Writers of Pro Football Prospectus 2008

10 Nov 2011

FO Mailbag: Time in the Lead

Peter M.: You guys are great at compiling and analyzing football stats. I was wondering if you wanted to take a stab at the most powerful stat of all: PERCENTAGE OF MINUTES IN THE LEAD. Athletes are like race horses -- some are front-runners, others are more comfortable when trailing. This statistic reveals which teams exert the most CONTROL over their games. The Pittsburgh Steelers led the league in "Percentage of Minutes in the Lead" the year they last won the Super Bowl. I'm not sure what compiling this statistic would entail, but I'm anxious to see what you come up with.

Sure. This doesn't end up on our drive stats page but we do track it as part of our tracking of drive and pace stats. Here's a table showing the average amount of time during each game that teams have been in the lead, tied, or behind this season, through Week 9:

Team Avg. Time
in Lead
Rk Avg. Time
Tied
Rk Avg. Time
Trailing
Rk
HOU 42:00 1 10:32 24 7:28 32
GB 41:55 2 7:08 31 10:57 31
CHI 33:19 3 7:51 29 18:51 24
NE 33:03 4 15:52 5 11:05 30
SF 32:19 5 11:47 16 15:55 28
BAL 32:16 6 13:04 11 14:40 29
PIT 32:04 7 8:59 27 18:57 23
PHI 30:24 8 9:31 26 20:06 21
MIN 29:58 9 12:06 15 17:57 26
DET 29:43 10 7:07 32 23:10 16
NO 29:11 11 8:04 28 22:45 18
ATL 28:53 12 11:20 18 19:47 22
BUF 26:28 13 16:56 3 16:37 27
OAK 26:00 14 11:03 20 22:57 17
DAL 25:32 15 16:19 4 18:10 25
TEN 23:59 16 12:25 13 23:36 15
Team Avg. Time
in Lead
Rk Avg. Time
Tied
Rk Avg. Time
Trailing
Rk
CIN 22:43 17 15:26 7 21:52 20
SD 22:18 18 12:23 14 25:20 12
MIA 22:05 19 13:31 10 24:24 14
CAR 20:50 20 14:33 9 24:37 13
NYJ 20:22 21 11:12 19 28:26 10
NYG 19:58 22 17:21 1 22:41 19
JAC 18:36 23 12:33 12 28:52 9
KC 18:28 24 7:50 30 33:43 7
ARI 17:40 25 15:29 6 26:51 11
WAS 14:45 26 15:11 8 30:05 8
DEN 14:07 27 10:57 21 34:57 6
IND 13:49 28 9:51 25 36:20 3
TB 12:45 29 11:36 17 35:39 4
CLE 12:25 30 10:52 22 36:43 2
STL 9:39 31 10:37 23 39:44 1
SEA 7:24 32 16:57 2 35:38 5

What do folks think? Particularly useful stat? I can't help but think of the memorable quote from Dan Duquette during one of those years where the Red Sox blew a big division lead, when he told a radio station that the Sox spent more days in first place than any other team. That's nice, but all 162 games count -- or in our case, all 60 minutes.

Posted by: Aaron Schatz on 10 Nov 2011

34 comments, Last at 14 Nov 2011, 4:38pm by steveNC

Comments

1
by Joseph :: Thu, 11/10/2011 - 3:33pm

I think it makes a good point of reference when looking for a team that might be better/worse in the 2nd half of the year, or in the next year--in other words, it's like looking at a team's estimated or Pythag wins. Interesting, but it pretty much tells you which teams have the best W-L records.

2
by Independent George :: Thu, 11/10/2011 - 3:33pm

Heh heh heh heh. It's not really appropriate anymore, but I love the fact that Minnesota's time in the lead is almost exactly 30 minutes...

3
by tally :: Thu, 11/10/2011 - 3:37pm

While it does have good correlation with W-L for obvious reasons, I think finding the teams whose Time in Lead do not match their W-L might indicate those teams who cannot hold onto leads or in the converse, teams who play better when they shift to a more pass-oriented catch-up offense.

