Interceptions- Where Are They Thrown?

I've been racking my brain trying to find a pattern as to where intercepted passes are most commonly thrown. Are they more often thrown to one side of the field as opposed to the other, more often down the middle, to the sidelines? A lot of questions, all meant to identify if a quarterback's target and location patterns could tell us more about why he throws so many interceptions.

My first theory: Quarterbacks who throw far more towards one side of the field than to the other become predictable, and thus, intercepted more often.

To examine this theory, I looked at the 2007 statistics of all quarterbacks and broke up the distribution of their pass attempts. Luckily, ESPN.com provides a breakdown of left sideline, left, right sideline, right, and middle. More frequently intercepted quarterbacks should have a more lopsided distribution from one side to the other. However, nothing really came of this. Regardless of whether the quarterback is extremely lopsided or even-handed to both sides of the field, the equity of distribution has no relationship to his interception rate.

For the record, the most even handed QB's are:
Philip Rivers, Drew Brees, Peyton Manning, Carson Palmer, Derek Anderson.
and, in case you're interested, the guys with a favorite side are:
Matt Hasselbeck (23% more to the right), Brett Favre (19% more to the right), Rex Grossman and Marc Bulger (18% each more to the right), and Tarvaris Jackson (13% more to the right).

My second theory: QB's who throw more passes to the middle of the field will be picked off less.

A QB who throws to the middle of the field will theoretically force the defensive backs away from playing only to one side and will require them to spread out. Furthermore, a quarterback who throws a lot to the middle likely has great confidence (either in himself or in his receivers), as opposed to a quarterback who dumps off or throws to the side a lot.

This theory yielded mild support. There is a significant, albeit modest relationship between the percentage of times a QB throws over the middle and the amount of times he's intercepted. Those who do it more often are actually picked off less, although the difference is not tremendous.

The five biggest up-the-middle gunslingers:
Brett Favre (19%), Jeff Garcia (17%), Kyle Boller (17%), Drew Brees (16%), Damon Huard (16%)

The five who rarely go up the middle:

Rex Grossman (9%), Carson Palmer (9%), Tarvaris Jackson (10%), Jason Campbell (10%), Marc Bulger (10%)

My third theory: QB's who throw more to the sidelines are more likely to get picked off.

This theory was more what researchers call an exploratory research question, or less euphemistically, a hunch only modestly based on evidence. I figured, hey, maybe if those who go up the middle are picked off less, those who go straight for the sidelines are picked off more. Remember, there are other areas aside from up the middle and the sideline. The sideline is the three-yard area near the sideline. The middle is inside the hashmarks. so, anywhere outside that range qualifies as left non-sideline or right non-sideline.

Anyway, this theory came up empty. There is no relationship between the percentage of sideline passes and interception rate.



Finally, I started thinking...I'm still sold on the idea that predictability will lead to higher interception rate. But it seems that interceptions have no relationship to how predictable a QB is related to where he'll throw the ball to. That begged the question- What if a QB is predicatble about who he throws the ball to?

Whether a receiver is lined up in the slot, wide right, wide left, some QB's simply have a favorite target, regardless of where he is lined up. Sometimes, though, the quarterback likes the target so much, that the defense can anticipate the two hooking up, leading to higher interception rates.

Eureka!

Final theory: The higher proportion of passes going to the team's leading receiver, the higher that quarterback's interception rate will be.

This theory checks out. There is a robust 0.36 correlation between the percentage of passes thrown to a team's leading receiver and the quarterback's interception rate.

Last year, four receivers were thrown more than 30% of their team's passes. Those men were Baltimore's Derrick Mason, Cincy's TJ Houshmandzadeh, Denver's Brandon Marshall, and KC's Tony Gonzalez. The QB's on three of those teams were veritable pick machines, with Kyle Boller, Carson Palmer, and Damon Huard each having an interception rate of over 3.5%, which is very high.

So, it appears that an over-reliance on a favorite target is a large hindrance to an offense. This should be kept in mind when evaluating how good a QB is at spreading around the ball, and thus, the defense.

Is Week One a Bellwether for the Rest of the Season? Part II

In Part II of this series, I look at whether or not you can tell how a running back will do on the year based on his Week One performance. Last week, we discovered that you can project much of a quarterback's season based on his opening week. Is the same true for running backs?

In a word, no. A running back's yardage opening day only has a 0.39 correlation with his per game average on the year. This correlation is not negligible, but it indicates that simply looking at a running back's performance opening day will not tell you all that much about his performance over the course of the year. Quarterbacks' QB rating, by contrast, has a 0.71 correlation with their overall year performance.

Another way to look at how unreliable opening week numbers are is by examining the degree to which that first game's numbers deviate from the average. On average, running backs' first game performances last year were 20.8 yards away from what their per game average ended up being. Some noteworthy examples: LaDainian Tomlinson- Week 1: 25 yards. Season average: 92.1 yards. Joseph Addai- Week 1: 118 yards. Season average: 71.5 yards. Jamal Lewis- Week 1: 62 yards. Season average: 89.6 yards.

What accounts for this phenomenon? Running back carries are far more vulnerable to strategy changes than are quarterback stats. A new running back could start performing well, and you don't have to bench the starter in order to make room. You simply give the new guy more carries. If a running back gets a minor injury, another running back can give him rest time. If a running back is slumping, his carries may decrease for a couple of games, only to be increased again later. In other words, the life of a running back has many ebbs and flows throughout the course of the year, whereas statistics demonstrate that a quarterback's season is more consistent.

Is Week One a Bellwether for the Rest of the Season? Part I

Welcome to a multi-part discussion that will take place this week on the importance of Week 1 in projecting a player's performance for the entire year. Today's topic: Quarterbacks.

First off, how much of an indicator is a quarterback's Week 1 performance with how he'll do the rest of the year? Well, actually, it says quite a lot. Over the past 3 seasons, a quarterback's opening day QB Rating has a very robust 0.71 correlation with his season rating. Therefore, a quarterback's opening day performance will generally go a long way in telling you what kind of year he'll have. This rule is not written in stone- Donovan McNabb and Brett Favre each had QB ratings over 89.0 last year after having opening day performances below 61.0. These exceptions aside, however, the numbers are compelling.

How about if a quarterback has an unbelievable opening day? Should you expect those unbelievable numbers to keep up? Simply put, no. More quarterback ratings of 100.00 or over are achieved in Week 1 of a season than in any other week. Overall, Week 1 is the best quarterbacking week of the year. Thus, outstanding opening numbers will almost assuredly not keep up their pace. In the past 3 years, quarterbacks who had a 100+ QB rating on Opening Week finished the year with a QB rating of over 90.0 only 50% of the time. However, while those quarterbacks' numbers did go down, they didn't usually drop too far. A quarterback who starts the year with a 100+ QB rating has a 79% probability of finishing the year with a top 10 QB rating.

As for bad opening day performances, the news is not good. Last year alone, seven quarterbacks had opening day performances with a QB rating under 60.0. A grand total of one of them finished the season with a QB rating above 80.0, and the average QB rating amongst them was 75.2. Hardly the stuff of legends. The names of those luminaries, Favre excluded? Matt Leinart, JP Losman, Rex Grossman, Damon Huard, Vince Young, Jason Campbell.

Now, quarterbacks do improve over their careers. They learn the systems and develop as players. However, the noticeable change in a quarterback's performance comes during the next season, not over the course of a season. Analysts may say otherwise, but the numbers don't lie.

So, whose opening day may be a signal of things to come?

First, the 100+ club for Week 1:
Ben Roethlisberger= 147.0
Jay Cutler= 137.3
Donovan McNabb= 131.0
Brett Favre= 125.9
Phillip Rivers= 125.1
Drew Brees= 124.9
Matt Cassel= 116.0
Aaron Rodgers= 115.5
JaMarcus Russell= 111.1
Tony Romo= 103.6
Jon Kitna= 103.1

Keep in mind the defenses that each quarterback played against, but the numbers still tell us something.

And those unfortunate souls that had less than 60.0 QB rating this week...

Tarvaris Jackson= 59.0
Matt Hasselbeck= 53.9
Vince Young= 45.6
Carson Palmer= 35.3

That number is actually a bit too typical of Jackson. He is the primary reason that I didn't pick Minnesota as a true contender this year. You simply cannot have that much uncertainty at quarterback.

Well, let's see how the rest of the season plays out. Next post, we'll cover running backs...

10 Unsung Players who Could Impact Their New Teams

Everybody's been talking about Favre, Jason Taylor, Jeremy Shockey and the like changing teams and having potential impacts in 2008. But there are a lot of other players who changed teams this year who are getting the same attention. Some of them could have huge impacts on their teams and influence the playoff picture. Based on the principles of Football Know-it-All, here are ten men flying under the radar who could help their teams this year:

1. Shaun Rogers, NT, Cleveland Browns- With Cleveland’s DLI a meager 20.1 last year, this mammoth All-Pro end-turned-nose tackle could be just what the Browns ordered in their conversion to a 3-4. With help from fellow Browns newbie Corey Williams at left end, Rogers could anchor a D-Line that propels the Cleveland defense to heights only its offense reached last year. Rogers could very well hold the key to an AFC North title for Cleveland.

2. Kawika Mitchell, OLB, Buffalo Bills- He is a tackling machine, and with MLB now a more solid position in Buffalo with Paul Posluszny’s return, Mitchell’s presence could help the Bills reduce the amount of Side Yards that elite runners usually run up against them.

