HAL9000

HAL9000
"It just isn't conceivable that you can design a program strong enough to beat players like me."

October 10, 2015

Android UCI engine update: Stockfish Dev 2015-10-08 JA

In parallel to TCEC-8 progressing stage by stage until an unexpected result, every stage becomes a pool on speculations whether Stockfish can take back the crown from Komodo.

No doubt, the head-to-head race makes the development compiles attracting. People are insterested to test new versions each and every day before the official update comes. Unstoppable incredible hunger...

Some days ago, we have seen 2015-09-07 version of Stockfish compiled by Peter Ă–sterlund. It's in my repository available for the most hungry ones :-)

After a little while, it's now too old as Jim Ablett compiled another dev't version dated 2015-10-08. This one is interesting because we have an arm7 compile from JA for the first time.

Remembering that a previous close tourney between Stockfish compiles revealed that JA compiles perform very well, the arm7 compile from JA might deserve attention.

If i have enough spare time remaining from endless working hours, i want to play another close tourney between Stockfish 6 official, dev 2015-09-07 and dev 2015-10-08.

Or should i push back my appetite and remain patient until Stockfish 7 official release comes?

Download Stockfish development 2015-10-08.JA arm7: HERE

Download Stockfish development 2015-10-08.JA arm5: HERE

New Android UCI engine: Laser 0.2 beta JA

Put your laser lights on! This one makes yet another Android engine. The fleet keep growing and my smile gets more visible.

Welcome then Laser 0.2, written by Jeffery An, compiled by Jim Ablett.

To use it with Chess for Android, don't forget to follow the uci2xb procedure and don't rename the ini file. You may use ini file settings however.

Download it: HERE

Android XB engine update: Chenard 2015.08.15 JA

Chenard, currently one of a few stable and trustable XB engines under 2000 ELO, just got an update over the previous version of July 2014.

I just can't wait to check and see how forward the gameplay has stepped.

Please note that the filename is based on the source code update date in the download link below, not the date of the compile as used by Jim Ablett. However the binary is the same: Commit number 5102f59f52

Download Chenard: HERE

Android UCI engine update: Bitfoot 15.09.22 JA

Bitfoot is the bitboard version of Clubfoot engine. Almost 200 ELO gap proves that the bitboards lead to stronger play.

Below binary compiled by Jim Ablett is based on the source code of September 22th. I've prefered to rename it according to the commit date instead of Jim's Oct 10th. This is always the same binary and the unique code of the source is d9aeb437d6 in github.

Enough said, HERE is the Android binary to download and try.

Android UCI engine update: Clubfoot 15.09.07 JA

Clubfoot is approximately 200 ELO wekaer than its big bro Bitfoot. However, its development is ongoing. Now, we have now a recent compile for Android based on the source code dated Sep,7th.

Once again, i've renamed the engine binary using the commit date instead of the compile date used by Jim Ablett. The exact commit number of the source is 048995de70.

Please note that Clubfoot only played fixed time per move games in Rapidroid. In case the new version plays Fischer clock games without problem it will become a passepartout engine and will use both time controls. That means the oppoonent portfolio will be enlarged and the accuracy will increase.

Download the new Clubfoot: HERE

Android UCI engine update: Sayuri 2015-10-01 JA

Another engine i could not succeed to run flawlessly in my Rapidroid experiment used to be Sayuri. It's a medium strength engine which is frequemtly updated by its author Hironori Ishibashi.

What makes Sayuri special to me is its ability to use multi threads. I'll be glad to recall it for duty and get it rated soon whenever the compatibility tests are okay.

Jim Ablett named binary according to his compiling date but i've choosen to rename it using the commit date of the source. It doesn't matter anyway unless nobody else compiles Sayuri. Let's note that the commit number is exactly 2b47cdd70a.

Download Sayuri for Android: HERE

Android UCI/XB engine update: GreKo 13.1 JA

Although i'm stuck on v12.5 in Rapidroid, GreKo is updated to 13.1 for Andoid by Jim Ablett. Previous 12.6 and 12.8 compiles were hanging randomly during my engine tournaments. I wish 13.1 doesn't repeat this, so that i can rate the most recent one.

Under Android, use GreKo only in Winboard protocol mode. Therefore only with Chess for Android, since it's the sole GUI compatible with it.

