Blog Feed

Usage Rate – Regular Season vs. Playoffs

When we look at games in the playoffs, we see completely different strategies employed by teams. Star players seem to be more relied on than they would in the regular season, while players with smaller roles seem to be less useful than in the regular season. This ‘hunch’ can be represented with a graph of usage rates in the playoffs vs in the regular season. Here is a graph (with a line created with a basic linear regression algorithm).

This graph’s line has a slope of 0.966, which basically means that overall, players normally do not deviate from their regular season usage rate. However, the R^2 value (which here is essentially a metric that evaluates how good of a line of best fit is) is 0.679, which isn’t that good (The optimal R^2 value in statistics is 1.0). Further, when we examine the graph, we see that players with considerably higher usage rate tend to be above the line of best fit.

To isolate player’s into different types, I decided to split players by the number of minutes. I split them into the following groups:

  • 35+ Minutes in the regular season
  • 25-35 Minutes in the regular season
  • 15-25 Minutes in the regular season
  • 5-15 Minutes in the regular season

35+ Minutes (Slope = 1.01, R^2 = 0.808)

The graph above is quite interesting as it shows that heavily relied on players do not typically get used more in the Playoffs. Rather, they get used about as often in the playoffs versus in the regular season. The R^2 value is also close to one, so the line is fairly accurate in predicting this underlying relationship.

25-35 Minutes (Slope = 0.938, R^2 = 0.695)

With the slope of this graph, we see that players who play 25-35 minutes seem to get the ball less often than in the regular season. However, interestingly, the spread between the points and the line is more than with the players with more than 35+ minutes (This is shown with the lower R^2 value as well). This means that players who have 25-35 minutes have more variation in their usage rate.
15-25 Minutes (Slope = 0.907, R^2 = 0.521)

Here, we see that there is an even lower slope but also a lower R^2 value. This means that variation is even higher than players with 25-35 minutes and on average players get less usage. With this low of an R^2 value, the line of best fit barely works as a guideline. This means we cannot really predict how much the player’s usage rate will change in the playoffs. In the next group, we will see an even better example of this.

5-15 Minutes (Slope = 0.877,  R^2 = 0.292)

When we look at this graph, we see that there really isn’t much of a trend in the data. This means that when a player’s minutes are this low, there isn’t a real correlation between a player’s usage rate in the regular season vs. in the playoffs. Instead, it requires more data (i.e a player’s points per minute, assists per minute, etc).

To see what actually determines usage at these lower percentages, I trained a neural network where I inputted some basic stats per minute, offensive rating, defensive rating, along with the player’s usage rate in the regular season. This got me a much higher R^2 Value for these lower minute value, which shows what teams really look for in these players who get fewer minutes. For the actual neural network code and the rest of the code used for this post look here.

Conclusion
When players are more relied on (play more minutes), they are more likely to keep the same usage rate. However, as the number of minutes that a player plays decreases, the variation increases tremendously. Thus, player efficiency is integral to determine how much a player will be used at these lower minute values.

Player Trajectories



Using the stats.nba.com API data and matplotlib, I recently developed a pretty cool visual to see how players’ points (adjusted by usage rate and minute) changed over time. This means that an excess of usage leading to extra points will not affect the quantified improvement over time. The graph above shows the average player who scored >15 pts at least once during his career. Let’s see some actual players on this graph.

The x-axis represents the players’ year in the league(ex. rookie season would be 1, sophomore season 2, etc.) and the Y-axis represents the players’ improvement relative to the base year. A negative value would mean that the player has gotten worse since the rookie season and a positive value means they have gotten better since their rookie year.

For the actual code used for this you can go here.

1. Pascal Siakam

For those doubting that Siakam has improved tremendously, the numbers don’t lie. This serious Most Improved Player candidate has shown that he is truly a force to be reckoned with. Now, does this graph mean he is a phenomenal player already? No.

This graph is relative on the player’s first year, which basically means it is measured based on the players’ improvement since he started in the league. Thus, it shows improvement since the base year, not how good the player actually is.

SeasonAge2P%eFG%FTFTAFT%ORBDRBTRBASTSTLBLKTOVPFPTS
2016-1722.515.5050.40.6.6881.22.23.40.30.50.80.62.04.2
2017-1823.612.5370.71.1.6211.03.54.52.00.80.50.82.07.3
2018-1924.602.5913.03.8.7851.65.36.93.10.90.71.93.016.9
Career.592.5641.41.9.7431.23.85.12.00.70.61.22.410.1

Data courtesy Basketball-Reference.com.

2. LeBron James

At first glance, this may look a little suspicious. Does this graph say that LeBron is worse than Siakam? No!

Again, this graph acts comparatively to the player’s first season. This means that the player’s point production, adjusted for usage rate and minute changes, is measured against his own point production at the beginning of his career.

So, what does this graph mean for LeBron? Surprisingly, LeBron has been pretty average in maintaining his averages on the average player graph. This means, his improvement has basically been perfectly modeled by the average 15+ point scoring player in the NBA.

