TCU has played much better in the second half of games this year. What other schools have seen dramatic shifts of play after halftime?
08 Jan 2004
by Aaron Schatz
That's the number to remember, as we enter the second round of what everybody now calls (in totally slavish Tuna-worship) "THE TOURNAMENT." Nine, as in nine wins by road teams in the second round since the playoffs went to six teams per conference in 1990. 9 wins, 43 losses. That's less than a win a year. Five times, all four home teams have won. Once, in 1995, two road teams actually won in the second round (9-7 Indy over 13-3 KC, perhaps an omen, and 11-5 Green Bay over 11-5 San Francisco). Seven times, one solitary underdog has triumphed on the second weekend of the playoffs.
So picking these games is somewhat silly. Of course, every home team should be favored. Of course, if you pick all home teams (straight up, forget the spread), you will likely go 3-1. One of these four underdogs is probably going to win, and probably only one. All four could be that team. Three of these games, if you toss out the advantage of the home field and the bye, are very evenly matched, and the fourth (which is the first chronologically) is no pushover either. For extra fun, the two AFC games match the top four teams according to our season DVOA (Defense-adjusted Value Over Average) ratings. So instead of picking these games, I'm going to begin each section with a short summation of why the underdog can win this game.
To digress, I'm watching a repeat of L.A. Dragnet while I'm writing this, and they are questioning a suspect. If you get arrested by the police, and they ask what you were doing on Sunday, and you say "watching football," and then they ask you "which game," the correct response is not "Yo, I didn't shoot nobody on Sunday." You have to know who played Sunday, smart boy. Football knowledge is important.
Anyway, instead of going through an explanation of our stats for the 100th time, I think I'm just going to link here to last week's playoff preview, and you can read the stat explanation that starts off that column, of just go to the good ol' methods column that explains most everything. Three notes:
If you're a fan of one of these eight teams and like this article, please let your fellow fans know about us by posting about it to your favorite message board. Last week we had a lot of visitors from Broncos and Ravens message boards. Hopefully they aren't so depressed that they didn't come back again this week...
WHY CAROLINA CAN WIN: St. Louis is one of the least consistent teams in the NFL, while the Panthers were this year's most consistent team from start to finish. Plus, you never know when Marc Bulger will go pick-iriffic.
As I have written a few times this year, the conventional wisdom about the Rams is lagging behind the reality. The onetime offensive juggernaut has become an average offensive team with a stifling defense, rated according to our statistics as the second-best in the league behind Baltimore.
That's the bad news for Carolina. Of all this weekend's games, this is the only one that is clearly a mismatch. The Carolina defense is basically a mirror image of the Rams offense -- 17th in passing, 12th in rushing. Unfortunately for the Panthers, the Carolina offense is basically a mirror image of the Oakland defense, and last time I checked the Raiders were not in this game.
The conventional wisdom that says the Rams' passing offense works much better in the Edward Whatever They Call It Dome seems pretty accurate, at least according to our stats. The Rams had the third-best offensive home-field advantage in the NFL (one of the two teams higher than them is also playing this weekend, so read on to find out who that is). At home, their offense scored +16.7% DVOA, but on the road they were almost opposite, with -15.9% DVOA. At home, the Rams' offense plays like Tennessee; on the Road, more like Buffalo or Miami.
The Rams biggest problem this year was inconsistency. You can see the chart of their performance is all over the place, and they were the #7 least consistent team in the NFL according to our VARIANCE statistic. Bulger's performance, in particular, bounced between glorious and grotesque. His 27 turnovers, including fumbles and interceptions, were four more than any other quarterback in the league.
Unfortunately for the Panthers, the Rams' inconsistency is in many ways related to this home field advantage. If you look at this chart of the Rams' game-by-game performance, nearly all of the peaks are home games, and all the valleys are road games. The Rams' worst home game performance was their 27-24 overtime win over San Francisco in Week 2 (-17.9%). Their best road performance -- by a huge, huge margin -- was their 33-21, nowhere near as close as it looked win over Pittsburgh in Week 8 (game seven on this chart). Except for those two games, every single Rams home game was a better total performance than every single Rams road game.
