Experience, Looked at Differently.

In the last post by Steve DiMiceli, Steve featured an article on the significance of experience.  As the resident math expert, I’m here to look at the data a little differently than Steve.

Looking only at the last basketball season (the 2012-13 season), data was collected on all 347 teams on percentage of minutes played by each class as well as the number of members of each class.  For example, Duquesne had two seniors, two juniors, six sophomores, and five freshmen on their roster.  The two seniors playing 23.9% of the total minutes.

Also, an average minutes played per class year was computed.  For the computations, a seniors minutes was multiplied by 4, a juniors by 3, a sophomore by 2, and a freshman by 1 to get an average minutes played per class.  Thus, a score of 4 means that seniors were getting every minute on the court, a score of one means that freshmen play every minute, and a score of 2.5 means that the average experience is that of a junior-sophomore combination or senior-freshman combination.

I looked at three major categories for some clarity on the experience vs. winning debate.  I looked at comparing experience to winning percentage, experience level to making the NCAA Tournament, and starting lineups for A-10 players.


1.  Stats by Winning Percentage

Experience per Minute     Winning Percentage
1.00 – 2.00                           42.8%
2.01 – 2.5                              45.3%
2.51 – 3.0                              54.0%
3.01+                                    55.5%

Experience per Minute     Conference Winning Percentage
1.00 – 2.00                           40.4%
2.01 – 2.5                              42.7%
2.51 – 3.0                              52.1%
3.01+                                    56.5%

As such, we can see that the winning percentages of teams that feature primarily upperclassmen are the teams that are more likely to be above 500, whereas the teams that have less than the half-junior half-sophomore mix at 2.5 tend to be the losing teams.  Obviously, there are outliers, but this was the general case.

Looking at the A10, there were five of the 16 teams that had an experience per minute score below 2.5.  Of those five (Fordham, Charlotte, Duquesne, George Washington, and Richmond), none had a winning conference record as those schools averaged a 34% winning record in conference play.  Looking some other non-BCS conferences, only UTEP, Boise State, and UNLV had winning conference records with a score below 2.5 from CUSA, the MVC, MWC, and WCC.

2.  Correlation Analysis:

There was a positive 23% correlation between conference winning percentage and experience as calculated above across the NCAA for last year.  While the correlation is low, there is a clear correlation between winning percentage and experience that is not likely explained by chance alone (less than 0.01%).  As such, Steve’s original question, “Winning and Player Experience:  Is It Connected?” should be answered with a we’re over 99.99% sure that they’re connected, but the correlation between them is low (meaning there is more correlation between winning percentage and other items).

Stats on Reaching the NCAA Tourney

Average years of experience of teams that danced:  2.833
Average years of experience of those that did not:  2.629

If you look only at the non-BCS high majors (A-10, CUSA, MVC, MWC, and WCC), then the numbers change to a similar

Average years of experience of teams that danced:  2.865
Average years of experience of those that did not:  2.653

The roughly two-tenths difference is about the difference in class status of one player being elevated by one grade level.

3.  A-10 Starting Lineups

I looked back over the starting lineups of A-10 teams that reached the NCAA Tournament over the last five years.  They consisted of the following:

Number of Seniors:  37
Number of Juniors:  31
Number of Sophomores:  21
Number of Freshmen:  1

For comparison, here is the Dukes starting lineups over the last five years:

Number of Seniors:  8
Number of Juniors:  6
Number of Sophomores:  6
Number of Freshmen:  5

76% of players that start on NCAA Tournament teams from the Atlantic 10 that are upperclassmen with only one starter for an A-10 team in five years.  Compare that to Duquesne’s numbers, which are much more spread apart amongst the classes.  Simply, you can’t rely on freshmen in A-10 play unless they’re extremely elite if you want to dance.  Like Steve said, this to me is one of the biggest reasons why Duquesne hasn’t been to the tourney.  The Dukes simply haven’t had enough veterans together on one team often enough to give themselves a shot and have relied on an average of one freshman per year, which hasn’t been a recipe for success in Atlantic 10 play.

Tags: Men Men's Basketball

comments powered by Disqus