Saturday, February 5, 2011

Defense wins Championships ...

... that might be true in Soccer, Hockey, Baseball, Basketball, Handball, Deep Sea Fishing or whatever, but not in Pro Football (nor Cycling).

by Karl Berthold

Pundits like Easterbrook, Prisco, TV Commentators and - Coaches tell us year in and year out that "Offense sells tickets but Defense wins Championships". But is that true, or is it a myth?

Many studies about Superbowl Champions were done by Brian Burke and others based solely on Regular Season Statistics which showed us, the smart readers of "Advanced NFL-Stats", that "Offensive-Teams" (read "Passing-Teams") have the edge to win the final Game of the Season.

But nobody has yet done a study of Playoff - Statistics of Superbowl Winners based on the (up to four) Playoff Games they had to win before winning the Lombardi Trophy. That means the above mentioned "Experts" could still say Playoff-Football is different and in cold January you need a good running game and great Defense to prevail in the end.

Well... no longer can that be said.

If anything, then those experts should say "Pass - Offense wins Championships". Here's why:

I compiled statistics of the Superbowl Winners in those 5 categories (explanation see below*): Y/PP (Yards per Pass-Play including Sacks), Y/R (Yards per Run), Y/P (Yards per Play), Passer-Rating and PPG (Points per Game) for both the Offense and Defense.

1.) The "Free Agent Phase" (1993-2009)
The 17 Teams winning the Superbowl performed on average as follows (remember, these Statistics are only from their Playoff-Games, no "mixing up" of regular season stats here);

Offense:
Y/PP; 14,9 % better than NFL-Avg. (0,88 Y/PP better than NFL-Avg.)
Y/R; 8,5 % WORSE than NFL-Avg. (0,35 Y/R WORSE than NFL-Avg.)
Y/P; 3,4 % better than NFL-Avg. (0,17 Y/P better than NFL-Avg.)
Passer-Rating; 19,5 % better than NFL-Avg. (15,5 Rating-Points better than NFL-Avg.)
PPG; 40,4 % better than NFL-Avg. (8,4 Points scored better than NFL-Avg.)

What we can see here is the same that we see in regular season games. There is no such thing as "Special Playoff-Football". Teams that win, perform extraordinarily well in the Passing - Game. Running the ball has no influence. Actually Superbowl winners perform below the NFL-Average there. The reason might be that the Superbowl winning teams were running out the clock in their Playoff-Wins, thus opposing defenses tried to stop the Run in the 2nd Half...
Even though points scored are "poisoned" by Return TD´s, it´s amazing how many more points those superbowl teams scored than the average. That can´t be explained solely on Special Teams and Defense Return TD´s but by sensational efficient passing.

Defense:
Y/PP; 8,3 % better than NFL-Avg. (0,48 Y/PP better than NFL-Avg.)
Y/R; 9,0 % better than NFL-Avg. (0,36 Y/R better than NFL-Avg.)
Y/P; 4,6 % better than NFL-Avg. (0,23 Y/P better than NFL-Avg.)

Somebody might be confused now, how both Y/PP and Y/R can be better than 8%, yet the Y/P is below 5%. Shouldn´t then the Y/P also be better than 8%?
Assume the NFL-Avg.-Y/PP is 6 (which comes close to the "real" averages), the Y/R-Avg. is 4 (that´s almost always true) and the Y/P is 5 (also almost always true). Now assume the Championship-Team performed 20% better in both, Y/PP (that would mean they allowed 4,8 Y/PP) and Y/R (would mean 3,2). The fact is that opponents of Superbowl-Winning Teams passed the ball heavily (with good reasons since they were behind most of the time) throughout the Playoffs. There is almost no exception through the decades. So assume the Championship-Defense faced 2/3 passes and 1/3 runs. That gives a Y/P of 4,27 (= "only" 14,7 % better than NFL-Avg.). The "worse" Y/P-Avg. simply comes because passing gains more yards than running.

Passer-Rating;
17,7 % better than NFL-Avg. (14,0 Rating-Points better than NFL-Avg.)
PPG; 20,9 % better than NFL-Avg. (4,4 Points allowed better than NFL-Avg.)