Bulger was a big part of this. He had a negative DVOA rating in five of seven road games but only three of eight home games (and one of those was just -0.8%). His interceptions, however, weren't that different: 12 in seven road games, 10 in eight home games.
Because of all the interceptions and sacks allowed, the Rams had an average pass offense despite the great performances of their wide receivers. Though Torry Holt was the best in the league this year by our ratings, you may want to watch out for Dane Looker. He has an overall negative DVOA, but he's positive on third downs (+16% DVOA on 36 passes) and when in opponent territory (+17% DVOA on 30 passes). Actually, Looker had a higher DVOA in both these situations than Isaac Bruce, and was a target just as often.
Last week I noted that the Panthers liked to throw long, ranking fifth in the league in complete passes of 30 or more yards. That was a great strategy against Dallas, which gave up tons of such passes. St. Louis, however, is one of the best defenses in the league at not giving up such passes. Not that Carolina won't try, of course. (By the way, if you still think of St. Louis as the speedy high fliers, you might be surprised that only 4.8% of St. Louis complete passes were for 30 yards or more, below league average.)
The Panthers should enjoy a nice advantage in field position from kickoffs and punts. As I noted last week, Todd Sauerbrun's punting rating looks much worse than he actually played because of the two blocks by the Giants in the final week of the season. Before that, the Panthers had one of the better punting units in the league. That's a good thing, because punt returns are one of St. Louis' weaknesses, and Carolina isn't going to be converting a lot of third downs. The Panther offense has more success on earlier downs, while the Rams were the third best defense in the league on third down. They were even better preventing successful plays on second down, not only the best DVOA of any defense in the NFL on second down but twice as good as the second-best defense (Baltimore).
While Carolina likes to go for that long bomb, they might be better off concentrating on seven and eight yard passes on first down that can put Stephen Davis in position to convert and get new first down after new first down. Carolina's best hope of winning is to capitalize on an early turnover or two, get an early lead, and then run the clock out by pounding away with Davis. Not only will this keep the St. Louis offense off that home field that gives them such an advantage, it also plays to Carolina's strength. For example, St. Louis is the third-best defense in the NFL at preventing third down success (-41.3% DVOA), while Carolina is one of the worst offenses in the league on third down (-21.6% DVOA). But that third down offense gets a lot better if you give the ball to Davis in a short-yardage situation.
That horrible 3rd-and-short passing number for Carolina is due to 20 passes, of which three were intercepted and only six went for first downs or touchdowns.
One last note: the Panther running game is a lot better than the stats say if they keep DeShaun Foster away from the field. He weighed down the team numbers by ranking last out of the top 53 most-used running backs in DVOA, far below his teammate Davis.
WHY TENNESSEE CAN WIN: They are just as good as the Patriots, their field goal kicker isn't going through a serious slump that could be aggravated by extremely cold weather, and Steve McNair is like that knight in Holy Grail who keeps getting his limbs cut off and responds, "It's just a flesh wound!" Plus, have you ever seen a man with a smoother head than Eddie George? Seriously, that's gotta be worth some points.
Look at that graph of the Patriots' week-to-week performance this year. That's a ridiculous graph. Look at how little the dots move up and down compared to the other seven graphs on this page. Notice that after the fourth game, not a single game is below average. That graph right there explains why the New England Patriots are the favorite to win the 2004 Super Bowl.
Much has been made of the Patriots' streak at home, not giving up a touchdown until Jacksonville got one in garbage time, three shutouts at home this year, fewer points in the last six games (22) than Tennessee scored on them in their first game (30). DVOA, however, says that conventional wisdom is wrong about why the Pats went 8-0 in Foxboro this year. The team's defense didn't play much better at home than it did on the road -- defensive DVOA at home was -21.7%, defensive DVOA on the road was -17.8%. No, offense was the place where Foxboro seemed to make a difference for the Flying Elvii, as they had the second-best offensive home-field advantage in the league behind Chicago. At home, their offense scored +18.1% DVOA, but on the road they were almost opposite, with -16.9% DVOA. At home, the Pats' offense plays like Minnesota or Seattle; on the Road, more like Buffalo or Detroit. Gee, that sentence sounded awful familiar. Even in the six Pats road wins, ignoring the Buffalo debacle and the Washington turnover-fest, the Pats offense totaled a -7.9% DVOA.
