A look at fourth-down decision making in 2014 highlights Sean Payton, Marc Trestman, and... wait, this can't be right... Jim Caldwell? Plus: Chip Kelly may be less aggressive than you think he is.
04 Dec 2003
by Aaron Schatz
Five games this week pit the first place team against the second place team in the same division. Boy, I'm sorry for those of you who are going to be stuck watching Detroit play San Diego. In the spirit of Football Outsiders, here's a look at different trends and splits for the teams involved in these five showdowns.
First, let's get the explanations of our innovative statistics out of the way for those who might be new to the website. VOA, which stands for Value Over Average, is the result of our detailed play-by-play database which compares every single play of the NFL season to the league average based on situation. It takes into account a number of variables, including things like down, distance to go, current score gap, location on field, and quarter. You'll find it explained further here. DVOA is VOA with an additional variable, the quality of opponents faced (which itself is sorted by variables like down and run vs. pass). Since VOA measures ability to score, a negative VOA indicates a better defense and worse offense, and a positive VOA indicates a better offense and worse defense. DPAR, or Defense-Adjusted Points Above Replacement, is a measure for individual players that combines quality above average with durability so that average players who are constantly involved in the offense (thus keeping the defense from covering other players) have value. You'll find it explained further here.
Unless otherwise mentioned, all stats below are for 2003. Individual stats are only through Week 12, team stats are through Week 13. You'll notice each section starts with a table splitting each team into rush and pass ratings for both offense and defense, color coded. Instead of putting one team on a line, I put one team's offense on top of the other team's defense and then vice versa so that you could see where teams matched up well or not so well.
If you like this article, and are a fan of one of these teams, please tell your fellow fans to come check this article out.
|IND OFF||43.3%||2||-3.6%||15||TEN OFF||46.5%||1||-18.9%||31||IND||-0.6%||23|
|TEN DEF||-2.7%||15||-6.2%||11||IND DEF||-11.8%||8||8.5%||30||TEN||0.2%||17|
Do you like passing? Well, then, this is the game for you. Ladies and gentlemen, the top two passing offenses in the NFL. The Colts would seem to get an advantage from having better pass defense, but then the Titans have better run defense while the Colts' run defense is abysmal. And the Titans have been playing better of late, until last week. Back in Week 2, the Colts embarassed the Titans 33-7, but that was in Indy and the Titans are a much better team now, while the Colts defense has faltered recently.
You can't talk about this game without talking about Steve McNair's astonishing ability to convert third downs. The Titans are the #1 team in the league in passing DVOA on third downs, +58%. They are the #31 team in the league in rushing DVOA on third downs, -37%. (The Patriots are last -- have I ever mentioned that they might consider drafting a running back next year?) The difference becomes even more pronounced on third-and-short. On third down with 1-4 yards to go, Tennessee's passing DVOA is +91%; rushing, they are about average at +4%. Heck, they are only that high because of a couple of nice runs on 3rd-and-4. Check the offensive line stats, and you'll see that the Titans are #29 in "power success," or conversion percentage when running on third or fourth down with 1-2 yards to go.
It's a good thing that McNair is so good on third down, since the Titans' rush game is dreadful. Will Indianapolis be able to counter McNair's third down prowess? Well, probably better than most teams -- that 8th ranked pass defense is 10th ranked on third downs, -29%.
As far as the Colts go, probably the most interesting trend is the Colts' record of playing better in the second half. I'm not sure if Colts fans will agree with me, but as far as I'm concerned the two most exciting Colts games of the year were the big comeback against Tampa Bay and the big comeback that didn't quite make it against New England. Indy's record of playing better in the second half exists on both offense and defense, and not only is it apparent this year, but last year as well (Q5 is overtime, by the way):
|Colts offense 2003||16.6%||27.4%||Colts defense 2003||6.9%||-10.7%|
|Colts offense 2002||-11.0%||11.5%||Colts defense 2002||18.0%||-1.0%|
By the way, for those of you who haven't been with us for more than a couple weeks, you'll notice that the television commentators are picking up on something I first wrote about in Week 10. Tennessee is not anywhere near being the best rush defense in the NFL this year. They have given up the fewest yards on the ground because nobody ever runs the ball against them, due to the fact that a) people think they can pick on the Tennessee pass defense and b) the opponents are spending a large part of nearly every game playing from behind anyway. Please, the Titans actually let Loser League MVP candidate Curtis Martin into the end zone for the first time since LeBron James was in high school. In fact -- I didn't even realize this until Al Michaels mentioned it on MNF -- Tennessee is threatening the all-time record for fewest rushing attempts against a defense. I have no idea who holds the record, if anyone feels like finding out please post it in the comments. Anyway, there have been only 243 rushes against the Titans this year. Only one other defense, Denver, is below 300. The Titans face 1.8 passes for every rush -- that ratio was actually over 2.0 until the Jets game.