However, recently (after his 13th season), he began to break off of this trend as he has surged greatly in the last few years, making these last few years some of his best in his career. In fact, instead of following the trend at year 13, James completely defied this trend, going the exact opposite direction. Although he did not make the playoffs, this year was probably one of his most efficient shooting years.

We would expect that LeBron James would take a major hit in the next 1 or 2 years. It is very likely that LeBron’s value will drop by a great amount next year, as his success is clearly not sustainable. However, he is LeBron James, so we can never know what will happen 😅.

3. Stephen Curry

Steph is a strange case. His career started off extraordinarily but after his first season, he was clearly on a downward trend. However, after becoming considerably worse and worse, Curry began to take his skills back during his 5th season in the league (disclaimer: he didn’t actually lose his skill; it was probably because he shot less because of Monta Ellis). Compared with other players, Curry reached his prime late, and it doesn’t seem like he is on a downward trend either.

Some More Players
I won’t explain these in detail, but you should be able to see how players improved/deteriorated.

4. Michael Carter-Williams

SeasonAge2P%eFG%FTFTAFT%ORBDRBTRBASTSTLBLKTOVPFPTS
2013-1422.440.4313.75.2.7031.44.86.26.31.90.63.53.016.7
2014-1523.428.4163.14.4.6941.04.45.36.71.70.53.82.514.6
2014-1523.412.4062.94.4.6431.05.16.27.41.50.44.22.515.0
2014-1523.457.4353.44.4.7800.83.14.05.62.00.53.22.614.1
2015-1624.472.4662.03.0.6540.94.25.15.21.50.82.83.011.5
2016-1725.401.3911.31.7.7530.52.93.42.50.80.51.52.36.6
2017-1826.365.3621.41.7.8200.71.92.72.20.80.41.01.94.6
2018-1927.424.4151.11.9.6040.71.82.52.50.70.50.71.74.8
2018-1927.429.4670.81.6.4620.20.60.81.30.60.40.61.74.3
2018-1927.419.3631.72.3.7411.33.44.84.10.90.80.81.85.4
Career.433.4232.33.3.7020.93.64.54.61.30.62.52.510.9
2 seasons.429.4223.44.9.6831.34.96.26.71.70.53.82.816.0
2 seasons.467.4552.43.4.7050.93.94.75.31.70.72.92.912.4
1 season.365.3621.41.7.8200.71.92.72.20.80.41.01.94.6
1 season.401.3911.31.7.7530.52.93.42.50.80.51.52.36.6
1 season.429.4670.81.6.4620.20.60.81.30.60.40.61.74.3
1 season.419.3631.72.3.7411.33.44.84.10.90.80.81.85.4

Data courtesy Basketball-Reference.com.5. Derrick Rose

SeasonAge2P%eFG%FTFTAFT%ORBDRBTRBASTSTLBLKTOVPFPTS
2008-0920.491.4822.43.1.7881.22.73.96.30.80.22.51.516.8
2009-1021.500.4953.34.3.7660.82.93.86.00.70.32.81.220.8
2010-1122.481.4855.96.9.8581.03.14.17.71.00.63.41.725.0
2011-1223.475.4735.06.1.8120.72.73.47.90.90.73.11.321.8
2012-1324
2013-1425.359.4022.73.2.8440.92.33.24.30.50.13.41.515.9
2014-1526.465.4503.03.7.8130.72.53.24.90.70.33.21.217.7
2015-1627.449.4482.22.7.7930.72.73.44.70.70.22.71.316.4
2016-1728.487.4773.54.0.8741.02.83.84.40.70.32.31.318.0
2017-1829.474.4541.61.8.8700.50.91.41.50.30.21.40.68.4
2017-1829.481.4622.22.6.8540.61.21.81.60.20.31.80.99.8
2017-1829.458.4350.60.61.0000.30.30.71.20.40.00.80.25.8
2018-1930.509.5182.73.1.8560.62.12.74.30.60.21.61.118.0
Career.481.4773.44.1.8240.92.63.55.60.80.32.71.318.8
7 seasons.478.4743.64.4.8130.92.83.76.20.80.42.91.419.7
1 season.487.4773.54.0.8741.02.83.84.40.70.32.31.318.0
2 seasons.505.5122.42.8.8610.61.82.43.90.60.21.51.016.2
1 season.481.4622.22.6.8540.61.21.81.60.20.31.80.99.8

Data courtesy Basketball-Reference.com.6. Kevin Durant

SeasonAge2P%eFG%FTFTAFT%ORBDRBTRBASTSTLBLKTOVPFPTS
2007-0819.455.4514.95.6.8730.93.54.42.41.00.92.91.520.3
2008-0920.486.5106.17.1.8631.05.56.52.81.30.73.01.825.3
2009-1021.506.5149.210.2.9001.36.37.62.81.41.03.32.130.1
2010-1122.504.5097.68.7.8800.76.16.82.71.11.02.82.027.7
2011-1223.535.5476.57.6.8600.67.48.03.51.31.23.82.028.0
2012-1324.539.5598.49.3.9050.67.37.94.61.41.33.51.828.1
2013-1425.549.5608.79.9.8730.76.77.45.51.30.73.52.132.0
2014-1526.565.5785.46.3.8540.66.06.64.10.90.92.71.525.4
2015-1627.569.5736.26.9.8980.67.68.25.01.01.23.51.928.2
2016-1728.608.5945.46.2.8750.67.68.34.81.11.62.21.925.1
2017-1829.565.5865.35.9.8890.56.46.85.40.71.83.02.026.4
2018-1930.587.5715.76.5.8850.45.96.45.90.71.12.92.026.0
Career.532.5426.87.7.8830.76.37.14.11.11.13.11.927.0
9 seasons.517.5307.28.1.8820.86.27.03.71.21.03.21.927.4
3 seasons.586.5835.56.2.8830.56.67.15.40.81.52.71.925.8

