And We Have the First Freeware Results!

[geots]

No reason for anyone to get their drawers in a wad. I am just merely making an observation- there is no blame- because NOTHING WRONG has been done. Let me say it once more: NOTHING WRONG has been done.

I spent quite a number of hours picking the engines for these 14 Groups of freewares that are playing in RRs. 212 freeware engines- to be exact. I went by CCRLs rating list because it is a good list, I know my way around in it, and I am comfortable with it.

I used the update they put out Aug. 14. If someone had taken my list, I promise you there would not have been one engine anyone would have found that "should have been in another group." Because I double checked- and then I triple checked.

Now this past Saturday (14 days later) they come out with the next update. So far, almost- and maybe all- engines I have checked from 2000 to 2500 have all had their elo increase somewhere between 50 and 100 points. In theory, there aren't enough computers in "Talk Chess" to borrow to play enough games at 40/40 to make the huge changes in all these elos that has taken place. I didn't say the huge elo changes stopped at 2450 or 2500. I said that was as far as I have checked.

And it is as far as I will check, because there are only so many hours in a day. I have to assume that some kind of change was made in their rating system- because trust me- nothing has changed by games played that should mean that now Quark and TRACE don't belong in the same group.

I still stand by my list from Aug. 14. Because it came from CCRL- obviously a group of guys I would trust with my life. I do not believe this will cause any problems in even 1 of our RRs, but if someone finds a few engines that since whatever changes- should be in another group- look, life aint perfect. Live with it is all I can say.

I will point out that most of the engines I mention so far haven't even played in the last 2 weeks, unless Graham was running them. And tho the huge changes, I have not seen ANY leap over others. They may be a little closer or a liitle more distant- but that is about it.

If you ask me if I am going to take this new list and compare it engine for engine with the update I used- don't even ask. I am going to play on like nothing happened, and by this time tomorrow 3 more RRs will be running on my xp pro quad, and maybe a 6th one all at once if I add another to this intel i5.

I merely mentioned this because if something unusual a bit comes up between 2 engines elos- I have told you why.

One last time: CCRL is a fine group- they have done nothing wrong. I am not asking because I don't want to know why all these elo changes. All I have to know is- if it was done- they had a good reason. And that is plenty good enough for me.


Until then,

george

[Modern Times]

Apologies for that. We have been trying different bayeselo parameters as suggested and discussed at length in the technical forum and here. We weren't happy with one of them (or maybe the combination of them) so all 3 lists were re-calculated last weekend with one particular parameter reverted back to the default value.

[geots]

Apologies for that. We have been trying different bayeselo parameters as suggested and discussed at length in the technical forum and here. We weren't happy with one of them (or maybe the combination of them) so all 3 lists were re-calculated last weekend with one particular parameter reverted back to the default value.

No problem at all, Ray. Like I told them- you guys don't do anything unless there is a good reason.


Best,

george