The Patriots-Titans game is going to be interesting, because the Titans and Patriots will have to decide whether to play to the weather or to team strengths. It is going to be cold, cold, bitterly cold, with the overnight low currently forecast at five degrees. That's five, a number so low that proper journalistic etiquette demands that I type it as a word instead of digits. That can't be good for the numerous Titans injuries, right?
The general consensus is that cold weather is rushing time, except that neither of these teams is good on the ground and both defenses seem to scare opposing coaches out of using the running game. According to DVOA, the Pats and Titans are ranked 25th and 26th in rushing offense. Based on our offensive line stats, they are 23rd and 29th. Last year, at least Eddie George was very good in third-and-short situations; this year, only Cleveland ranked lower than Tennessee in "power success." The Titans also were last in the league in getting "stuffed," meaning that on 31% of running back plays they were kept to zero yards on first down and less than one-quarter of needed yards on other downs.
No, I'm not sure how Tennessee ran for 165 yards on Baltimore, the #1 rush defense according to our stats. But that was more yards than the Titans got on the ground in any other game this year. That performance was no more indicative of the Titan ground game than the 161-yard fluke that the Pats busted out on the Titans during their first meeting in Week 5, which was the Patriots' best rushing performance all year.
That was also the year's worst performance by the vaunted Titans run defense. As I've mentioned numerous times, I believe that the Titans run defense is somewhat overrated because of the fact that nobody ever runs against them. The Titans faced only 333 rushing plays this year, and only 36.4% of the plays against them were runs. And who is right behind the Titans, with the second lowest ratio of passes to rushes faced? New England, against whom only 37.4% of plays were runs. Perhaps nobody runs against them because they are that good -- something made Baltimore sit Jamal Lewis -- but nobody runs against the Patriots either. When teams do run against the Titans and Patriots, the Patriots are better at preventing success.
I've been asked to break down the Titan run defense to show why our stats have them middle-of-the-pack, or just slightly above, instead of number one. Skip down if you don't care. On first down, the Titans allow an average of 4.0 yards per carry. That's better than league average, which is 4.3 yards per carry. Their opponent, New England, allows 3.8 yards per carry. Adjust for variables like opponent, field location, and score situation, and the Titans have a -3% DVOA on first down which ranks #23 in the league (pass plays have a better rate of first down success than runs, which is why almost every team has a better rush DVOA on first down than second or third). The Pats have a -13% DVOA on first down and rank #17 in the league. The best run defense on first down was Dallas, -40%. Of course, the Titans faced fewer first down rushes than any other team in the league.
On second down, the Titans run defense fares much better. They allow an average of 4.1 yards with 7.6 to go. The league average is 4.3 yards with 7.1 to go. The Pats allow an average of only 3.3 yards, but with only 6.5 average to go, so the Titans and Pats both have -3% DVOA on second down which ties for #8 in the NFL. The best run defense on second down was St. Louis, -32%. Of course, the Titans faced fewer second down rushes than any other team in the league.
On third down, the Titans allow an average of 3.4 yards rushing with 4.7 yards to go. The league average is 4.5 yards rushing with 4.9 to go. The Pats allow 4.0 yards with an average of 4.9 to go. The Titans have -5% DVOA on third down, the Pats -4% DVOA on third down, and they rank #9 and #10 in the NFL. The best run defense on third down was Denver, -21%. The Titans actually did not face fewer third down rushes than any other team in the league, but they had the fourth-fewest.
Part of the reason our stats think the Titan run defense is overrated is that they faced a very easy schedule. The poor rushing offenses Tennessee faced included Pittsburgh (#32 in our rankings), New Orleans (#29), New England (#25), Houston twice (#24), Carolina (#22), Miami (#21), and Oakland (#20). Yes, Deuce McAllister and Ricky Williams were that low, on a play-by-play basis; Carolina is lower than you might think because DeShaun Foster was lousy. Continuing with Tennessee's schedule, you have the middle of the pack teams, as the Titans played Jacksonville twice (#15), Indy twice (#16), Tampa (#17), and Buffalo (#18) That means the best rushing offenses the Titans faced all year long were the Jets (#11) and Atlanta (#5). Well, gee, no wonder they looked so good.
