TCU has played much better in the second half of games this year. What other schools have seen dramatic shifts of play after halftime?
10 Dec 2005
by Aaron Schatz (and the FO staff)
Time for another look at the Football Outsiders mailbag. We get a lot of e-mail, and there are a lot of comments on the discussion threads, so I apologize if your question doesn't get answered. There simply are too many good questions that require well thought out answers. The best way to get your question answered at this point is to use the contact form. If it is a question not related to the DVOA stats, it is more likely to be answered if you send it to one of the other writers, not me.
Be aware that we reference plenty of our innovative FO stats here, not to mention their unfamiliar terminology, so if you are a recent addition to the readership you might want to read this first.
I mentioned in this week's DVOA ratings thread that I would run week-to-week graphs on Indianapolis and Cincinnati to see how the Bengals could end up slightly higher than the Colts in total DVOA for the season, even though the Colts are a) undefeated and b) number one in weighted DVOA.
So here's the graph over on the right. I've put Cincinnati and Indianapolis on the same graph, and as long as you have a color monitor you should be able to tell which is which. The three Cincinnati losses are marked "L."
As you can see, Cincinnati's wins in Weeks 2 and 3 really stand out compared to the rest of the Bengals' season until last week's victory over Pittsburgh. The ratings for those wins have moved up a little each week because Minnesota and Chicago keep winning each week. I know many people will say, "How can you give the Bengals a bonus for beating the Vikings back when they were terrible," but for now our opponent adjustments get retroactively updated each week based on each team's play over the entire season. I don't want to make a subjective judgment on when a win over a good team "shouldn't count," and while I've thought of doing "rolling opponent adjustments" to take into account the improvement and decline of teams during the season, I just haven't been able to work on that yet.
It is obvious from this graph just how consistent the Colts have been this year. They have only one game that even slightly dips below 0% after opponent adjustments, the 31-17 win over Houston in Week 10. The Colts have only one other game below 20%, the 13-6 win over Cleveland in Week 3. By comparison, Cincinnati has five games at 20% or lower. Like the Colts, the Bengals have one win below 0%, and like the Colts, that win came against Houston.
The other week that stands out is probably Week 11, because that was the week the Colts beat the Bengals, and yet the Bengals actually end up with a slightly higher DVOA. Did you know the Bengals actually gained 7.6 yards per play in that game, the Colts just 6.4 yards per play? In a close game like this, sometimes you end up with a weird little adjustment result. The Colts barely beat the Bengals in non-adjusted VOA. But because of the specific mix of rushing and passing plays and the way the adjustments work, the team that is technically number one gets a bigger boost for playing well against the team that is number two than the team that is number two gets for playing well against the team that is number one. It's not worth getting hung up on this, these little quirks of a percentage point or two end up washing out when we look at 12 games rather than one -- and like a lot of things related to the Colts, it will look different after Indianapolis has played three more teams in the top ten over the next three weeks. To me, it's pretty obvious from that graph that Indianapolis has been the better team since Week 4.
As long as we're looking at week-to-week graphs, how about two more AFC teams, the ones ranked third and fourth? The Denver and San Diego weekly ratings are over on the left -- both teams this time have some losses, so the orange Ls are Denver losses and the blue Ls are San Diego losses. In last week's FOXSports.com power rankings commentary, I noted that the Colts are on pace to become only the second team to finish in the top five in DVOA but the bottom five in VARIANCE. Guess what -- the Chargers are the third. They are nowhere near as consistent as the Colts, but they are currently 28th in VARIANCE.
How hard has San Diego's schedule been this year? Three of the four San Diego losses actually have DVOA ratings above 0%! The 20-17 loss to Philadelphia is the only exception, and that's only because the Eagles have fallen apart -- for much of the season, the opponent adjustment had that game above 0% also. (San Diego does have one other game slightly below 0%, Week 9's 31-26 win over the Jets.) Meanwhile, see if you can pick out Denver's loss to Miami and win over Philadelphia. Those games sort of stand out, don't they?
Onto some questions, with answers from a number of different FO writers...
Duane: Unluckiest. Team. Ever. That will go a long way to assuage my concerns about the Packers, except for that nagging feeling that, down the road, â€œthe 2005 Packers will look a lot more like the 1981 Patriots than they will the 1979 49ers.â€? But who has been the Luckiest. Team. Ever?
