Earlier this season, I wrote a post about how San Francisco did an excellent job limiting their opponent's starting field position. After a similar dominant performance against Pittsburgh in Week 15, I decided to revisit this topic. This year, the 49ers lead all teams in average starting field position for their opponents; their opponents, on average, start at their own 24.4 yardline. The next best teams? New England Patriots and Green Bay Packers. In fact, this year, there is a 0.73 correlation between win percentage and average opponent starting field position.
But let's go back. Since 2000, we see a 0.45 correlation between win percentage and average opponent starting field position.
In fact, for every 3 yards closer to their opponent's end zone, a team will add 2 wins on average. Kickoff rules have changed the game, however. Since 2000, average starting field position was around the 31 yardline. This year, it is the 28.4 yardline. It is this change that may be the cause of the noticeably higher correlation in 2011 - starting field position is now at a premium. The previously highest correlation in a single year since 2000 was in 2000 at 0.63.
But, how can a team limit their opponent's starting field position? There are three main factors: Kickoffs, Punts, and Turnovers. Using the assumption that kickoff and punt coverage are relatively normally distributed - more specifically, they are not predictive or consistent on a season-to-season basis - we are left primarily with punter ability, decisions to punt, and turnovers.
There is a negligible correlation to actual punting efficiency but a -0.36 correlation to average line of scrimmage for a punt. The negative signifies that as your punt line of scrimmage is closer to the opponent's end zone, the further your opponent's average starting field position will be from your own end zone.
The big kahuna is offensive turnovers. Turnovers, on average, swing field position around 40 yards and as a result, they are the largest contributing factor to opponent field position. Since 2000, the correlation between total turnovers and average opponent starting field position is -0.54 (again, the negative symbolizes the fact that fewer turnovers means a higher total yards from your end zone).
So far this year, there is a -0.50 correlation between the two, with the 49ers (9), Packers (9), and Patriots (15) leading the league in fewest turnovers. Here are this season's numbers:
I also looked at overall offensive and defensive efficiencies to see if this had a direct effect on opponent starting field position; this resulted in a 0.38 and -0.16 correlations respectively. It seems reasonable, however, that the majority of the offensive efficiency contributing to opponent starting field position would be turnovers.
Is limiting your opponent's field position important? Yes. Are efficient punts and kickoffs important in doing so? Yes, on a per event basis. But, more important than both is taking care of the ball. The 49ers, Packers, and Patriots all know this and their records reflect it accordingly.
Keith Goldner is the creator of Drive-By Football, and Chief Analyst at numberFire.com - The leading fantasy sports analytics platform. Follow him on twitter @drivebyfootball or check out numberFire on Facebook.
Even more reasons to go for it on 4th down and goal to go. If you kick the gf, they get the ball on their 20. If you go for it and fail, they take over deep in their territory.
The strong correlation doesn't necessarily mean that field position is important. Imagine if the rule was that, after a score, the opponent gets possession at their own 1 yard line. Then there would be an extremely strong correlation, because teams that score a lot (good teams) would have a lot of opponent possessions starting at their own 1 yard line. But that would not be a benefit of field position - it would be due to the fact that having a good offense tends to give your opponent bad field position.
Moving the kickoff line to the 35 shifted things in that direction - now scoring gives your opponent worse field position than it did last year - which is probably why the correlation is higher this year. Good offenses have fewer turnovers (turnovers give the other team great field position), score more (kickoffs now give the other team worse field position), and are more likely to punt after stalled drives rather than 3-and-outs (which means punting from better field position, which gives the opponent worse field position). On good offenses the punter is more likely to be punting from around midfield and trying to pin the opponent inside their 20, rather than booming a punt from deep in their own territory - as Brian discussed, that leads to worse average field position for the opponent but also a shorter average punt distance for the punter.
***Using the assumption that kickoff and punt coverage are relatively normally distributed - more specifically, they are not predictive or consistent on a season-to-season basis....***
This seems like a very large assumption. I would not concede it. Certain special teams coaches routinely field better units wherever they go, like Brad Seeley. Certain punters -- Lechler and Lee are the ones I'm most familiar with -- routinely have higher gross and net averages. This seems like one advantage that would be extremely consistent.
I see how you strategically only mentioned the top-3 in terms of feild position since Seattle of all teams is 4th.
Actually, he strategically "left out" every team except for the top 32 in this analysis.
In addition to the offensive efficiency issues Vince links to above, I think there's a problem with the method of analysis.
The most turnovers on here for any given team is ~50. Seeing that the average team gets ~10 drives per game, you're discarding somewhere between 65 and 95% of drives for most teams. You're then correlating an average field position from 100% of the data to the remaining 5-35% of data. (Additionally you haven't accounted for TOs in a rate-dependent fashion.) Turnovers are a fundamentally different way to give the ball back, so the distribution of TO field position is also likely to be very different from kickoffs, which are probably different from punts.