If we compare the Offenses to the Defenses it´s obvious that the Offense, more often than not, was the key to Championship Football. One could argue now that there are no boundaries for Offense-Performance (You can throw 14,0 Y/PP thus performing 127 % better than the 2010-Average NFL-Team, but you can´t do better than 0,0 Y/PP on Defense. So the maximum is 100 there). To solve this problem, I compared the units of each Years Champions.

The advantage of the Offense isn´t as clear anymore, but it´s still there:
Y/PP; 10-7 (10 Teams had a better Pass-Offense than Pass-Defense compared to the NFL-Avg.)
Y/R; 5-12
Y/P; 9-8

Passer-Rating; 11-6
And, even though this Stat is "hated" on this site (including by me), Passer-Rating is very helpful here. In this System there are boundaries for Offensive-Performance. For example no more than 12,5 Yards per Attempt are "allowed", the maximum Rate is 158,3 etc. Since the Rating hovers around 80 (above 80 in later years, high 70s in earlier years), both the Offense and Defense can´t top the Avg.-Rate by more than about 80 Points.... And still the Offenses outperformed the Defenses, no matter how you look at it.

PPG; 10-7

Conclusion: We can call this Era an "Offense-Era", where running the ball had no influence on winning the Championship. The myth of "you need to run the ball to be successful in the Playoffs" is refuted.

2.) The "Golden Era" (1978-1992)
15 Superbowl Winners were crowned. Here are the Numbers;

Offense:
Y/PP; 30,5 % better than NFL-Avg. (1,79 Y/PP better than NFL-Avg.)
Y/R; 2,8 % better than NFL-Avg. (0,11 Y/R better than NFL-Avg.)
Y/P; 13,0 % better than NFL-Avg. (0,64 Y/P better than NFL-Avg.)
Passer-Rating; 42,7 % better than NFL-Avg. (31,3 Rating-Points better than NFL-Avg.)
PPG; 53,8 % better than NFL-Avg. (10,8 Points scored better than NFL-Avg.)

Astonishing! Passing-Football at it´s best. Aikman, The Hogs giving Rypien 10 seconds per Pass to find the open Receiver, Montana, "The-Greatest-Quarter-in-History" (1987-WAS), Simms being perfect, Bradshaw in beautiful "Air-Battles"... Defense wins Championships?? Not really in the "Greaties".

Defense:
Y/PP; 17,0 % better than NFL-Avg. (1,00 Y/PP better than NFL-Avg.)
Y/R; 13,9 % better than NFL-Avg. (0,56 Y/R better than NFL-Avg.)
Y/P; 12,8 % better than NFL-Avg. (0,64 Y/P better than NFL-Avg.)
Passer-Rating; 23,1 % better than NFL-Avg. (17,2 Rating-Points better than NFL-Avg.)
PPG; 39,5 % better than NFL-Avg. (8,2 Points allowed better than NFL-Avg.)

Unit-Comparison:
Y/PP; 11-4
Y/R; 7-8
Y/P; 8-7
Passer-Rating; 9-6
PPG; 8-7

Again, the numbers are not as "eye popping" as when we look at the pure percentages. But still the Offenses outperformed their Defenses; especially in Passing-Efficiency (Y/PP, the Stat which correlates best and is the reason for winning, as the readers of this site know).

3.) The "Pundits Era" aka "Dead Ball Era" (1966-1977)
The first 12 Superbowl Champions performed as follows;

Offense:
Y/PP; 18,0 % better than NFL-Avg. (0,98 Y/PP better than NFL-Avg.)
Y/R; 2,1 % WORSE than NFL-Avg. (0,08 Y/R WORSE than NFL-Avg.)
Y/P; 3,4 % better than NFL-Avg. (0,15 Y/P better than NFL-Avg.)
Passer-Rating; 36,9 % better than NFL-Avg. (23,9 Rating-Points better than NFL-Avg.)
PPG; 20,9 % better than NFL-Avg. (4,1 Points scored better than NFL-Avg.)

Defense:
Y/PP; 11,3 % better than NFL-Avg. (0,63 Y/PP better than NFL-Avg.)
Y/R; 9,6 % better than NFL-Avg. (0,38 Y/R better than NFL-Avg.)
Y/P; 7,9 % better than NFL-Avg. (0,37 Y/P better than NFL-Avg.)
Passer-Rating; 29,0 % better than NFL-Avg. (18,9 Rating-Points better than NFL-Avg.)
PPG; 43,0 % better than NFL-Avg. (8,5 Points allowed better than NFL-Avg.)

