25 replies to this topic

[Dibs]

Yesterday I read a post stating the old axiom of “Defense wins championships”.  I wondered if this is still a truism in the modern NFL or not.  It also got me thinking about similar theories that I have heard over the years and whether there is any specific area that one needs to be dominant in to win a SB.

Subsequently I started analysing data from past SB winning teams in order to try and determine obvious patterns.  As I progressed, the data started to intrigue me on additional levels….so I kept collating.  I feel that I may have overdone it a bit….but for those interested, this is what I found.


I ended up selecting the season rankings of the last 26 SB winning teams(2012 season for the 2013 SB champ etc). I broke these stats up into…
Total Offense/Defense
Passing Offense/Defense
Rushing Offense/Defense

As there are different schools of thought on which stats best represent these categories, I collated two statistics for each one.
Total Yards O&D and Points For/Against
Passing Yards O&D and Passer Rating O&D
Rushing Yards O&D and Rushing Yards/Attempt O&D

I also noted which SBs were won with/without a Star QB.  This was of course opinion based and anybody is free to argue the validity of certain choices made.  I put a “?” for Flacco as there is still hot debate whether he can be considered a great QB or merely a good one.


Data in attachment:  How to win a Super Bowl1.doc   80.5K   24 downloads
(Word doc)

One thing that visually stood out was the single digit rankings.  There seems to be a clear change from the year 2000.  Every category has a higher(often much higher) number of “Top 10” years prior to the 2000 season.  Due to this apparent division I decided to study 26 years(double the 13 years from 2012-2000).  For those interested, I actually looked at 36 years of numbers with years 27-36 showing similar results to years 14-26.

To further study the numbers, I summarized the results into categories, and the number of SB winning teams for each category.

Categories are….Top 5, Top 10, Bottom 5, Bottom 10

I also decided to apply a points grading system.  I did this so that each individual rank could have effect on the study as well as to get a clear summarized number for each category in order to more easily compare them.  I gave a #1 ranked category 28 pts, #2 27 pts, #3 26 pts etc.  Zero being lowest points earned.
Furthering to that I also averaged out the grading numbers for the two sub-categories in each category.  Example: Total Yards O + Points For.


Data in attachment:  How to win a Super Bowl2.doc   56K   9 downloads
(Word doc)

Looking at these summaries one can clearly see a distinction between the 2012-2000 years and the 1999-1987 years.  It seems that prior to 2000, teams needed to be good-dominant in nearly all areas as well as not being poor-bad in any area to win the SB.  2000+ teams however are nearly always lacking in several areas, with teams that are poor-bad in a certain area being able to regularly win the big one.

The only thing that I can think of to explain such a disparity between the eras is the flow on effects of Free Agency(1993)….but I am dubious that it would take 8 years for this to effect the SB winning teams to such a drastic measure.


Comparing the results with the old axioms and sayings is quite telling….

“Defense Wins Championships”
Appears to have been true pre-2000 with 7/8 of 13 teams having a Top 5 D with another 4/5 being in the top 10.  It is however misleading as 12/11 of 13 teams also had a Top 5 Offense.

In relation to 2000+ teams, it is perhaps appropriate to say that “A good D helps win Championships” as only 5/7 had a Top 5 D with only another 3/1 being in the top 10.  Total Offense however takes a huge step backwards in importance in the modern NFL.  Only 2/3 were Top 5 plus 3/5 Top 10.  A quick look at the first table shows many teams being of average or below average in the Offensive Yards/Points categories.

“Run the ball & Stop the run”
I won’t list the numbers as before(look in attachment 2) but it seems that prior to 2000 this was quite true(though the same could be said for the passing game)….but since 2000, the importance of the run game seems to have been greatly diminished.

“In the modern NFL you need to be able to stop the pass”
Fairly True.  9 of 13 teams had a Top 10 D for Passer Rate….but that figure was 10 for pre-2000 teams so the theory should not be limited to the “modern” NFL.


I won’t go through any more common theories as there really isn’t any perceivable pattern to suggest a formula for victory from the modern data.

There is however one area that clearly has commonality with modern SB winning teams.  The Star QB.