OK, enough Titans run defense talk. Let's talk Pats defense instead. Bringing back memories of the bend-don't-break Pats defense of the 2001 Super Bowl run, the Patriots allow less and less success as the offense approaches the goal line, until you get to the red zone where they were the second-best defense in football in 2003 (behind Miami). On the other hand, Tennessee's offense sputters a bit when it gets into the red zone, possibly because their high-powered air attack doesn't have as much room to maneuver:
|DVOA by location||TEN 1-20||TEN 21-39||Between 40s||NWE 21-39||TEN in Red Zone|
The fact that the Pats had the best red zone defense of the remaining eight teams brings up an important point. You may be aware that near the end of the year we introduced a statistic called the "FOREST INDEX" that combined various DVOA ratings to give an estimate of the number of wins a team should have (normalized by schedule strength). Although the Pats finished fourth in total DVOA, in a virtual tie with Seattle and St. Louis, they were the top team in estimated wins with 12.3. The reason is that the Pats were particularly good in situations that help you win more games than your points or yards would otherwise indicate. Consistency is part of the formula, and the Pats were the league's most consistent team over the last 13 weeks. Having a great red zone defense is huge. So is having a better DVOA in the red zone (+12%) than on the rest of the field (-2%). And the final part of the formula is second half defense when the game is within a touchdown either way. At -29% DVOA in these situations, the Patriots were fourth in the league and best of the eight teams remaining.
Of course, the Titans aren't slouches in these late and close situations, especially Mr. McNair. As I remarked last week in discussing the Titans-Ravens game, "Tennessee's pass offense only got better at the end of the game. They were top of the league with +64.4% DVOA passing in the fourth quarter, almost twice as good as the next best team (Kansas City). They also led the league with +65.6% DVOA passing in the second half when the game was within a touchdown either way." And don't forget third down, where the Titans were fourth in the league with a great +25.2% DVOA, including +42.1% passing. All these stats get even higher when you don't consider the games that Steve McNair did not play.
I know Bill Belichick spends hours studying film and learning the Titan tendencies. But as viewers, don't think you have any clue who the Titans are throwing to in the end zone. For the second straight year, the Titans spread around their red zone passes more than any other team in the NFL. Five different Titan receivers -- Bennett, Calico, Kinney, Mason, McCareins -- had between 10 and 12 red zone passes. Of these five receivers, McCareins was the best (66 yards, +84% DVOA) and Mason, surprisingly, the worst (29 yards, -4% DVOA) -- although he had the same number of touchdowns on red zone passes as McCareins, four.
By the way, if you are wondering why Steve McNair is listed so low on our QB rushing ratings, the answer is "three lost fumbles." Discard those fumbles, and he hasn't been a great rushing quarterback this year, but he's been closer to average. Passing, of course, he's been unreal. But can the warrior overcome injury, cold, a rested opponent, an angry crowd, and the Coach of the Year? I'm a Pats fan, so what do you think my answer is? My one worry is Vinatieri; in a cold weather game that is likely going to have a number of field goal attempts, a shaky kicker enjoying his worst season is about as trustworthy as Pete Rose in a room full of bookies.
WHY INDIANAPOLIS CAN WIN: Kansas City's defense has plummeted faster than XFL ratings, and Peyton Manning is undefeated in the playoffs.*
Do you like offense? This is your game. The top two offenses in the NFL go at it. Both teams are ok playing the pass and terrible playing the run.
I'm not sure what to say about this game. I tried breaking down nearly every split. They all basically come out the same. Scoring, scoring, scoring, scoring, and more scoring. These teams are loaded with great offensive players who score just as highly in our statistics as they do in the standard NFL numbers.
Are there any places to look for actual defense here? Well, as I noted before the Denver-Indy game, the Colt offense goes from super on first and second down (+25.5% DVOA) to merely average on third downs (+5.9% DVOA),. Of course, who cares what you are doing on third downs when you are launching 87-yard bombs to Brandon Stokley on first down.