3. Antwan Odom, DE, Cincinnati Bengals- Observers have been waiting for the Bengals defense to develop under supposed guru Marvin Lewis. After 5 years, Lewis may finally have a pass rusher to at least make his defense respectable. Odom racked up 8 sacks last year and pressured the quarterback frequently. With rookie LB Keith Rivers, the Bengals may not have an elite defense, but perhaps they can move to the middle of the pack.

4. Jerry Porter, WR, Jacksonville Jaguars- He may have the reputation of a malcontent, but especially with the underachieving Matt Jones gone, it is crucial for David Garrard to have a big target. Porter averaged 16.1 Yards/Catch last year. As mentioned in previous posts, Long Pass Plays have a 0.75 correlation with winning percentage. Receivers than can spread the field help the offense substantially, and the Jaguars may have gotten a bargain in Porter.

5. Chris Myers, C, Houston Texans- Myers centered a Denver O-Line that is famous for making big holes for running backs, and that could come in handy for rookie Steve Slaton and vet Ahman Green.

6. Gibril Wilson, SS, Oakland Raiders- One of the leaders in the NFL among DB’s in solo tackles. This means that Wilson is excellent at taking down ball carriers in the flat and not allowing them to gain extra yards while help arrives. In case the linebacker corps does not improve as projected, Wilson’s tackling ability will be what keeps the Raiders from giving up as many big plays as they did last year.

7. Jeff Faine, C, Tampa Bay Buccaneers- Faine was a highly sought-after center from New Orleans, and an emerging Ernest Graham will thank him for coming to Tampa. However, Faine’s strength is in picking up the blitzes up the middle, and that skill will be especially crucial with a 38-year-old Jeff Garcia in the pocket.

8. Muhsin Muhammed, WR, Carolina Panthers- Bringing Muhammed back returns that possession receiver to pair with Steve Smith and really opens up the passing game. Also, keep an eye on 2nd year man Dwayne Jarrett, the lanky big-play wideout from USC who got off to a slow start in 2007 but may emerge.

9. Randall Gay, CB, New Orleans Saints- A lot of the attention went to the acquisitions of Jonathan Vilma and Jeremy Shockey, but Gay may be just as important. In the past two years, Gay only allowed a 100-yard receiving game twice. With Mike McKenzie aging, Gay may be covering #1 receivers before long.

10. Justin Smith, DE, San Francisco 49ers- Smith can get to the quarterback, and he also gets a large amount of tackles on run plays. He is so fast, he can move laterally and seriously reduce the Side Yards of many running backs. However, he can just as easily converge on the middle to clog up the guard and tackle running lanes. Also capable of playing OLB, Smith could almost single-handedly bring the Niners’ defense into respectability.

16 Predictions for 2008

*I love predictions. They are often so wrong, and you get to look back on them after the season and mock those silly enough to make them. That said, I invite you to keep this post and remind me in February how I did. If I get more than 10 right, I deserve some sort of prize. If not, I guess I deserve whatever messages you'd like to send me.*

1. The Minnesota Vikings will have the NFL’s best defense. The Viking D-Line was already the best run-stopping group in the league last year. With the addition of Jared Allen, this unit will get to the quarterback more, as well. Football know-it-All, as you know, puts a lot of emphasis on D-Line performance. Chad Greenway and EJ Henderson are underappreciated linebackers who make tackles all over the field. The Vikes updated their safety corps with a solid Madieu Williams, and Cedric Griffin is entering prime age.

2. A new running power will emerge in Carolina- The Panthers drafted Jeff Otah in the first round, and he will allow Jordan Gross to move back to the left side. This can give Carolina a top-tier pair of tackles. Combine that with their other top draft pick Jonathan Stewart, who will now team with the criminally overlooked DeAngelo Williams. As discussed on this blog, tandem running is crucial in today’s NFL, and this looks to be a potent 1-2 punch.

3. The Buffalo Bills will have a resurgent defense. In the linebacker corps, Paul Pozlusny was having a great start to his rookie season, with 25 tackles in 2.5 games before a season-ending injury. Plus, Kawika Mitchell, who gathered over 50 tackles in the second half of 2007, joins the squad. Marcus Stroud should improve a run defense that was porous last year, plus take double-teams away from Aaron Schobel. If Leodis McKelvin plays even close to expectations, he can give the Bills a good second cornerback.

4. The fight for the NFC East crown will be between the Eagles and the Cowboys. The Cowboys add Pacman to a crew that already excelled in virtually all categories. Plus, their tandem of Marion Barber and Felix Jones may be really special. The Eagles, meantime, look to dominate on the lines, improving on both rushing and pass defense.

5. Philly’s lack of a big-play receiver gives the edge to the Cowboys. This dearth of a playmaker to stretch the field has been evident for many years, with the exception of TO. There is still nobody there to help out, unless undersized rookie DeSean Jackson can become an instant phenom.

6. David Garrard will solidify himself as an elite quarterback. With a big-play target upgrade in Jerry Porter in for Matt Jones, Garrard will have an extra weapon. Plus, Marcedes Lewis is entering his third year, a year where receivers and tight ends often significantly improve statistically. Plus, Garrard’s key stats, completion percentage and TD/INT ratio demonstrate discipline that often leads to maturation rather than regression.

7. Brett Favre is in for a rude awakening. Favre owed a good portion of his resurgent year to a couple of factors. First, the receiving corps led the league in YAC and had very few drops. Also, the offensive line was superb, creating a fantastic running game and letting Favre hit the turf a conference-low 15 times. The Jets are weak in all of those categories. The offensive line is not as upgraded as some think. Allan Faneca turns 32 this year, and Damien Woody is not a natural tackle. Plus, the Jets receivers won’t bail Favre out as much as Jennings and Driver did.

8. The Seahawks will get beaten this year by…jet lag? Mike Sando at ESPN.com once mentioned how much more difficult it is for West Coast teams to travel east than vice-versa. Statistical evidence backs that up. Over the past five years, West Coast teams have played significantly worse when travelling to the east coast than East Coast teams heading west. Seattle this year has the bad fortune of playing both the AFC and NFC East division. Their winning percentage in such games is dismally below their norm, and they now have four of them to play this year. They could be a long-shot to finish .500. Hey, they may still win their division.

9. The New Orleans Saints will return to NFC contention. One strength that made New Orleans so good in 2006 was the mileage that they got out of Reggie Bush with Deuce McAllister taking the majority of rushing attempts. McAllister’s injury force Bush into a more prominent rushing role and exposed him as a tandem runner at best. He can settle back into that role this year. Plus, with Jeremy Shockey added to the offense, Bush will get even more open looks in the slot. The Saints’ offense could be amazing. Meanwhile, the Saints got to use a top-10 pick on DT Sedrick Ellis, who should solidify the run defense. Jonathan Vilma is a statistician’s dream, as he devours tackles and now anchors the MLB position. This Saints team is a real threat.

10. A youth movement will catapult Pittsburgh- The Stillers made two draft picks that play right into Football Know-it-All’s idea of offense. They drafted a top running back, Rashard Mendenhall, to tandem with Willie Parker. Justin Hartwig taking over at center should also offset the loss of Faneca at guard. Plus, they drafted big-play target receiver Limas Sweed, who may be the most promising rookie receiver. If these two rookies start quick, this could be a huge year for Ben Roethlisberger.

11. If the Jets offense does perform, an unlikely candidate will be the savior. Jets fans think that their running game hinges on Thomas Jones, but Jesse Chatman will be a good partner, one who may even take the starting job. Last year, Chatman was top 10 in both Side Yards and 2nd half Yards. If Favre can be convinced to let the running game take a larger role late in close games, Chatman should make the Jets glad they signed him.

12. The Bears will be fighting for the #1 draft pick. There is every reason to believe that this team is heading farther down their spiral. Their starting receivers are now slated as Devin Hester and Brandon Lloyd, neither of whom has ever shown any big play ability. The below-league-average running backs show no signs of improvement, and their QB situation is still a mess. On defense, the Bears were far worse than their points allowed indicated last year. Their yards allowed were over 355/game, but 33 takeaways saved them from giving up more points.

13. The Atlanta Falcons may beat them to it. If Matt Ryan will start this year, he has the fight of a lifetime on his hands. His line is positively sadistic, letting their varied quarterbacks hit the ground more than 50 times last year. That will also make it hard for Michael Turner to become a lone feature back, since he’s only succeeded before in San Diego’s running-back-friendly scheme. The D-Line is also the same, meaning very bad.

14. If a superstar is born out of that sorry scene, it will be Roddy White. White had 82 receptions last year for over 1,200 yards (top 10 in the NFL). In his three years in the league, White has averaged over 15 yards per reception, and has done while his QB’s were as varied as Michael Vick, Byron Leftwich, Joey Harrington and Chris Redman. Now at the age of 26, White appears on the verge of his prime. Rookie qb’s are known to latch onto and keep returning to a feature wideout, and that should be White for Ryan. 2008 could be the year the fantasy owners take notice of him.

15. A rising special teams unit will make Minnesota even more dangerous. When Adrian Peterson went down last year, Aundrae Allison took over as the kick returner and revealed that he is one of the finest in football. He virtually always returns the ball more than 20 yards and average 28.7 yards per return, more than 3 yards over league average per possession. In other words, Allison gets a team more extra yards per possession than most running backs.

