Engine changes: passing rework, shot balancing, and rebound tuning

Just deployed a batch of engine changes — did my best to avoid unintended side effects, but let's see how it performs in real games. If you spot anything unexpected, let me know.

• ENGINE Changed one element of the engine architecture: passing is no longer a team-level skill. The engine now picks "the last passer" purely based on position and role — the last passer's passing skill now affects the shooter's efficiency on assisted shots. This came out of tough questions from @Lolek and @dcarra
• ENGINE Slightly reduced randomness in field goal chance calculations. This should reduce the number of great games from otherwise poor players, though it won't eliminate them entirely
• ENGINE Capped 3-point attempts for players set to shoot mostly threes — capped at 50% of FGA (down from 70%), with a slight reduction to 3P FG% as well. Reported multiple times by @grobarbrate
• ENGINE Noticed some leagues had rebound leaders pulling 20+ reb/game. Rebounds are now distributed more evenly — a cosmetic balance change only, doesn't affect overall team performance

Keep delivering ideas, and together we can build the best online basketball manager ever.