WELCOME to ETOC-G Engine Test Operating Center-Gutweiler Last update : April 21st, 2026 (21:15) Still running : Junior 13.3.00 HCE Reasons for ETOG-G: ------------------- 01. To have own results for my engine-overview of "TOP-250" engines! A very simple idea for testing engines without focusing on Elo. 02. The aim is to avoid high draw rates. 03. The issue is no inflation of playing strength. 04. I would finally like to see a list, where all engines had the same opponents, the same hardware and the same conditions. 05. Achieved results should not become “old” and “uninteresting” in the list. Long-term improved comparability of results achieved. 06. Easy parameters regarding speculation about playing styles in the first, in my opinion the most important phase of the game. The number of quick wins / losses, average number of moves as indicators. The many extremely high ratings are mostly due to the transition to the endgame and, above all, the very important pawn structures. 07. ETOC-G is applicable to engines that reach min. approximately humans super grandmaster level. Sorry, I can't test over 1.500 available engines alone in homework. 08. Fun for me and I hope for my visitors in the foreground! 09. Easy to copy for anyone. Perhaps someone will start an ETOC-T (ETOC-T = Tokyo) at some point. On such an easy test-system, many people can work to the same time. 10. Avoid headaches and sleepless nights when your thirst for knowledge about engines is not quenched?! :-) You know, point 10 is most important? The insidious "Computer-Chess-Syndrom"! *** DIARY *** 030. April 18th, 2026 (15:00 - 17:00): AMD Ryzen™ 9 5950X maintenance! For this time the still running Alexandria 9.0.0 test-run stopped. 029. April 16th, 2026 (20:45): Due to other commitments, I’m sometimes unable to update daily. In that case, I run two or more test-runs simultaneously, distributing the 16 cores among them. I've used it for the first time. I love the organized chaos of testing. Once everything has been tested, it unfortunately gets a bit boring. 028. April 13th, 2026 (14:00): I am back, earlier as await??! During one of those sleepless nights, I was thinking about my ranking system?! With success: A new Rank, yay a new one!?? 00# General (Ret.) Ret for Retired ... Honorary title! Both, at first a Honorary title for retired persons do a lot - over many years - for our beloved hobby and as second for outstanding contributions. Prof. Robert (Bob) Hyatt : He was allways there, so many years! Crafty accompanied us on the Lord’s path. Stefan Meyer-Kahlen : UCI protocol, very nice GUI (the mother of UCI-GUIs), many times Computer-Chess-World-Champion - often engines comes and go - often after the first 1-3 updates - often engines with many power, that's sad - often programmers lost interest for I think different reasons (often because of people like me) ... I'm innocent! But still, all the works are always available, and we can enjoy them. We don't have to look for them on Oak-Island (Nova Scotia, Canada). Over 250 treasures don't need to be dug up. Your PC is in wait-position, start your engines! Go over my overview with the gold detector (mouse) and listen to the song of the engines. Or something like that!? Damn it, that doesn't mean Bob and Stefan's programs get a free pass. Crafty and Shredder have to compete against the 26 "Standard-Engines", too. An order for Bob and Stefan. A hard, very hard, harder than hard, even harder order you know. I wouldn't want to be in their shoes. Later ... Nothing's running away here; we're not on the run. OK, I am - I'm Speedy Conzales moving at a snail's pace. 027. April 01st, 2026 (09:00): Nothing to do here - no April Fool's joke! ETOC-G usually runs automatically in the background. I don't want to change anything else right now. The errors on the page have also been corrected. Time to organize the next Tipp-Kick tourneys with my friends. I hope my favorite football club (SG Wattenscheid 09) goes one league higher, the next games are very important. Here the move-average / quick-wins (Goal difference) can be decisive. My bike is also ready to go for some exciting rides. Spring cleaning has been done for a few months now. I just quickly made a few corrections to my diary. My goodness, I wrote such a mess here. You know ... if the day is long!?? I'll be back soon with my next entries, around the year 2033 or a bit earlier. Go Wasp go ... (TOP-Secret, is my favorite engine). 