The Daltons is an AI trying to play Four-in-a-Row.
Updated known positions.
Added a known winning and losing positions to the search tree.
Reduced. the value of strong threats.
Hugh improvement on speed (benchmark on 10 ply: 3.4 seconds to 25 ply instead of 7.5). Also added detection on direct winning multiple threats and threats on two different columns.
Different values for different threats.
Fixed strong threat bug, and skipped 2 out 4 now.
Tweaked evaluation to also take 2 out 4 into account.
Speed improvement with different sorting.
Further tweaking of time consumption.
Redone Branching.
Tweak in evaluation and time consumption.
Made lowest winner less winning, and made branching bigger.
Fixed evaluator bugs and added odd/even winning position detection.
Added double threat-detection and huge speed improvements.
Made book moves less forcing.
Just play 3 at the start.
Don't look beyond ply 9 until we are there.
Also keep losing nodes.
Start calculation on move 1.
Totally refactored the search tree, and simplified it.
Tweaked book.
Added an opening book.
Time fix, and fix for forced moves because of single open column.
Tweaked and improved evaluation. Branching restriction. More time-consuming to start with.
Static evaluator for the board position.
Struggling with cleaning up.
Clear all cache of previous rounds.
Added sorting tweaks and a premature book.
Fix for not be able to send winning moves.
Fixed tons of issues with the search tree.
Fixed tons of bugs, in parsing, ply-determination, and evaluation.
Removed debug output.
Because of missed communication changes.
First attempt. Not to smart yet.