Starting QBs of winning SB teams…
(As mentioned earlier, I am calling Flacco a “?” QB.  Therefore I will assume 2011-1987 as a 25 year sample for the purpose of this study).

Last 25 seasons
20 Stars
5 non-stars

Last 20 seasons
18 Stars
2 non-stars

Unfortunately for us Bills fans, it seems the only common factor in the modern era for winning a SB is having a Star QB.  Fingers crossed we find one this off-season.

[Billsrhody]

This is awesome. Great analysis!

[BringBackFergy]

The 5 non-star winning QB's during past 25 yrs are Hostetler, Williams, Dilfer and who?? I'm drawing a blank and too lazy to look it up.

[silvermike]

BringBackFergy, on 05 February 2013 - 04:08 PM, said:

The 5 non-star winning QB's during past 25 yrs are Hostetler, Williams, Dilfer and who?? I'm drawing a blank and too lazy to look it up.

Brad Johnson would be one.  Maybe Mark Rypien would go in that category, too.  Jim McMahon?  Phil Simms?  Joe Theisman?

[NewEra]

Doubtful we find one this year.  Maybe next season. Sad, but true.  You can't suck blood from a stone. I see a lot of stones out there.

[Triple Threat]

NewEra, on 05 February 2013 - 05:16 PM, said:

Doubtful we find one this year.  Maybe next season. Sad, but true.  You can't suck blood from a stone. I see a lot of stones out there.

So you think there will be 6 obvious star franchise QB's coming out next year?

[jeremy2020]

What about making the playoffs?

I'd also be interested in how many of these QBs performed well in in their first 3 years with a poor running game and/or a poor defense.

Edited by jeremy2020, 05 February 2013 - 05:40 PM.

[dave mcbride]

Dibs, on 05 February 2013 - 02:22 PM, said:

Yesterday I read a post stating the old axiom of “Defense wins championships”.  I wondered if this is still a truism in the modern NFL or not.  It also got me thinking about similar theories that I have heard over the years and whether there is any specific area that one needs to be dominant in to win a SB.

Subsequently I started analysing data from past SB winning teams in order to try and determine obvious patterns.  As I progressed, the data started to intrigue me on additional levels….so I kept collating.  I feel that I may have overdone it a bit….but for those interested, this is what I found.


I ended up selecting the season rankings of the last 26 SB winning teams(2012 season for the 2013 SB champ etc). I broke these stats up into…
Total Offense/Defense
Passing Offense/Defense
Rushing Offense/Defense

As there are different schools of thought on which stats best represent these categories, I collated two statistics for each one.
Total Yards O&D and Points For/Against
Passing Yards O&D and Passer Rating O&D
Rushing Yards O&D and Rushing Yards/Attempt O&D

I also noted which SBs were won with/without a Star QB.  This was of course opinion based and anybody is free to argue the validity of certain choices made.  I put a “?” for Flacco as there is still hot debate whether he can be considered a great QB or merely a good one.


Data in attachment: How to win a Super Bowl1.doc
(Word doc)

One thing that visually stood out was the single digit rankings.  There seems to be a clear change from the year 2000.  Every category has a higher(often much higher) number of “Top 10” years prior to the 2000 season.  Due to this apparent division I decided to study 26 years(double the 13 years from 2012-2000).  For those interested, I actually looked at 36 years of numbers with years 27-36 showing similar results to years 14-26.

To further study the numbers, I summarized the results into categories, and the number of SB winning teams for each category.

Categories are….Top 5, Top 10, Bottom 5, Bottom 10

I also decided to apply a points grading system.  I did this so that each individual rank could have effect on the study as well as to get a clear summarized number for each category in order to more easily compare them.  I gave a #1 ranked category 28 pts, #2 27 pts, #3 26 pts etc.  Zero being lowest points earned.
Furthering to that I also averaged out the grading numbers for the two sub-categories in each category.  Example: Total Yards O + Points For.


Data in attachment: How to win a Super Bowl2.doc
(Word doc)

Looking at these summaries one can clearly see a distinction between the 2012-2000 years and the 1999-1987 years.  It seems that prior to 2000, teams needed to be good-dominant in nearly all areas as well as not being poor-bad in any area to win the SB.  2000+ teams however are nearly always lacking in several areas, with teams that are poor-bad in a certain area being able to regularly win the big one.

