did a "Tech Curve" estimate.
Something like "I'm 32 times faster than my baseline speed, so I'll add 200
points for each doubling of speed.", and it instantly went from 2700 to
3600.

It clearly shows how you can take a program with a real rating and then
start estimating strength increases (due to clock rate or some other factor)
and end up with an absurd result.

(For the record, the current SSDF list shows it at 2678 when run on their
baseline Athlon 1200mhz system. I would expect modern Core2 systems to run
it fast enough to get another 100+ points of it, though.)


I think most people in here can probably figure out why I bothered to post
this....




----== Posted via Pronews.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.pronews.com The #1 Newsgroup Service in the World! 100,000 Newsgroups
---= - Total Privacy via Encryption =---