Usage of the ini file and the book is not mandatory. However, the ini file works and can allow selecting higher hash sizes at least.

Downlaod GreKo 13.1: HERE

Android UCI engine update: Protector 1.9.0 TCEC JA

Protector is no joke at all. It's one of the leading engines and it's capable to beat monsters like Houdini 4. TCEC-8 is the proof!

Unfortunately, i could not rank the previous Android compile 1.8.0 for frequent bad termination issues. I hope that new compile heals the problem.

To use Protector under Droidfish, use the UCI engine directly and for Chess for Android, follow the uci2xb method explained previously.

Download and try Protector 1.9.0 TCEC for Android: HERE

New Android UCI engine: Mchess2 1.02.00

Mchess2 is a UCI compatible chess engine written by Michael Jorgensen from Denmark. I have no idea yet how strong it is. Probably not too much at this step but even under 1500 ELO, i'll be happy to include it in Rapidroid in case it works without problem sÅŸnce an accurate rating list needs engines from any ELO range, especially balanced among all ranges.

You can download a compile made by Jim Ablett HERE and give a try on your Android device.

Android UCI engine update: Gunborg 1.39.JA

Gunborg 1.0 is updated to 1.39. Thanks to Jim Ablett we have now a recent version in hand.

The working structure is the same: If you want to use it with Droidfish, the engine itself is OK. If you prefer Chess for Android, proceed the uci2xb method and avoid renaming the ini file.

Download Gunborg 1.39.JA: HERE

October 6, 2015

Android UCI/XB engine update: Galjoen 0.30.2.JA

Galjoen just got an update compiled by Jim Ablett for Android. I remember that the previous one had issues with Fischer clock and it could not play fixed time per move either, using too much time beyond toleration. Therefore, it wasn't certified for Rapidroid.

Now it seems things changed. The updated version, after the first two games where it faced the native engine of Chess for Android is confirmed OK for Fischer clock.

Fixed time per move on the other hand is still open to speculation. It started by using 16 seconds vs 15 selected but with every further move the usage increased constantly up to 20 seconds. I didn't yet decide whether i should certify it for that time setting or not.

Probably it's wise to avoid tolerating excess time usage beyond 10%. We'll see.

The engine doesn't look strong yet, It won above mentioned match by 1.5 to 0.5, meaning that it plays way below 2000 ELO, maybe around 1600.

Download Galjoen 0.30.2: HERE

Android UCI engine update: Stockfish Dev 2015-09-10 (towards Stockfish 7)

Each hot championship is a source for new public demands. Since TCEC-8 has started last month, chess freaks didn't stop asking for a new release of Stockfish. Version 6 was released months ago and development version have already progressed a lot, although it's not yet considered enough to release an official update.

On the other side on the medallion, Komodo 9.2 has already fulfilled the expectations so far during TCEC-8. The reptile currently leads the Stage-2 ahead of the fish, somehow making me smile because i still remember the speculations stating that TCEC had to be delayed by the organisers just to make sure a recent version of Komodo stronger enough to cope with the latest Stockfish version becomes available.

Be it true or not, the tournament is in place, running at full speed and i'm not too concerned about possible commercial timings and manipulations. Both teams deserve respect, no matter the status, commercial or open source.

Computer chess can't progress without competition after all. Besides, i wish there could be more candidates to chase the leaders. Unfortunately there's only two engines and all others following far behind. Isn't it sad to see Houdini 4 still playing for 3rd place after so long time?

Regarding Android side of the two musqueeters on top, Komodo 9.2 is already available while Stockfish-6 was the latest among the fishes, until i heard of an alternative development version compiled by Peter Ă–sterlund, the author of Droidfish and his own engine Texel.

I see no reason for that development version play weaker than v6, given that Peter's previous compiles performed very well and that it's based on arm7 or 8 already.

I will schedule a head-to-head match between that dev't version and v6 to see what's the gap as soon as i can. Rapidroid ranking will wait for an official release anyway, regardless of the result.

Meanwhile you may download the package which includes all possible versions compiled and test for yourself: HERE


October 4, 2015

Schachcomputer.info rocks!

If i suggest people to check out Rapidroid for many programs based on Android, i would recommend Scachcomputer.info for any kind of information related to table top chess computers, from the oldest one to the most recent.

This is a very old network of dedicated chess computer freaks, mainly from Germany, establishing tourneys and games between their computers since years, online or over the board. A great dedication i admire a lot!

