# How useful (or useless) are preseason statistics for rookies?

Zion Williamson has been phenomenal this preseason for the New Orleans Pelicans. This has led to various opinions in the basketball world on how Zion will perform in the regular season.

Some say that Zion is going to be an All-Star in his rookie season. In fact, Stephen A. Smith made the bold claim that Zion’s rookie season will mirror Shaquille O’ Neal’s, based off Zion’s unparalleled efficiency in the paint.

To examine this idea more objectively, I attempted to look at a general case rookies in the preseason and the regular season and isolate some key statistics. One thing to take note of that affect interpretation of these results:

Teams have extreme variability of schedules in the preseason. For instance, one team could be playing a strong Lakers team each game, while another team plays teams like the Shanghai Sharks. Thus, it is tough to generalize anything from preseason to regular season. However, for the purposes of this article, we will assume that this will not affect player/team statistics by a wide margin.

I looked at three different basic statistics (points per game, assists per game, rebounds per game) in the regular season vs. the preseason. I plotted these values on separate histograms for regular season and preseason statistics.

These stats show that show that the overall spread of points scored during the regular season is skewed more right than the preseason points per games for rookies.

To examine this further, I wanted to look at whether players were less efficient during the regular season, or whether one of the main causes for this was because of less opportunity during the regular season.

In this scatter plot, the size of the dot corresponds to the rate of the increase of time played. Mathematically, it is simply expressed as

$(MPG_{regular season})/(MPG_{preseason}) * k$

This constant value is just so that one can distinguish between the sizes of the dots.

If you look closely at this plot, there is a clump of small dots below the line. (which is the line that represents the values staying the same). This means that most of these players received far less minutes during the regular season, which will in turn, skew the points per game histogram more to the right.

Thus, we can assume that rookies maintain fairly similar point averages across the preseason and the regular season. We can also generalize this to assists, with the same type types of plots.

We can see from this plot that regular season assist averages are also more skewed right than preseason assist averages.

And again, we see the same trend as we did with points. However the data is far more clumped near zero, which makes sense. Most players do not have the ball in their hands in order to facilitate too much.

When we look at rebounds, however, we can see that the distributions remain fairly similar across the regular season and the preseason.

When we look at the scatter plot for this graph, it looks like the number of points above the y=x line and the number of points below the line and above the line seems around the same.

There is obviously some correlation between preseason stats and regular season. The main bottleneck for rookie players which causes regular season statistics to dip is that rookie players do not get as many opportunities in the regular season as they do in the preseason.

However, we should expect Zion to continue to be the main man for the New Orleans Pelicans, and with his body frame and ability to score at will, don’t expect too much of a drop off in regular season statistics.

## One thought on “How useful (or useless) are preseason statistics for rookies?”

1. Karankumar Mageswaran says:

very insightful. well written!

Liked by 1 person