Hacker News new | past | comments | ask | show | jobs | submit login

From the Wikipedia page on Chess960:

> Hence, on average, a Chess960 starting position is actually 18.2% more balanced than the standard starting position.

I'm also interested in the underlying distribution (not just the average). For each of the 960 starting positions, what is known about the first-player advantage? (I'm pretty sure these would just be estimates because a full solution is still infeasible.)




On average. Some starting positions are much less balanced than regular chess.


>> I'm also interested in the underlying distribution (not just the average).

> On average. Some starting positions are much less balanced than regular chess.

Yes, I expect variation, which I why I asked. :confused-face: At the risk of over-explaining, when I said "distribution" I meant https://en.wikipedia.org/wiki/Probability_distribution

What does "much less" mean, quantitatively? For each position, what is the white-winning probability if you used e.g. Stockfish or some suitably adapted tool?

I would love to see e.g. a histogram where the x-axis buckets the estimated-advantage-to-white and the y-axis counts how many of the 960 starting positions fall into the bucket. What shape might it take? Lacking any particular insight, I would guess normal.




Consider applying for YC's Summer 2025 batch! Applications are open till May 13

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: