Comments/Ratings for a Single Item

In computer self-play without opening book Capablanca Chess has only about half the draw rate of orthodox Chess (about 16%, vs 32%). These are all games that start well within the draw margin, though. So as play approaches perfection, they should eventually get a 100% draw rate. A large body of opening theory can be viewed as an attempt to play perfectly at least for the first 20 moves, or so, reducing the game phase in which result-changing errors can occur accordingly.

I'd like to add something on this topic, especially for you HG.

A while ago, around 2 years, when starting the design of the apothecary games, you had recommended me extra major pieces (it was actually not extra minor pieces but towards the idea that a higher percentage of strong pieces leads to clearer outcomes). This for the simple reason that in the endgame best moves become much more specific especially when it comes to trades with strong pieces. When the board is empty but the material is very strong, tempo matters a lot :)! I still don't like ridiculous strong pieces. I tend not to use amazons, for example in my game ideas.

I think in designing chess variants this makes the drawing chances smaller. With perfect play it will still be a draw as long the rules allow it (in Arimaa obviously this is not possible). That for studying purposes.

This becomes more interesting when studying games that can be strictly longer then omega0 (the smallest infinite ordinal). But we don't have many such games probably, I know of no good one :(!

Other nicer finite cases are games with different armies. Spartan is a very good example. With perfect play I think it can be a black win. Here Zermelo's theorem (https://en.wikipedia.org/wiki/Zermelo%27s_theorem_(game_theory)) does not apply in it's classic sense as the players have different moves, but it can probably (although I have not did it) easily be proven that the game is either a win or a draw.