4
by EsotericEnigma :: Thu, 11/10/2011 - 3:53pm

Of the top 15 in "Avg Time in Lead" only two have a losing record - Phi and Min. Both teams have had highly documented troubles in the second half, and the 4th quarter specifically. And there's the Giants way down there at 22 despite being 6-2 - because the Giants have won a lot of games coming from behind late.

I'm not sure then how useful the stat is. We already know the information it provides. Good teams often have leads because they're good and win games. Bad teams don't often have leads. The Eagles and Vikings blow games late, the Giants do the opposite. This just puts a lot of numbers to what we already know.

17
by tuluse :: Thu, 11/10/2011 - 6:33pm

Even if the stat doesn't tell us anything we didn't know it can still be useful. Just quantifying things we know makes it easier to compare.

20
by Revenge of the NURBS (not verified) :: Thu, 11/10/2011 - 7:08pm

I agree, more or less, that this particular table of stats doesn't tell us much we didn't already know. I'm not ready to say that the stat will always be redundant, but in this case, yes, it basically just tells us what we already know.

Being ahead correlates strongly with winning -- this is perhaps the most obvious observation in all of sports. It also confirms the narratives that teams like MIN have blown leads, and teams like TB and NYG have made late game comebacks.

It also strikes me that with only 8 games on the books, a game like Miami just experienced has the potential to skew the results dramatically. That one game probably moved their "time in the lead" results by 5 minutes in the positive direction.

22
by bravehoptoad :: Thu, 11/10/2011 - 8:28pm

Both NY teams really pop out.

5
by zenbitz :: Thu, 11/10/2011 - 3:58pm

The obvious weakness is that being down by 1-2 is treated the same as down by 14. Probably WPA is better for this. Time winning = WPA > 0.60, Time losing = WPA < 0.40, time "tied" between 0.4-0.6 (or other cutoffs).

21
by JasonG (not verified) :: Thu, 11/10/2011 - 7:24pm

Or weight it. 5 minutes leading by 1 = 300 Lead Seconds. 5 minutes leading by 7 = 2100 Lead Seconds. All tied seconds = 0 LS. Losing seconds are just negative. Sort by LS. How would that list differ from this list? How would it differ from a point differential list or won-lost list? How would it differ from the DVOA rankings? Would teams that lead big early then fade or struggle early but come back stand out somehow? Inquiring minds want to know.

29
by Jerry :: Fri, 11/11/2011 - 9:45am

If you care that much, you can go to the play-by-plays and derive it yourself.

33
by Anonymous_Marver (not verified) :: Mon, 11/14/2011 - 5:23am

I peed in Jerry's cheerios this morning. Sorry, guys.

6
by andrew :: Thu, 11/10/2011 - 3:59pm

Only six teams have trailed less than the Vikings. Yet here they are at 2-6.

7
by Rhombus (not verified) :: Thu, 11/10/2011 - 4:22pm

Giants lead the league in time tied. Considering their penchant for "gutting" their way through games, this seems appropriate.

8
by QQ (not verified) :: Thu, 11/10/2011 - 4:41pm

Still amazing GB is on a streak of 14 consecutive games without trailing in the 4th Quarter.

34
by steveNC (not verified) :: Mon, 11/14/2011 - 4:38pm

I bet they didn't lose many of those.

9
by krugerindustria... :: Thu, 11/10/2011 - 5:11pm

In order to fully test the premise, don't you also have to show the DVOA for teams while in the lead, tied and behind? Or perhaps that's all that's important. I'd also add the categories ahead/behind by >7pts.

Not sure what the value would be though, if a team plays great while behind, but terrible when ahead? and they meet a team that is the exact inverse, you would hope they would suck in the first half? I dunno, interesting thought though.

11
by Aaron Brooks Go... :: Thu, 11/10/2011 - 5:15pm

That basically describes a Vikings-Packers game. Neither team can hold a lead, the Packers just jump out earlier to larger ones.