The only thing that I can think of to explain such a disparity between the eras is the flow on effects of Free Agency(1993)….but I am dubious that it would take 8 years for this to effect the SB winning teams to such a drastic measure.


Comparing the results with the old axioms and sayings is quite telling….

“Defense Wins Championships”
Appears to have been true pre-2000 with 7/8 of 13 teams having a Top 5 D with another 4/5 being in the top 10.  It is however misleading as 12/11 of 13 teams also had a Top 5 Offense.

In relation to 2000+ teams, it is perhaps appropriate to say that “A good D helps win Championships” as only 5/7 had a Top 5 D with only another 3/1 being in the top 10.  Total Offense however takes a huge step backwards in importance in the modern NFL.  Only 2/3 were Top 5 plus 3/5 Top 10.  A quick look at the first table shows many teams being of average or below average in the Offensive Yards/Points categories.

“Run the ball & Stop the run”
I won’t list the numbers as before(look in attachment 2) but it seems that prior to 2000 this was quite true(though the same could be said for the passing game)….but since 2000, the importance of the run game seems to have been greatly diminished.

“In the modern NFL you need to be able to stop the pass”
Fairly True.  9 of 13 teams had a Top 10 D for Passer Rate….but that figure was 10 for pre-2000 teams so the theory should not be limited to the “modern” NFL.


I won’t go through any more common theories as there really isn’t any perceivable pattern to suggest a formula for victory from the modern data.

There is however one area that clearly has commonality with modern SB winning teams.  The Star QB.

Starting QBs of winning SB teams…
(As mentioned earlier, I am calling Flacco a “?” QB.  Therefore I will assume 2011-1987 as a 25 year sample for the purpose of this study).

Last 25 seasons
20 Stars
5 non-stars

Last 20 seasons
18 Stars
2 non-stars

Unfortunately for us Bills fans, it seems the only common factor in the modern era for winning a SB is having a Star QB.  Fingers crossed we find one this off-season.

Good post.

[Dibs]

silvermike, on 05 February 2013 - 04:16 PM, said:

Brad Johnson would be one.  Maybe Mark Rypien would go in that category, too.  Jim McMahon?  Phil Simms?  Joe Theisman?

Yep.....Brad Johnson & Trent Dilfer were the 2 in the last 21 years.
The 3 in the 5 previous were....Mark Rypien, Jeff Hostetler & Doug Williams
There were another 4 in the 5 year before that Phil Simms, Jim McMahon, Jim Plunkett, Joe Theismann,

Winning with a journeyman QB who has a good season or two seemed to be very common.
7 in a 10 year span.....2 in the following 20 years.

[Dibs]

jeremy2020, on 05 February 2013 - 05:39 PM, said:

What about making the playoffs?

Most of us Bills fans have forgotten how technically easy it is to get into the playoffs.  The odds have changed over the years due to number of teams/playoff spots....but I've done a calculation & the average for the purposes of this exercise is very close to the current chance to make the playoffs 37.5%  

Using 37.5% as the average.....
Last 21 years:

Playoff appearances for Star SB winning QBs (In years they started 11+ games).

Troy Aikman:  8 in 12 years  (66.7%)
Steve Young:  7 in 8 years    (87.5%)
Brett Favre:  12 in 19 years   (63.2%)
John Elway:  9 in 15 years   (60%)
Kurt Warner: 5 in 6 years     (83.3%)
Tom Brady:  10 in 11 years   (90.9%)
Roethlisberger: 6 in 9 years  (66.7%)
P. Manning:  12 in 14 years  (85.7%)
E. Manning:  5 in 8 years (62.5%)
Drew Brees:  5 in 11 years   (45.5%)
Aaron Rogers: 4 in 5 years  (80%)

Flacco:  5 in 5 years (100%)


Playoff appearances for non-star SB winning QBs (In years they started 11+ games).