16. Sleeper Pick- The Carolina Panthers. Their quarterback play was why their offense was so wretched. Jake Delhomme, unknown to even himself, had need ulnar surgery for three years and finally got it after his elbow popped last year. He has looked better in the preseason than he has in years. Combine that with a newly revamped running game that was already above average and the reunion of Steve Smith and Muhsin Muhammed, and you have a revitalized offense. The O-Line has been upgraded as well, as I note in Prediction 2. A high-tackling linebacker, Landon Johnson has been added, as well as depth on my favorite unit, the D-Line. The Panthers and the Saints will both be back this year, and one of them may end up winning the NFC.

Projecting Linebacker Units for 2008 Season

The amount of tackles per starting linebacker that a defense has is a significant predictor of a team's defensive ranking. Tackles Per Starting Linebacker (TPSL) is therefore a stat worth looking at. One should note, however, that DLI (Defensive Line Index, as discussed in previous posts) is a stronger predictor of defensive performance. In other words, better to have a great line than a great linebacking corps, generally speaking. The correlation between TPSL and Yards Allowed Per Game is 0.34, whereas DLI's correlation with Yards Allowed Per Game is 0.52.

Good news for prognosticators is that a linebacker's tackles correlate very well with each other from year to year. Therefore, the adjustments made to project last year's tackles onto this year's tackles are simply age and previous injuries. Furthermore, I looked at past first-round picks and saw how their tackle rate converted to NFl numbers in their first year. There are certainly other factors, but these three adjustments should provide a good estimate. The big changes, therefore, will come from teams who made major acquisitions, draft picks, or from teams whose linebackers are entering their prime years.

Some teams have made major upgrades and should improve their defense considerably as a result. Other teams have lost personal and may backslide.

Thus, with those adjustments made, my projection for Tackles Per Starting Linebacker for 2008 are:

Carolina Panthers- 111
San Francisco 49ers- 106
Buffalo Bills- 103
Minnesota Vikings- 100
Chicago Bears- 100
Detroit Lions- 97
New Orleans Saints- 96
New England Patriots- 96
Atlanta Falcons- 93
Houston Texans- 92
Oakland Raiders- 92
Tampa Bay Buccaneers- 91
Baltimore Ravens- 91
Seattle Seahawks- 90
Washington Redskins- 90
Indianapolis Colts- 89
San Diego Chargers- 87
Pittsburgh steelers- 87
Tennessee Titans- 86
Philadelphia Eagles- 85
New York Jets- 84
Denver Broncos- 80
Kansas City Chiefs- 76
Dallas Cowboys- 75
New York Giants- 75
Arizona Cardinals- 73
Cleveland Browns- 70
St. Louis Rams- 68
Miami Dolphins- 65
Cincinnati Bengals- 64
Jacksonville Jaguars- 62

On Thursday, I'll post my 16 Predictions for 2008.

Does "Clutch" Exist in Quarterbacks?

It has been the subject of debate for a long time- Does clutch actually exist, or is it a figment of our imagination? Famous baseball stats guru Bill James is famous for asserting that clutch hitting is all fiction. The statistics, however, have not been so completely supportive of that statement.

One thing is certain, and that is that people often see "clutch" and "choke" performers where they do not exist. Alex Rodriguez, for instance, has borne the reputation of a choke artist, even though altogether, his late-inning, close-game numbers are just as good as his overall numbers.

First of all, how do we even define, Clutch? Well, I felt that a pretty apt definition of a clutch situation is the 4th Quarter, with the two teams within 7 points of each other. These are the late-game, high-pressure situations. Only quarterbacks are being tested. The test as to whether a quarterback is clutch is whether his stats are statistically so different from his normal stats that random chance can be ruled out. The stat being used is quarterback rating. As discussed in an earlier post, QB rating has gotten unfairly criticized bvut is actually a very valid measure of QB performance.

Ok, with the tests done, guess what?...Clutch does exist. Whereas the majority of quarterbacks play at the same level in normal situations as they do in clutch situations, there are 11 quarterbacks who are dramatically different players in crunch time.

Not only that, but clutch performance seems to be a persistent trait. For the players affected by clutch, I went back at previous years of their careers and found that clutch situations consistently produced different results in them than normal situations. Another, somewhat surprising result- Vulnerability to clutch situations does not appear to be something that guys get over. Those who had it early on seem to always play differently in the clutch.

Before we get to the players affected by clutch situations, a word about the others. Most quarterbacks, it turns out, are not affected by clutch situations. Peyton Manning, Tom Brady, and Tony Romo, to name three, play at a very high level in both normal situations and clutch situations. On the, uh, other end of the spectrum, the same is usually found. Most mediocre quarterbacks remain that way in the clutch. They don't become good, nor do they shrink and become terrible.

Ok, now for the quarterbacks who play differently in clutch situations. Of the eleven, 8 of the quarterbacks play significantly worse in the clutch than they do normally. The other 3 quarterbacks actually rise to the occasion and play substantially better in the clutch than they do normally.

So, without further ado, these are the three quarterbacks who perform better in the clutch than they do normally:

------------------------Overall QB Rating Clutch QB Rating
Matt Hasselbeck 91.4 -------------99.9
Kurt Warner 89.8-----------------------96.5
Sage Rosenfels 84.8-----------------------116.9

I know, right? Sage rosenfels, who woulda thunk it? Now, I know what some of you are thinking. Rosenfels only played half a year, and these numbers are just a fluke. Well, not necessarily. His sample size is smaller than others, but it is still large enough to draw a statistical conclusion, given the disparity between the two QB ratings. Also, a review of Rosenfels's (admittedly brief career) demonstrate that year after year, he performs significantly better in clutch situations than he does overall. He has done this for so many years that, even though the sample size is always modest, it's difficult to ignore.

As for Hasselbeck, he is excellent virtually all the time and becomes even better in the clutch. This pattern, too, has repeated itself throughout Hasselbeck's career. Warner had an underappreciated splendid season last year. His one real bugaboo, which Qb rating does not cover, is a propensity to fumble.

So, who would I want as my quarterback with the game on the line? Tom Brady or Peyton Manning. as I noted earlier, most quarterbacks are the same in the clutch as they are normally. Manning and Brady are two such quarterbacks. In the clutch, they are their normal brilliant selves and are still better than even Hasselbeck, who rises to the occasion.

Below are the eight quarterbacks who, for whatever reason, become shells of themselves when the score is close late in the game:


---------------------------Overall QB Rating Clutch QB Rating
Brett Favre 95.7----------------82.0
*Drew Brees 89.4-----------------------76.8
Jay Cutler 88.1-------------------68.7
Carson Palmer 86.7--------------------63.7
Chad Pennington 86.1--------------------58.3
Jon Kitna 80.9--------------------73.8
Jason Campbell 77.6--------------------66.6
Marc Bulger 70.3--------------------35.6

That last number is not a typo. Bulger's numbers in clutch situations were that bad this year. He has not always been as bad a clutch performer, but a review of his career demonstrates that he consistently performs worse in the clutch than he does normally.

Cutler and Campbell have less than two full seasons under their belt, so that have slightly more chance to change. I would not hold my breath though. Statistical analysis demonstrates that cllutch performance cannot really be learned by the time you get to the pros. By then, it's either in you or it's not.

Carson Palmer had one anomalous season a few years ago, in which his clutch performance was significantly better than his normal performance. However, that seemed to be one exception in a career that has seen consistenly bad clutch performance.

Drew Brees is really the enigma of the bunch. As I have said, clutch performance is normally an enduring trait, and everybody else on this list has demonstrated consistency in their clutch performance throughout their career. Brees is the exception to this rule, which is why I'm giving him the asterisk. Some years, he's fine, and in some years, he's even better in the clutch. Oh well, not even science can explain everything. For all other players on this list, however, their 2007 numbers are indicative of a career-long trend.

I am also aware that over the past decade or so, the fastest growing religion in the United States has been the Worshippers of Favre. I am not saying that he has not had a great career or that his records are not impressive. all I'm saying is that, 1. he's overrated, and 2. His legion of followers, especially in the media, have created a myth that when the chips are down, there is nobody like Favre. This belief is patently false.

Not only has Favre been less than stellar throughout his career in the clutch, but his clutch numbers are actually pretty woeful compared with other situations. i'm not even talking about his well-documented flops in the postseason, like the terrible game he played in this year's NFC championship, getting shut out of the endzone at home against the Falcons, or blowing a divisional game multiple times against the Eagles.

In Brett Favre's career, his clutch performance over a season has only been as good as his overall performance three times. Every other year it was worse. A whopping 10 times, the difference between his clutch numbers and his real numbers was significantly different, statistically speaking. Even in his glorious Super Bowl-winning year, his qB rating in the clutch was an unbelievable 21 points lower than his normal QB rating. seven times, his clutch QB rating was more than 15 points lower than his normal rating.

It is time to put the myth to rest. Brett Favre is not clutch. He never has been. Yes, he may have had some great late-game performance, but as the numbers demonstrate, those were the exception, not the rule. Ok, I think I've pre-empted the Favre defenses, but feel free to let loos anyway.

so now, the truth is out there. Clutch exists, and not everybody has it.

What Teams Will Have the Best Defensive Lines in 2008?

Exhaustive statistical research has indicated that the defensive line is the most important part of a defense, and it has the most predictive value in forecasting a team's defensive performance. With that in mind, I now look ahead to 2008 and forecast which teams will have the best defensive lines.

