So before the meltdown in Happy Valley, we were talking about our previous source of basketball angst: turnovers. One point raised was that it would be nice if we had more detailed data on turnovers. I can think of about a dozen types of turnovers:
Getting the ball stolen while dribbling/holding it
Throwing the ball to the defense
Throwing the ball out of bounds
Stepping out of bounds
5-second violation attempting to in-bound the ball
3-second lane violation
Over and back violation
Shot clock violation
- Offensive foul
Unfortunately, this sort of detailed information does not exist. One possibility would be manually tracking MSU’s turnovers on a game-by-game basis. This would give us some insight in to what’s causing their turnover problems, but we’d have no baseline to compare it to so the value would be somewhat limited.
There is one distinction we can draw with the turnover data currently available: whether the ball was stolen by the defense or turned over in some other manner. In terms of the list above, this distinguishes the first two types of turnovers from the rest of the list.
I think, in an ideal world, you’d want to be able to look at whether turnovers were forced or unforced. Forced turnovers would result from an inability to handle defensive pressure. Unforced turnovers would result from either (1) a mental error or (2) a failed attempt to create a basket. Arguably, though, any of the types of turnovers listed above could be forced or unforced, so even more detailed data might not get at the cause of turnover problems.
In terms of the steal-vs.-nonsteal distinction, my intuitive guess is that steals are more likely to be forced turnovers and nonsteals are more likely to be unforced turnovers. But I’m certain that the correlation is well below 100%. For example: An offensive player might make a bad, unforced pass that goes directly to a defender for a steal. On the flip side, defensive pressure might cause a player to travel with the ball.
The split between steals and nonsteals is about 50/50. For Big Ten teams this season, 47-48% of turnovers are the result of steals. The remaining 52-53% are nonsteals.
To explore this further, let’s look at the current Big Ten team turnover data–both offensive and defensive. Be forewarned that the analysis of this data will mostly fall under the category of “thinking out loud.” All data are through games of this past Sunday.
No clear pattern in terms of style of play for the teams that turn the ball over the most. Iowa runs a motion offense. MSU pushes the ball some, but relies on a lot of set plays. Minnesota tends to play more creatively on offense.
At the other end of the spectrum, you have less athletic teams that hold on to the ball, but don’t necessarily create great shooting opportunities for themselves. (Note: When I say “less/more athletic,” it’s obviously based on my own subjective analysis. But I think it would hold up if you looked at, for example, the recruiting ratings of the players in each program.)
% of Offensive TOs on Steals
The teams that turn the ball over more on steals than nonsteals are less athletic. There’s no correlation with overall TO%. Northwestern has the best overall offensive TO%; Iowa has the worst.
The teams that turn the ball over more on nonsteals are the more athletic teams. This could be the result of (1) a lack of mental discipline or (2) more aggressive individual attempts to create shot opportunities off the dribble, some of which result in turnovers like traveling and offensive fouls.
At the top of the list, Minnesota and Northwestern both play defensive schemes designed to create turnovers (so does Michigan, they’re just not very good at it). Purdue and Wisconsin just play very good man-to-man defense.
At the bottom, Michigan State and Iowa are man-to-man teams. Izzo has never focused on creating turnovers. Ohio State plays mostly the 2-3 zone–which is a fairly passive defensive scheme focused on not giving up shots near the basket.
% of Defensive TOs on Steals
Minnesota and Northwestern have, by far, the highest defensive steal percentages in the league (15.3% and 13.4%, respectively). This is why they appear at the top of both the defensive TO% list and this list. The next three teams on the list (IU, MSU, Ohio St) are arguably the three most athletic teams in the conference. So a tendency to create more turnovers off of steals than nonsteals may result from either (1) pressure defenses or (2) superior athleticism.
The teams at the bottom of this list are arguably the least athletic. And the bottom four all play mainly man-to-man defense. Purdue and Wisconsin create a lot of turnovers off nonsteals. Playing solid man-to-man defense may result in opponents forcing things on offense and committing unforced turnovers (traveling, offensive fouls, etc.).
I’m pretty sure I’m guilty of some combination of (1) oversimplification and/or (2) overanalysis here. But my take on the data above is to draw the following conclusions:
1) MSU has the second highest offensive TO% in the league, but the lowest % of turnovers off steals. This is a function of the fact they lead in league in the % of their possessions that end in nonsteal turnovers–while ranking only 7th in the % of their possessions that end in a steal by the other team.
2) Intuitively, this seems to indicate that a large percentage of MSU’s turnovers are unforced–the result of mental errors, rather than an inability to deal with defensive pressure. It’s possible it’s a result of more aggresive offensive play, which results in things like traveling and bad passes to teammates that go out of bounds. But that doesn’t jive with MSU’s offensive structure, which doesn’t allow for a lot of one-on-one play.
3) The Big Ten data imply that a higher percentage of turnovers resulting from steals tends to indicate either (a) the defense is more athletic than the offense and/or (b) the defense plays a scheme designed to create turnovers. Conversely, a lower percentage from steals would indicate (a) the offense is at least as athletic as the defense and/or (b) the defense does not play high-pressure defense.
4) The implications stated in (3) make sense for MSU, as it’s clearly one of the most athletic teams in the league and has turned the ball over in higher percentages against a number of teams that do not utilize defensive schemes designed to create turnovers.
5) Final statistical observation: The way that an opponent tends to create turnovers doesn’t seem to affect MSU’s propensity to cough it up. MSU hasn’t posted offensive TO percentages abovet 25% against the two school who rely most on steals to create turnovers (Minn, NW), as well as to the three school that rely least on steals to create turnovers (Pur, Ill, Iowa).
So I’m not sure I’ve accomplished a whole lot here in terms of examining’s MSU’s turnover problems. It appears to be a problem of discipline/execution–not an issue of being overmatched. But I think we knew that. As noted previously, this is a team that has pretty good individual ball-handling skills–including three Big Ten quality starting point guards.
The more abstract question remains: Is correcting the problem a matter of “toughness” or “intensity”? Or is there something fundamentally incompatible between MSU’s offensive system and the current group of players that results in those players making bad decisions with the ball when they’re not playing within the structure of a set play? After a season and a half of persistent turnover problems, I’m starting to lean more toward the latter.
Again, this isn’t to say the benefits of MSU’s offensive philosophy don’t outweigh the costs. But the costs can’t be ignored and, increasingly, it’s hard to see that the costs created by turnovers can be reduced in time to be an elite team by March.
This has turned out to be a pretty long-winded post. (I’ve pretty blatantly ignored the standard advice to writers about knowing what you want to say before you start writing.) If you made it this far, feel free to chime in with your own thoughts. At minimum, hopefully I’ve provided a little more context for observing MSU’s turnover tendencies in future games.