Again, no evidence here that "Defense wins Championships", not even in the beloved Era of the Pundits, who try to explain the game to us 40 years later. Passing-Efficiency ruled as always and it was better on Offense than Defense, no matter if you look simply at Y/PP or Passer-Rating. Maybe back then there were no calculators available, so the Pundits made the old mistake, looking solely at cumulative Stats (in this case: Points scored). If they did, yes they are right. Games were low scoring, therefore one can say "Defense won championships".... But could it be that coaches were already smart enough to run down the clock once they led because of efficient passing, thus scoring fewer points but saving the games? It seems likely. The rushing numbers show that too. Also seen at ...

The Unit-Comparison:
Y/PP; 7-5
Y/R; 4-8
Y/P; 5-7
Passer-Rating; 6-6
PPG; 4-8

4.) All Eras combined (1966-2009)

Offense:
Y/PP; 21,1 % better than NFL-Avg. (1,22 Y/PP better than NFL-Avg.)
Y/R; 2,9 % WORSE than NFL-Avg. (0,12 Y/R WORSE than NFL-Avg.)
Y/P; 6,7 % better than NFL-Avg. (0,33 Y/P better than NFL-Avg.)
Passer-Rating; 32,1 % better than NFL-Avg. (23,2 Rating-Points better than NFL-Avg.)
PPG; 39,6 % better than NFL-Avg. (8,0 Points scored better than NFL-Avg.)

Defense:
Y/PP; 12,1 % better than NFL-Avg. (0,70 Y/PP better than NFL-Avg.)
Y/R; 10,9 % better than NFL-Avg. (0,43 Y/R better than NFL-Avg.)
Y/P; 8,3 % better than NFL-Avg. (0,41 Y/P better than NFL-Avg.)
Passer-Rating; 22,6 % better than NFL-Avg. (16,4 Rating-Points better than NFL-Avg.)
PPG; 33,3 % better than NFL-Avg. (6,8 Points allowed better than NFL-Avg.)

Unit-Comparison:
Y/PP; 28-16
Y/R; 16-28
Y/P; 22-22
Passer-Rating; 26-18
PPG; 22-22

The bottom line is this: "Defense does not win Championships". Buddy Ryan, his sons and all those who jumped on this band wagon might still think it because "Da Bears" won 1985 behind the best Playoff-Defense (and of course one of the best Regular Season Defenses) ever.
But Buddy & his buddies might have forgotten that Jim McMahon was (almost) healthy for one Season which also led to the 2nd best (Regular-Season) Offense that year. And McMahon´s effieciency hit new heights in the 3 Playoff-Games (7,87 Y/PP, 100.7 Passer-Rating).
A good QB with one kidney (in combination with a good OL and WRs), more often than not, is at least as important as 100 Walter Paytons combined with a good Defense. It doesn´t matter if that QB wears gloves and snow is falling on the turf while the cold wind blows punts away. Passing still rules. As bad as it might sound, the above statistics show that.

And the journalists and "pundits" should listen more to their hearts...
Every year they elect an Offensive-Player as Regular-Season- and/or SB-MVP (owed to their sentiments), yet they try to be realistic when talking about Championship/Playoff-Football. That ways they fail in both categories...

Finally the complete Playoff-Game-Statistics of all 44 Champions:
They won the Y/PP-"Battle" in 78 % (106-30-1) of their Games,
but were not as close in the Y/R-"Battle" (59% by going 81-56),
outgained their opponents in 77 % (105-32) of the games, and
won the "Turnover-War" in 80 % (99-17-21).
All Game-Stats were dominated in the important categories which led to wins in the "Turnover-Battle", but only with little help from the running game.

So even if the Steelers win tomorrow, that doesn´t mean that the old myth becomes reality. Just look at the winner from the year before...



(* How the Stats were calculated:
For example we look at the 1999-Rams Offensive-Y/PP Stats;
They achieved a great 8,38 Y/PP in 3 Playoff-Games. The NFL-Avg. for that year was 5,85. That means they performed 43,2 % above the Avg. (8,38 divided by 5,85). The Net-Output was 2,53 Yards per Pass above NFL-Avg. (8,38 minus 5,85). Since they "outperformed" their own Pass-Defense (they were 1,2 % better than NFL-Avg.) a "win" for the Pass-Offense is scored in the Y/PP-Unit-Comparison. All other Stats were calculated the same way.)