In order to make these forecasts, I took each team's Defensive Line Index (DLI) from 2007 and made adjustments based on personnel changes. (refer to the previous post for a more detailed description of DLI)

a. Points were added if a team was upgrading with a player who performed well in the criteria used for DLI.

b. Points were subtracted when a team lost a player high in DLI stats and replaced him with someone worse.

c. Points were added if a defensive line's average age moved closer to the 26-30 age range, which a random sampling demonstrated are the best years of a defensive lineman's career. In other words, teams got points for maturing. Teams that got a lot younger suffered.

d. Points were added on a sliding scale if a team drafted a defensive lineman in the first three rounds, with more points being awarded to higher draft picks.

Thus, The projected 2008 rankings of NFL defensive lines:


1. Minnesota Vikings- 42.6
2. Philadelphia Eagles- 40.7
3. San Diego Chargers- 39.9
4. Seattle Seahawks- 38.9
5. Tampa Bay Buccaneers- 37.9
6. New York Giants- 37.5
7. Dallas Cowboys- 37.0
8. Jacksonville Jaguars- 35.4
9. Green Bay Packers- 34.2
10. Arizona Cardinals- 33.7
11. Baltimore Ravens- 32.4
12. New England Patriots- 31.2
13. Miami Dolphins- 30.8
14. Detroit Lions- 30.1
15. San Francisco 49ers- 29.8
16. Chicago Bears- 28.3
17. Tennessee Titans- 27.1
18. New Orleans Saints- 26.8
19. Indianapolis Colts- 26.8
20. St. Louis Rams- 26.4
21. Washington Redskins- 25.9
22. New York Jets- 25.8
23. Buffalo Bills- 25.6
24. Houston Texans- 25.3
25. Denver Broncos- 25.2
26. Kansas City Chiefs- 24.8
27. Cleveland Browns- 23.1
28. Pittsburgh Steelers- 21.9
29. Carolina Panthers- 19.8
30. Atlanta Falcons- 18.4
31. Cincinnati Bengals- 16.8
32. Oakland Raiders- 15.4


**Now, keep in mind, that DLI is still a new metric, and I'm working on perfecting it. Furthermore, I'm also working on developing better predictive metrics to forecast future performance. Does that sound like hedging? Okay, I'm a little bit guilty. But hey, that's better than those prediction gurus that promise the moon. The bottom line, though, is that a number of reliable statistical measures went into this formula, and I feel good that, barring injuries, the numbers will shape up something like this.**

P.S. As for the Giants, this formula assumes that both Osi Umenyiora and Michael Strahan will be missing. If Strahan returns, their score should improve.

Defensive Line Rankings- A New Ranking System

As has been asserted here several times, line play in the NFL is vital. Many people have said for a long time that games are won and lost on the offensive and defensive lines. Well, it's rarely that simple, but it is true that how good your line is will certainly go a long way in predicting how good your team will be.

The Football Know It All has developed a system using three categories to determine the quality of a defensive line. Today, I will lay out the criteria and list the defensive line rankings of each team for 2007. In my next post, I will take into account acquisitions, key losses, aging, and draft picks to give a best estimate for how the defensive lines will perform in 2008.

The new defensive line rating will be called Defensive Line Index (DLI). The three categories that make up DLI are:

1. Sacks- I wish that QB pressures was an official stat, because I believe that it is a fairly more accurate assessment of how much pressure a D-line puts on a quarterback. However, sacks are still a good measure of how much pressure a D-Line exerts on quarterbacks. Furthermore, teams that sack the quarterback a lot give up fewer points and win more games. There is a 0.55 correlation between the number of wins a team has and its sack total. The top five teams in sacks in 2007 all made the playoffs, and the top 2 teams faced off against each other in the Super Bowl.

2. Rushing yards/carry in the first half- Why only look at rushing yards in the first half? Because in the first half, all offenses are trying to score points, whereas in the second half, teams with big leads simply run more and try to eat up time. This leads to lower yards/carry. Thus, teams that are often behind in the second half give up fewer yards/carry in the second half, because the run is more predictably coming. This phenomenon helps them put up better run defense numbers than they deserve. Incidentally, yards/carry allowed in the first half has a -0.31 correlation with wins, whereas overall yards/carry allowed only has a -0.08 correlation. Clearly, First Half Yard/carry is the better index.

3. Percentage of tackles by D-Line- The better the D-Line, the more often it will tackle the ball carrier. Lesser D-Lines miss more tackles and depend more on linebackers and defensive backs to tackle ball carriers. % D-Line Tackles had a 0.23 correlation with winning percentage.


Using these three criteria, I used a regression for each category relative to its relationship to win percentage and adjusted weights in order to account for the differing numerical values of different categories. Below are the results for the 2007 rankings for DLI. The average score was 27.4, with a standard deviation of 7.0:

1. New York Giants- 44.5
2. Seattle Seahawks- 38.9
3. Minnesota Vikings- 37.6
4. Dallas Cowboys- 37.0
5. San Diego Chargers- 35.9
6. Baltimore Ravens- 34.4
7. Tampa Bay Buccaneers- 33.4
8. Green Bay Packers- 32.2
9. Philadelphia Eagles- 31.7
10. New England Patriots- 31.2
11. Chicago Bears- 28.3
12. Tennessee Titans- 28.1
13. Washington- 27.9
14. New Orleans Saints- 27.8
15. Jacksonville Jaguars- 27.4
16. Kansas City Chiefs- 25.8
17. Houston Texans- 25.3
17. San Francisco 49ers- 25.3
19. Detroit Lions- 25.1
20. Pittsburgh Steelers- 24.9
21. Indianapolis Colts- 24.8
22. Arizona Cardinals- 24.7
23. Carolina Panthers- 23.8
24. St. Louis Rams- 23.4
25. Miami Dolphins- 22.8
26. Denver Broncos- 22.2
27. New York Jets- 20.8
28. Buffalo Bills- 20.6
29. Cleveland Browns- 20.1
30. Cincinnati Bengals- 16.8
31. Oakland Raiders- 15.4
32. Atlanta Falcons- 14.4

The DLI, by the way, has a very strong correlation of 0.50 with winning percentage and a correlation of -0.52 with points allowed. These correlations demonstrate that DLI has high predictive validity as a predictor of overall defensive and team performance.

How Running Back Tandems Help Win Games

One truth that has been uncovered by the Football Know it All is that great running backs really distinguish themselves in the 4th quarter. In the first three quarters of a game, there is less variance in yards per carry, but in the fourth quarter, elite running backs like LaDainian Tomlinson and Adrian Peterson really pull away from the pack.

The ability to run the ball late in the game is a key determinant not only in who are the great running backs. Yards per carry in the fourth quarter is also one of the strongest predictors of which team will win a Close Game, previously defined as a game decided by 8 points or less. There is another stat that is important to predict the outcome of games. Team yards per carry after the 20th carry of the game have a 0.32 correlation with the overall outcome of the game. A 0.32 correlation indicates that while there are certainly other factors, yards/carry after the 20th carry have a significant impact on the game's outcome.

Only 3 running backs averaged more than 20 carries per game last year, however. So, in order for a team to enjoy this advantage, they must have a potent running back tandem. Teams that have a second running back who had over 300 yards rushing and averaged over 4.0 yards per carry enjoyed the late-game running advantage and had the inside track on winning close games. Below is a list of #2 running backs who helped their teams by gaining 300 yards rushing and averaged 4.0 yards per carry. The backs in bold backed up running backs who also average 4.0 yards per carry:

Jerious Norwood- Atlanta Falcons
Fred Jackson- Buffalo Bills
DeAngelo Williams- Carolina Panthers
Travis Henry- Denver Broncos (released)
TJ Duckett- Detroit lions
Indianaplois Colts- Kenton Keith
Maurice Jones-Drew- Jacksonville Jaguars
Jesse Chatman- Miami Dolphins (now on Jets)
Chester Taylor- Minnesota Vikings
Sammy Morris- New England Patriots
Derrick Ward- New York Giants
Leon Washington- New York Jets
Correl Buckhalter- Philadelphia Eagles
Najeh Davenport- Pittsburgh Steelers
Michael Turner- San Diego Chargers (now on Falcons)
Maurice Morris- Seattle Seahawks
Chris Brown- Tennessee Titans (now with Texans)

As you can see, running back tandems have become the norm in the NFL. Moreover, it has become increasingly common for team to use a draft pick in the first three rounds on a running back, even when they already have an elite back on the roster. This trend demonstrates that smart teams are investing in running back pairs. Even in their rookie years, strong backs who work in tandem with another strong back have seen their teams increase their win totals by three or more games. Recent examples in just the last three years are:

1. New Orleans Saints
2. Minnesota Vikings
3. New York Jets
4. Tennessee Titans
5. Jacksonville Jaguars
6. Miami Dolphins

Expecting a rookie to be your lone feature back has not worked, however. Teams seem to have only benefited immediately with drafting a running back when there is another start back in place with which to team. Below are the high-drafted rookies who will join established stars in the backfield:

Darren McFadden- 4th overall pick- Oakland Raiders
Jonathan Stewart- 13th pick- Carolina Panthers (DeShaun Foster no longer around; will team with Williams)
Felix Jones- 22nd overall- Dallas Cowboys
Rashard Mendenhall- 23rd overall- Pittsburgh Steelers
Ray Rice- 55th overall- Baltimore Ravens (as a side note, having seen Rice a lot, I see a lot of upside teaming him with McGahee)
Jacob Hester- 69th overall- San Diego Chargers
Steve Slaton- 89th overall (will team with Chris Brown and Ahman Green)

Ranking NFL Quarterbacks- A More Accurate Passer Rating

Before I delve into adjusting passer rating, one issue needs to be cleared up. The QB rating system, as it stands now, is not that bad. True, there can be some adjustments, two of which I will make in this post, but overall, the QB rating fairly accurately measures a quarterback's performance.