Their two rating lists are based on tournament time controls and also rapid chess which they call Activschach for speed chess. FIDE had forgotten long ago that speed chess term but it still counts for computer chess, using 30 seconds per move.

One thing to keep in mind is that they never include any engine other than those installed inside a touchable, dedicated chess unit! That's the main reason there's nothing like 3000+ ELO in there.

The lists are comparable to SSDF in terms of calibration, therefore they should reflect traces of human vs machine games of the past. Similarity with FIDE scale is no surprise either.

For a very detailed forum and a trustable ranking, don't forget to:
1) Bookmark the link of THEIR WEBSITE
2) Take a look at the latest rating list releases based on SPEED CHESS and TOURNAMENT TIME CONTROLS.

RAPIDROID RELOADED: September-2015 with Komodo 9.2


Although this release took a few days of delay, it was just because of a hard working week. Anyway, i believe the improvements compensate largely.

The highlights of this release are:
* UPDATED Komodo 9.1 to Komodo 9.2
* UPDATED Zurichess Basel to Zurichess Fribourg
* ADDED Fridolin 2.0.0
* ADDED Rybka 2.3.2a MP as new Windows engine
* ADDED The King 3.50 x64 as new Windows engine
* ADDED Novag Secondo as first tabletop computer engine
* ADDED 50 ELO to all Android engines for better calibration

THE TOPPER:

Komodo 9.2 is clearly stronger than 9.1, in contrast to my initial expectations. After a short hesitation, i replayed all previous Komodo games and 9.2 proved itself. It seems the reptile will stick to the top for a while, at least until Stockfish 7 kicks in.

As mentioned previously, the list needs good calibration vs CCRL and SSDF to be as readable as possible. I always knew it's very complex without the same hardware configuration for Windows engines and use them as trustable calibers.

AMD Athlon x2 4600+ used by CCRL and Intel Quad Core Q6600 used by SSDF have both disappeared. The only solution remaining is to find an acceptable correlation between my Intel Core i7-620M, which has also disappeared from the market long ago.

Past comparisons in benchmarking sites and some available tools show that i7-620M performs better than Athlon and slightly lower than Q6600 with MP. Therefore Windows engines must collect as much ELO as in CCRL but not more than SSDF when tested against Android engines, simply and logically.

Remote connection feature of Chess for Android worked well between Samsung Galaxy Note II and my notebook except for the gauntlets. No way to run like this.

I've had to alter my pairing method to increase automation, by playing vs each Android opponent one by one in separate gauntlets using a collection of related openings instead of consecutive rounds with various opponents using same opening for whole round.

The network engine feature works okay but needs improvement to become more effective. That's a topic i'll come back to in details in a future post.

WINDOWS CALIBERS:

Both Rybka and The King seemed to have some flaws in the calibration task:
Rybka uses hypertrading in MP mode. HT divides each physical core to two logical threads and virtually doubles the total number of threads. Unfortunately, the increase in nodes per seconds doesn't deliver the expected strength increase. HT is not efficient when used with chess engines. I've chosen to tolerate the efficiency loss in Rybka case and keep following because on my notebook there's no option to disable HT via BIOS.
The second caliber is The King. It works well under Arena as a Winboard engine via KingUCI.exe which acts as wb2uci interface. Threads and hash are selectable via wb2uci.eng but strangely the engine hogs the CPU for no reason unless i run it with only 1 thread ouf 4 available. When i use all 4 threads, 3 of them stay loaded after King makes its move. I made sure it's not because of pondering because Fritzmark running in parallel hits the usual score. There must be something wrong keeping the "freed" threads busy. Since it's not comfortable to hear the cooling all the time, relying on 1 CPU rating of CCRL looks better. However, i suspect 1 thread out of 4 on my notebook is not exactly the same as 1 core out 2. It should be something between 0.5 to 1 CPU.

Finally, i've decided to add 50 ELO to all Android engines. Rybka and The King are both ranked well within acceptable error margins for the moment.

In next steps, i'll need some single thread Windows engines to bypass the hyperthreading effect and compare better with other rankings.

SWEET MEMORIES:
If one wonders why Novag Secondo is in the list as one the weakest dedicated chess computers ever sold, i must state it's a very old friend to me. It was my first chess computer in 1993 and still works without problem in 2015!!!