9 comments:

Anonymous said...

Mr. Berthold I like the analysis but have trouble with the counter-intuitive notion that the running game is much less imprortant. Less important that the pass - I understand. MUCH less important - I have trouble with. You rely heavily on yards per rushing attempt (Y/R) to represent the rushing disipline and its impact on games. Your relaince on this stat is what that may be exagerating the rift between the passing game and the running game. This of course, does not take away from your overall analysis and use of simple terms to describe this incredible sport we love. Outstnading job. This is why we all keep coming back to this site.

Anonymous said...

"is at least as important as 100 Walter Paytons combined with a good Defense." Ouch! Easy with the rhetoric. Please take that out. McMahon's passing success cannot be understood outside the context of Walter's pass blocking, check-down receiving, and decoying. Historians see the 85 team as DEFENSE+Walter with a healthy touch of McMahon (both passing and leadership) = dominant championship. Nice job though.

Anonymous said...

"Defense does not win Championships". Ever?
Good to see we all "hate" the passer rating.

Anonymous said...

Walter Payton was a great blocker, receiver, and even passer. What i wanted to say is that the Bears had the same dominating Defense in at least 84 and 86. And Payton was on the team with all his 1.000-Yard-Rushing-Seasons long before 1985. But one thing to succeed was always missing: The passing game for one full season. When it was there, nothing could stop the Bears. Arguably the greatest and funniest team in the history of Pro Football...

Yes i know SR is more important than pure Y/R Stats. The problem is we only have those Stats since some years. So i had to use Y/R-Stats.

No, Defense alone never won Championships. But that is exactly what the pundits declare. My personal opinion is that you have to be good on both sides of the ball.

Thanks for the replys so far, Karl

Jim Glass said...

A good QB with one kidney (in combination with a good OL and WRs), more often than not, is at least as important as 100 Walter Paytons combined with a good Defense.

Interesting data, but don't hedge with your conclusion, tell us what you really think. :-)

Anonymous said...

Hi Jim,

since i first got into Football in 1985, all i saw was Offenses having the edge.

We had a great team over here (Berlin Adler), which scored at will, because they had a american QB who really could throw.

But the funny thing is: Commentators and journalists told me its the Defense which makes the difference. I was confused, so i started investigating. Over the past 25 years i got more and more convinced that indeed Offenses make the difference. The 1st paper about the importance of Y/PP was written around 1991 in SI. Since then more and more studies came up with the same results.

Finally i looked at Playoffs and Superbowls (because there were no studies): Same or even more convincing data (scoring is up, up, up, comes January & especially the Superbowl).

So i really annoys me to get lied at repeatedly by some high paid writers.

But one thing also is sure (SL in 2000 showed me that in a painful way): You can´t win without at least an Avg.-Defense (+ some luck), no matter how great the Offense is.

BTW: GB´s numbers this year in the playoffs showed an edge in all the categories i described abovwe. The Offens-Shows continued in 2010/11.

As Theisman said in the Live-SB-Coverage: "You can´t defend a perfect pass". That pretty much sums it up.

Enjoy the Off-Season, Karl, Germany

Anonymous said...

I think the research shows that good defense alone won't win a championship. You have to be balanced. Everyone is so good these days that if you have a single weak spot, it will be identified and exploited against you. The 80's Bears had good D and rushing, but it wasn't until the passing game came around that they won it all. The 2000's Rams didn't win a title until their D came around.

To win, you need to be balanced. Defense affects field position, turnovers, and opponent's strategy. Good passing offense puts the pressure on the opponent's defense and allows you to comeback if your defense fails. The two are so interrelated that you can't separate their effects from each other.

Matt W. said...

I just think it's hilarious that a guy in Germany who uses commas in place of decimal points understands this better then the guys on ESPN.

Anonymous said...

Thanks Matt W. ;-)

But here in germany we use commas. Didn´t know it´s different in USA.

BTW, all playoff numbers are again way up this playoff season, even by those teams who are "defense" teams (except BAL). It´s amazing.

Karl, Germany

Post a Comment