I hear pundits and commentators laugh and sneer all the time about the rating, because it requires some convoluted formula that few people understand. Well, if they were to look at the actual criteria that make up the quarterback rating instead of loking for something to complain about, maybe they wouldn't always be so quick to complain. The QB rating in the NFL is based on four different categories:

1. Completion percentage
2. Touchdown rate
3. Interception rate
4. Yards per attempt

Are any of those four categories unreasonable? Are there better categories for getting to the core of what a quarterback is supposed to do? I dunno, these four categories seem pretty sensible to me. Plus, the reason that the formula is so complicated is because the four categories produce very different numbers that cannot simply be added up. Sophisticated weighted scales and formulas must be constructed in order to give all categories equal weight.

A final defense of QB rating before I make my moderate adjustments to it: Look at the top ten QB's in passer rating in 2007:

1. Tom Brady- 117.2
2. Ben Roethlisberger- 104.1
3. David Garrard- 102.2
4. Peyton Manning- 98.0
5. Tony Romo- 97.4
6. Brett Favre- 95.7
7. Jeff Garcia- 94.6
8. Matt Hasselbeck- 91.4
9. Donovan McNabb- 89.9
10. Kurt Warner- 89.8

Now, you may have some slight qualms with a name on the list here and there, but one would be hard-pressed to argue that this list was not a pretty good reflection of QB performance in 2007.

That said, there are a couple of adjustments that can be made to make the list a little more representative of QB performance. There are three factors I see that the QB does not control that get factored into QB rating:

1. Dropped passes by the receivers lower a quarterback's completion percentage.
2. Yards after catch by a receiver give the quarterback extra yards per attempt.
3. Sacks affect the quarterback in numerous ways.

As for sacks, some of them are his fault, while others are the offensive line's fault. I will try to figure out how to distinguish between these better, but as of now, I do not have a good strategy. Therefore, I will not address sacks now.

I will address YAC and Dropped Passes. I will factor out the YAC from Yards per Attempt and factor out Dropped Passes from Completion Percentage to come to what I called QB Rating Plus. Below are the 2007 rankings for QB Rating Plus:
Rating Rating Plus
BUF Edwards 70.4 69.4
MIA Lemon 71.0 70.6
NYJ Clemens 60.9 59.4
NE Brady 117.2 116.9
IND Manning 98.0 99.7
TEN Young 71.1 70.8
HOU Schaub 87.2 86.1
JAC Garrard 102.2 102.7
CLE Anderson 82.5 83.7
PIT Roethlisberger 104.1 104.0
BAL Boller 75.2 75.6
CIN Palmer 86.7 86.1
KC Huard 76.8 77.5
SD Rivers 82.4 81.1
OAK McCown 69.4 69.7
DEN Cutler 88.1 87.5
NYG Manning 73.9 76.8
DAL Romo 97.4 97.0
PHI McNabb 89.9 89.5
WAS Campbell 77.6 78.6
TB Garcia 94.6 96.1
NO Brees 89.4 90.1
ATL Harrington 77.2 78.2
CAR Testaverde 65.8 67.7
DET Kitna 80.9 81.3
CHI Griese 75.6 78.4
GB Favre 95.7 93.0
MIN Jackson 70.8 71.8
SEA Hasselbeck 91.4 90.4
SF Dilfer 55.1 58.6
ARI Warner 89.8 90.2
STL Bulger 70.3 71.8

As you can see, the differences are not tremendous, but for several quarterbacks, the changes alter the rating enough to changes their ranking in passer rating by several spots. The biggest winners under the new system are: Eli Manning, Peyton Manning, Jeff Garcia, Brian Griese, Mark Bulger, Vinny Testaverde

The biggest losers under the new system are: Brett Favre, Kellen Clemens, Philip Rivers.

During the 2008 season, I will be studying sacks. I believe that this will provide the best improvement yet in measuring quarterback performance.

Last Year's Close Games Can Help Predict This Season's Records

In one of the Football Know It All's earliest posts, I discussed how year-to-year records have a faily weak correlation. One of the reasons for this is parity in the league. Turth be told, aside from the elite five or six teams and the five or sixe teams residing at the the bottom of the NFL barrel, there is not a tremendous amount of variance between the rest of the NFL teams.

This relative equality in the NFL leads to a large amount of Close Games, defined as games decided by 8 points or less. In fact, over the past five seasons, 48%, or roughly half, of the NFL's regular season's games have been Close Games. What decides the outcomes of these close games? There are many factors, and in fact, in my next post, I will talk about the kind of running game that helps teams tremendously in close games.

One factor, however, is random events. When teams are within a few points of each other, a number of factors, from a gust of wind, a penalty, or a dropped pass can affect the outcome. Supporting this assertion is that the standard deviation between NFL teams is significantly greater for overall record than for record in Close Games. This figure implies that in Close Games, the difference in the teams' qualities starts becoming less important while the impact of chance creeps in.

These things make victims of even the best teams in football. For instance, the Colts have had a 79% winning percentage over the past five seasons. However, in Close Games in that time, their winning pecetnage has been only 63%. Now, 63% is still very good, but it is substantially worse than 79%. One thing that happens is that, when they let teams stay close late into a game, they allow for the possibility of random events having more of an impact. The mark of good teams, however, has been that they rarely allow inferior teams to close in and make it a game.

The thrust of all of these figures is that random events do play a role in the outcomes of close games. An interesting notion to consider is that almost all teams play between 7-9 Close Games per year. So, each year, the random events that occur in close games can have a small, yet noticeable, impact on a team's record. As we've mentioned before, random events even out over time. It has been demonstrated that a team that loses a disproportionate number of Close Games one year will often improve in its Close Games record, and often in its record overall.

Thus, teams that lost a disproportionate amount of Close Games in 2007 may, by the simple power of probabilities, see an improvement this year. To keep an eye on this trend, here are a collection of teams that lost a large percentage of Close Games last year:

Pittsburgh Steelers
San Diego Chargers
Indianapolis Colts
Minnesota Vikings
Seattle Seahawks

Some teams that fared better in Close Games than they did overall and thus may not fare as well this year include:

Carolina Panthers
Denver Broncos
Buffalo Bills
Tennessee Titans
San Francisco 49ers

Remember, though- Just because random events take more of a role in close games does not mean that they are the sole predictors. Far from it. Even in close games, the odds still overwhelmingly favor the teams that are designed to win such games. Next post, I'll discuss what type of running game is needed to improve a team's chances in close games.

Age and Defensive Performance

A quick post today...

Today we take on two pieces of conventional wisdom about how age affects defensive performance.

Conventional Wisdom #1: While you want veteran leadership, you don't wasn't a defense to be too old. After all, older players will get injured, and they get worn out by the end of the season.

The truth: Actually, there is a very strong correlation between age and a defense's performance. Namely, the older the defensive unit, the better the defense performs over the course of the entire season. This truth holds even when taking injuries into account. In fact, there is no relationship between how old a defensive player is and how many games he will miss in a season due to injury.

NFL defensive schemes are a complex thing to learn, and some gray hairs among the defensive unit are clearly a good thing, not a bad thing. So, if your team had upgraded on defense but you worry about the new acquisition's age, relax. The odds are likely that he will do more good than harm.

Conventional Wisdom #2: Penalties are a killer, and younger, less disciplined players are more likely to commit them.

The truth: Well, for starters, defensive penalties are not a killer. There is no relationship between the amount of penalties a defense commits and its defensive ranking. Nor is there a relationship between defensive penalties and the amount of points a defense allows. While a penalty in any particular game can obviously change the course of a game, over the course of a season, a defense will not be harmed if it is prone to penalties.

The second point to make about this addresses age and penalties. You often hear commentators calling penalties, "rookie mistakes." You also hear them mention how as a player matures, he will smarten up and stop making so many penalties. Well, according to statistics, this is nonsense. When it comes to committing a lot of penalties, it appears that players are who they are. There is zero statistical decline in a player's penalty rate as his career progresses. Furthermore, there is no relationship between years of NFL experience and the amount of penalties that a player gets.

Which Stadiums Give Their Teams the Best Home-Field Advantage

We usually assume that home field gives the host team a distinct advantage over the road team. In fact, announcers will usually talk about the disadvantage that a team has because they have to go on the road to play. However, when you look at the evidence, not every home field gives its team an advantage. Consider this:

1. There is no correlation overall from year to year for home records. In other words, many teams in the NFL that performed 3 games better at home than on the road one year, for instance, fare no better at home than on the road the next year.

2. When comparing each team's home record against each team's road record, there were 10 teams for which there was no statistical difference whatsoever. Those 10 teams, in other words, performed just as well on the road as they did at home. In fact, there are four NFL teams that actually perform better on the road than they do at home over the past five years!

Okay, so those points made, the are an additional handful of teams that over the past five years have enjoyed a consistent, moderate homefield advantage. Finally, there are an additional dozen or so teams that perform so much better at home than they do on the road, that these teams are considered to have the best homefield advantage in all of the NFL.