So i could not keep myself from putting it into the pit after all these years. After all, my retro companion needs more weak opponents to move forward a little bit.

BAYES RATINGS AFTER 12088 GAMES PLAYED BY 121 PROGRAMS
Rnk Name                     T O/S  Elo   +   - Gam Sco Oppo Dra
001 Komodo 9.2               4 A32 3287  45  42 236 82% 3037 28%
002 Stockfish 6              4 A32 3273  44  41 236 80% 3039 30%
003 Critter 1.6a             4 A32 3108  37  36 240 58% 3055 47%
004 Sting SF 4.8.4 JA        4 A32 3078  39  39 240 59% 3007 29%
005 Firenzina 2.4.1 xTreme   4 A32 3074  38  38 220 51% 3063 46%
006 BlackMamba 2.0           4 A32 3053  37  37 238 51% 3047 45%
007 Rybka 2.3.2a mp          4 W64 3009  90  91  40 45% 3051 35%
008 Texel 1.05               4 A32 3008  39  39 242 52% 2991 32%
009 DeepSaros ver.2.3f       1 A32 2995  38  38 246 55% 2957 32%
010 Senpai 1.0               4 A32 2956  40  39 236 54% 2931 28%
011 RobboLito 0.085e4l       4 A32 2937  40  40 230 50% 2938 30%
012 Cheng 4.39               4 A32 2922  39  39 234 50% 2928 32%
013 Hiarcs 13.71             2 IOS 2917 115 111  28 61% 2838 21%
014 Shredder 1.7.0           2 IOS 2883 109 108  26 54% 2863 46%
015 Gaviota v1.0-d           4 A32 2858  37  37 256 49% 2872 29%
016 Hakkapeliitta 3.0        1 A32 2846  38  38 248 50% 2857 31%
017 Arasan 18.0              4 A32 2841  38  38 240 48% 2864 32%
018 Exchess_7.71.beta.JA_xb  4 A32 2838  39  39 228 49% 2861 34%
019 Grapefruit 1.0           4 A32 2816  36  36 250 49% 2822 41%
020 Cyclone 3.4              4 A32 2789  37  37 240 49% 2799 37%
021 DiscoCheck 5.2.1         1 A32 2781  38  38 246 48% 2796 33%
022 Deep Saros 0.9           4 A32 2759  37  37 244 48% 2769 33%
023 Toga II 3.0              1 A32 2752  38  38 232 47% 2776 33%
024 Deuterium v14.3.34.130   1 A32 2733  37  37 238 55% 2695 37%
025 Bobcat 6.4b              4 A32 2723  38  38 246 49% 2734 23%
026 GNU Chess 5.60-32        1 A32 2714  39  39 230 54% 2686 28%
027 Doch32 1.3.4 JA          1 A32 2710  38  38 236 46% 2741 32%
028 Chess Pro 2016.02        2 IOS 2705 112 114  22 45% 2733 55%
029 Fruit reloaded 2.1       1 A32 2703  38  38 220 52% 2695 39%
030 Scorpio_2.7.7.JA_xb      1 A32 2688  42  42 198 51% 2690 31%
031 Murka 3 ARM UCI          1 A32 2687  35  35 258 50% 2689 36%
032 Strelka 5                1 A32 2658  38  38 230 54% 2629 30%
033 The King 3.50 x64        1 W64 2653  48  49 144 41% 2717 31%
034 IvanHoe 9.46b            4 A32 2652  39  39 232 44% 2700 28%
035 Crafty_24.1.JA_xb        1 A32 2623  38  38 244 51% 2614 25%
036 TheMadPrune 1.7.04       4 A32 2620  41  41 204 46% 2649 29%
037 Rodent 1.7 build 1       1 A32 2603  38  38 234 48% 2617 31%
038 Tucano_5.00.JA_xb        1 A32 2595  42  42 190 47% 2616 36%
039 CNVCS 1.2.0              2 IOS 2584  94  95  36 46% 2607 36%
040 RedQueen 1.1.97          4 A32 2561  39  40 232 46% 2590 22%
041 Bison 15.1               1 A32 2561  38  38 236 52% 2547 30%
042 Rhetoric 1.4.1           1 A32 2534  38  38 236 50% 2538 30%