Aaron Schatz: Here's the top ten (post-1950), not including the 1982 strike year. Two of these teams will look very familiar. You'll also find a 2005 team that is on pace to finish among the "luckiest" of all time -- and if Green Bay is one of the "unluckiest" teams, the identity of this team makes perfect sense:
Hi. We're the 1992 Indianapolis Colts. You may remember us from such Football Outsiders articles as Too Deep Zone: Short Runs and Pythagoras on the Gridiron. The 1992 Colts were outscored by 86 points, in part, because they couldn't manage three yards per carry, and they couldn't manage three yards per carry, in part, because they were busy running out the clock in a bunch of close, lucky wins. That team was 1-15 the year before, then shockingly went 9-7, and then went 4-12 the year after.
You may be surprised to learn that a lot of these teams were fine the next year. The 1963 Steelers were 7-4-3. The 1990 Steelers went 9-7 again. The 2000 Titans improved to 14-2, even though they did not return to the Super Bowl. That 1960 Bears team is really weird, because that overachieving 5-6-1 year came between an 8-4 year and an 8-6 year and in both of those seasons, the Bears pretty much equaled their Pythagorean projection.
On the other hand, both of the historically strong overachievers from 2004 are in danger of not making the playoffs. Nobody expected the Steelers to go 15-1 again, but this is part of the reason we were so down on Atlanta in PFP 2005.
And, of course, two three-point losses to the Vikings are a big factor in Green Bay falling so far below its projection, so it makes sense that Minnesota would be playing greatly above its projection. Minnesota has won six of its last seven games -- but the average margin of victory in those six wins is just 6.5 points, and four of Minnesota's five losses have come by 20 points or more.
Kevin Rich: Regarding the Reggie Bush Bowl - I see that my favorite team, Green Bay is tied with two other teams for second behind Houston for the 1st pick in the NFL draft. I went to the official NFL site to see what the tiebreaker would be for the 1st pick. It is based on the strength of schedule tie breaker. My problem is that the site did not explain how strength of schedule works. So I am asking you how is the strength of schedule computed?? I hate to say it, but I think it would be best for Green Bay to lose every single game until the season is over and hope for the 1st pick.
Tim Gerheim: Strength of schedule is just the aggregate winning percentage of all the teams a team played in the season. Take all 16 teams the Packers played (counting division rivals twice), add up their wins, and divide by 256 (16 x 16). A team that played poorly against a worse schedule wins the tiebreaker. There are still ties, since there are so few games, so then you go to division tiebreaker if they're in the same division, or conference tiebreaker if they're in the same conference, and if those don't work it's a coin flip. It seems like a draft order tie gets broken by a coin flip every year.
At this point, Green Bay is ahead of all the other 2-10 teams but behind both 3-9 teams. And if the Packers can make up ground on the Texans they'll definitely have that tiebreaker. Plus, if the Texans know what's good for them (cough ... as well as I do ... D'Brick-cough-shaw) they'll be looking to trade down rather than take Reggie Bush.
Speaking of whether or not the Texans should take Reggie Bush...
Bob Mozitis: Mike, fellow La Salle grad here. I just don't get the whole Reggie Bush thing. Can you address this please? Bush, at best, could be as good as Barry Sanders, and Barry never even approached a Super Bowl because a lack of talent around him. Am I crazy to think that the Texans should trade the #1 pick for all sorts of wealth this year and completely overhaul their team??? This is the best season ever to have the #1 pick, but not to draft Bush. Instead, the value you could get in a trade would seem to be much much better than single player, whether it is Leinart or Bush. Besides not paying #1 pick money, you could get a handful of players and picks in return, since teams seem so desperate to impress their local media that they would do anything to get the top pick.
Mike Tanier: I believe that most of us at FO feel that, while Bush will be a phenomenal pro, the Texans have so many needs that he wouldn't make that much of an impact on the team. If they end up with the #1 pick, it would probably make sense for them to trade down, presumably to get two first-round picks. If the Niners or Packers wind up with the top pick, maybe it's a different story.