Data courtesy Basketball-Reference.com.7. Russell Westbrook

SeasonAge2P%eFG%FTFTAFT%ORBDRBTRBASTSTLBLKTOVPFPTS
2008-0920.415.4144.35.2.8152.22.74.95.31.30.23.32.315.3
2009-1021.438.4284.05.1.7801.73.14.98.01.30.43.32.516.1
2010-1122.451.4546.57.7.8421.53.14.68.21.90.43.92.521.9
2011-1223.482.4815.26.3.8231.53.14.65.51.70.33.62.223.6
2012-1324.466.4705.67.0.8001.43.95.27.41.80.33.32.323.2
2013-1425.482.4805.36.4.8261.24.55.76.91.90.23.82.321.8
2014-1526.457.4558.19.8.8351.95.47.38.62.10.24.42.728.1
2015-1627.503.4895.87.2.8121.86.07.810.42.00.34.32.523.5
2016-1728.459.4768.810.4.8451.79.010.710.41.60.45.42.331.6
2017-1829.485.4775.27.1.7371.98.210.110.31.80.34.82.525.4
2018-1930.481.4684.16.2.6561.59.611.110.71.90.54.53.422.9
Career.465.4655.77.1.8011.75.37.08.41.80.34.12.523.0

Data courtesy Basketball-Reference.com.8. James Harden

SeasonAge2P%eFG%FTFTAFT%ORBDRBTRBASTSTLBLKTOVPFPTS
2009-1020.424.4842.63.2.8080.62.63.21.81.10.31.42.69.9
2010-1121.514.5183.54.2.8430.52.63.12.11.10.31.32.512.2
2011-1222.579.5825.06.0.8460.53.64.13.71.00.22.22.416.8
2012-1323.477.5048.610.2.8510.84.14.95.81.80.53.82.325.9
2013-1424.515.5297.99.1.8660.83.94.76.11.60.43.62.425.4
2014-1525.480.5118.810.2.8680.94.75.77.01.90.74.02.627.4
2015-1626.494.5128.810.2.8600.85.36.17.51.70.64.62.829.0
2016-1727.530.5259.210.9.8471.27.08.111.21.50.55.72.729.1
2017-1828.531.5418.710.1.8580.64.85.48.81.80.74.42.330.4
2018-1929.528.5419.711.0.8790.85.86.67.52.00.75.03.136.1
Career.506.5257.38.5.8570.84.55.26.21.60.53.62.624.3
7 seasons.506.5248.810.2.8610.85.16.07.71.80.64.42.629.0
3 seasons.506.5293.64.3.8350.52.93.42.51.10.31.62.512.7

Data courtesy Basketball-Reference.com.9. Giannis Antetokounmpo

SeasonAge2P%eFG%FTFTAFT%ORBDRBTRBASTSTLBLKTOVPFPTS
2013-1419.440.4631.82.6.6831.03.44.41.90.80.81.62.26.8
2014-1520.511.4963.24.3.7411.25.56.72.60.91.02.13.112.7
2015-1621.537.5203.75.1.7241.46.27.74.31.21.42.63.216.9
2016-1722.563.5415.97.7.7701.87.08.85.41.61.92.93.122.9
2017-1823.554.5456.58.5.7602.18.010.04.81.51.43.03.126.9
2018-1924.641.5996.99.5.7292.210.312.55.91.31.53.73.227.7
Career.557.5394.66.2.7421.66.78.34.11.21.32.63.018.8

Data courtesy Basketball-Reference.com.10. Klay Thompson

SeasonAge2P%eFG%FTFTAFT%ORBDRBTRBASTSTLBLKTOVPFPTS
2011-1221.460.5201.21.4.8680.42.12.42.00.70.31.62.312.5
2012-1322.437.5091.61.9.8410.43.33.72.21.00.51.92.916.6
2013-1423.464.5331.82.3.7950.52.63.12.20.90.51.72.918.4
2014-1524.481.5552.93.3.8790.42.93.22.91.10.81.91.621.7
2015-1625.510.5692.42.8.8730.43.43.82.10.80.61.71.922.1
2016-1726.516.5652.42.8.8530.63.03.72.10.80.51.61.822.3
2017-1827.526.5851.11.3.8370.43.43.82.50.80.51.81.620.0
2018-1928.516.5531.72.0.8160.53.43.82.41.10.61.52.021.5
Career.491.5511.92.3.8480.43.03.52.30.90.51.72.119.5

Data courtesy Basketball-Reference.com.