test result with a difference

yesterday i tested ivan b63 b10c vs firebird 1.2

5 min games .no books.win-7 32 bit, core2duo

in the test ivan b63b10c shows good result with firebird 1.2

but when the piece value of firebird changes like queen value 900,
bishop pair value 46,knight value 300 ,bishop value 310..
with this change value firebird plays better than ivan b 63 b10 c ..

plz test it yourself....also to say that this mean that we can either change ivan value to this also to get good performance too i never changed it ..but any one can test it to find whether ivan too play better with those changed value also to say that while making engine these value has an effect on engine performance so while loking at it it means ..with the changed value firebird playes better and its also that ivan too can play better with the changed ....so i hope the complier must look at the changed value and more precise value to make their engine work strongest and better..


thank and regards

I have noticed that Firebird 1.2 is very tunable, and produces good results when tweaked as mentioned above. Just be careful, because tweaking in the wrong direction can really make it bad too. I found that Bishop pair higher seems to be better, along with a slightly higher Bishop value compared to knight. Yet, this is favored also in more open positions. When playing very closed positions, it becomes worse. Choose your opening book carefully when playing tournament!

dont understandes your changes.why bishop value 310 and not 318 and same thing knight value why 300 and not 310 and how many games you played with those changes?is that based on any roles?

There is not enough statistical evidence that any setting is better than another. It is perhaps better for 10 games, but for the next 10, it might not be. This is why it is necessary to test 1000's (and more) of games before we can draw any conclusion. I think we are at a point were small changes in settings favour certain postions (open vs. closed) and if you say version X is better than version Y based on small amount of games, we are making fools of ourselves.

good posts , nice to see constructive debate from members at last...more please