PHCEL

[Rubinus]

Finale C ended.
Serie Nr.4 games/partie here.
PH-ELO

Today night launch finale B (8 cycles, 56 rounds)

[Rubinus]

Finale B ended.
Serie Nr.4 games/partie here.
PH-ELO

Today night launch finale A (12 cycles, 84 rounds)

[Rubinus]

Finale A ended.
Serie Nr.4 games/partie here.
PH-ELO

So far gauntlets LCZero id 359, 390, better new. Next bonus finale A 12+5, 14 rounds.

[Rubinus]

Finale A 12+5 ended.
Serie Nr.4 games/partie here.
PH-ELO

So far gauntlets LCZero id 495. Next launch serie 005.

[Rubinus]

Serie 005-2018. http://phcel.cz/ph-005-2018-2m2s/
CM personalities qualify. Promotion to swiss B Conqueror - 9/14, Leonidas and Aiglos II the same, but less SB.
games

[Rubinus]

Serie 005-2018. http://phcel.cz/ph-005-2018-2m2s/
CM personalities qualify. Promotion to swiss B Conqueror - 9/14, Leonidas and Aiglos II the same, but less SB.
games

CM_GM_personalities qualify. Promotion to swiss B Botvinnik and Capablanca.
games

[Rubinus]

Start serie 005-2018 swiss B. Test first round.
New: Nemorino, Baron, Deuterium, Vajolet, GNU Chess, Maverick, Mephisto Roma 32 Bit, WChess 1.6, AICE, RubiChess.
Test 1st round, in the event of a problem, participants may be replaced.
http://phcel.cz/
19 rounds, 48 promotion to swiss A.

[Rubinus]

Swiss B end.
table
games
phcel-ELO
Surprisingly stable tournament, no single interruption of crashes sw. I only had to change the result twice because some WB engines playing blacks outlined the rating from their point of view, and occasionally the GUI ended the part with the wrong result, which I did not think of Scidlet and another weaker engine.
I was most interested in LCZero, who was on the edge of the first half, but the second half managed well. The resulting rating is weak. LCZero as a tactical leak, but the locally aggressive engine against the weaker always loses, but against the better it uploads more. So it's likely to increase in A tournament, 501 is a pretty good network.
The first forty-eight runs, the order is according to ChessBase 14 with the parameters of the middle BH, BH, Progres, SB. Everyone with 9.5 points and more and nine points has the best escort criteria for E.T.Chess. 49.Rubichess, because I want the second LCZero in the tournament with a newer network. And as shown in the LCZero test, version 0.16 is bad for my computer so the new network will run with version 0.10. I do not want to rewrite the table here from the GUI 11 / Deep Rybka, but where the order is different, I write the red number ... Despite the fact that MCS was turned on, Scorpio was doing well, with previous version 2.8.4 making gross errors. What is the surprise about, I would probably not expect in the top 48 of Hermann and Aic - I remember from that time as a relatively fast fool. BamBam, for example, had unexpectedly hit points. On the other hand, I was expecting a better Mephisto, but who knows what the UCI emulator really is ...

[Rubinus]

Starting A_swiss 005-2018. It's mostly played overnight.
Tournament status

[Rubinus]

Aswiss_ph_2018_005_2m+2s
games
PH-ELO

32 qualifiers from the previous series (AE finalists, or their successors - newer or other versions), 48 of them advancing from Swiss B. The wild card received the best version of LCZero-id573 at the time of the start and the even number of participants was happy RubiChess 0.9 MB , 49th from the bull.
After the experience of the previous series, he returned to the tried 19 rounds. The red numbers in the order indicate the difference against the table generated by Rybka 3 (Fritz 11). He does not have a rating, so the actual ranking is found in ChessBase 14 and ranked by midBH, BH, progress and SB. Surprisingly, when the Stockfish version came out of 19 games, the rating was over 3700. This will of course be corrected after a long A finale. Direct participation in final A was won by Ethereal at the expense of Shredder. Enough new releases recorded a lower rating than the old one, such as Komodo, Raubfisch, Fire, Andsacs, Laser, but if there are 50s, guaranteeing the H finale with a theoretical chance to progress to A, there will be a chance to fix it. There have also been some jumpers for whom it will probably be difficult to defend the PH-ELO, except for Arasan, including Ethereal. Against the fall in the last series, Rybka improved. Finally LCZero. Even his two nets were much better than last but it was not enough to advance to the final group even though there was only one Buchholz in the better 573.
serie 005/2018 www
TOP 30