Teams' Steals, 2001-2016; Have Things Stabilized?

[connie]

Here are the charts for NCAAWB Average For All Teams (And UConn), And Standard Deviation. Years 2000-01 to 2015-16. All Div. I teams except those that were reclassified.

X-axis=year
Y-Axis=steals/game

As you can see, the Div. I average has stayed pretty flat since 2001, with a slightly downward trend the last several seasons. Less than 8 steal/game by end of 2015-16. Are offenses learning their lessons?

During that timeframe, UConn average steals/game increased, at least since 2008-09.

Next, standard deviation:

X-Axis =year
Y-Axis= standard deviation (average number of steals that each team differed per game from Div. I average).

This shows us how far, on average, teams were from the Div. I mean. Lower number means all teams were closer to the league average -- more parity. Higher number means all teams were further from the league average -- less parity.

If anything, there is a very slight increase in std. deviation, from 2006-07 to the 2014-15, with a slight decrease the last 2 seasons. This suggests an increase in lack of parity over the past ten years, with a bit of narrowing the last two seasons.

Difficult to draw conclusions. But arguably this data shows a kind of stabilization with respect to steals, with offenses perhaps learning to protect the ball just a bit better. Defenses (except UConn's) do not appear to stealing more.

 

[NVUConnFan]

I like statistics and math. One of my favorite subjects in college was stochastics (used in many different fields, particularly where stochastic or random processes are used to represent systems or phenomena that seem to change in a random way, such as image processing, computer science, physics and astronautical engineering). The above data does not appear random at all or nearly not so. However in any one game (being a random sample), there can be great variation both from the data above (even with the small standard deviation) and also between teams. Interesting...

 

[UcMiami]

Here are the charts for NCAAWB Average For All Teams (And UConn), And Standard Deviation. Years 2000-01 to 2015-16. All Div. I teams except those that were reclassified.

View attachment 19941
X-axis=year
Y-Axis=steals/game

As you can see, the Div. I average has stayed pretty flat since 2001, with a slightly downward trend the last several seasons. Less than 8 steal/game by end of 2015-16. Are offenses learning their lessons?

During that timeframe, UConn average steals/game increased, at least since 2008-09.

Next, standard deviation:

View attachment 19942
X-Axis =year
Y-Axis= standard deviation (average number of steals that each team differed per game from Div. I average).

This shows us how far, on average, teams were from the Div. I mean. Lower number means all teams were closer to the league average -- more parity. Higher number means all teams were further from the league average -- less parity.

If anything, there is a very slight increase in std. deviation, from 2006-07 to the 2014-15, with a slight decrease the last 2 seasons. This suggests an increase in lack of parity over the past ten years, with a bit of narrowing the last two seasons.

Difficult to draw conclusions. But arguably this data shows a kind of stabilization with respect to steals, with offenses perhaps learning to protect the ball just a bit better. Defenses (except UConn's) do not appear to stealing more.

Interesting stuff again, but re UConn or any specific team, the specific players playing the majority of the minutes is such a huge variable that I am not sure there is any conclusion to draw - Stewart was a great shot blocker, Tuck was a solid defender, but neither had the speed or quickness (or desperation) of Gabby or Napheesa - I added desperation, because if a good cebter actually catches a pass in a good low post position against either, they are likely toast. Stewart and Tuck could defend those same [;ayers straight up. The plus side of course was Moriah who was very fast and quick.
The numbers on all teams are based on a large enough universe to probably tell something, but to my untrained eye it isn't a very distinct story and could be a simple aberration - 0.5 difference in steals per game over 15+ years out of 8.5 steals per game is not all that significant. The team deviation is all over the place and again gets into more issues with specific players playing for specific coaches. I imagine Rutgers steals per game are about half what they were in 2005 when CViv had great players that fully bought into her defensive system.

 

[connie]

Interesting stuff again, but re UConn or any specific team, the specific players playing the majority of the minutes is such a huge variable that I am not sure there is any conclusion to draw - Stewart was a great shot blocker, Tuck was a solid defender, but neither had the speed or quickness (or desperation) of Gabby or Napheesa - I added desperation, because if a good cebter actually catches a pass in a good low post position against either, they are likely toast. Stewart and Tuck could defend those same [;ayers straight up. The plus side of course was Moriah who was very fast and quick.
The numbers on all teams are based on a large enough universe to probably tell something, but to my untrained eye it isn't a very distinct story and could be a simple aberration - 0.5 difference in steals per game over 15+ years out of 8.5 steals per game is not all that significant. The team deviation is all over the place and again gets into more issues with specific players playing for specific coaches. I imagine Rutgers steals per game are about half what they were in 2005 when CViv had great players that fully bought into her defensive system.

I agree that without more, making sense of this data is guesswork.

Still, we are talking about a lot of basketball minutes here. 30+ games of 40 minutes for each of 330+ teams over a 14 year period. That's over a half million minutes of basketball. And on average, someone gets a steal about every five minutes of play. That seems to have remained fairly constant. Don't know what it means, except that steals constitute a change of possession and therefore it takes away an opportunity to score while creating one.

I think a more interesting and potentially meaningful statistical study would compare overall team points per game average and margin of victory. It would be interesting to see if Standard deviation for margin of error shrinks (indicating increased parity).