All the stats, of course, don't consider the wild card in this game, and that is the condition of Steve McNair. We all know that McNair is tough he would take the field after having his legs amputated, but he didn't seem to be quite himself in the loss to the Jets. Who knows how he feels right now, and who knows how good Billy Volek is if McNair can't play.
My pick: Out of the five showdowns, in my opinion, this is the closest matchup. I am going to go against conventional wisdom and take Indianapolis. This is about as silly as the New York Times constantly trying to pick final scores, but I think a reasonable storyline for this game goes like this: Tennessee takes an early lead. and holds it through halftime. At some point in the third quarter, hits from Dwight Freeney begin to take their toll on the already injured McNair. The Titans can't keep the ball with a running game, and the Colts find their way back. Mike Vanderjagt field goal wins it. If I'm right about all this, I'm starting my own psychic hotline.
|CIN OFF||21.1%||7||0.2%||12||BAL OFF||-31.5%||31||-16.3%||27||CIN||-3.2%||31|
|BAL DEF||-27.6%||2||-20.2%||4||CIN DEF||12.3%||21||8.4%||29||BAL||5.7%||1|
Otherwise known as the "Who Would Have Thought This Was For A Division Title" Bowl. For all the hype about Cincinnati's turnaround this season, Baltimore's turnaround has been even stronger. They might not have been as bad as Cincinnati in the past few years, but they certainly were as bad as the Bungles at the beginning of this season, except for Week 2 against Cleveland. Cincy is the #10 team in DVOA since Week 8? Well, Baltimore is the #6 team in DVOA since Week 8. Cincinnati also doesn't rank as high according to our ratings because they have played an easier schedule than Baltimore. Except for the upset of the Chiefs, the Bengals surge has come against lower-rated teams like Houston, San Diego, and Pittsburgh, while the Ravens have played stronger teams like St. Louis, Seattle, and Miami.
The story in this game is the (man, is this weird to say) explosive Jon Kitna-led passing game of the Bengals against the Ravens, who this week became our #1 rated defense. The Bengal passing game has gotten better as the season has moved along, but so has the Raven pass defense. Remember the statistic about Tennessee being the #1 team in the league passing on third downs? Cincinnati is actually third, +45% DVOA (New Orleans is second, +46%). But the best pass defense on third downs is on the other side of the field is Baltimore, -90%.
One place where the Bengals offense and Ravens defense don't match is first half vs. second half. Cincinnati's offense declines in the second half, going from +20% in the first half to +1% in the second half and overtime. Baltimore's defense gets better, going from -12% DVOA in the first half to -36% DVOA in the second half and overtime.
While the Bengals offense has been great and getting better, the Bengals defense has been lousy and staying lousy. The Bengal defense is much worse than its reputation, or the reputation of head coach Marvin Lewis as a defensive expert. Then again, wasn't Brian Billick supposed to be an offensive guy before he came to Baltimore? Anyway, I would say that the Baltimore offense against the Cincinnati defense was a battle of the resistible force against the moveable object, but that was before the past two games. The Anthony Wright offensive explosion may be the most unexpected development of the season so far, and the Baltimore pass offense ranking above doesn't mean much if this strange phenomenon continues. But lost in the excitement is the fact that the Baltimore running game has disappeared. In fact, the Baltimore running game has been dropping all season long. Here's a graph of Jamal Lewis' DPAR for each game this season (including Week 13, I figured it especially for him, sorry the font in the graph looks a little goofy but you get the idea):
Eek. Although (and I didn't notice this until Al pointed it out in this week's Scramble for the Ball) Jamal Lewis' strength matches perfectly with Cincy's weakness. Baltimore rushing (mostly Lewis) is #2 in the league in adjusted line yards going up the middle; Cincinnati's defense is last in the league in the same statistic.