026. March 30th, 2026 (05:15): Until now, I had been including wins of 49 moves or fewer in the statistics as quick wins. I have now changed this to include wins of 44 moves or fewer. So I think that's a better way to describe a short game. By the way, after 72.150 games: Wasp 7.00 played the fastest quick-win game, winning in 18 moves! CSTal 2.0 played the second fastest quick-win game, winning in 21 moves! 025. - maintenance - March 29th, 2026 (23:30): Maintenance has been completed. I found a lot and made the corrections! 024. March 28th, 2026 (12:15): "Breaking News" as a small idea? On weekends, I come up with a surprise, a deviation from my “test plan.” I have to find a way to keep my visitors on the edge of their seats. That’s not so easy these days. Especially not with just one system I am using. Anyway, the main thing is that it’s fun. The many Generals will help me find a solution for my "Visitor-Problem". So, potential Generals will be play, holding what amounts to a symbolic autograph session. The truth is that anyone can organize something exciting on their own using a simple system. That’s why one of my main goals is to encourage people to do just that. Animation = progress; progress means development ... again and again! 023. March 28th, 2026 (11:00): Following the latest test-runs of the "Super-Generals", the question arises: How can we reduce the number of draws in the future without driving up the average number of moves? The answer is: I don’t have one! I’m slowly starting to lose faith that there will be any significant further improvements in playing strength. Well, the history of computer chess teaches us otherwise. Things are getting exciting again! Perhaps too many draws are being conceded against significantly weaker engines? Perhaps there is room for further improvement here? Should a neural network try to figure out during a game just how strong the opponent actually is, so as not necessarily to play the best move as, for example, World Champion Lasker did? It’s probably best not to burden my mind with such questions any longer. So I’m going to turn my attention back to the cats in my village. 022. March 27th, 2026 (10:45): Currently I offer 295 Shredder 12 GUI engine configurations for my download file. Well, that also means I plan to run 295 test runs. I'm very confident about this and trust in my energy. From this perspective: 39 test-runs are ready + 26 “Standard Engines” have been played a tourney! Of course, ongoing updates will be add from time to time. I'm going to introduce “update blocks” here. This currently corresponds to approximately 65 × 100 : 295 = 24.53%. That's a lot after such a short time ETOC-G is still running. That's looks like a new rating list. Actually, that's not the case. Rather, ETOC-G attempts to divide playing strength into 20 groups so that I can display a brief summary of this information in my engine overview. Always keeping in mind that the engines also run smoothly under the GUI being used. Honestly, when I'm watching, I prefer time controls where a game lasts about 30 minutes to an hour. That makes it much more interesting to watch. But that's way too long a thinking time for what I'm trying to do here. I am quite sure, that my ETOC-G system isn't so strong to measure playing ability very accurately. The Elo range is to high. Means, for the strongest and weakest engines among the TOP-250 it will certainly be slightly less accurate. In summary, however, as I mentioned earlier, I am very satisfied with the “new” system. 021. - maintenance - March 25th, 2026 (07:00): Right now, I'm focusing on finding errors on my few pages, most of which were caused by the many updates. I also found and corrected errors in the engine configurations. As part of a test run, I naturally check the configurations that have already been made. For example, Dirty can now use Nalimov, I had completely overlooked that. I’m always grateful for any feedback on errors, of course. https://www.amateurschach.de/main/_contact.htm 020. March 23rd, 2026 (21:15): ETOC-G is very young. 46 out of 264 (17,43%), each programmer(s) / et. al. is allowed one rank! 46 different engines tested so far. Can be see on "rankings". It will take me more than a year to assign a military rank to all the engines in my engine overview. With all these results I can thinking about it to make ETOC-G better. Maybe increase the number of "Standard-Engines" from 26 to 30 (1500 games instead of 1300), or maybe replace some "Standard-Engines" with others. I'm not even familiar with many of the programs I'll be testing yet. Maybe I reduce the Elo range from 900 to 500? All this is unclear so far! The main thing is that that stupid AMD computer doesn't take up so much space for no reason. Mini PCs are better! But for now, we're focusing on gathering results. The Panini soccer stickers will be back soon, just in time for the World Cup. Men always have to collect something. Isn't that terrible, and all this with 59 years. 