If Charley Casserly stays on at GM, he will probably look to deal. Casserly is good at wheeling and dealing when it comes to draft picks: he landed LaVar Arrington and Chris Samuels when he was in Washington and has pulled off some miracles in his career. (Drew Henson for a third-round pick?)
Tom: The stats you guys have here are amazing, what a great site. I'm trying to help out a professor who is writing a book here at Michigan State. Who coined the phrase "Red Zone" in football? It wasn't around in the 80s, but throughout most of the 90s and all the time today you hear this phrase. We have searched and searched, maybe we are just missing something that is simple to find. Any information you could offer would be greatly appreciated.
Al Bogdan: It might have been Joe Gibbs. I found a Washington Post article which suggests that, if it wasn't Gibbs that coined the phrase, it at least was a novel term when Gibbs first started throwing it around in 1982.
"In the last three weeks, the Redskins have scored only two touchdowns, and one was a 65-yard pass Joe Theismann threw to Charlie Brown. The only touchdown that resulted from a sustained drive in which the Redskins moved inside the opponent's 20--called the red zone by Gibbs--and then scored was on a 17-yard pass to Brown."
-- Paul Attner, "Recent Dearth of Touchdowns Concerns Gibbs," Washington Post, Dec. 14, 1982, at C1.
The phrase wouldn't appear in the New York Times, at least applied to football, for another two years.
"Noise levels did not rise again until midway through the second quarter, when the Colts stopped the Giants in the so-called red zone, between the 20-yard line and the end zone."
-- Craig Wolf, "Colts Defeat Giants in Indianapolis Debut," New York Times, Aug 12, 1984, at S11.
The phrase didn't appear in SI until 1989, in a Peter King article:
"Says Bengal offensive coordinator Bruce Coslet, "In 1987 we'd get into the red zone [inside the opponent's 20-yard line], and we couldn't score. So, going into 1988 we stressed scoring every time we got into the red zone. That was the whole difference in our team. Every possession was important."
-- Peter King, "Inside the NFL," Sports Illustrated, Sep. 18, 1989, at 69.
Jim: How many times has a team chosen, when playing extra time, not to receive the ball first after winning the coin toss.
Michael David Smith: Of course, one of the answers is the Lions, which is why Aaron forwarded this question to me. Marty Mornhinweg was an idiotic coach of historic proportions, but, believe it or not, he's not the only coach to take the wind in overtime. Entering the 2005 season, there had been 377 regular-season overtime games. Of the 377 coin toss winners, 368 chose to receive and nine chose to have the wind at their backs. Teams taking the wind are 4-5.
Bruce Bates: Do you know the average recovery stat for a team when they fumble? I now this is pretty much dumb luck so I would figure offenses get it back near 50 percent of the time. The Redskins have had unbelievably bad luck at recovering their fumbles this year.
Also, I notice the Redskins have ranked 30th or worse at acquiring fumbles per drive for the last two seasons. This seems impossible to me. But when you don't force them, it's hard to recover them. I wondered if there were any stats to show that Gregg Williams defenses have always struggled at forcing and recovering fumbles?
Aaron Schatz: The average recovery rate is about 53 percent, but it is different for different types of fumbles:
Washington's fumble record this year is really wacky, and represents more than half of that huge difference between Washington's VOA (0.0%) and DVOA (17.6%). Washington's rating if we removed all the opponent adjustments, but still considered all fumbles as equal, would be 9.6%.
Do Gregg Williams defenses have any particular history of not recovering fumbles? I don't think so. Over to the right is a table showing the Washington defense for the last two years and the defenses in Buffalo, where Williams was head coach, for the three years before that. Last year the Redskins had perfectly normal luck recovering fumbles, they just caused fewer of them.
On offense, by the way, Washington also recovered 50 percent of its fumbles in 2004, and just 30 percent of its fumbles in 2005. Here's a look at all the Washington fumbles this season through Week 13:
2 aborted snaps, 1 recovered
8 sacks, 2 recovered
6 on rushes, 1 recovered
4 on receptions, 2 recovered
4 aborted snaps, 1 recovered
6 sacks, 2 recovered
4 rushes, 1 recovered
2 receptions, 0 recovered
I really don't know if there's a pattern here, I think it really just is blind luck. Or unluck, in this case.
73 comments, Last at 16 Dec 2005, 1:29pm by Mr Shush