The hidden factor in this game, and the one that may make the difference, is the special teams. With Kansas City's special teams receding -- when is the last time you heard the name "Dante Hall?" -- Baltimore moved into the top spot this week. Cincinnati had the worst special teams in the league last year and only the replacement of Neil "Loser League Diety" Rackers with Shayne Graham has kept them from the bottom position in 2003. I determine special teams rankings by estimating how field position impacts the probability of scoring the next points (more explanation here), and here are my estimates of how many points, compared to average, the different aspects of special teams have been worth to the Bengals and Ravens this year:
Eek again, this time from the Cincinnati side. That's a lot of lost field position.
My pick: While the decline of Jamal Lewis is a concern, I think that the Baltimore pass defense trumps the Cincinnati pass offense, especially since the Ravens should easily win the field position game. I'll take Baltimore.
|DAL OFF||-6.0%||23||-9.7%||23||PHI OFF||-3.8%||18||14.1%||3||DAL||-0.2%||20|
|PHI DEF||12.4%||22||-2.0%||19||DAL DEF||-9.0%||10||-22.2%||2||PHI||1.7%||7|
Looks close, don't it? 9-3 against 8-4. Two teams that match strengths and weaknesses. Dallas' below-average offense matches up against Philly's below-average defense. Philadelphia's very good rush offense and mediocre passing game match up with Dallas' very good rush defense and mildly above-average passing defense. Back in Week 6, Dallas won a close game 23-21.
Except that what this game really matches up are two teams going in completely opposite directions. Remember how in this week's VOA commentary I split the season into two halves, Weeks 1-7 and Weeks 8-13, to show how certain teams had declined while others had improved? In Weeks 1-7, Dallas was the #4 team in the league while Philadelphia was #28. In Weeks 8-13, Philadelphia was the #4 team in the league while Dallas was #27. That's a switcheroo worthy of Judge Reinhold and Fred Savage in Vice Versa. Let me show you what's happened, broken down into a bit more detail. Here's the table above, but only giving the data (and ranks) for Weeks 8-13:
|DAL OFF 8+||-21.4%||26||-9.6%||24||PHI OFF 8+||50.6%||2||17.0%||2||DAL 8+||-1.6%||25|
|PHI DEF 8+||8.2%||19||5.0%||25||DAL DEF 8+||12.9%||21||-22.2%||10||PHI 8+||-1.0%||23|
Well, when Dallas has the ball things stay pretty much the same. But when Philadelphia has the ball, it ain't gonna be a happy time in the Lone Star State. The Dallas defense, which looked so great at first against all those terrible teams, has come back down to earth even after taking into account the harder opponents. At the same time, Donovan McNabb has put himself back together, and the Philadelphia passing offense is clicking on all cylinders, which is pretty incredible since I believe McNabb is still throwing to a receiving corps consisting of two members of Allen Iverson's posse, the corpse of Uday Hussein, and that guy from Laugh-In who said "Verrrry Iiiiiiiiiiiiiinnnnteresting."
Actually, looking closer at the numbers, that's not really far off. Does Allen have friends who can play wide receiver? McNabb is doing this almost entirely with passes to running backs. I haven't had a chance to run receiving numbers for Week 13 yet, but here are the DPAR (defense-adjusted points over replacement) numbers for Philadelphia receivers in Weeks 8-12 (min. four passes):
Believe me, Pinkston and Thrash were even worse during the first six games. The other thing you should know about the Philly offense is that their high rushing rank comes from McNabb, not the running backs. Even when he was struggling early, McNabb was the best running QB in the league, and for the season he's far beyond everyone else. According to our stats, the difference between McNabb (10.3 DPAR) and the second most-valuable running QB this year, Jeff Garcia (6.1 DPAR), is larger than the difference between Garcia and the #13 rushing QB of the year, David Carr (2.1 DPAR).