019. March 20th, 2026 (08:45): I added now the next test-runs in ETOC-G details site. Yay, Frank has a plan. Not really, but it kind of looks like he does. I have to tell everyone in the neighborhood right away. 700 people in Gutweiler only, you know. 1. The priority is on commercials, 2. MCTS, 3. and “very old” engines that could make it into the TOP-250. 4. Next come GitHub "archived engines" 5. and the Java-based engines. 6. Next are the programmers who offer two engines that rank in TOP-250. There are a few other engines I'd like to see move up in my list. 7. After that, I'll check to see how many of the engines I've already tested have been updated (July 2025 will be the first month of updates). What I find interesting is comparing ETOC-G Elo to my older rating lists. Shredder 12 should be have around 2800 Elo = 12,5% of possible points. Shredder 13 should be have around 3125 Elo = 37,5% of possible points. I am using Nemorino 6.11 JA for Ordo-Elo-Calculations. It seems that 3275 is 15-20 Elo to high, later ... With this structure in place, the list is well-organized for testing the remaining engines one by one. ETOC-G will then take its first break from around middle of December 2026 to the end of January 2027. There is a lot to do before I go into hibernation. I’ll probably get my first round of snarky comments when I post the first list on TalkChess after point 7 is ready. Many computer chess players don’t get along very well, and these days there are hundreds of so-called experts. They always think they know better and often they actually do. It's not easy for me to get my old stories out there :-) Oh my good, all the Generals have to play vs. Spark 1.0 HCE ... loud thinking! No one would do that to themselves. Note: I may choose to use a specific engine at any time, such as an update to a “Standard-Engine” or a bug fix version. 018. March 18th, 2026 (10:00): Incidentally, the “Standard-Engines” were also selected based on the criterion of keeping the move-average low. In other words, the selection wasn't based on aggressiveness in the attack or specific strengths such as the impressive transitions into the endgame often seen in SmarThink. An move-average of 88 is considered very high in ETOC-G ... in my opinion, far too high. 017. March 17th, 2026 (03:30): To be honest, I don't know anything about computer chess. But people who don't know much about something always have a lot to say. Every day I spot an error somewhere in my engine overview. I then fix it on a secret mission. This absolutely must stay between us. I'll look into it over the next few days, I promise. 016. March 16th, 2026 (01:00): Now that I’ve repeated the tournament among the "Standard-Engines", the question arises again: How should I proceed with testing? I have a plan for the next ~ 25 days or so; see the ETOC-G details page. It makes sense to test the older engines first to avoid running too many test runs on engines that are currently under development. I will test the group of generals and possible generals soon. I haven't figured out yet how to quickly bring the generals into the action. Most of users have interest to see only the strongest engines :-( But right now, I’m still struggling with the question of whether the list of engines below 3000 Elo is even accurate. For the time being, however, everything is going according to plan, and my still unproven system seems to be working quite well. 015. March 15th, 2026 (09:45): - has been changed - I made my engine-overview smaller. Listed are now engines only, I like to test with ETOC-G. ETOC-G is designed that an engine with approximately 3000 Elo achieves the rank of “Private First Class.” If Nemorino 6.11 JA is playing with 3275 Elo? ... who knows that? ... the ETOC-G range is: 904,09 Elo! Standings after the repeated "Standard Engines" tourney: Program Elo Games Score Draw Points Error 01. Stockfish 18 JA 3577,46 1300 83.4% 32,9% 1084,0 17,51 08. Gaia 4.1.3 3439,94 1300 72.1% 49,8% 937,5 14,48 09. Zangdar 6.1 3431,39 1300 71.3% 45,4% 927,0 14,07 --. Stockfish 200731 HCE dev 3404,84 2550 71.3% 43,6% 1817,0 11,00 14. Wasp 7.14 dev 3353,38 1300 63.5% 45,6% 825,5 13,81 --. Wasp 7.00 3344,12 2550 65.6% 41,8% 1672,5 10,04 