Novag Diamond II - Introduction

This is the Novag Diamond II from 1997, its a dedicated chess computer. Its stated rating on the box is 2420 ELO, but in actual play its around 2108, still very good for 1990's hardware. In Dec 2013 I played Novag Diamond II Level-FD3 against StockFish1.7 Level-1 from a Tablet. SF won 6-3-1.

[@sjp34christ]

Hello Paul, I really like your video featuring the Novag Diamond II chess computer. I am a collector of vintage and modern (few that they are) dedicated chess computers. I used to play chess and I was a member of the USCF for many years but more recently, less playing and more collecting. I guess it's the nostalgia of remembering an earlier time when computers were not so dominate in the field of chess. At that time, most computers whether dedicated, running on PC hardware or even special purpose mainframe could be easily beaten by the worlds best Grand Masters. Now the best engine on PC will tromp a Grand Master or even World Champion on a frequent basis. If you ever want to sell the Novag Diamond II, I'd be glad to own it. Take care and have fun actually playing chess with your computer(s). SJP.

[@paulc1562]

would you believe this "Old Chess Machine emulator" website was just created in 2021, and they have emulated over 400 chess machines, and 50 Novag machines, but they don't have my Novag Diamond II. You being a collector of machines I thought you might find it interesting, you have to download a zip and extract. en.chessbase.com/post/the-wonderful-world-of-chess-machine-emulators

[@velocityjockey1272]

​@@paulc1562I just found your comment from over a year ago. Thanks for the link, I'll check it out. By the way, do you, or anyone you know, have schematics or other technical information on the vintage Mephisto line, including the Senator and Munchen models? I have two dozen Mephisto's I imported from Germany about 10 years ago. Most are pristine, a few need some repair. Appreciate your time.

[@fabianromeroaparicio8157]

I remember when i smile if she was happy or wrong from the sound

[@paulc1562]

Yes back in the 1970's and 1980's we all get a good smile, and a happy feeling when we outsmart the computer, but now with Stockfish even Magnus Carlsen ( World Champ ) cannot beat the computer anymore.

[@dgontar]

The Kasparov RISC 2500 was estimated at 2400. It's extremely unlikely this computer is 2100 if it uses RISC hardware. 300 pts wouldn't be a little inflated. It would be gigantically inflated.

[@paulc1562]

I dont understand your logic, if they both have RISC hardware should they not have similar ratings? But if Kasparov RISC elo is 2400 and Novag elo is 2100, do you know for a 300 point difference, the higher elo will win 85% of the games? This 2108 elo rating in my description is from actual game play in human tournaments, so its accurate rating for the Novag Diamond II. And from Wiki I see that the Saitek Kasparov Risc 2500 elo is only 2190.

[@billmorrigan386]

That's a very good point. I think chess engines and old chess computers have their Elo numbers underestimated by 100-200 points on average and Wiki Elo ratings are incorrect. They even say they are not directly comparable to human ratings. Of course, they are not. They are skewed completely, and should be almost 200 points above to be comparable with FIDE Elo of human players. The thing is old chess computers play end game badly and it looks like they are very weak. Any pro chess player will think like that. However by playing a few games, it will become apparent that it is not often easy to reach end game vs computer without being destroyed in the middle game. Engines and old chess computers are tactical beasts. They should not be underestimated. Paul C, the author, of this video is also correct claiming that it plays like 2100. Indeed, it looks like that, as I already said. Yet, it is only like that until you play a lot of games against it. Then you realize that the machine is a tactical beast and cannot be beaten by amateur players 99% of the time. Only occasionally an amateur will eke out an honest victory without taking moves back or consulting opening theory to find the antidote against the built-in opening book of the old computer. What's more, this RISC processor is an old advertising gimmick in my opinion. It's just to sell more chess computers. Alas, these old machines still disappeared from the market. There's still a new King Performance machine. However, I believe older ones were of higher quality. They play slightly differently than a modern engine would play when you try to simulate Elo. Most cannot even simulate Elo properly. They totally suck (Lucas chess, Chess Explorer) by alternating terrible moves with good ones (I hate them). That said, there are some good ones: Deep Fritz and Shredder simulate human play much, much better. Still, older machines give some diversity in terms of playing style. At some point I thought old chess computers were meaningless but I was in for a surprise. They are not. They give some diversity when you want to play chess. Old or weaker engines that you can add to, say, Fritz program, are neither really old, nor weak!!! They don't let enjoy chess!!! They are horrible engines.

[@dgontar]

@@paulc1562 The Kasparov RISC was 2400. I don't know what source you're using. perhaps you are talking about a different version. But the one I played was definitely over 2300. I also played the machine back in 1993, and I had then at least an expert's rating. It was extremely strong.

[@dgontar]

@@paulc1562 You can see its rating here: It's about 2380. www.spacious-mind.com/html/risc_2500_2mb.html#:~:text=This%20extremely%20high%2Dpowered%20RISC,variety%20of%20exciting%20special%20features.&text=RISC%202500%20is%20also%20extremely,its%20built%2Din%20sensor%20board.

[@dgontar]

@@billmorrigan386 The Kasparov RISC is officially measured at 2387. www.spacious-mind.com/html/risc_2500_2mb.html#:~:text=This%20extremely%20high%2Dpowered%20RISC,variety%20of%20exciting%20special%20features.&text=RISC%202500%20is%20also%20extremely,its%20built%2Din%20sensor%20board.