The problem for the Colts is that Kansas City is actually pretty keen on third downs, with a -24.0% DVOA that matches the third down defense of (gasp!) the Patriots. The problem for the Chiefs is that they allow +3.6% DVOA on first down and +26.0% DVOA on second down.
OK, I know what you are saying. "Gee, that looks like in total the Chiefs defense isn't so bad. I thought they were worse than that guy singing 'Sweet Home Alabama' on FOX's endlessly run commercials for American Idol 3." Well that's where the Kansas City's inconsistency enters the scene, along with their gradual fade. After the first three weeks of the season, the Chiefs had the fourth best defense according to the DVOA system, -28.5%. Their combined defense in Weeks 4-17, however, ranked as the fifth-worst in the league, allowing offenses a +11.3% DVOA. And that includes the Week 8 Sunday night pasting of Buffalo, which according to our stats was the single best total performance by any team in one game this year: +61% DVOA offense, -89% DVOA defense.
(Yeah, a few weeks ago, I said Denver's win over San Diego in Week 11 was the best game of the year. My goof, that one was actually number two.)
Anyway, take the first three weeks of the season out of the equation and all of a sudden those respectable defensive numbers become a little less respectable. Weeks 4-17, Kansas City actually got better on first down (-3.6% DVOA) but was really bad on second down (+36.4% DVOA) and below average on third down (+9.3% DVOA).
Another warning sign for Kansas City -- remember from the Patriots' comment, the importance of second half defense in close games? The Chiefs had the worst second half defense in close games of the eight remaining teams, and the third-worst in the league, allowing +20% DVOA late and close. Only Arizona and Washington were worse. Indy's defensive DVOA in these situations? Above average at -12%.
On the other hand, the Colts' defense isn't going to challenge the 1985 Bears any time soon either. The pass defense is reasonable, but the rush defense is abhorrent, and somehow I don't think the strategy of "hope the opposing star running back becomes irrelevant because his offensive line is getting penalized every ten seconds" is going to work two straight weeks.
If trends through Week 12 (the last time I had a chance to update directional rushing statistics) hold true, expect a lot of runs left by both teams. Indianapolis and Kansas City rated as the top two teams in the league running to the left by our adjusted line yards stat. That same side is the Colts' biggest rush weakness, as they rank last in the league. I guess Dwight Freeney is strictly a pass guy. The Chiefs shouldn't feel much better, they rank 30th in the league stopping runs left.
For all the talk about the poor run defenses, Kansas City pretty clearly has the better run offense. Of particular interest is that Indianapolis is poor both short and long. They ranked #30 this year in "power success" (thank you, Willie McGinest) and also ranked #30 in breaking runs over ten yards. Of course, no team in the league gave up more runs over ten yards than Kansas City did, so we'll have to see if Edge can break one (and if so, if he remembers what to do when he finally gets downfield).
Strangest statistic of the year: Priest Holmes, one of the best receiving backs in football, not have a single receiving touchdown this year. That's not the strangest statistic, though. The strangest statistic is that Holmes was not the intended receiver of a single Kansas City red zone pass in 2003. In the red zone, it's all about Tony Gonzalez, target of more red zone passes than any other tight end and a man with a +96% red zone DVOA.
Oh, and special teams. Kansas City had the best in the league this year. They have a colossal advantage when Indy is forced to punt, and a pretty big one on both kickoffs and kickoff returns. Indy will play America's favorite new game show sensation, "kick away from Dante Hall." On the other hand, Mike Vanderjagt is riding the longest string of hitting the target since the Kingpin hired Bullseye to take out Daredevil.
This game is going to be a lot of fun to watch. I hope it stays close all the way. I was one of the people who thought the Chiefs might be "one and done," but after looking closer at the matchup of these two teams I think the Colts -- as great as they are -- are going to have a harder time than many people expect overcoming that bye week advantage.
*(Note: Only includes games in which the opposing defense forgets to touch the receivers when they are on the ground.)
WHY GREEN BAY CAN WIN: The Eagles can't stop the run, and the Packers have one of the league's top backs. The second-best offensive player on the Eagles is out for the playoffs. Brett Favre's thumb is better now than it was when the Pack lost to the Eagles 17-14 on Monday Night Football, and whatever the heck is going on with his emotional momentum is still going on.
