40 ms·
i actually tried. i couldn't get enough interested parties and had a cofounder that wasn't very motivated. it would take your game history logs, analyze them,
by noodle 16y ago
i actually tried. i couldn't get enough interested parties and had a cofounder that wasn't very motivated.
it would take your game history logs, analyze them, and let you know where your most common mistakes were based on context (i.e., ICM for SNGs)
i still think its a good idea, for various reasons, i've just moved on.