Trent Dilfer:   1 in 5 years (20%)
Brad Johnson: 4 in 7 years (57.1%)


The results above were pretty much expected IMO.

jeremy2020, on 05 February 2013 - 05:39 PM, said:

I'd also be interested in how many of these QBs performed well in in their first 3 years with a poor running game and/or a poor defense.

Though I don't really see a relevance to it.....I looked through while I did the previous stats....

Of the 11 different Star QBs who won the SB in the last 20 years, 6 of them performed well in their first 3 years without a good(top 10) running game or defense.
These were P. Manning, Elway, Favre, Warner, Brady & Rodgers(his was 4th year+ as he was backup for first 3 seasons).

[RealityCheck]

Who is our QB again?

[JÂy RÛßeÒ]

Nice work, Dibs.  Have you applied for a position in the Bills' new football analytics department yet?

[Cash]

I don't have much to add, but I do want to thank Dibs for all the effort.  Very interesting read and much appreciated!

[mrags]

Thanks for that Dibs. I've actually been wanting to do this very same thing for quite some time now. Work and life just won't allow it. Much appreciated.

I'd add a few things here, I haven't seen the lists yet but going off everyone's comments about the SB winning QBs and being elite. It's a tough one to do. And I know ill get flamed for this as I have in the past, but you can't go by the player as they stand now and put them into either argument of that list. My biggest examples in the past have been Brady in his first 2, maybe 3 SuperBowls and Big Ben in his first were definitely not Elite QBs at the time these games happened. Of course it won't tip the scales massively in either direction, but there's more names out there that could add to the other side of the argument and make it less reliant on QB play.

I always have and always will believe "Defense wins Championships" because if you have a better defense you can stop anybody.

Maybe your next experiment should be of the teams playing in the SBs for the last 25 or so years, of the winning teams, who had the better defensive game? Who had the better statistics as QB? Who had more rushing yards?

I'd bet the better defensive team wins the majority of that time.

[l< j]

This is great and just solidly researched and argued. Kerry Byrne at SI tries to do the same with a slightly different (smaller) set of metrics that might be of interest. (One interesting point he makes is that "the passing game is LESS important now than it was in the 1940s, 1950s or 1960s.")

http://sportsillustr...l-quarterbacks/

But I like the points here better. Thanks.

kj

[Dibs]

mrags, on 06 February 2013 - 09:53 AM, said:

Thanks for that Dibs. I've actually been wanting to do this very same thing for quite some time now. Work and life just won't allow it. Much appreciated.

I'd add a few things here, I haven't seen the lists yet but going off everyone's comments about the SB winning QBs and being elite. It's a tough one to do. And I know ill get flamed for this as I have in the past, but you can't go by the player as they stand now and put them into either argument of that list. My biggest examples in the past have been Brady in his first 2, maybe 3 SuperBowls and Big Ben in his first were definitely not Elite QBs at the time these games happened.

Thanks mrags(and everyone).  I do these studies for my own personal interest.....but it is still nice to get a pat on the back for it.

In relation to your "elite QB" comments.....I tend to look at it that a QB with the "it factor" has that special something even before they have accumulated enough experience to show themselves as being an "elite QB".  Brady isn't considered elite because he won several SBs early in his career....he is considered elite because of his consistent elite level play.  He obviously has the "it factor".
Using Roethlisberger as an example....I was quite dubious of his abilities after his first 2 seasons.  He had a great run game & D to work with and was obviously being reigned in and limited in how much he contributed to the team.  I had these reservations about him until I saw him play in his 3rd(?) season.  In the game that I watched he did some absolutely amazing things which kept drives alive and helped earn them the W.

"Elite QB" is obviously unquantifiable.  Early Elway shows us that it is not just stats.  Everybody who saw him play knew he was special, even though his statistics were average at best.  It isn't just SB wins....Kelly & Marino are testament to that.  It is having the "it factor".  In the list of the 11 Star SB QBs(post #10), the only one that I personally have reservations about is Eli Manning.  For my mind he is too erratic and streaky to be considered an "elite QB".  I have not however seen a great deal of his play.....his potential was recognized pre-draft as great....he led his team to 2 SB victories, one of which was with the worst running game in the league and a bottom 10 Defense....and he is widely considered "elite"....so what do I know lol.

mrags, on 06 February 2013 - 09:53 AM, said:

.......but there's more names out there that could add to the other side of the argument and make it less reliant on QB play.