My pick: Philadelphia is the most likely winner of any of the ten teams involved in these games. Which means, of course, that the Cowboys will probably win just to spite me. I'll pick Philly anyway.
|MIA OFF||-16.0%||26||-12.0%||25||NWE OFF||13.4%||9||-18.1%||29||MIA||-0.8%||24|
|NWE DEF||-19.9%||5||-6.1%||12||MIA DEF||-22.1%||4||-16.8%||7||NWE||1.4%||9|
Apparently, the line makers have forgotten that Miami always, always chokes in the cold in December. Let me tell you, it is freezing out there today, and have I mentioned that we're expecting snow on Sunday? If you think that somehow favors the powerful Miami running game, let me point out that a) the Miami running game has been horrible this year and b) I have this neat DVD on my shelf called Super Bowl XXXVI Champions: Official Enhanced DVD Edition. That's right, people, this is no ordinary championship DVD. This DVD is enhanced with extra Lonnie Paxton snow angels. Maybe I'll stick it in at halftime of the 1 p.m. games and relive the Snow Bowl as inspiration.
One of the images that people have of the Patriots is that they are a lucky team, winning games they shouldn't win. Actually, the only two games this year where the Patriots have a negative DVOA -- meaning that our stats say they played below average and deserved to lose -- are the two games where they actually lost, Week 1 to Buffalo and Week 4 to Washington. The game against Miami was pretty much tied in DVOA, as befits a game that ended so closely that it was determined by missed field goals (and almost by massive official errors like blowing the coin toss and an obvious step out of bounds by Derrius Thompson).
Both Miami and New England suffer in our ratings because of huge fluke losses in Week 1, to Houston and Buffalo respectively. What happens to the rankings of the Dolphins and Patriots if you take that game away from each team. I removed Week 1 from the DVOA tables -- and changed the weight of Week 1 in the weighted DVOA formula to zero -- and the results look like this:
|NWE season to date||-1.5%||16||-14.1%||7||14.0%||8||18.0%||6|
|NWE Weeks 2-13||4.0%||12||-16.4%||6||22.1%||5||34.4%||1|
|MIA season to date||-13.9%||27||-19.7%||4||5.1%||12||7.3%||10|
|MIA Weeks 2-13||-10.2%||25||-23.9%||2||12.7%||10||26.6%||6|
I didn't list special teams here because Week 1 doesn't change much, but it is added into total DVOA like usual. You can see where each team improves when you don't consider that first week, but the most shocking number is that "1" for the Patriots under WEIGHTED DVOA. Now, that's not quite fair to the other Super Bowl contenders -- they didn't have their worst games in Week 1 -- but it is a pretty good indication that the Patriots are for real and that first game (you know, the one which had writers sending Buffalo to the Super Bowl and Bill Belichick to the unemployment lines) was a massive, massive fluke. In case you are curious, the teams between #1 New England and #6 Miami when you take WEIGHTED DVOA and zero out Week 1 are Kansas City, Tennessee, Baltimore, and then St. Louis.
Yes, as we all know by now, Jay Fiedler is better than Brian Griese. Our stats list Fielder as being worth 4.0 DPAR while Griese was an egregious -17.5 DPAR. By the way, that's only through Week 12 -- after last week, considering how well the Dallas defense had been prior to Thanksgiving, Fiedler is probably worth about a zillion points. So, do the Dolphins pass with this quarterback. Um, no.
Remember when I said that the Titans face 1.8 passes for every rush? It turns out the Patriot defense is second in that statistic, facing 1.6 passes for every rush. Their run defense, which was awful last year, has been pretty good this year, and they now have Ted Washington back. On the other side, Miami's offense is one of only three that have run more rushing plays than passing plays (the other two are Baltimore and Carolina). Miami has done this despite the fact that the rush has been completely unsuccessful for them. It is very likely that Dave Wannstedt will decide "snow = run," and if that's the Miami game plan they are going to be in trouble. I also expect Wannstedt to come out in serious K2 Survival Gear, while Bill Belichick wears Bermuda shorts and a tank top.
The other fun part of snow? Special teams become more important, and both of these teams have good kickers going through inexplicable slumps. There are going to be a lot of field goal chances, especially for Adam Vinateiri. Miami's defense is by far the best in the league in the red zone this year. They allow -92% DVOA in the red zone, but -12% DVOA over the other 80 yards of the field.
