I do not agree with that.

Other engines are not supposed to be changed the TC dependent of te openend and the time control selected.

So default has to be used.

Yes exactly !

Silvian, please dont dinky weep.  *wipe your tears away* ... its only 120 dollar.

The point is, that R4  seems not the result of two years of work. I have the feeling that playing strength was  intentional reduced because of future plans with rental rybka.

Check the CCRL or CEGT rating lists and see the uncertainty for numbers of games between about 50 and 90, and you'll get a pretty good idea.

Yes, though many of the statistical uncertainties present in computer play are negated by extreme preparation and concentration during the games.  Computers don't have this "sense", so I think that the true uncertainty of human matches is much less than one would obtain from the statistical analysis.

In my view human world championships are for mere entertainment only, they're pretty much worthless for determining who the stronger player is.  Even if these championships ran for 1000 games you cannot know who the stronger player is from that match alone.  Just because B beats A doesn't necessarilly mean B is stronger than A.  B could perform worse against C, D, E, F, and G, making A the superior player.

> Rybka 4 was released with some settings, why not test like that, what we bought ?

Didn't you buy the option to change the settings? Do you always leave the hash size as it is set by default

I think it's important to know as much as possible how the default settings perform, but results for different settings are also interesting (as long as they are based on enough games). However, different settings are hard to evaluate unless you know how the standard version performs under the same conditions.

> You dont have to read my post if they are so anoing to you.

It's nothing personal, but generally speaking I just think some people are complaining too much. People have the right to complain and offer constructive criticism, of course, but I think some of these complaints are unreasonable. Sometimes I'm left with the feeling that certain people here believe they can make demands from Vas, as if he was owing them something...

> I bought it because I trusted in Rybka blindly since version 1.

I can understand why you feel disappointed if you expected a much bigger improvement, but perhaps your expectations were too big? The kind of improvement you saw going from R2 to R3 isn't commonplace, and shouldn't be expected. Besides, purchasing a product blindly isn't always a good idea...

> I also read the posts where it was said that it should be at a minimum about 60 ELO better.

I think it's too early to draw conclusions. So far we have (AFAIK) only blitz games to base our estimates on. And these results vary from around 60 (IPON) to over 100 points (CCRL). More results at longer time controls will become available over the next few days, and should give us an indication of how strong R4 is at tournament levels.

If it turns out that R4 is actually 60 points stronger than R4 at tournament level, it would still represent a nice improvement over R3 in my opinion.