I'm not sure what you meant here.

mrags, on 06 February 2013 - 09:53 AM, said:

I always have and always will believe "Defense wins Championships" because if you have a better defense you can stop anybody.

Though this used to be the case(nearly all SB winning teams had a top 10 if not top 5 D prior to 2000), it is no longer a truism.
Having a strong D is certainly still a key component for ones chances to win the SB but it is no longer a "must have".

SB winning ranked Ds....
2012 Ravens 17/13
2011 Giants  27/25
2009 Saints  25/20
2007 Giants  7/17
2006 Colts   21/23

It however certainly seems to be a "must have" if you don't have an elite QB....the 2 SBs won by non-elite QBs in the last 20years were won with a #1 & #2 ranked D.

mrags, on 06 February 2013 - 09:53 AM, said:

Maybe your next experiment should be of the teams playing in the SBs for the last 25 or so years, of the winning teams, who had the better defensive game? Who had the better statistics as QB? Who had more rushing yards?

I'd bet the better defensive team wins the majority of that time.

Since every single time in the past 20 year an elite QB has defeated a non-elite QB.....and half the time the losing team has an elite QB as well, putting together an elite defense or running game is obviously a distant secondary consideration when thinking how one is to build a team in order to win a SB.

For interest sake though, here is a list of the SB match ups since 2000 with their defensive rankings. (Am only going back to 2000 as the stats above clearly show a changing point in the NFL at that time).
(Highest ranked D listed first....red for winner)....(Star QB underlined....except 2012)

2012: 49ers(4) v Ravens(17)
2011: Giants(27) v Patriots(31)
2010: Steelers(2) v Packers(5)
2009: Colts(18) v Saints(25)
2008: Steelers(1) v Cardinals(19)
2007: Patriots(4) v Giants(7)
2006: Bears(5) v Colts(21)
2005: Steelers(4) v Seahawks(17)
2004: Patriots(9) v Eagles(10)
2003: Patriots(7) v Panthers(8)
2002: Buccaneers(1) v Raiders(11)
2001: Rams(3) v Patriots(24)
2000: Ravens(2) v Giants(5)

Results are basically even.  7 wins for the better D....6 wins for the worse D.
In games where the D rank difference is 10 or more(6 games).....3 wins each.

You likely hate the concept but.....the NFL has changed to the point where it can no longer be said that Defense wins Championships.

[#34fan]

Fantastic research.

However, it was one hell of alot of work to tell us what we already knew. -We'd better find a QB to have any shot at becoming a championship franchise.   -Easier said than done, this year anyway.

In the '13 draft, the Defensive "cornerstone" players are there for the taking. -Franchise Quarterbacks not so much.

Why not take what's given rather than trying to reach for something that clearly is not there?

[Tcali]

Dibs, on 06 February 2013 - 01:45 AM, said:

Yep.....Brad Johnson & Trent Dilfer were the 2 in the last 21 years.
The 3 in the 5 previous were....Mark Rypien, Jeff Hostetler & Doug Williams
There were another 4 in the 5 year before that Phil Simms, Jim McMahon, Jim Plunkett, Joe Theismann,

Winning with a journeyman QB who has a good season or two seemed to be very common.
7 in a 10 year span.....2 in the following 20 years.

theisman and plunkett were stars. simms was very close to being one. so disagree there--but it only proves your point more.

[mrags]

Thanks again Dibs. Your right, I prolly won't agree with it. Well, not completely. But very nice job and I do agree with quite a bit.

[RuntheDamnBall]

Some good work here.   I do wonder a bit about "stars" and our definition of them, though.

First, it's all built on past repertoire and not a pure indicator of future success with injuries and aging curve factored in.

Second, there are players one could characterize as stars like Wes Welker, where the presence of other stars (Brady) "made" the guy into a star.  As in, he might not be one on his own merits with another team.  Also, having stars at positions like QB or MLB is more important than having a star CB (just in terms of how the player alters the landscape around him).