[@billmorrigan386]

Good machine and a good review. I'm giving it a like. Yet, I disagree with Wiki Elo numbers. I think they are totally incorrect. Try giving this computer to a 2300 FIDE Elo player and see the result of the games!!! If and only if they survive the middle game without total destruction, only then they will most likely beat it in end game. To fully understand its strength, we need to now how many nodes (positions) per second it computes (5000, 10,000 nodes, or 20,000 nodes), as old programs were all similar in strength (+-50-100 Elo points). Nowadays it is very complicated with modern engines. They are totally different and while computing far fewer nodes they are almost infinitely more powerful and selective. There are also very old PC programs. For example, I still have Deep Fritz 6. Even that old engine is superstrong like Mangus Carlsen, probably (on a powerful computer, on an old legacy PC it will be like 200 points weaker).

[@paulc1562]

Hello Bill you make a good point on "nodes per second" and I do note your "nodes per second" theory in another video I made, I will link it below. But essentially the point is with both machines at a setting of "2 seconds per move" Novag Diamond II calculates 4197 nodes/sec while Stockfish calculated 508,000 nodes/second. kzbin.info/www/bejne/npm2eHVmf8mAjbs

[@billmorrigan386]

@@paulc1562 Thank you, Paul. I watched that video and posted a comment there too. Hopefully it is not an irritating comment since I totally disagree with Wiki Elo numbers for engines. The CCRL system rates engines but it has not been calibrated vs humans for a long time now, and by that I mean properly calibrated. So, they (CCRL) say their ratings are 'not directly comparable to human ratings'. However, the reality is their ratings are just undervalued by around 200 points. As to only 4000 nodes per second for Novag, well, it is surprisingly low for the machine with the specs it got. My old chess computer (Saitek Travel Champion 2100 = Saitek GK2100 in compact form-factor) computes around 3000 nodes per second. It runs on a meager 10MHz 8bit CPU with a small ROM and RAM. I thought the difference would have been much more than 30% (4000 vs 3000 N/s). I thought Novag Diamond II would be like 10,000 nodes per second. Thank you for providing this info. Now I know how things stand. I mean it's no go for me to upgrade to Mephisto Chess Master or Novag Star Diamond, as Elo and strength will improve only by a mere 50 points (would be useless to me). I sometimes play vs Deep Fritz and Mobile Shredder engines at their settings of Elo 2000 (=Fide 1900 Elo in my opinon, although I haven't played in OTB tournaments for 20 years now!!!). Why I use old engines and dedicated machines? Well, they play a bit differently and give me some diversity of play styles as opposed to only playing Fritz (rated games at Elo 2000) or games vs Shredder at its internal Elo setting of 2050. I also found that these simulations (shredder/Fritz Elo numbers are tremendously off at low and high settings, i.e., they play way too strong at Elo 2600 and way too weak at Elo 1500). As to PC emulators of old dedicated machines, I'm not very fond of them. Perhaps, I will need to buy the King Performance in the near future. As for now, the Saitek G2100 is still strong enough for me. I have started to play chess recently after many years of not playing and I'm encroaching on my old chess computer in terms of strength.

[@kasparov937]

@@billmorrigan386 My friend would you sell deep fritz 6? I had that 20 years ago, but it was stolen when I repaired my PC and had the disc in there. Back then I would compete with a Dual Pentium 3 1ghz rig on yahoo chess 😀

[@billmorrigan386]

​@@kasparov937 Do not buy an old engine that you cannot run on modern Windows! That said, you can get a whole array of old engines like that on Deep Fritz 6, 7, 8 all the way to Fritz 11 and later on ebay. Some newer Fritz programs (8 - 11) may have legacy engines: Fritz 5.32, Fritz 6, Come, Fruit, etc. but the latest (Fritz 15-17) ones don't offer many old engines, if I'm not mistaken. After Fritz 13, the newer Fritz doesn't even use Franz Morsch engines. The advantage of Franz Morsch program: it is not too advanced. At lowest settings, pro players can play against legacy engines natively, for example vs Fritz 7 but you need to turn off the permanent brain hidden deep in the menus. By the word 'native' I mean the computer does not dumb down its moves! The engine's strength is constrained only by the time you allow it to compute moves. No, sorry. I don't sell mines. But you don't need a lot of engines unless you want a lot of diversity and unless you aspire to be a pro. In that case, Fritz and Chess Base programs (even older ones) are topnotch. You can select Elo with a slider an play rated games. You also can get a large opening book library. My newest Fritz is Fritz 13. It's already obsolete but I still use + I got a database with some several million games in it, including even Napoleon games vs some madam! My oldest (Fritz 6) didn't have that much. And what I love the most is that customization is totally awesome as is graphics (but some folks say it's too complicated/advanced for them -- what a dumb world!) On mobile phones and tablets Shredder should suffice and it's dirt cheap!!! Everyone should start with Shredder. It's enough for everyone except for those who want to be pros. Yet programs/engines don't make you a great player but a lot of practice in classical time controls does!!! Just grab some NOT TOO OLD and NOT TOO NEW Deep Fritz on ebay if you are super-serious about chess. That's my advice. Sorry, I don't sell mines. Very old programs might give you headaches or not work at all as modern Windows don't support them. To Run the old Fritz 6 (2001) on Windows 7, I had to jump through a lot of hoops!!! Make sure you can run old software on your operating system.

[@kasparov9]

@@billmorrigan386 Ok no problem, is there a chance to copy Deep Fritz 6 somehow for me? At a price.