When trying to pick the outcome of a close matchup and you want to take homefield advantage into account, it may be wise to consult the table below. As you'll see, not all teams enjoy the same advantage at home. Let's break down the list and try to examine why some stadiums help their teams more than others.

First, here are the teams for which their home field does not help them at all. The number in parentheses in the margin by which their winning percentage at home is better than their winning percentage on the road:

Carolina Panthers= -10
New Orleans Saints= -8
New York Giants= -3
Philadelphia Eagles= -3
Tennessee Titans= 5
New England Patriots= 5
Dallas Cowboys= 5
Atlanta Falcons= 5
Indianapolis Colts= 8
Green Bay Packers= 8

A few observations:

1. As my wife actually pointed out, the Giants and Eagles being on this list is not surprising. I live in New York and she lived in Philly for four years, and fans of those two teams are, shall I say, brutal to their teams. Maybe this shows that constantly berating your own team may just have a negative impact on their performance.

2. New England and Indy also are not a surprise. Over the past 5 years, they've won so consistently on the road that their home record simply had little room to be much better.

3. Two myths shattered in one list- The Saints actually play worse in the Superdome than they do on the road. Plus, the vaunted Lambeau Field gives the Packers absolutely no discernible advantage over their visitors. Go figure.

4. Carolina fans, care to weigh in with a theory?

Next, the following teams enjoy a moderate advantage playing in their own stadium:

Miami Dolphins= 10
Washington Redskins= 13
San Diego Chargers= 15
Cincinnati Bengals= 15
Oakland Raiders= 18
Cleveland Browns= 18
St. Louis Rams= 18
Tampa Bay Buccaneers= 20
New York Jets= 20
Buffalo Bills= 20
Pittsburgh Steelers= 20

Observations...

1. Weather and geography, at first glance, don't appear to play much of a factor. There are just as many warm weather teams on the list as there are cold weather teams, and they seem to be spread around the country.

2. The Jets, who play in the same stadium as the Giants, enjoy a significant advantage at home, while the Giants are actually worse. In my opinion, it's always been clear that Jets fans have been more loyal than Giants fans (probably because they have to be). But can that really be the sole explanation? I'm open for suggestions.

3. The Redskins are the only NFC East team that actually has a homefield advantage. I'm not sure if I should make anything of that, or if it's simply weird.

Finally, here are the teams with a tremendous homefield advantage. Before looking, do you think you can guess #1?

Houston Texans= 23
San Francisco 49ers= 23
Denver Broncos= 23
Jacksonville Jaguars= 23
Detroit Lions= 25
Chicago Bears= 25
Kansas City Chiefs= 28
Minnesota Vikings= 28
Arizona Cardinals= 35
Seattle Seahawks= 38
Baltimore Ravens= 43



Observations...

1. First of all, just astonishment at that number by the Ravens. Playing with a winning percentage 43 points better at home than on the road, they are like two completely different teams.

2. Something Seahawks fans already know- As the playoff race gets tight, they must keep an eye on home and away game left, because at home they are excellent. On the road, they're not even in the top half of NFL teams.

3. I've read many times about how loud and unfriendly Arrowhead can be for visitors. Now I see empirical proof of it.

4. Now we know what it will take to turn around the Cardinals- more home games.

How to Tell If Your Team's Defense Will Improve This Year

The most commonly used way to measure the quality of a defense is actually a pretty good and valid system. That measurement, which determines defensive rankings, is the amount of yards that a defense gives up. It's simple, and it's to the point. After all, what else is a defense asked to do than stop the other team from gaining yards? Points given up is a slight bit trickier. After all, some defenses often start with worse field position than others and thus, could give up few yards but still yield some points.

For the most part, however, a good defense will give up fewer points than will a bad defense. The data supports this assertion. There is a staggeringly high correlation of 0.768 between the amount of points and the amount of yards given up by a defense.

However, there are many instances in which, over the course of a season, good defenses will give up slightly more points than they should, and bad defense will give up slightly less. One key determinant of how many "extra" points a good defense gives up or how many points a bad defense "saves" is the amount of takeaways that a defense has. A regression analysis was performed using the variables of yards allowed and takeaways as independent variables and points allowed as dependent variables. In other words, I ran an equation to see if both yards allowed and takeaways would each be significant predictors of the amount of points that a defense allowed. The results indicated that both factors are indeed significant predictors of the amount of points that a defense allows.

So, what determines how many takeaways a defense has during a season? Would you believe luck? Well, that appears to be the answer. There is nor correlation whatsoever from year to year over how many takeaways a team will have. In other words, there is no known way to predict even approximately how many takeaways a team will have the next year. For instance, Baltimore led the league in takeaways in 2006 with 37. In 2007, they finished 27th with 23. Meanwhile, San Diego had only 27 takeaways in 2006 but led the NFL in 2007 with a whopping 48 takeaways. The complete lack of consistency implies that chance is the primary determinant of how many takeaways a defense will have.

The law of averages dictates that eventually, everybody's luck evens out. Therefore, if chance is the primary factor in takeaways, a team with an otherwise good defense that got unlucky with few takeaways should experience an improvement in points allowed the next year. On the other hand, some teams that gave up a lot of yards but got a lot of lucky takeaways should see its yards allowed go up this year.

Of course, offseason personnel and coaching changes can naturally change a defense, as well. However, a team bringing back basically the same defensive squad should fall prey to these laws. Below are three teams whose defensive performances were better than their points allowed because of bad luck in takeaways. All other factors being equal, these teams may expect to allow fewer points this year.

Baltimore Ravens- 302 yards allowed/game; 23 takeaways; 24.0 ppg allowed
New York Giants- 305 yards allowed/game; 25 takeaways; 21.9 ppg allowed
Kansas City Chiefs- 319 yards allowed/game; 22 takeaways; 20.9 ppg allowed

The three teams below had poor defensive years but allowed fewer points than expected because of a large number of takeaways. Expect these defenses to give up more points this year:

Buffalo Bills- 363 yard allowed/game; 30 takeaways; 22.1 ppg allowed
Minnesota Vikings- 335 yard allowed/game; 31 takeaways; 19.4 ppg allowed
Chcago Bears- 355 yard allowed/game; 33 takeaways; 21.8 ppg allowed

Rating the Best Running Backs- A New Way of Measuring Them

As I've mentioned before, much of what running backs do depends significantly on their surroundings. Therefore, simply looking at the yardage leaders to determine the best running backs can be misleading.

However, there are some things that do distinguish the best running backs from others. Below are three stats that can help determine just how good a running back is:

1. 4th Quarter Yards per Carry- There is significantly more variance between running backs in 4th quarter yards/carry than in overall yards/carry. There are several possibilities for this. Perhaps one true test of a running back's value is how well he can handle a beating over the course of a game and remain effective. Maybe the reason is that as lines get tired, better running backs have more opportunity to differentiate themselves by maneuvering more. Whatever the reason is, the data demonstrate that there is less similarity among runners in the fourth quarter of football games than there is in the rest of the game. The final period, according to the numbers, is where the best runners separate themselves.

2. # of 10+ Yard Runs- As stated above, offensive lines have a lot to do with the success of a rush. This fact, however, is primarily true for the beginning of the rush. Once a running back gets more than 5 yards downfield, a lot of his success depends on his own abilities to break tackles and elude defenders. Therefore, the best running backs would be the ones with many 10+ yard rushes.

3. Side Yards Per Carry- By this I mean the yards per carry a running back gets when running to the left or right, as opposed to up the middle. The offensive line and lead blocker will usually reside in the middle, and when sent out to the sides, a running back will have to rely more on his own skill and elusiveness. Obviously, this statistic cannot be perfect, as blockers sometimes accompany runners to the side as well. However, it is much more common for a runner to be alone on the sides than he is up the middle, and thus, this stat is a clearer indication of his individual performance than running up the middle.


Combined, these three statistics can give us a better picture of who truly are the best running backs in football. Consider however, that these statistics only measure their value as rushers and not as blockers or receivers. Therefore, some running backs who also catch passes may be undervalued in this system. In order to rank the running backs, I awarded them inverse points to their position in each list. For instance, in a list of the 20 top running backs in 4th Quarter Yards/Carry, I awarded 20 points to the top running back and 1 point to the 20th. I added up the scores of all three categories for a top 10. What is revealing is how many of the running backs on this list appeared in all 3 of the lists. While I have not done studied any further yet, the close relationship that the three statistical categories have with each other and the lack of a relationship that they have with some other prominent rushing categories suggests that these three categories may form a common metric of how to measure true individual rushing performance. Look for deeper analysis at a later date.

Another note- Running backs have their best years between the ages of 25-29. Players between that age range can be expected to maintain their effectiveness barring any extraneous circumstances. Younger backs can be expected to improve, while older backs are probably due for a decline soon. Running backs' ages are included in parentheses. Without further ado, the Top 15 Running Backs are:

15. Joseph Addai (25), Indianapolis Colts- Addai is looked at as one of the best offensive weapons in football, much of his prowess comes from his receiving numbers. The skill set that Addai possesses are very much in the Marshall Faulk mode, but as an actual runner, he is merely slightly above league average. The only category in which he broke the Top 10 was 10+ Yard Runs, in which he placed 10th. Many of the stats that Addai excels in may be largely attributable to a great line and offense, which helped Edgerrin James look every bit the superstar. Since going to Arizona, by the way, James' numbers don't earn a spot anywhere near this list.

