Death By Turnover pt. 2
Now, the first part of this discussion, in which I emphasized not just that turnover margin is important but that winning the turnover battle is very closely correlated to winning the football game, leads me to my second point, in which I will discuss how turnovers relate to our oft-maligned defense, as well as our recently-maligned offense.
First, I’ll tackle the offense. With regards to turnovers, all the offense can do is not cough up the ball. Of course the primary goal is to score points, but an offense that is concerned with trying to win the turnover battle will necessarily become rather conservative, running the ball often and throwing lots of short, high-percentage passes. This does not mean that the offense gets predictable or uncreative; it simply means that risk, wherever possible and prudent, is downplayed.
Of course, even if the offense protects the ball perfectly, if the defense fails to make any plays and recover a turnover here or there, the turnover margin will be even at 0. In cases like this, where mistakes are minimized by both teams (note that this also includes penalties and such), the better, more talented team will almost always win. Thus, more talented favorites will tend to execute conservative, by-the-book offenses, while less talented underdogs will resort to high-risk/high reward offensive strategies to try and equalize the talent gap. I remember Tedford calling lots of trick plays and such back in 2002, when the Bears didn’t have the talent level they do now, but now that Tedford has players he feels can simply outplay many of their opponents, Tedford’s play-calling has gotten, in my completely unresearched opinion, more conservative.
On defense, well, that’s a different story. If Ken Crawford is to be believed, more than 90% of turnovers are the result of offensive mistakes; yes, defenses do ‘force’ turnovers sometimes, but not very often. If this is the case, a defense that wants to maximize turnovers should simply lay back and wait for them to happen. In fact, the ideal defensive philosophy to accomplish such a goal would be the much-criticized "Bend But Don’t Break". Simply put, if offenses are going to commit a turnover every [small percentage] of plays, the defense should focus on forcing the offense to run as many plays as possible. If an offense takes 15 plays to march down the field every time it gets the ball, it’s 3 times more likely to turn over the ball than an offense that’s able to score every time in only 5 plays.
So, if we think that winning the turnover battle will, in a large part, help us win the game, and are thus concerned with maintaining as large a turnover ratio as possible, what we come up with is a rather conservative, by-the-book sort of game plan. If we feel confident in the superior athletic skill and training of our players, we simply put them in low-risk situations and let our superior talent win out, which the percentages say will happen over the course of a football game. Now, I’m not saying that winning the turnover battle is the primary motivation behind Cal’s game plan. What I am saying is that, looked at in this context, one can see how their game plan, given the athletic talent available, would give Cal a great chance to win every football game that they play. Does the game plan need to be adjusted in-game, reacting to the tendencies of the opponent? Of course. Do risks need to be taken in order to beat good football teams? Sure. Do players still need to execute the game plan for it to be effective? Absolutely. But such failures should not cause one to tear down to its foundation what, at its root, is a solid game plan.
—–
Now, Ken Crawford took issue with one of the conclusions of my last post, so much so that he went and wrote an entire post about it. In response to his reasoned and thoughtful argument, I thought I’d take an entire post to respond.
One of the main points of Mr. Crawford’s post was that he believes more than 90% of turnovers are the result of offensive mistakes. Personally, I think the percentage of turnovers that are ‘forced’ is quite a bit larger than 10%. If turnovers were largely a function of offensive mistakes, we would see that turnovers recovered would be mostly dependent on the schedule a team played. Accounting for variations in ‘forced’ turnovers and luck, we would also find that teams in the same conference, which play mostly the same schedules, would recover, in a large part, the same number of turnovers. However, the statistics don’t bear that out. Last year in the Pac-10, Oregon State recovered 16 fumbles, while Washington, which played mostly the same schedule, recovered only 4. Perhaps OSU benefited from being able to play Washington, while UW suffered from having to face Oregon State? Nah, as OSU also gave up 16 fumbles, while Washington only gave up 6. In fact, in their game last year, Oregon State gave up one of the 6 fumbles Washington recovered, while Washington didn’t turn over the ball once. A look at the interception statistics tells much the same story: some teams are clearly better than others at recovering turnovers. Yes, the offense has to make mistakes for turnovers to happen, but it’s the good defenses that will make them pay for their mistakes, stepping in front of a lazily thrown pass or punching a poorly-protected ball out of a running back’s hands. To some degree at least, a large percentage of turnovers are, in fact, ‘forced’.
Now, I definitely agree with Mr. Crawford’s assessment of Oregon State; their large number of turnover were masking what is actually a very good defense, and the very poor turnover margin turned out to be somewhat of a red herring. In general, however, I think that turnover margin is a characteristic of the team, indicating how sloppy the offense is with the ball and how well the defense gets after it. While these are things that can be worked on, it’s not an easy, overnight fix. The problem with Oregon State was that they’re so inconsistent with regards to turning the ball over that if they play one of their ‘good’ games (as they did against Cal), they’re very tough to beat. However, regarding Mr. Crawford’s comment, I think I see a different trend concerning USC.
When you talk about turnover margin, the conversation pretty much has to start with USC. During Pete Carroll’s first five years at USC, the Trojans finished in the top five nationally in turnover margin every year, including first in 2004 and second in 2003 and 2005.
USC Turnovers by Year
| Year | Rank | Games | TO Gained | TO Lost | Margin |
| 2007 | 89 | 7 | 13 | 17 | -4 |
| 2006 | 38 | 13 | 22 | 18 | 4 |
| 2005 | 2 | 13 | 38 | 17 | 21 |
| 2004 | 1 | 13 | 38 | 19 | 19 |
| 2003 | 2 | 13 | 42 | 22 | 20 |
| 2002 | 5 | 13 | 36 | 18 | 18 |
| 2001 | 5 | 11 | 33 | 19 | 14 |
| 2000 | 113 | 12 | 17 | 36 | -19 |
These guys are the gold standard of turnover margin. Or, at least, they were up until last year. Last year, they continued to avoid turning the ball over, but after 5 straight years of recovering at least 33 turnovers per year, recovered turnovers dropped to only 22 last year. This year, not only are turnovers gained still down, but the offense has started coughing up the ball at an alarming rate; they’ve already given up the ball 17 times this year, as opposed to 18 times all of last year. More than the offhand line I tossed into my last post, this is why I think USC is in trouble this year; for whatever reason, they no longer are able to force lots of turnovers, and now they’re increasingly less able to hold onto the ball.
What this suggests to me is that USC is a team that is getting by on talent and a soft early schedule; the precise, well-oiled football machine of the Trojans’ halcyon days is gone, or at least dormant. Yes, they’re still über-talented, but look at who they’ve played: a consistent doormat in Idaho, a fading and overrated Nebraska, the worst 4 teams in the Pac-10 (Washington State, Washington, Stanford and Arizona), and an awful, awful Notre Dame squad. Jeff Sagarin rates their schedule thus far as only 91st in the country; in terms of what they’ve accomplished on the field so far this year, USC’s résumé looks a lot like another 6-1 team, UConn. Is anyone quaking in their cleats at the thought of facing the University of Connecticut Huskies (football team, at least)?
Bottom line, USC is still a very good, dangerous football team. However, they just aren’t doing all the little things that made them turnover machines during 2001-05, the little things that got them within a Vince Young of three consecutive national titles. To me, this makes them much more vulnerable than they have been in a long time, and that’s why I’m looking forward to Cal’s November 10th matchup with them.