Well, in case you can't tell from that long write-up, or the fact that I took Ahman Green with my first pick in our Football Outsiders playoff fantasy draft, I agree with the conventional wisdom that says this is the most likely upset of the weekend. Not that Green Bay should be favored, but they are probably an even money bet to win this game.
For all the Favre talk, this game is likely going to depend on the run game. You know that Indianapolis has one of the worst rush defenses in the NFL this year. You know that Kansas City has one of the worst rush defenses in the NFL this year. What you may not know is that Philadelphia also has one of the worst run defenses in the league this year, ranked right behind those two. The Eagles D may spend much of the day cursing and whining, "Aw, man! That guy ran over us again!"
In our adjusted line yards stat, Green Bay and Philadelphia have the top two run-blocking offensive lines in the league. (Green Bay's DVOA is a bit lower due to fumbles.) The difference is that while the Green Bay's run defense is middle of the pack, Philly's is poor -- and the Eagles' best rusher, Brian Westbrook, is out with injury. Check our running back DVOA stats, and you'll find that Westbrook ranks second in DVOA, though he's listed a bit lower in Points Above Replacement because of fewer attempts. He's also fourth in receiving DVOA. But don't worry, Eagles fans, Duce Staley to the rescue! Staley also had a very good season, ranking #8 in rushing DVOA and #3 in receiving DVOA. He could be due for a very nice game assuming he can overcome his natural inclination to hold out for more money.
Another place Westbrook will be missed is special teams. That #9 Philly punt return ranking is almost all Westbrook, and that great #2 Philly kick return ranking is about half Westbrook. His injury leaves James Thrash with the responsibility of taking advantage of the poor Green Bay kickoff coverage crew.
Back to the running game -- when the Pack runs, you can't use the directional stats to see which direction Philly should overstack, because Green Bay is one of the league's best offensive lines running in any direction. On the flipside, according to the directional stats (through Week 12 only, alas), Philadelphia's strength running up the middle (third in the league) is countered by the fact that the Packer rush defense is better up the middle (ranked #5) than to the sides (#15 left, #24 right).
Another important thing to watch for -- not like it isn't always, but particularly here -- is which team can convert more third downs. Both teams have been more successful on first and second down than on third down on offense, and the opposite on defense. This trend will be stronger when Green Bay has the ball, as Philly's offensive decline on third down isn't really that bad.
|Packers have ball||1st Down||2nd Down||3rd Down||Eagles have ball||1st Down||2nd Down||3rd Down|
|GNB OFF||13.9%||6.7%||-16.6%||PHI OFF||7.1%||29.6%||-2.5%|
|PHI DEF||10.6%||19.5%||-25.3%||GNB DEF||-1.3%||-8.9%||-21.8%|
The exception is when Green Bay gets it third-and-short, as they are ranked #2 in our offensive line rankings for "power situations," third and fourth down with one or two yards to go.
Another interesting stat: Both teams are better in the red zone than on the rest of the field, on both offense and defense. Green Bay's offense improves by 38% DVOA in the red zone, while Philadelphia's offense improves by 26% DVOA. But Green Bay's defensive rating drops by 26% DVOA in the red zone, and Philly's drops 48% DVOA (with a lower number meaning fewer points scored).
Here's another interesting stat. I've talked a lot about Philly's dependence on screen passes, and the poor statistics of their wide receivers. Well, Green Bay happens to defend passes to running backs better than almost any other team in the NFL...
|RECEIVING DVOA||pass to RB||pass to TE||pass to WR|
As I noted last week, the Packers had a backwards road-field advantage on offense in 2003. Only the two New Jersey teams, oddly enough, played better offense on the road compared to home. Meanwhile, the Eagles had pretty much no defensive home-field advantage... and on offense, they had a backwards road-field advantage, like the Pack. Which will be nice if they move on to St. Louis, but is another negative for this week.
Facing all the advantages of home teams in the second round -- the crowd, the rest -- Green Bay has an uphill battle. But of the four teams playing on the road this weekend, I think they are most likely to make the summit. Which would likely send them to St. Louis. I wonder what Brett's guardian angel could do about getting that roof torn off the Edward Whoever Dome?