Who wouldn't want a simple, concrete, credible, and accurate "number" to attach to a player? This can’t resolve controversy but might enlighten it.
The Holy Grail (HG) would reflect contribution to winning, impact on teammates, team offense and team defense.
It would not 'curve fit' to stratify players to conform to an analyst who "likes" someone's game and has misgivings about someone else's. And it won't say whether so-and-so is the fifth best guard in a league or the tenth.
The supporting cast also impacts a player's effectiveness. If you're on the floor with better cutters and shooters, then you might have more assists. If you're on the court with 'selfish players' then you won't get as many touches. Offensive and Defensive ratings depends on those on the floor with you as well.
As a basketball analyst, I'd rather have 2026:
- On/Off data
- Assist-to-turnover data
- TS%
- Defensive rebounding
- Steals
- and minutes played
than a single 'contribution' number. That partially reconstructs the "scores and stops" framework.
Let's use a concrete example, Jalen Brunson, NBA Finals MVP and LaMelo Ball of the Charlotte Hornets.
I asked ChatGPT Plus to develop a tabular comparison:
| Category | Jalen Brunson | LaMelo Ball |
|---|---|---|
| Scoring Volume | ⭐⭐⭐⭐⭐ | ⭐⭐⭐⭐ |
| Scoring Efficiency | ⭐⭐⭐⭐⭐ | ⭐⭐⭐ |
| Playmaking | ⭐⭐⭐⭐½ | ⭐⭐⭐⭐⭐ |
| Turnover Control | ⭐⭐⭐⭐⭐ | ⭐⭐ |
| Rebounding | ⭐⭐ | ⭐⭐⭐⭐ |
| Offensive Impact | ⭐⭐⭐⭐⭐ | ⭐⭐⭐⭐½ |
| Defensive Reliability | ⭐⭐⭐ | ⭐⭐ |
| Decision Making | ⭐⭐⭐⭐⭐ | ⭐⭐⭐ |
| Availability/Durability | ⭐⭐⭐⭐ | ⭐⭐ |
| Coach Trust Factor | ⭐⭐⭐⭐⭐ | ⭐⭐⭐ |
| Overall Total Contribution | ⭐⭐⭐⭐⭐ | ⭐⭐⭐⭐ |
There's no precise "apples to apples" comparison available but analytics staffs are working on that.
One statistic I'd love to invent:
Knowing your interest in "Coaches' Total Contribution" and holistic evaluation, I'd create a metric called Possession Value Added (PVA).
Rather than asking: "How many points did he score?," I'd ask, "How much better is every possession because this player was involved?"