14. Frank Gore (25), San Francisco 49ers- While getting off to a bad start, Gore followed up a sensational 2006 season with a respectable 1100 yards in 2007. This system's stats also put his performance at respectable, if not quite special. Gore, like Addai, was a top 10 performer in Big Runs, and also did reasonably well in Side Yards. However, Gore's main problem is that he slowed considerably near the ends of games, with production dropping especially steeply in the 4th quarter. Hopefully, the 49ers can get him more support, because he doesn't appear to have the stamina to carry more than 20 times per game.

12(t). Willis McGahee (26), Baltimore Ravens- McGahee is one of the most explosive running backs in the game. He had 32 Big Runs in 2007, which was sixth in the NFL. However, McGahee's problem, going back to his days in Buffalo, is a woeful lack of stamina. His 2nd half numbers are fairly bad, and his fourth quarter numbers are atrocious. In 2007, McGahee averaged less than 3.5 yards per carry in the fourth quarter. McGahee also seems lost somewhat outside the tackles, as his numbers drop significantly when running outside.

12(t). Marion Barber (25), Dallas Cowboys- Barber is another explosive runner, albeit one who is severely hampered running in the outside lanes, as he doesn't even crack the top 20 in Side Yards. His real strength is bursting across the middle and breaking tackles, as he has broken free for 27 Big Runs. Barber also places in the top 10 in 4th Quarter yards per carry.

11. Justin Fargas (28), Oakland Raiders- While he is not particularly flashy or evasive, Fargas is one of the most consistent running backs in the NFL. He is tireless, as his yards per carry actually increase considerably as the game wears on. Only starting in 7 games in 2007, he still made it over the 1,000 yard mark. His numbers may drop this year as he shares the backfield with Darren McFadden, but Fargas's contributions could make the Raiders' offense more productive, especially late in games, where Oakland has suffered recently.

10. Chester Taylor (27), Minnesota Vikings- Even with the arrival of phenom Adrian Peterson, Taylor remains one of the most thrilling open-field runners in football. He is sixth in the NFL in Side Yards, and he also puts up a respectable amount of Big Runs. Teamed with Peterson and a great offensive line, the Vikings ground game should take off enough pressure to let its passing game grow.

9. DeAngelo Williams (23), Carolina Panthers- I hate to brag (sort of), but when Williams was drafted two years ago out of Memphis, I said he'd be the best running back of his class. Forced to play second-fiddle to the inferior DeShaun Foster, Williams still ran for 700 yards on less than 150 carries. He had one of highest yard/carry averages in the NFL, and in the 4th quarter, nobody beats Williams. In the 4th quarter, Williams averages an amazing 6.6 yards/carry, easily the best in football. He is also improving his open field running, and based on how much playing time he got, his 23 Big Runs are excellent. Look for Williams to become an elite back in the next year or two.

8. Willie Parker (27), Pittsburgh Steelers- Parker is an outstanding open-field runner who gets by largely on his blazing speed. When he gets just a bit of daylight, few are better an exploding into the hole to pick up 8 or 9 yards where others may have only gotten five. Often, he gets enough acceleration to accumulate big gains in chunks. This energy seems to wane, however, as his efficiency drops sharply in the 4th quarter. That, however, is his one weakness in an otherwise strong skill set.

6t. Maurice Jones-Drew (23), Jacksonville Jaguars- Jones-Drew is not a particularly flashy runner, and he is not prone to making highlight reels. He is, however, in extraordinary shape, and as defensive linemen and linebackers get worn down late in the game, Jones-Drew maintains his stamina and racks up loads of yards in late, close situations. He is also outstanding at breaking to the sideline ahead of defenders, which is how he finished 4th in the NFL in side-yards. At only 23, he's only getting better.

6t. Brandon Jacobs (26), New York Giants- Jacobs has earned the reputation of a big power runner, which he is. In fact, watching him carry defenders on his back as he fights for first downs is a sight to behold. His ability to simply bowl over linebackers is why he has managed to gain so many yards outside of the tackles. He also keeps his production high late in games, as his sheer power and size overwhelm tired defenders.

5. Brian Westbrook (29), Philadelphia Eagles- If this list were about overall offensive threats, Westbrook might fare even better. It is his overall value, ironically, that actually makes him underrated as a runner. His receiving and returning yards make people forget that only two players have more Big runs than Westbrook and nobody has more Side Yards.

4. Jamal Lewis (29), Cleveland Browns- Many left him for dead before last year, but Lewis proved that there's still something left in the tank. Lewis churned out a ton of side yards and Big Runs, just like Westbrook. The question is, at 29, does Lewis have it in him to do it again?

3. Fred Taylor (32), Jacksonville Jaguars- Talk about being left for dead. Very few believed that Taylor truly had another great season in him, but at 31, he may have had his best one yet. He was able to rack up Side Yards and good 4th Quarter efficiency, but it was his trademark Big Run ability that gave him such a special year. He's 32, but with a talent like Jones-Drew lightening the load, Taylor may be able to shake out a few more good years.

2. LaDainian Tomlinson (29), San Diego Chargers- What's not to love? He gains yards up the middle, down the left, and down the right. He has more Big Runs than anybody, stacking up a phenomenal 44 of them last year. Maybe most noteworthy of all, his late-game performance is amazing. His 4th quarter numbers are highly impressive, and he averages an unheard-of 7.6 yards/carry after his 20th carry of a game.

1. Adrian Peterson (23), Minnesota Vikings- Peterson actually placed second in every one of the system's categories, which indicates that he is outstanding across the board. His explosive speed and ability to break long runs have been witnessed by any viewer of SportsCenter. He is also an extraordinarily graceful runner, whose ability to dodge defenders and fit through holes may be unmatched since Barry Sanders. Plus, the guy's only 23. Imagine if this is just him warming up.

P.S. Deuce McAllister was not eligible for this list because of his 2007 injury. However, expect him to be a top 10 back in 2008. His historical numbers in the above categories suggest that he belongs there.

Undervalued Kick Returners and How They Change Games

Most fans don't really notice kick returners unless they break for a touchdown. After all, it seems that the vast majority of kick returns end at about the 25 or 30 yard line and are almost uniform. Aside from the rare spectacular return, most kick returns seem to be pretty much the same.

Kick returns, it should be known, play a huge part in games. There is a reason that a Buffalo Bills team that finished 26th in offense and 29 in defense last year finished with 7 wins instead of, say, 3 wins. Their special teams were amazing, often getting them 5-8 yards more per possession that would an average kick return unit. Those yards put them in vital field goal position for a team that won 4 of its 7 games by 5 points or less.

Buffalo Bills aside, what would you say if someone told you that your quarterback who threw for 3,000 yards last year would throw for 3,400 hundred yards with no extra interceptions? Or, if your team's lead running back will run for 1,500 yards this year instead of his usual 1,100? That's right, the top five kick returning teams gained, on average, 400 yards more in kick returns than league-average kick returning teams. A net increase of 400 yards is a substantial difference over the course of a season. Who cares if they're the first yards of a drive instead of the last yards?

A few more facts:

- The league's top 5 kick returners last year averaged 29.0 yards per return, which means that they contributed 29 yards to each possession on which they returned a kickoff. League average among returners who returned at least 30 kicks was 24.5. So, elite returners will give you an extra 4.5 yards per possession. Over the course of a game, 4.5 yards per possession will likely lead to multiple extra scores, as either field goals or touchdowns. In other words, those yards are valuable.

- The top five rushers last year averaged 13.7 yards per possession. League average for lead rushers was 10.0 yards per possession. Elite rushers, in other words, gave a team an extra 3.7 yards per possession, which is less than what returners give you.

- The top five receivers averaged 13.3 yards per possession. League average for a team's top receiver was 10.2 yards per possession. Thus, an elite receiver will gain a team an extra 3.1 yards per possession.

So, now that we've touched upon how valuable a great kick returner can be, who are currently the best returners in the NFL?

5. Darren Sproles, San Diego Chargers- Sproles didn't make a lot of highlight reels, largely because he only had one touchdown. Touchdowns, however, are an overrated stat for returners, as very rarely does anyone have more than 2 or 3 in a season. Sproles, however, did average 27.2 yards per return. He also had returns of more than 20 yards on 73% of his returns. He returned the ball more than 40 yards on five different occasions, tied for second in the league. Big returns like those on such a frequent basis can change games, and often.

4. Leon Washington, New York Jets- Washington is simply explosive. He averaged 27.5 yards per return, bringing back 5 of them for more than 40 yards. Three of those returns happened to wind up in the end zone. Thirty-five of his returns gained more than 20 yards, which is 9th in the NFL. Washington's 1,291 return yards were a rare bright spot in the Jets' 2007 season.

3. Aundrae Allison, Minnesota Vikings- One of the league's undiscovered treasures, Allison does not get the attention of his fellow Viking returner, super-rookie Adrian Peterson. However, when Peterson went down to injury, Allison had his shot to shine. He only returned 20 kicks once he was installed, but he averaged an outstanding 28.7 yards per return, which is more than 4 yards above league average. Even with those rare chances, he still returned 3 kickoffs for more than 40 yards. That is a rate more than twice of even elite return men. If Allison becomes the full-time man in Minnesota, Adrian Peterson can get more rest and the Vikings could have themselves an elite weapon in Allison.