043 Alfil 12.10              1 A32 2533  38  38 232 49% 2539 29%
044 Gull 1.2 arm JA          1 A32 2520  38  38 240 53% 2491 30%
045 Chess Genius 4.0.00      2 IOS 2507 180 198  10 30% 2677 20%
046 Cheese 1.7               1 A32 2489  38  38 242 44% 2531 31%
047 Daydreamer 1.75 JA       1 A32 2477  37  37 238 48% 2486 33%
048 GarboChess 3             1 A32 2472  39  39 236 51% 2464 25%
049 Rotor 0.8                1 A32 2471  38  39 236 46% 2506 28%
050 Fridolin 2.00            4 A32 2461  37  37 238 52% 2442 30%
051 Glaurung Mainz           1 A32 2457  45  45 168 46% 2486 29%
052 Chess Genius 2.6.4       1 A32 2455 224 239   4 38% 2505 75%
053 Danasah_5.07.JA_xb       1 A32 2442  40  39 222 53% 2418 26%
054 BBChess 1.3b JA          4 A32 2417  38  38 238 51% 2406 26%
055 Dirty_030411.JA_xb       1 A32 2405  40  40 222 54% 2376 27%
056 Sloppy_0.23.JA_xb        1 A32 2401  38  38 224 51% 2394 33%
057 Phalanx_XXIV.JA_xb       1 A32 2374  40  40 230 47% 2401 17%
058 GreKo_12.5.JA_xb         1 A32 2367  40  40 222 55% 2328 25%
059 Pepito v1.59             1 A32 2364  37  37 238 50% 2363 33%
060 Pawny_1.0.JA_uci2xb      1 A32 2349  40  39 228 55% 2317 25%
061 BetsabeII_1.47.JA_xb     1 A32 2339  41  41 224 52% 2324 20%
062 Maverick 1.0 JA          1 A32 2314  39  39 244 47% 2336 22%
063 Chess Wise 3.1.7         2 IOS 2298 311 508   2  0% 2507  0%
064 Ifrit_m1.8.JA_uci2xb     1 A32 2297  39  39 224 54% 2262 29%
065 Typhoon_1.0.r358.JA_xb   1 A32 2273  39  39 236 56% 2222 25%
066 Diablo 0.5.1b JA         1 A32 2263  39  39 236 47% 2285 24%
067 Amy_0.8.JA_xb            1 A32 2245  38  39 244 48% 2259 25%
068 Zurichess Fribourg       1 A32 2239  59  60 104 45% 2287 23%
069 Olithink_5.3.2.JA_xb     1 A32 2238  40  40 228 52% 2220 21%
070 Myrddin_0.86.JA_xb       1 A32 2226  42  42 228 51% 2206 21%
071 TJchess 1.1U             1 A32 2203  38  38 250 49% 2204 24%
072 MangoPaola_1.1.JA_xb     1 A32 2188  40  40 234 51% 2177 22%
073 Sungorus 1.4 JA          1 A32 2180  39  39 234 47% 2198 22%
074 Natwarlal_0.14.JA_xb     1 A32 2178  40  40 230 51% 2164 23%
075 KmtChess_1.21.JA_xb      1 A32 2111  41  41 234 49% 2124 20%
076 Bitfoot 1.0.norev        1 A32 2108  39  39 246 51% 2098 25%
077 Scidlet_2.61b2.JA_xb     1 A32 2099  41  41 234 52% 2076 13%
078 NGplay_9.86.JA_xb        1 A32 2086  40  40 230 51% 2079 24%
079 Rattate_Nosferatu.JA_xb  1 A32 2084  41  41 234 49% 2095 18%
080 Resp_0.19.JA_xb          1 A32 2077  40  40 240 52% 2061 19%
081 DanChess_1.04.JA_xb      1 A32 1999  41  41 226 49% 2009 18%
082 Kurt 0.9.2.2 JA          1 A32 1967  40  40 236 51% 1958 21%
083 Robocide 281214.JA       1 A32 1930  41  41 234 47% 1953 17%
084 AdroitChess0.4 JA        1 A32 1910  42  42 226 51% 1897 17%
085 Witz_Alpha21.JA_xb       1 A32 1906  41  41 222 46% 1941 20%
086 Knightcap_3.7F.JA_xb     1 A32 1895  43  42 216 55% 1859 15%
087 BikJump v1.8             1 A32 1895  42  42 224 48% 1908 19%
088 Clubfoot_r2293.JA        1 A32 1893  42  42 220 50% 1898 18%
