On this news page, You can read all my postings about mainly SCCT (3m+1s) MP Rating
30.01.2022: As last entry is Cheng 4.41 engine...managed to get 3050 Elo..not bad... Invictus r382 test is cancelled...due to often crashes...no any idea why.. Plus SlowChess 2.83 test is cancelled e.g without NeuralNets file suffers..
29.01.2022:
I have good news, the current chess engine rating is already resumed... Depending on my free time, I may test new engines, but not so frequent In short, in case of often engine releases..I can test in every 4-5 months And as last new entry is Combusken 2...managed to be in Top 10 ranks Congrats to M. H. Bartkowiak really not so easy to gain so many points! Btw, in other ratings: NON-NNUE ranks are much lower than deserved... Due to many chess engines are played via Evalfiles, not so fair for sure.! And as usual, only those engines are allowed which are capable to ponder10.01.2022:
A new engine competition, but this time as NON-NNUE, for more Details Really impressive...as we see, Stockfish via Contempt 40 is much better.. Plus via Cont. 40, my used engines (in book tours) produce less draws !! Sure.. I'm not expert on engine programing...perhaps I have to give a try..) And in case...instead of C40, I am considering to call it as Stockfish SC )) Because to gain over 40 Elo than default...it is not so easy as it looks...!) And what a pity that no one still congratulated Me..sad...on other hand: My fingers are already tired...e.g congratulating the other Number Ones)) Btw, there is a such Russian saying:, which perfectly suits over this issue: Сам себя не похвалишь, никто не похвалит that means in English: If you don't blow your own horn, someone else will use it as a spittoon. Ok...for right now current competition is paused, due to I started CS LIX09.01.2022:
As usual, I am pleased to announce that,
I started to work over new exciting project: Results
For anyone missed... and as I stated earlier, during this year:
As far as possible, I plan to test mainly NON-NNUE engines!
Some Notes,
My main target is to test the Top engines, but with less help!
To be more clear, via huge assistance: any fool can manage)
As you may see, as usually I prefer short openings in lines,
4-MEN EGTB...but where this time the Evalfiles are excluded...
In short, via current new conditions: just pure original chess!
Otherwise, would not be so much boring ?! still not so clear?
Forget NNUE, e.g via deeper openings, more endgames etc.
Mostly of the Top engines perform much better for sure...so,
Giving some assistance is all-right..but when too much..sorry..
Btw, just one example came on my mind...not about chess,
But I think that perfectly suits over these assistance issues...
I have another hobby (since childhood): Racing Pigeons!
And they are as one of most intelligent birds on the planet!
Continuing...e.g when I race my pigeons from long distances..
Via testing method: all together (all in once) then as usually,
Mostly pigeons are easily oriented..due to all flying together...
Because the pigeons from each other are getting assistance...
Plus in this way, it's not so clear which one comes as 1st..!)
But if they are released one by one, in every 15 minutes etc.
Then many have troubles to be oriented (I mean in short time)
And what does it mean ? sure it means a lot..to see in reality:
Which one is best: definitely less assistance/help required!
Otherwise, not so logical and what's point of these testings?)
Meanwhile,
Let me say please a few more words in case of via Evalfiles...
I see NNUE's revolution something like as Lenin's revolution )
E.g nowadays all things are more equalized, and I think that,
The online players with high-end machines are affected a lot...
But who are owners of low-end machines, probably are happy )
And according to my testings (if Top engines play each other),
To see a win, something like, looking for a needle in a haystack!)
Yes, it's very difficult to appear a win...sure under serious cond.
But just in case if we use 1 core, ponder off, weak openings etc.
Or via ultra fast time controls or vs weaker opponents...we can
See more wins...but unfortunately for nowadays I have more
Important things to do...that's why once more I wish to say:
Impressive, but boring... and thanks for your understanding...
For more information over these issues (about NNUE): Details
Ok..let's come back to current results of 1st days of 2022 year
Note that so far, all of the engines are 64-bit, where this time
Instead of BMI2 versions, only POPCNT vers. are used to play,
Due to some of the games are played on AMD Ryzen 7 4800H
Plus in this way is more fair right ? no much speed advantages...
E.g AMD's games are at 2m+1s, due to is approx. 25% faster...
Btw, I included some of previous played games too e.g: Results
And as alternative, some matches played under Fritz 14 GUI too
Anyhow, mostly are played on 2x E5-2686, Arena GUI, via 3m+1s
I picked as participants, e.g which are less than 60% in similarity
Exception, it's allowed...but just only in case of 100+ Elo stronger...
And the current engines are less than 5 MB (uncompressed size)
Note also that all engines are played via default Contempt settings
Exc. only Stockfish 310720 engine is used to play via Contempt 40
Later, I realized to test Stockfish 310720 via default Contempt too
And as we see, the influence is huge, sure in favor for Contempt 40
Best Regards, Sedat Canbaz