2. Andre Davis, Houston Texans- The 30.3 yards per carry should say it all. One concern with Davis, however, is that only 67% of his returns are for more than 20 yards. All that means, however, is that when he does break for more than 20 yards, he breaks a big one. His three touchdowns are testament to that. A returner with such breakaway speed can have his team at midfield in no time.

1. Josh Cribbs, Cleveland Browns- Many people may remember his circus-like touchdown return against the Stellers last year, but Cribbs is more than just a one-hit wonder. In fact, Cribbs is the most consistent returner in the game. An amazing 83% of returns, he gains more than 20 yards. He also leads the NFL with 7 returns of more than 40 yards. An average of every other game, Cribbs busts a game-changing, colossal return. The Browns benefited to the tune of 1,809 yards from Cribbs' kickoff returns in 2007. Many returners who break off so many long runs have fumbling problems, but Cribbs only coughed up the ball once all season.

Okay, you're wondering about Devin Hester? Well, as a punt returner, he's phenomenal. But as a kickoff returner, he's actually not too much better than average. His return average of 21.7 is actually almost 3 yards less than league average. Furthermore, he doesn't even advance half of his returns more than 20 yards. His 2 touchdowns fool a lot of people, but when everything else is analyzed, the five above definitely do more.

More Valuable- An Explosive Running Back or a Big Play Receiver?

To those of us raised on the maxim that running and defense win championships, it may seem obvious that an elite running back is vital to a Super Bowl run. It may also seem obvious that a great running back can be the core of a team, whereas an elite receiver cannot.

Based upon what I've seen, it appears to me that elite, big play receivers will generally add more to an offense than will elite running backs. This post is not meant to denigrate the value of players like LaDainian Tomlinson or Adrian Peterson. Those guys undoubtedly add a ton to their teams. However, consider some quick facts:

1. Running backs' performance are reliant on more outside variables than are wide receivers' performances. As will be demonstrated in coming weeks, the performance of running backs are tied extremely closely to the quality of their offensive lines, especially the center and guards. Wide receivers tend to produce based more upon their own ability to get open quickly and make catches. One indicator of this is that a wide receivers' statistics stay more consistent when he moves teams, whereas when a running back moves to a new team, his numbers generally correlate more closely with those of the running back who preceded him on his new team than his own previous stats. This finding indicates that running backs are at least in large part a product of their surrounding cast (see James, Edgerrin). Be sure that Steve Hutchinson, Matt Birk, and co. got nice Christmas gifts from Adrian Peterson.

2. The amount of receiving yards per catch that a team averages is almost 4 times better a predictor of whether a team will win than a team's rushing yards per attempt.

3. Having a receiver that can stretch the field helps teams win, period. The four teams that had the most big completions, worth 25 yards or more last year? New England, Indianapolis, Dallas, and Green Bay, which just happened to be the four teams who had playoff byes. Cleveland and Arizona were also high on that list, and those two teams saw very encouraging progress last year. Something to think about for this year, perhaps...

4. The vast majority of NFL rushing attempts end up gaining between three and five yards. One of the traits that makes elite running backs stand out is how often they can break for a gain of more than five yards. Upon examination of the data, however, there is no correlation between the number of 10+ yard rushes a team has and the amount of points it scores, or the amount of wins that it gets. Yes, those big bursts are the ones that you will see on the highlight reels, but don't be fooled- the teams that get those big runs more often are actually no more likely to win games.

There is a mountain of additional data that I will get to, but when contemplating whether a team's offseason pickup of a receiver or runner will help a team more, let these facts be some food for thought.

Sacks- The Most Damaging Offensive Stat?

I have examined the relationships between dozens of offensive stats and how many points an offense scores. It is surprising how little of a relationship some statistics have with how much a team scores (e.g. rushing yards per carry). Other stats, however, are shockingly relevant to how many points a team scores.

Of all of the statistics I have looked at, nothing is linked more closely with how many points a team scores than the stat of how often its quarterback gets sacked. Now, it shouldn't be surprising that sacks are detrimental to an offense's performance. What is eye-opening is just how damaging they are. In 2006, 11 teams allowed fewer than 30 sacks. Ten of these teams made the playoffs, with the exception of Tennessee. Of the 12 teams that allowed more than 40 sacks, only one, Seattle, made the playoffs, and that was with a 9-7 record. Last year, only one of the 10 teams who allowed 40 or more sacks made the playoffs.

It's clear, then, that teams whose quarterbacks hit the turf a lot simply don't win. In fact, New England's worst offensive performance last year was in their Super Bowl loss, when Tom Brady was sent down five times. But in order to make this stat meaningful, we need to dig deeper: Why do some quarterbacks get sacked more than others? The question is not as straightforward as you might believe. Many people rush to the conclusion that the offensive line stinks. That is certainly one explanation, but it is not the only one. First of all, some teams play in divisions with better defensive lines. Of course, the quarterbacks sometimes can influence how many times they get sacked. Quarterbacks better at moving in the pocket or releasing the ball more quickly can save themselves a trip to the turf. It is probably no coincidence that Brian Griese, while making more attempts than teammate Rex Grossman, suffered only two-thirds as many sacks as Grossman in 2007. Also, receiver matchups and game plans can come into play. A quarterback with no open receivers or viable options is more vulnerable to getting sacked.

So, what are some of the best ways for teams to avoid racking up sacks? I think I'll save that for another post. One thought that keeps coming up, though, is that conventional thinkers may be right about something, after all. Line play is absolutely essential to winning. Expanding on that, in some of my next posts, I'll reveal why I really do believe that an offensive guard or center is more important than a star running back.

Strength of Schedule- Does It Really Matter?

A budding tradition in preseason NFL projections is factoring in which teams have the hardest and weakest schedules. A Lexis search demonstrated that last summer, over 50 articles were published in major newspapers in which prognosticators took made their playoff predictions based at least partly on what the teams' opponents records were the year before. The Seattle Times warned that due to their easy schedule, the 49ers would challenge the Seahawks for the NFC West Crown. Based on the 49ers record, the Kansas City Star also projected a big year for San Francisco. Meanwhile, my beloved Bills were expected to finish the year in the bottom five of the NFL, having drawn the "most difficult" record in the NFL.

Needless to say, the Niners finished a putrid 5-11, never seriously making a run at either the NFC West crown or even the playoffs. The Bills, well, they missed the playoffs. However, they finished 7-9, far from the bottom of the league, and achieving the same exact record as the year before, when their record was said to be easier.

Two examples, I realize, does not prove or disprove an argument. But if one were to examine not just these two teams but the entire league, it will become clear that a team's strength of schedule has virtually nothing to do with its success. From year to year, a team's record change by, on average, 2.5 games. In other words, statistically speaking, a team that finished 8-8 one year should be expected to finish anywhere from 5-11 to 11-5. Year after year, the average correlation that an NFL team's record has to its previous year's record is 0.26, which means that there is a modest relationship between the two records at best. Think about it. Pick your favorite team, and chances are that over the past five years, its record has been on something of a roller coaster ride.

Considering how much records change from year to year in the NFL, why do sportscaster's and columnists continue to emphasize teams' strength of schedule? Before the 2007 season, football analysts and experts were picking apart the Dallas Cowboys' upcoming season game-by-game on ESPN Radio, discussing which ones were winnable and which ones would be tough. Among the NFC out-of-division games, they focused on the Bears, a team that won 13 games the previous year, would probably be the major challenge of the years, whereas the Packers, coming off an 8-8 season with a young team, should be easily winnable. Of course, the Cowboys trounced a Bears team that would only win 7 games all year and fought tooth-and-nail with a Packers team that ended up winning thirteen. Potentially tough games against the Rams and Panthers, 8-8 teams from the year before, turned out to be games against doormats, neither team winning more than 4 games in 2007.

One expert even said that the 'Boys hardest interconference test would be against the upstart Jets, a team full of optimism after a surprising 10-6 year in 2006. Well, the Jets surprised yet again, this time by sliding near the AFC cellar with a 4-12 mark. These examples exist all across the league. Yet still, I guarantee that you will endure more nonsense this year right before the season about how your team's strength-of schedule will affect them.

You wanna get straight to the bottom-line statistic? I decided to analyze the relationship that a team's strength-of-schedule has with the amount of games that it will win that year. The correlation? A ridiculously small 0.018, which means that there is absolutely no relationship whatsoever between a team's strength-of-schedule and its performance.

What is The Football Know It All?

Welcome to my first post. To whomever is reading this, I am a graduate student who has a) a decent amount of statistical knowledge, b) a tremendous (fanatical) passion for pro football, and c) the curiosity to delve into the numbers and figures to get beneath the surface and look at NFL games in a whole new way.

Now, statistical analysis in football is nothing new, and even blogging about football stats is certainly not breaking new ground. There are also some fantastic sites out there that break down a plethora of football statistics that you may never have even thought of before, and these sites are a great asset to info-hungry fans, myself included.

What I am setting out to do here is to create a blog in which with each new blog, I will discuss a topic in pro football and look at it in a new way. I will use a number of stats and also analyze them, explaining their implications in the real-world NFL. For instance, I will conduct statistical analyses to determine questions like: Does strength of schedule really have an impact on a team's record? What offensive stats are most important to a team scoring points? What are the most under appreciated positions on the field? Who are truly the best players at each position? what stats should you look out for when predicting a team's wins and losses?

Hopefully, the posts will be entertaining and fun to read for any football fan who wants to take a new look at the game. It's still the offseason, but with a month to go before kickoff, this is the perfect time to start digging to get a closer look at the upcoming season. Till next time...