2

Way back when, Hans Berliner worked for IBM. I met him at an IBM data center one time, but he worked in the Federal division and just used the center for a demo. I had known him from the Washington Chess Divan earlier.

I know he had done a computer program to help with chess. It was the first as far as I can tell. This was in the days when the s/360 was ruling the market, and there were no PCs, and pdp8s and the like were not up to the job although they did cause IBM a lot of grief by their cheap price and what they could do for small special-purpose jobs.

The question I always had was did Hans use the IBM computers to assist his game when he won the ICCF world championship? I know that Hans was very good as a player, but was he so good that he could have won the ICCF world championship without the computer's help? It is certainly possible, but having a really good computer analysis on some tough positions had to be helpful.

2 Answers 2

16

Short of asking a deceased man, I would say that the answer is "no way".

He won the 5th World Correspondence Chess Championship in 1965. Computer chess, if it existed at all, was not strong at all. Even with mainframe-type hardware back then, it paled on comparison to the first Chess Challengers of the early 1980's, and they were only rated about 1200 then.

My first chess computer in 1985, and it was the strongest and most expensive on the marker at the time, was only about 1600. There is no way that you could win against correspondence Grandmasters with that type of help.

I would say, fairly confidently, that is was all his work.

16
  • 4
    Again, I doubt it. The difference between the first chess computers 15 years later, and what it takes to be helpful to even someone at that level is so great, I would say it is impossible. Short of someone coming up with absolute proof that he did, which I doubt, you are probably trying to prove a negative, which is impossible. It is probably going to be just speculation...against it. Commented Dec 14, 2019 at 19:29
  • 4
    Even as a blunder check, the computers at this time would be almost useless. At most the computer could point out a one move tactic.
    – Mike Jones
    Commented Dec 14, 2019 at 19:39
  • 4
    @yobamamama I worked at the National Security Agency in 1988, and we had acres of mainframes. There was a hallway in the basement between all the mainfraimes that was in the "Guinness Book of World Records" because you could drive a semi tractor trailer down if for a quarter mile to give you an idea of how big we are talking about. All of them combined today are nowhere near as powerful as ONE PC. You are talking about almost 30 years before. Just because NASA had it then, it does not mean it was that strong relative to chess. Commented Dec 14, 2019 at 23:32
  • 5
    Sorry, but it is clear that you have no real conception of what they could do at that time relative to chess, which was next to nothing. Commented Dec 15, 2019 at 1:23
  • 9
    You asked a question that you are not willing to take the answer for. I was a very early adopter of chess technology; I have owned a computer company for 25 years; and I have seen up close the best computers the government had to offer in 1988; and even then, they were not capable of strong play. I qualify as an expert. Commented Dec 15, 2019 at 1:27
5

Humans VS machines in the 60s-70s: one way of putting this accomplishment in perspective is by regarding the results of the match between Bobby Fischer and MIT Greenblatt computer program (Running Mac-Hack, a software written by the great Richard Greenblatt himself) in 1977.

I do believe that in 1977, Fisher was still very representative of what humans can do at their best at chess (even though he was retired), and so was the MIT Greenblatt computer for the machines. Still, Fischer obliterated the machine (3-0). This was 12 years after Hans Berliner won his title.

Now going back to your question, it seems fairly obvious that the gap between Greenblatt's computer and any computer 12 years older is very considerable. However, it appears to me that the gap between Fischer and Berliner is not that big. This is enough evidence for me, that, as a master, using/trusting a computer in 1965 was completely inappropriate, if not to say suicidal.

From the programmer's perspective: Berliner entered Carnegie Mellon in 1969 (4 years after his title) to do his PHD, where he came with the idea of programming HiTech, a chess engine. From Wikipedia:

It performed well, but only until it ran into transitions, that is, points in the game when the balance between the players changed. This led Berliner to conclude that HiTech was weak in board evaluation.

He then gave up on solving chess and focused on another simpler problem: Backgammon. His first prototype was released in the early 70's. It took him nearly 10 years to achieve a strong level after he applied principles from fuzzy logic. Once again, Wikipedia:

in 1979, BKG 9.8 was strong enough to play against the ruling world champion Luigi Villa. It won the match 7–1, becoming the first computer program to defeat a world champion in any game [..] Berliner states that the victory was largely a matter of luck, as the computer received more favorable dice rolls.

He then goes back to trying to solve chess, and re-develops HiTech (with other big names of chess software, like Murray Campbell) and he finally obtains a computer program with master strength (2500~ elo) in 1985.

Temporality of events does not line up, according to his own words in the early 70's "HiTech was weak in board evaluation." It's only 20 years after his title, (which corresponds approximately to two eternities regarding advances in computer science) that he obtains a software having the strength of a master. Said software would have probably not won the world ICCF championship at the time, as it was still way weaker than the super GMs of the 80s.

If he used a computer in 1965, either the computer was garbage and would have been completely ineffective, in which case Berliner would have been stupid to use it, or he had a 2600+ strong software at the time, and then faked the next twenty years of his life by developing 1800+ software to hide his secret.

TL;DR No.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.