AI · Theory

Alpha-Beta Pruning

Theory

27/215

Alpha-Beta Pruning

Alpha-beta pruning skips branches that cannot change the minimax decision. Alpha is the best max can already force. Beta is the best min can already force. If a node becomes worse than that window, stop expanding it. Same answer as minimax, fewer nodes.

Order matters: good move ordering prunes more. Viva: “same result as minimax, less work”. Don’t say it makes a different (smarter) strategy.

Alpha-Beta Pruning — output — three lines. Not a different strategy — less work.

Exam tip

Same result as minimax + why order helps.

Example

# Alpha-beta idea
print("minimax value stays the same")
print("alpha/beta skip branches that cannot change the decision")
print("better move order → more prune")

Alpha-Beta Pruning — output: three lines. Not a different strategy — less work.

Short notes

DefPrune useless minimax branches.
RuleSame value, fewer nodes.
Rememberalpha / beta window.

AI · Theory

Alpha-Beta Pruning

← All stacks

Theory

27/215

Alpha-Beta Pruning

Order matters: good move ordering prunes more. Viva: “same result as minimax, less work”. Don’t say it makes a different (smarter) strategy.

Alpha-Beta Pruning — output — three lines. Not a different strategy — less work.

Exam tip

Same result as minimax + why order helps.

Example

# Alpha-beta idea
print("minimax value stays the same")
print("alpha/beta skip branches that cannot change the decision")
print("better move order → more prune")

Alpha-Beta Pruning — output: three lines. Not a different strategy — less work.

Short notes

DefPrune useless minimax branches.
RuleSame value, fewer nodes.
Rememberalpha / beta window.

Alpha-Beta Pruning

Example

Short notes

Questions

Alpha-Beta Pruning

Example

Short notes

Questions