Consistency and winning in the NBA -

Consistency and winning in the NBA

Do you have or know of any models showing players’ consistency or inconsistency? For example, I’d be curious to know how the shooting percentages and usage rates of various teams’ starters or top scorers compare with one another, and whether consistency (or possibly inconsistency) correlates to wins for the team, all-star appearances for the individual, etc.
-Justin, via the comments

Justin’s comment got me thinking – how well do top scorers produce night in and night out? And do top teams tend to have consistent scorers – or big scorers? To keep the scale of this inquiry manageable, I limited it to team’s top scorers and looked at just the last 2 years. Conclusive? No. Interesting? I’d certainly say so.

Here’s the data – scroll through and look it over. Then, a couple of graphs for you.

First up – Points per game and winning percentage.

Points per game and Winning Percentage in the NBA

Leading Scorer Pts/game and Winning %

No surprise here – having someone who can score a lot of points helps you win. Last year, the Bobcats leading scorer was Stephen Jackson, with 20.6 pts/game (for the full season, including time with the Warriors). Plugging that into the equation that fits the data – the Bobcats would be expected to have a winning percentage of 48.3%, a 40-win team; only 4 games off reality, and if you consider their extreme home/road win splits, maybe pretty reasonable.

Now – the standard deviation of those points charted against team’s winning percentage:

Points per game standard deviation and winning percentage in the NBA

Standard Deviation of Pts/game and Winning %

Meh – not much to look at here – a big standard deviation of the leading scorer’s nightly points totals still means he scores a lot of points.

Next – TS% of leading scorer and winning %:

True Shooting % and Winning % in the NBA

True Shooting % and Winning %

Time for Captain Obvious – having a leading scorer who scores efficiently will help you win…

Lastly, standard deviation of the leading scorer’s TS% and winning %:

Standard deviation of TS% and winning % in the NBA

Std Dev of TS% and Winning %

And having a leading scorer who is wildly inconsistent in how well he scores those points does not.

How does this apply to the Cats? Well, so far this year, Gerald is their leading scorer – and wildly inconsistent. However, Stephen Jackson seems likely to regain that title as the season progresses – and his last year numbers don’t exactly spell great things for the Bobcats success: I already mentioned his points per game as projecting the Cats as losers – and it is similar when looking at his Std Dev of TS% – it projects to a winning percentage of 50.8%, almost 42 wins. Mediocre. The fact that Jackson is a year older (and struggling so far) does not seem to bode well, unless Crash is ready to the crown.