Turnovers are also intrinsically linked to a loss of offensive scoring opportunities. How does correlation of field position relate to wins differently than changes in EPA based on turnovers? That statistic seems to already capture the point you're trying to make.
Vince -
I'm not saying field position is the most important thing, but it is important because it limits your opponents probability of scoring.
Anonymous -
I think that is a fair assumption. While there are some better punters/kickers (which was addressed above in the punt efficiency) the coverage itself is extremely variable. You rarely see punt/kickoff coverages as predictive from year to year.
G-
I understand what you are saying about treating turnovers as a rate statistic but I don't think this is as important of a distinction.
The general concept I was trying to convey is that opponent field position can have an effect on winning games. The biggest determining factor in that is turnovers. Offensive efficiency is somewhat important given the change in kickoff rules, but the largest contributor is turnovers. Just another big effect turnovers have on the game.
Also, addressing those comments about me leaving out everything but the top 3, look at those teams on the left and the right of the graph. Pretty clear which side has the noticeably better teams.
But you didn't answer the most important question. How much of the 49ers opponent starting field position is luck and how much is repeatable? In other words, what would you expect the 49ers opp starting field position to be next year? What was it for the Packers and Patriots in the previous two years? What about the league general year-to-year opp starting field position correlation for all teams?
If it is not repeatable, then maybe a lot of it explains the 49ers' great record. And thus an expected regression to the mean tomorrow.
The top teams in 2010: ATL, KC, PHI, BAL, NE, NYJ (all solid teams). SF and GB both turned the ball over 21 times, fairly average.
To answer your question, a lot of turnovers are due to normal variability, Brian has discussed this. However, there are certain things teams or players can do to limit their turnovers. Players like Aaron Rodgers do a great job with this which would be somewhat predictive in terms of starting field position.
I would expect a regression to the mean next year for the 49ers given Alex Smith's track record.
I don't get the point of this article, esp. given the title. It starts out trying to talk about field position, but then quickly backtracks to say that really only turnovers matter (in determining avg. field position), and even they are mostly random. I think any reader of this site already knew this.
Oops, most of my comment didn't get published, Here is take two.
Assuming the validity of the hypothesis that there is a correlation between average starting line of possession and number of wins, as a coach I would always punt on fourth down. Punting the ball will always (assuming no blocked kicks) have my opponents starting furtherest from my goal line.
Readers of this site "know" that frequently the optimum decision is to go for it on fourth and short from the opponents 41 yd line. This article offers an insight into why coaches punt in these situations. As a coach I chose to punt as it gains me 21 yards in field position.
We know that punting surrenders 3% - 8% of WP in this situation. (See The Best Gift of All.) It would be interesting to see how much WP field position gains.
Perfect game to illustrate this - Steelers loss to 49ers this year. Steelers outgained the 49ers by 100 yards and had something like 8 explosive plays over 20 yards to 3 for the 49ers - but the Steelers only scored 3 points and the game wasn't remotely close.
Average starting pos for SF drives was near the 30 yd line, while PIT averaged near their own 15.
To use Brian's terminology, it seems like starting field position is, like time of possession, an "intermediate outcome". It's the product of being good at the fundamentals of football (passing efficiency, running efficiency, limiting turnovers, etc.). But a team shouldn't try to be "good" at starting field position as a goal to itself. Otherwise, it would lead to sub-optimal decisions, like punting on your opponent's 37 on 4th and 5 as opposed to going for it.
Here's the link to the time of possession article, which explains the concept better than I could:
Whoops. Here is the link.
http://www.advancednflstats.com/2009/09/time-of-possession.html
***I think that is a fair assumption. While there are some better punters/kickers (which was addressed above in the punt efficiency) the coverage itself is extremely variable. You rarely see punt/kickoff coverages as predictive from year to year. ***
Evidence to support this?
I'm not a statistician, but if you look at Brad Seely's ten years with New England, for instance, you can see that over that time kicks, punts, and coverage were all well over average.
There are several nice points made. The first is a comparison of Field position to probability to score. This makes a defense look potentially stronger.
The second not correlated but mentioned in this thread but has data available from this site is Field position to probability of turnover. The longer the drives the more likely a turnover occurs. Field position is no doubt a factor as to why SF had the highest takeaway ratio. The last point is special teams that can alter field position. A single booming punt can change position by 10 yards or more. I have seen several games where offenses are pinned inside their 20 only to flip the field with a booming punt. Again, Having the best punter in the league substantiates the starting field positions.
It would be nice to see a correlation of scoring by team to oponents field position and a correlation of starting field position to scoring. The presumption is that those that score, puts the opponent in poor field position and thus by default, them in better field position.
I think HJim Harbaugh recognized this advantage as one reason to choose to kickoff instead of receive to start games.