How will future simulators model basketball offense?
Just as chess pieces have a virtually infinite combination of space, time, and movement, so does basketball offense.
Because Fergus Connolly's construct asserts truth across sports, recall that it informs four 'states':
- Initial space/formation
- Player movement
- Ball movement
- Scoring action (scoring moment)
At the root of the scoring moment is ability to score off the catch-and-shoot, off the dribble, and off complex interplays of screening and cutting.
And what circumstances might help use 'multiple action' assembly to construct attacks?
- Lego pieces fit together in nearly endless combinations
- A roll of the dice generate a limited number of possibilities
- Shuffling and picking cards generates 'random' actions
- A wheel spin (e.g. Game of Life) produces a number
- End-game video (e.g. close and late) of elite teams of elite players and coaches fills a catalog ("It's a copycat league.")
Lagniappe.
Don't apply NBA statistics to lower level play. Few high school teams are making threes at 35% clips. Learn to make midrange shots off the dribble.
Lagniappe 2. NC State ties it up in OT with a perfect flare screen three. This reminds everyone of the taking the foul debate up 3 late. One might legitimately ask why Williams was laying off the point guard allowing an easier pass.