18. Nemorino 6.11 JA 3275,00 2550 58.8% 41,6% 1499,5 10,26 34. Critter 1.6a HCE 3017,38 2550 33.6% 29,8% 857,0 10,94 35. Demolito 251216 HCE 3014,84 1300 28.4% 33,7% 369,0 15,56 44. Spark 1.0 HCE 2795,83 2550 16.5% 19,8% 421,0 14,07 45. KnightX 4.8 HCE 2764,17 1300 11.1% 16,8% 144,0 21,16 47. Belette 3.0.0 HCE 2673,37 1300 7.3% 9,9% 95,5 25,26 Have a look on the draw quote with an equal opening book :-) Have a look in Stockfish, current version compared to the latest HCE version, probably without Elo inflatations. With around 72% = ~ 3435 Elo, an engine is rank 06# = "Colonel" With around 28% = ~ 3005 Elo, an engine is rank 19# = "Private First Class" and achieves its first promotion! With around 7% = ~ 2650 Elo, an engine can participate in ETOC-G and will be listed in my engine-overview. 2650 Elo is probably place ~250 in the World, speaking from available engines I know 03/2026. In my opinion, the ErrorBar achieved is good enough for “only” 1300 games played to give the engines a military rank. Everything is wonderful and good for my sleep statistics :-) In German: Ich mach' mir die Welt Widdewidde wie sie mir gefällt (Pippi Langstrumpf). In English: I make the world the way I like it (Pippi Langstrumpf). You know: Computerchess people like that, exactly this one to do! A military rank for engines? I must be crazy. But chess is a war game and we don't need Grandmasters here, what we need are Generals, Colonels and heated discussions about it until the sparks fly. 014. March 05th, 2026 (16:45): The question is whether or not to use JA compiles. The same question for dev versions. My answer is very clear. Of course JA compiles and definitely dev versions. But often the devil is in the details. Fizbo 2.0 is clearly stronger as Fizbo 2.0 JA. For the ETOC-G "Standard-Engines" I made four changes since my ETOC-G preparation tourney on my "Test" Intel i9-10900k system has been finnished. For my first 25 test-runs I am using the "rigth" versions. One more reason to repeat the tournament between the "Standard-Engines." At the moment, I'm producing more of an organized mess. The journey is the destination! During the time the tourney between the "Standard-Engines" is repeated I have time for the cats in the village. This one I have to repeat: PRE-T (preparation tourney on Intel i9-10900k): https://www.amateurschach.de/download/gmames/etoc-g_pret.7z = 14,476Mb Note for the HET: "dev" does not mean devil. dev means development version of engine, you know?! "HET" = Hardcore-Engine-Testers 013. March 03rd, 2026 (01:00): I don't think that a lot of engines I listed in my engine-overview are strong enough for my ETOC-G system. After several result simuulations, a test run should generate at least 6-7%. If the result is less than 10% (~2750 Elo), the error bar is too wide for my purposes! Now the French engines are tested, I can repeat the "Preparation-Tourney". But before I do that I like to see Clockwork. On this engine a lot of programmers are still working. A very interesting project! 012. March 01st, 2026 (21:00): For users who prefer to see Elo ratings rather than military ranks, I have included an Ordo (Elo calculation program) batch in the ZIP file of the ETOC-G database of all games played. 011. February 28th, 2026 (12:30): Sorry, I forgot to offer the opening book for download again. You can find the download link on the ETOC-G detail site under parameters. Thanks to an user from Peru for this hint. ETOC-G detail site: https://www.amateurschach.de/main/_etoc-g.htm Direct download (2.554Mb): https://www.amateurschach.de/download/feobos-v20.1-contempt_3-5_tuned-v2.zip Have a look on the old FEOBOS project: https://www.amateurschach.de/download/_feobos-v20.1.zip (54,287Mb) 010. February 26th, 2026 (02:30): Now that my system is slowly getting up to speed, starting a new test-run is a matter of 15-30 minutes (including updates to the HTML web pages). I have automated many things. It takes much more work to keep the pages up to date or to look for new powerful engines. I have automated a lot here too, but I still estimate that it takes me an average of 30-45 minutes of work per day. Why am I writing this? I have been asked this question several times now. Apparently, my engine overview is loaded very often. I don't have any statistics on this, nor am I interested in them. This site should be viewed purely as a pastime. The motto is: Every day a different engine! It's annoying that I have to repeat the tournament between the "Standard-Engines". ETOC-G is an old idea. I've been testing engines