089 Woodpecker_2.11.JA_xb    1 A32 1886  43  43 218 49% 1892 13%
090 Sjeng_1.12.JA_xb         1 A32 1853  42  42 218 50% 1854 18%
091 Leonidas_r83.JA_xb       1 A32 1852  42  42 216 47% 1877 17%
092 ZCT-0.3.2500             1 A32 1837  41  41 232 45% 1876 15%
093 Faile_1.44.JA_xb         1 A32 1821  41  41 206 49% 1830 30%
094 Cilian_4.14.JA_xb        1 A32 1813  39  39 230 50% 1810 27%
095 Samchess_JA_xb           1 A32 1808  43  43 210 47% 1833 16%
096 Ecce rev. 508            1 A32 1768  42  42 224 49% 1782 17%
097 Colchess_8.0.JA_xb       1 A32 1740  44  44 190 50% 1741 24%
098 Surprise_4.3.b13.JA_xb   1 A32 1738  58  58 122 54% 1706 12%
099 Claudia v. 0.5           1 A32 1731  61  60 118 56% 1681 17%
100 Smash 1.03.JA            1 A32 1730  43  43 230 52% 1694 13%
101 Zzzzzz_3.5.1.JA_xb       1 A32 1646  45  45 178 54% 1589 29%
102 Gunborg_1.0.JA_uci2xb    1 A32 1621  42  42 228 54% 1556 21%
103 Hoichess_0.12.1.JA_xb    1 A32 1621  42  43 218 52% 1581 20%
104 Colossus 4.0 100X        1 C64 1616 236 208  10 80% 1243 20%
105 Rocinante 2.0 JA         1 A32 1608  44  44 218 55% 1522 17%
106 Kitteneitor_060513.JA_xb 1 A32 1606  45  45 176 49% 1576 36%
107 Jester_0.84.JA_xb        1 A32 1604  50  50 168 53% 1538 11%
108 Tscp_1.8.1.AB_xb         1 A32 1594  47  48 176 48% 1587 17%
109 Chenard_2014.07.11.JA_xb 1 A32 1594  47  48 190 42% 1651 12%
110 Pulse 1.5-cpp            1 A32 1579  43  43 222 55% 1481 27%
111 VIRUTOR CHESS 1.1.1      1 A32 1415  48  48 220 52% 1353 10%
112 Superpawn b108           1 A32 1371  55  56 152 49% 1388 16%
113 Chess for Android        1 A32 1371  47  48 227 47% 1359 15%
114 K2 v.075                 1 A32 1363  61  61 146 51% 1341  8%
115 Chess Titans             1 W64 1281 211 202   7 57% 1252 29%
116 Trappy_Beowulf_2.0.JA_xb 1 A32 1163  51  53 216 39% 1248 10%
117 Colossus 4.0             1 C64 1118 169 186  14 36% 1203 14%
118 Byak 8.10.14.JA          1 A32 1060  57  60 159 24% 1317 16%
119 Xadreco_5.7.JA_xb        1 A32  972  64  70 154 16% 1334  8%
120 Novag Secondo            1 TTC  821 344 -39   5 40%  900  0%
121 OliveChess 0.2.7         1 A32  556 353-300 128  0% 1438  0%


Rapidroid test platform:
* GT-N7100 1.6 * 4 + 256MB hash: All Android progs,
* Codegen Novatab 1.4* 4 + 256MB: All Android progs,
* Polypad 1010IPS tablet 1.61 * 2 + 128MB: Single thread Android progs
* HTC Diam 528Mhz, 16MB hash: Windows Mobile,
* i7 M620 2.67 Ghz dual + Arena 3.5 + 2GB hash: Windows 64
* iPhone5S A7 1.3 Ghz * 2: iOS programs
* DosBox 1.74: DOS programs,
* WinVICE 2.24: Commodore-64 progs,
* Messtiny UCI adapters or CB-Emu2014: Mephisto programs,
* Openings: 20 ply from Adam Hair or 16 ply from TCEC, no Q exchange, +0.15 to +0.40 evaluated by Stockfish and Komodo at depth 20 minimum, played twice each color.
* Repeating openings and twin games not allowed between two programs
* Tablebases and pondering off
* Time control: 10 to 30 sec/move or 600+0 to 1800+5 or closest known by both programs.