# Equilibrium Points of an AND-OR Tree: under Constraints on Probability

arXiv: https://arxiv.org/abs/1401.8175 · HTML: https://mlsanity.com/paper/1401.8175/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 31 Jan 2014.

## Built on by (most cited first, 1 of 1)

- [Separation and Collapse of Equilibria Inequalities on AND-OR Trees without Shape Constraints](https://mlsanity.com/paper/2405.20138/) (2024)

## Abstract (arXiv)

We study a probability distribution d on the truth assignments to a uniform binary AND-OR tree. Liu and Tanaka [2007, Inform. Process. Lett.] showed the following: If d achieves the equilibrium among independent distributions (ID) then d is an independent identical distribution (IID). We show a stronger form of the above result. Given a real number r such that 0 < r < 1, we consider a constraint that the probability of the root node having the value 0 is r. Our main result is the following: When we restrict ourselves to IDs satisfying this constraint, the above result of Liu and Tanaka still holds. The proof employs clever tricks of induction. In particular, we show two fundamental relationships between expected cost and probability in an IID on an OR-AND tree: (1) The ratio of the cost to the probability (of the root having the value 0) is a decreasing function of the probability x of the leaf. (2) The ratio of derivative of the cost to the derivative of the probability is a decreasing function of x, too.