that interest me for a long time. Basically, there's nothing special about this at all. What bores me is the pursuit of Elo or Elo itself as a system for calculating playing strength. It makes you lose sight of the powerful moments of the engines. Any work, whether it's an engine or some other form of assistance, is good for our community. It should never be forgotten that even weaker engine results still represent super grandmaster playing strength. The playing strength of the best engines is far from good and evil. We are way ahead of our time! 009. February 23rd, 2026 (01:30): I have finalized the organization for the next test-runs (subject to change). Priority will initially be given to updates from the “Standard-Engines” or second engines from the programmers of the “Standard-Engines”. First, however, I will repeat the tournament between the “Standard-Engines”. At least I have a plan for around the next 45 days. Can be see on ETOC-G detail site. Perhaps as next the new and older commercials, I don't know. It looks strange when I play with older versions of the “Standard-Engines,” so I will have to prioritize the programmers of the “Standard-Engines” when new updates come out (if not to often). Honestly, not a secret ... most of the "Standard-Engines" are my personal favorits. Good for my own motivation for ETOC-G, because I can see every day what they do. I have never used many of the engines that are to be tested at ETOC-G over the next 1 1/2 years. I am therefore completely unaware of their strengths and weaknesses. It will therefore be very exciting for me. But at first the French engines, I am not ready yet :-) 008. February 16th, 2026 (01:15): I added a buglist on engine overview and ETOC-G detail site. In additional you can see also the number of repeated games here. Draws under 20 moves and games over 300 moves will be repeated. The first 7 engines tested ran without any technical problems. 007. February 15th, 2026 (05:15): One test-run need around 20-22 hours. Of course, it depends on the average number of moves. With 16 matches running, the power consumption of the hardware used is approximately 103-106 watts. An AMD Ryzen™ 9 5950X clocked at 4.0Ghz is used. The conditions / parameters can be found on ETOC-G detail page. 006. February 13th, 2026 (09:00): I have so many other ideas for making ETOC-G more interesting. But every time I implement an idea, the amount of work involved increases. It is precisely this amount of work that I want to keep to a minimum. I think I can start a test-run on about 250 days a year, but of course I don't always have time for this hobby. Anyone can test their favorite engine themselves in 20 hours with such a system ... really, it's very easy and simple! 005. February 13th, 2026 (07:45): With so many engines to test, I don't really have a plan for where to start. The priority is not on the strongest engines or newly released engines. I think I'll start with the group of French programs. After I will repeat the "Standard-Engines" tournament on the AMD hardware currently in use. That's an event, what I mean is watching a different engine every day in the year. Because all of the engines listed are probably more powerful than 99.99% of humanity. In the pursuit of meaningless Elo ratings, this is often forgotten in the computer chess community. 004. February 12th, 2026 (11:30): I added PNGS stats (move-average statistics) on ETOC-G detail site. PGNS is an old tool by Ferdinand Mosca (Philippines) he created for me. 003. February 09th, 2026 (21:00): I started ETOC-G with Stockfish 17.1 JA as first engine! Sure, Stockfish is our number 1! The question is how long? In these turbulent times, with so many "Generals" on the scene, Stockfish’s top spot is no longer secure. 002. February 09th, 2026 (01:00): After spending a whopping nine months getting an overview of the available engines again, I now know about 3.5% of what's going on. That means I can join in the conversation again, even if no one wants to hear me anymore. It's time to start ETOC-G on ~ Tuesday, February 10th, 2026. What bothers me about my own configuration is that I had the "Preparation Tourney" run on another hardware as I plan for using for ETOC-G. I will therefore repeat the "Preparation-Tourney" among the 26 "Standard Engines" in the first half of 2026 - very soon -. 001. January 27th, 2026 (14:30): ETOC-G is starting with a delay. The reason is: "Spring-Cleaning" at home! The first test-run is expected to start on ~ February 10th, 2026. Frank Quisinsky, Gutweiler (Germany)