Neil described the exact methodology in his quarterbacks post, so I won’t waste time repeating it. However, I wanted to look at coaches over an even longer period, and 1950 sounded like a good cut-off.1 Since we don’t have point-spread data for games from 1950 to 19772, I simply used the projected point spread based on the differential between each team’s SRS ratings and by awarding the home team three points. So for pre-1977 games, coaches are credited with wins over expectation based on the SRS, and for post-1977, for wins over expectation based on the Vegas line. Here are the results.
|Rank||Coach||First||Last||Teams||G||Act W||Exp W||Index|
|32||Jim Lee Howell||1956||1959||nyg||4||2||1.6||0.4|
|67||Jack Del Rio||2005||2007||jax||3||1||1||0|
- By clicking on the Index column, you can sort the table in reverse order. After doing so, you can see why this is called the Schottenheimer Index. Marty lost two games where his teams were 10.5-point underdogs (1985 in Miami, 1991 in Buffalo), but what really hurt him was an 0-8 record in games featuring a three-point (or smaller) spread. This includes The Drive, The Fumble, a game where his third-string quarterback threw a game-clinching 4th quarter interception when trailing by 5, and loss in Miami when Dan Marino threw two 4th quarter touchdowns to give Miami a 17-16 win. Schottenheimer also lost three games as five-point favorites: the Lin Elliot game against the Colts, in overtime against the Jets,3 and the famous Marlon McCree game against New England.
- Joe Gibbs had two bad playoff games. While Washington may have been only a two-point favorite against the Raiders in Super Bowl XVII, the Redskins were the defending Super Bowl champions and just set the record for the points scored in the regular season. They were embarrassed 38-9. In their next playoff game eleven months later, Washington lost as nine-point home favorites to the Bears. In Gibbs’ 22 other playoff games, Washington went 17-5, with all five losses coming in road games where the Redskins were at least 3.5-point underdogs (and in four of them, the Redskins were getting at least a touchdown).
- Tom Coughlin led two of the five biggest playoff upsets since 1950, at least according to Vegas (or the SRS). The five largest underdogs to win: the Jets (+18) in Super Bowl III, the Patriots (+14) in Super Bowl XXXVI against the Rams, a bad 1977 Vikings team (SRS of -1.6) on the road against a Rams team that beat them 35-3 earlier in the year in Los Angeles, the Giants (+12.5) in Super Bowl XLII against the Patriots, and the Jaguars (+12.5) against the Broncos after the 1996 season. But Coughlin incredibly won four other games despite being touchdown underdogs, and is 4-1 when the spread is within three points.
- John Harbaugh is mostly here thanks to this great postseason run, as winning in Denver and in New England was worth 1.45 wins. If the Ravens beat the 49ers, he’ll jump into the third slot on the list.
- Obviously all of Vince Lombardi’s games came before 1977, so I used the SRS for most of those lines. In 10 playoff games, he was a favorite 8 times. The two exceptions came in 1966 at Dallas (+0.2) and 1967 at home against a Rams team (+1.2). In six of his playoff wins, the SRS (or, for the two Super Bowls, the Vegas line) put the Packers as at least 9-point favorites. Lombardi’s teams were too good to give him that much credit for exceeding expectations, although a seventh-place finish isn’t too shabby.
- Bill Parcells with the Giants: +2.9. With the Patriots: +0.1. With the Jets: 0 (two playoff games, won as a 9-point favorite, lost as a 9-point underdog). With the Cowboys: -0.9. Obligatory Tony Romo reference.
- I can’t add anything to the Bill Belichick discussion that would differ from what Neil wrote about Tom Brady, so I’ll just note that with the Browns, Belichick won as a 3-point favorite against New England and lost as a 3.5-point underdog against Pittsburgh.
- Paul Brown is an interesting case. The NFL does not officially recognize Cleveland’s records from their time in the AAFC, when Brown led them to five playoff victories and four championships (although the Pro Football Hall of Fame does). That caveat aside, Brown fares pretty poorly here. In the NFL, his playoff record was 4-8, and half of those wins came in 1950, the Browns first year in the NFL. Even ignoring his time in Cincinnati (0-3), he went 1-4 in games with a three-point spread and also lost to a far inferior Lions team 17-16 in the 1953 title game. Cleveland started that season 11-0, lost a meaningless final game, and then lost in Detroit. The 1951 Browns were just as good, but also allowed a fourth-quarter comeback to lose the title to the Rams. And despite being better than Detroit again in 1957, Cleveland was throttled 59-14. On the plus side, Cleveland did exact some revenge against Detroit in the middle, winning 56-10 in the 1954 title game.
Some final thoughts.
Doug came up with this idea six years ago, when he called it the Dungy Index (itself a revival of an earlier post). 4 As a reminder of the environment we were in at that time, consider Doug’s closing thoughts:
In the span of five weeks last year, Bill Cowher turned from A Coach Who Can’t Win The Big One to unquestionably one of the best coaches of his generation. I don’t have inside knowledge, but I can only assume that happened because he radically changed everything about the way he coached. I mean, what else could possibly explain how a Coach Who Can’t Win The Big One could win the big one?
As of right now, it appears that Tony Dungy might have completely and totally changed his coaching style and philosophy too, but we can’t be sure about that until next Sunday.
Spoiler: Dungy won on Sunday, and now he’s no longer a choker, either.
One final note. Please keep in mind that this measures a coach’s (team’s) postseason performance relative to his (team’s) postseason expectations. Knowing that Marty Schottenheimer scores poorly here does not tell us whether he deserves blame for underachieving in the postseason or credit for overachieving during the regular season. I also note that while I find this concept interesting, I make no claim that it holds any predictive value (to the extent such a concept makes sense when talking about retired coaches).
- Note that coaches, like Paul Brown, who coached before 1950 are included, but their pre-1950 stats are not. [↩]
- One other piece of fine print: for the Super Bowls, I used the actual Vegas lines, since those are readily available. [↩]
- Some have understandably called this the Nate Kaeding game, since he missed a 40-yard field goal in overtime that would have won the game. I won’t do that for two reasons. One, the Chargers ran LaDainian Tomlinson three times for no gain once they got to the 22, a move that was unnecessarily conservative. But more importantly, the game never should have gone to overtime. Trailing 17-10 with 24 seconds left, San Diego had 4th and goal from the Jets two-yard line. Well, Drew Brees’ pass for Antonio Gates was incomplete, but the drive was kept alive by a roughing the passer penalty on Eric Barton. The call was correct, but it was a gift for the Chargers just to get to overtime. [↩]
- Then I came up with my first Schottenheimer Index six years ago, although with a crude formula that does not hold up well on repeat reading (at least, in my opinion). [↩]