Having enjoyed my time contributing to both CEGT & AEGT testing groups, it is time for me to branch out and do some testing of my own! My main testing objectives are interest and  F.U.N  so you will find a selection of testing and tournaments that achieve these goals for me. Hopefully there may be something here that interests you also! .

ChessMaster Grand Blitz Tourney  (Started - 26 December 2005)

< now on a dedicated page > 

Fritz versions test -     5 August 2006

Fritz has always been one of my favourite engines, (I was first introduced to Fritz when I bought Fritz 2 at the time of the Kasparov/Short World Championship match). Fritz in Bahrain (a special and later edition of Fritz 7, released to celebrate the 'Brains in Bahrain' man v machine match) always interested me as I saw in several rating lists that it seemed to perform really well (anecdotally I had always felt that the original Fritz 7 was just as good if not stronger than Fritz 8, especially at blitz time controls).

So, I thought I would run a little tourney with Fritz 7 and Fritz in Bahrain both running a gauntlet of the same 10 engines in a 20 game match (ie making a total of 200 games). For fun, and to assess progress (and the benefit of an additional CPU!), I also put Deep Fritz 8 through the gauntlet.

A tourney of only 200 games is not enough to make statistically sound conclusions but it was interesting to run and you can find the results here.

Shredder 7.04 processor test -   21 July 2007

I still enjoy playing with and testing older versions of both commerical and freeware engines. It is perhaps relatively little known fact but prior to the promotion and separate release of multi-processor (or 'Deep') versions, some engines already had the capacity to use more than one processor - Shredder 7.04 (the free upgrade to Shredder 7) was one of these engines.

I have long been meaning to run a short test to see the impact of the additional processor on the performance of Shredder 7.04. Also it would be interesting to see for a relatively early version of a 'Deep' engine how effective the 'scaling' impact is in terms of overall performance strength. Generally a 50 - 80 elo improvement might be expected in moving from 1 to 2 processors. So how will good old Shredder 7.04 fare?

Usual test conditions apply - in this case, Shredder plays two gauntlets, first with a single processor and then using two processors. Games are played against the same 12 opponents in a 20 game match, making a total of 240 games (of course more games would provide an even clearer result!).