Glad to see you’re back. I’d thought you’d taken your show to the abyss where many a good podcast ever return.

In this episode, you repeatedly refer to a game being “solved”. I was wondering if you could clarify that.

Do you mean there is an optimal strategy for either side?

Theoretically, every game of perfect information (including chess) has a saddlepoint. That is that each side has an optimal strategy, and if all sides adopt an optimal strategy (and why wouldn’t they?) then the outcome is fixed. To my knowledge, what that outcome is for chess has never been determined, much less what those optimal strategies are.

The fact that a computer could beat a human (or fail to beat humans) does not prove anything – either theoretically or intuitively. AFAIK, the best algorithms are still non-deterministic, relying on heuristics instead. That’s a far cry from proving that any position has an inevitable outcome.

Thanks for the show -
-Joel

I agree with your recommendation of Yasser Seirawan’s series. I do have one small update–it’s no longer published by Microsoft, but is now published by Everyman Chess. Furthermore, he has written a seventh book recently which was never published by Microsoft (much to my delight for the book and chagrin that I now have one mismatched cover in the series).