10
by Aaron Brooks Go... :: Thu, 11/10/2011 - 5:14pm

Other than Minnesota, the NFCN doesn't waste much time opening the game with points to someone. Chicago, Detroit, and GB are 29th, 31st, and 32nd in time tied -- and there's only been two games played between the three!

12
by akn :: Thu, 11/10/2011 - 5:38pm

Very surprised to see the Bears #3 in time spent ahead.

My completely unfounded explanation is that the Bears manage to put 3 quarters together pretty well. It's the 4th where we generally drop off (not counting the Philly game), yet usually manage to hold on against most teams.

15
by Kal :: Thu, 11/10/2011 - 6:20pm

I'm not. Think about it - this is average, not median. So in the 3 games the Bears have lost they've been in them most of the game (Detroit was competitive to the half, GB was as well, NO was about the only one that got away later). Of the 5 other games they had the lead or were tied almost the entire game. That skews the results.

So realistically the Bears tend to stick around one way or another and then win. They don't do a lot of back-and-forth games either.

18
by tuluse :: Thu, 11/10/2011 - 6:35pm

Here's an alternate explanation, the Bears are good until Martz goes full Mad Mike Martz mode and orders Cutler to pass the ball 60 times in a row.

Also, the Bears are good and have won most of their games and winning means you were leading at least at some point.

13
by LionInAZ (not verified) :: Thu, 11/10/2011 - 6:00pm

I think this stat tells us pretty well who the winners and losers are, but beyond that, nothing much. Narrow this down to whether teams maintain or lose leads in the 2nd half and you might have a useful stat.

25
by RickD :: Thu, 11/10/2011 - 10:23pm

We might call that "the standings."

14
by TomTom (not verified) :: Thu, 11/10/2011 - 6:10pm

And THIS is why it's hard to be a Viking fan these days.It's painful to watch the Vikings, knowing they'll find a way to lose the game.

16
by kbukie :: Thu, 11/10/2011 - 6:23pm

So, it would appear the NFC North teams like having the lead. I wonder what the compiled statistics are for the division when not counting games against each other.

19
by QQ (not verified) :: Thu, 11/10/2011 - 6:44pm

GB not being tied for long is unsurprising considering they have allowed TDs on the opening drive like 4 times this year

30
by justanothersteve :: Fri, 11/11/2011 - 1:09pm

Probably because you know whatever the score is, one of the teams will be scoring again soon. Hard to stay tied when the Packers will likely either score or give up a score in the next 3 minutes.

23
by RickD :: Thu, 11/10/2011 - 10:01pm

Giants lead league in time spent in a tie game.

I guess that halftime 0-0 score against the Pats makes sense now.

24
by nat :: Thu, 11/10/2011 - 10:02pm

A fun stat if not a useful one. There's a place for fun stats. No, really. I mean that in a good way.

Dan Duquette. *shudder*

26
by ChrisZ (not verified) :: Thu, 11/10/2011 - 10:32pm

I think this proves that we're gonna have a GB-HOU superbowl. No one else comes close!

28
by Mr Shush :: Fri, 11/11/2011 - 6:00am

The latest DVOA playoff odds report does indeed have that as the most likely scenario at 8.26%, just ahead of GB-BAL at 7.63%.

27
by Tim W (not verified) :: Fri, 11/11/2011 - 4:23am

What are the stats for past years? Would be interesting to see how well the eventual champions/teams that made the playoffs did

31
by Oldcat (not verified) :: Sat, 11/12/2011 - 12:00am

The problem with this stat is that it conflates 'good" and "bad" results on the same directions on the axis.

A team that blows out other teams is in the lead a lot, but so is one that doesn't finish games and blows leads late. A 'plucky' team that comes from behind a lot will be behind or tied a lot, as will a hopelessly bad team.

A quick solution would be to separate the stats into two bins "In wins" and 'In losses" for each team.

32
by BSK :: Sat, 11/12/2011 - 9:50am

Another stat showing the Eagles are FAR better than their record.

And all of it means bubkis if they can't make the tourney.