Our problem on defense is not offensive turnovers

I’ve been thinking a lot about this question ever since Frank Martin said the following at a press conference:

Our team defense has been not great, but it’s been good enough to win with. The problem is we’re giving teams so many free points because of our inability to play offense.

Quite frankly, I think that’s wrong.  Or, at least, 99% wrong.

I took a swing at this question on Twitter the other day, but decide to spend some time today digging deeper into the question.  I started by seeing how our turnovers correlated to the points per possession of the opposition.  Then I wondered if maybe steals were a more effective question, since those were more likely to result in fast break opportunities.  The answer?  Well, neither.  At all.  In fact, the more turnovers USC commits, the lower its opponents PPP thus far (the correlation is -0.63, and surely insignificant).  Steals similarly told no story, with a correlation of -0.19.

Opp. Opp. PPP Opp.
Stl %
USC TO%
UWM 94.9 8.8% 21.3
Morgan St. 92.2 14.7% 25.2
Rider 104.1 12.0% 24.1
Elon 95.6 17.4% 31.8
Missouri St. 91.8 13.7% 32.7
UALR 92.5 16.2% 30.8
St. John’s 136.9 11.9% 17.9
Clemson 101.6 14.9% 28.2
Jacksonville 96.1 8.9% 21.6
Correlation: -0.19 -0.63

OK, so if that’s not where the difference comes in, where does it?

My next thought was that surely this should show up in points off of turnovers.  So, I took the same data and added that information for each game.  Then, I separated out possessions that end in turnovers from possessions that do not end in turnovers, and took a look at the efficiency of each of our opponent’s based on each type of possession:

Opp.
Opp. PPP
Opp. Pts.
USC Poss.
USC TOs
USC TO%
Pts. off TO
Opp. PPP (non-TO)
Opp. PPP (w/ TO)
Delta
UWM
94.9
75
80
17
21.3
15
95.2
88.2
(7.0)
Morgan St.
92.2
71
75
19
25.2
20
91.1
105.3
14.2
Rider
104.1
76
75
18
24.1
21
96.5
116.7
20.2
Elon
95.6
65
69
22
31.8
18
100.0
81.8
(18.2)
Missouri St.
91.8
67
73
24
32.7
26
83.7
108.3
24.7
UALR
92.5
62
68
21
30.8
14
102.1
66.7
(35.5)
St. John’s
136.9
89
67
12
17.9
10
143.6
83.3
(60.3)
Clemson
101.6
64
67
19
28.2
23
85.4
121.1
35.6
Jacksonville
96.1
74
79
17
21.6
20
87.1
117.6
30.6

So let’s break this down a little bit.  Over the course of the season thus far, South Carolina has turned the ball over 169 times, and conceded 167 points immediately after those turnovers (for an opponent PPP of 0.988).  On the 484 defensive possessions that did not immediately precede a turnover, South Carolina has conceded 474 points (for an opponent PPP of 0.979).  That’s a difference of 0.17 points per game, or 0.009 points per possession.  Or, nothing.

Simply put, offensive turnovers do not hurt the defense (because of course after I did the research, I went online and found out that Ken Pomeroy had already figured this out 6 years ago).

So, does this mean we’ll shut up about turnovers?  Absolutely not. How do turnovers hurt you?  Well, they prevent you from having an opportunity to shoot, which is something the Gamecocks aren’t half bad at doing.  And if you turn it over before you get a shot off, it also prevents you from the chance to get an offensive rebound, and thus another shot.  Turnovers are (not surprisingly) bad for your offense.

So how bad are they for Carolina’s offense?  Well, on possessions where the Gamecocks have not turned the ball over, they’ve been scoring 1.382 points per possession, an excellent rate.  Of course, on possessions ending in a turnover, they’re scoring a disappointing 0.000 ppp, which can be quite harmful if you’re turning the ball over on almost 26% of your possessions.

In fact, given that South Carolina (on an adjusted basis) averages 66.3 possessions a game, those almost 18 turnovers a game are costing the Gamecocks 24.7 points a game on offense.  Alternatively, based on what we learned earlier, the 0.009 extra points the Gamecocks give up per possession on possessions that end in a turnover (as opposed to those that do not: 0.988 less 0.979) cost South Carolina, on average, per game, 0.0016 points.  Or, nothing.  Absolutely nothing.

So what did we learn today?  Well, I learned three things.  First, offensive turnovers and defensive efficiency, contrary to our experience, have nothing to do with each other.  This is a myth we can lay to rest (or at least try to lay to rest, I’m sure your mileage will vary on that).

Second, we are on the cusp of having a very good offense.  No one would suggest you can completely eliminate turnovers, but if South Carolina could even get down to the national average of 21.0% of possessions, that’s an extra 3.5 possessions a game where the Gamecocks have the ball.  Multiply that by the 1.38 points we score every possession we don’t turn the ball over, and you’ve just swung our offense upward from an adjusted efficiency of 102.1 (121st in the nation) to 106.7, which would be good for 59th in the nation.  That’s not going to win you championships, but for this team, it’d be a nice result, and the best offense we would have seen at Carolina since the 2008 team that was led by Devan Downey (who interestingly, never turned the ball over – only 15.3% of their possessions).

Which leads us to the third thing – there’s now one less excuse for our putrid defense, currently giving up an average of 1.006 ppp against horrid competition (when adjusted for SOS, KenPom.com says the average NCAA team would score 1.033 ppp against us, which is “good” for 254th in the nation).

What really has me confused about this is that defense is where Martin’s teams normally shine: at K-State,  he never put a defense on the floor that gave up over 0.93 ppp (adjusted for SOS again), which came to 42nd in the country.  This isn’t just the worst defense Martin’s ever put on the floor in college, it might be if you go back and start including some of his high school teams.

We’ll see if there are creative ways to parse out what’s causing that as we go forward on the blog.  For now, cut out the turnovers, guys.  It may not be hurting our defense, but it’s really bringing down our offense.

About these ads

About marvinnedick

Blogging from the mid-Atlantic on Gamecock sports, as well as general musings on sports theory otherwise.
This entry was posted in Uncategorized and tagged , , , , , , , . Bookmark the permalink.

2 Responses to Our problem on defense is not offensive turnovers

  1. Pingback: Better Know an Opponent: Manhattan |

  2. Pingback: What are the problems with our defense? |

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s