The Patriots also had a horrible punter. Had, because Ken Walter got cut this week. Hallelujah. According to my calculations, Walter's punts lost the Patriots 22.4 points worth of field position this year. No other team was above 16 points. However, the Pats are only listed as losing 6.6 points worth of field position on punts, because their punt coverage teams did a good job of preventing returns on all those crappy punts. Oh, and did I mention that the Miami punt return unit scores as the sixth-worst in the league, worth -10.1 points in field position? Welcome to the NFL, Brooks Barnard!
One more fun note: Despite the second half meltdown against the Colts, the Patriots are a second half team -- and the Dolphins are a first half team. If Miami is up at halftime, fellow Pats fans, panic not.
|NWE offense||-10.8%||5.9%||MIA offense||-6.4%||-20.8%|
|MIA defense||-29.9%||-9.5%||NWE defense||-1.2%||-27.2%|
My pick: New England. A snowy night game against a warm-weather team with a left tackle made of tissue paper? Please.
|KAN OFF||31.3%||3||13.2%||4||DEN OFF||-5.1%||22||2.3%||9||KAN||4.8%||2|
|DEN DEF||-13.3%||7||-21.2%||3||KAN DEF||-17.0%||6||7.2%||27||DEN||0.7%||12|
Technically, this qualifies as a first place vs. second place division showdown, but it won't really feel like one. The only way that Denver can win the AFC West is if they run the table and Kansas City goes on a four game losing streak. Carol Moseley Braun has a better chance of winning the Democratic nomination. Heck, Tom DeLay has a better chance of winning the Democratic nomination. Nonetheless, this is still a very important game for the AFC playoff picture. Baltimore, Cincinnati, and Miami fans will be rooting for Kansas City, hoping to knock Denver further down in the race for the second wild card spot. Tennessee, Indianapolis, and New England fans, especially the ones watching at my house, will be rooting for Denver, hoping to knock the Chiefs out of the top spot and the home-field advantage throughout the playoffs.
Looking at the table above, Kansas City's Achilles heel seems pretty clear: rush defense. And, wouldn't you know it, Denver has one of the league's top running backs. Look a little closer, though, and it isn't quite the slam dunk for Clinton Portis that you might think. Very little of that success against the Chiefs rush defense is obtained when the game is still close and early. In the first quarter, with the game still close, Kansas City has a very good rush defense. The problem comes when the Chiefs have the lead. Our numbers adjust for the fact that teams often give up more yards per carry when they have a second-half lead (since they're expecting the offense to pass), but still KC is particularly bad at defending the rush when the other offense is attempting a comeback. Now, Denver hasn't played a lot of games where they fell way behind, so we don't have much data on Portis in that situation. In close games, however, Portis this year has faltered in the third quarter before righting himself in the fourth quarter. Here are the numbers so you can judge for yourself just how much success Portis is likely to have. Because the Chiefs and Broncos have both played from behind so little this year, I've added together Q3 and Q4 for those numbers:
9 or more
9 or more
9 or more
9 or more
Where else might Denver find a weak point to attack the Chiefs? Get the Chiefs to third down. The Kansas City offense has achieved most of its success this year on first and second downs, while the Denver defense is extremely stingy on third downs:
|Down||1st||2nd||3rd||4th (not punt)|
|KAN offense DVOA||+35%||+29%||-18%||+74%|
|DEN defense DVOA||-9%||-9%||-52%||+53%|
On the other side of the ball -- this will sound very odd -- look for Denver on second downs:
|Down||1st||2nd||3rd||4th (not punt)|
|DEN offense DVOA||-11%||+25%||-17%||-76%|
|KAN defense DVOA||-6%||+23%||-41%||-121%|
So, those are some things to look for from the statistics, but there are always those dreaded "intangibles." Denver has a huge home-field advantage (as do Denver teams in every sport) and they've beaten Kansas City the last two years at home. AFC West interdivision games always seem to be close no matter how good the teams are otherwise. Kansas City is the better team, but Denver is hungrier -- the Chiefs are playing for a first-round bye, but the Broncos are trying to get into the playoffs altogether.
My pick: Kansas City, especially if I'm picking against the spread, but I can easily see Denver -- on this one Sunday -- being the better team.