Writers of Pro Football Prospectus 2008

Most Recent FO Features


» The Deep Ball Project

Guest columnist John Kinsley breaks down the tape of every deep pass in the NFL in 2017 and comes away with a shocking conclusion: even without Andrew Luck, the Colts had the best long-ball quarterback in the league.

17 Feb 2004

Guest Column: Was Super Bowl XXXVIII the Greatest of All Time?

Guest Column by Vincent Verhei

"I was in a group of three or four writers talking with [Adam] Vinatieri after the game. Someone asked him: 'Do you think this was the best Super Bowl of all time?' Vinatieri said he would leave it up to others to decide.

'Best Super Bowl of all time,' I said. He smiled and said that was good enough for him."

— Peter King, SI.com, Feb. 2, 2004.

So there's one man's opinion as to the best Super Bowl ever. Others will cite the Patriots' other win over the Rams in SB XXXVI. Personally, I'll always remember The Rams' victory over the Titans in SB XXXIV, and Mike Jones tackling Kevin Dyson one… yard… short.

We could argue about which Super Bowl was the most exciting all day long, but we'd never reach a satisfying conclusion. After all, you can't come up with a way to measure the excitement of a football game… can you?

Well, as a firm believer that anything can be quantified, I set out to attempt to measure the excitement level of a football game. While some fans may love high scoring games, or lots of passing, others (like me) love to watch solid defense and running. So I didn't want to look at styles of play.

I decided there were three elements to any exciting football game:

• A close finish. Obviously, a game that comes down to the final play is more exciting than one that's over by halftime.

• A big comeback. Is there anything better than watching a team struggle against its opponents and the clock, and pulling it out?

• A close game throughout. This may seem somewhat contrary to a big comeback, but a game in which neither team can build a big lead is a game that never gives fans a chance to catch a break.

Before I explain my methods, I'd like to give any non-math types a chance to skip four paragraphs down and get to the results. Now, how to measure these factors? The close finish was easy: I simply used the game's final margin. In case of overtime, I set the “final margin” to 0. (Obviously, there has never been an overtime Super Bowl, but I wanted my system to work for any football game, Super Bowl or otherwise.) The closest margin in any Super Bowl was in SB XXV, when the Giants beat the Bills by a single point, 20-19. The widest margin came in the 49er's 55-10 blowout of the Broncos the year before in SB XXIV, a 45-point margin.

The comeback factor is slightly trickier. Suppose a team falls behind by 10 in the first quarter, then roars back and wins by 32, as Washington did against Denver in SB XXII. Technically Washington came back from a 10-point deficit, but that comeback was overwhelmed by the offensive onslaught later in the game. I decided to measure comebacks by taking the largest lead in the game, at any point, and subtracting the final margin. The biggest “comeback score” for any Super Bowl came in SB XVI, when the Bengals fell behind the 49ers 20-0, then came flying back, only to lose 26-21.

To measure whether a game was close throughout or not, I use a metric I call “average margin.” I weight each score of the game by the amount of time that score was in effect. I total these products, and divide by 60 (or the total time of the game in the case of overtime games). For example, suppose a game is scoreless for 30 minutes, then the home team kicks a field goal as the final gun of the first half sounds, and there is no scoring in the second half. This game was scoreless for 30 minutes (30 X 0 = 0) and there was a 3-point lead for 30 minutes (30 X 3 = 90). Totaling those two figures (0 + 90 = 90) and then dividing by 60 gives us an average margin of 1.5, which I then round up to 2. That's a very low figure, which demonstrates that neither team was ever more than one play out of the game. The highest average margin in Super Bowl history came, again, in SBXXIV, when the 49ers jumped ahead early and never looked back. The record for lowest average margin in Super Bowl history was actually broken this year; the average margin of SB XXXVIII was a mere 2 points.

Now it's time to put it all together. I give each game a base score of 100. I subtract the average margin. Then I subtract the final margin. Finally, I add the game's “comeback factor.” I call this total, for lack of a better term, the excitement factor. I'll list the top five games with brief comments, then list the rest:


SB XIII, Pittsburgh 35, Dallas 31 — Excitement factor: 105.

Often overlooked in the great games discussion, possibly because it occurred before the Super Bowl was the international event it is today. Terry Bradshaw passed for 318 yards and 4 touchdowns, but it was nearly all for naught as Roger Staubach threw 2 touchdowns in the final three minutes before running out of time.

2. SB XXV, New York Giants 20, Buffalo 19 — Excitement factor: 104.

This game had two things going for it that go unaccounted for in my ratings. At the time, it appeared to be a huge upset. Given the benefit of hindsight, I'm not so sure it was. More importantly, the game took place at the peak of the first Gulf War, and patriotic fervor was running wild. I'm sure we all remember Scott Norwood's 47-yard field goal falling just to the right.

3. SB XXXVIII, New England 32, Carolina 29 — Excitement factor: 103.

The system says Peter King was wrong. Now, if Vinatieri had missed that kick…

4. SB XXXVI, New England 20, St. Louis 17 — Excitement factor: 102.

I'd imagine the average hair color in the Boston area has gotten a lot grayer in the past three years.

5. SB XVI, San Francisco 26, Cincinnati 21 — Excitement factor: 101.

This was the first one, with Ken Anderson vs. Joe Montana. As noted above, the 49ers were ahead 20-0 at halftime (including two field goals in the last fifteen seconds) then holding the Bengals off to win.

And the rest:

Rank Super Bowl Winner Pts Loser Pts ES
6 X Pittsburgh 21 Dallas 17 100
7 V Baltimore Colts 16 Dallas 13 97
8 XXXIV St. Louis 23 Tennessee 16 96
9 XXXII Denver 31 Green Bay 24 92
10 (tie) III New York Jets 16 Baltimore Colts 7 91
  VII Miami 14 Washington 7 91
12 XXIII San Francisco 20 Cincinnati 16 90
13 XXVI Washington 37 Buffalo 24 88
14 IX Pittsburgh 16 Minnesota 6 86
15 (tie) XIV Pittsburgh 31 Los Angeles Rams 19 85
  XVII Washington 27 Miami 17 85
17 XXX Dallas 27 Pittsburgh 17 84
18 XXI New York Giants 39 Denver 20 80
19 XI Oakland 32 Minnesota 14 79
Rank Super Bowl Winner Pts Loser Pts ES
20 (tie) II Green Bay 33 Oakland 14 78
  XXVIII Dallas 30 Buffalo 13 78
  XXXIII Denver 34 Atlanta 19 78
23 XXXI Green Bay 35 New England 21 77
24 VIII Miami 24 Minnesota 7 75
25 (tie) XV Oakland 27 Philadelphia 10 74
  IV Kansas City 23 Minnesota 7 74
27 XII Dallas 27 Denver 10 73
28 VI Dallas 24 Miami 3 70
29 XXXVII Tampa Bay 48 Oakland 21 69
30 XXIX San Francisco 49 San Diego 26 67
31 XIX San Francisco 38 Miami 16 66
32 I Green Bay 35 Kansas City 10 65
33 XXXV Baltimore Ravens 34 New York Giants 7 62
34 XVIII Los Angeles Raiders 38 Washington 9 56
35 XXVII Dallas 52 Buffalo 17 51
36 XXII Washington 42 Denver 10 50
37 XX Chicago 46 New England 10 49
38 XXIV San Francisco 55 Denver 10 32

Obviously, the system isn't perfect, but I think it does a great job of putting the good games on top and the bad games on bottom. I am pleased.

Posted by: Guest on 17 Feb 2004