NFL should tackle statistical silliness (1 Viewer)

[onionsack]

link

NFL should tackle statistical silliness

12:17 PM CDT on Tuesday, August 5, 2008

Middle linebacker Patrick Willis of the San Francisco 49ers was the NFL Defensive Rookie of the Year in 2007.

If you trust his statistics, he should have been the NFL's Defensive Player of the Year as well. Maybe even the NFL Defensive Player of the Decade.

Willis was credited with 226 tackles by the 49ers – a staggering 42 more than the league's next most prolific tackler, Ray Lewis of the Baltimore Ravens.

Willis collected 117 more tackles than the starting NFC middle linebacker in the Pro Bowl (Lofa Tatupu, Seattle Seahawks) and 99 more than the AFC starter (Demeco Ryans, Houston Texans). Was Willis that much more active, that much more proficient than any other defensive player in the NFL?

But I don't trust the statistics – not tackle stats, anyway.

Each NFL team employs a stat crew for its home games. That crew's final statistics are deemed official by the league office: yardage for rushing, passing, receiving and various kick returns plus field goal and punting distances, sacks, interceptions, touchdowns ...

But 28 of the 32 NFL teams do not acknowledge the press box tackle stats as official. On the Monday following games, NFL defensive coaching staffs break down the films on their own and award a new set of tackle numbers. Those are listed by each team as its "official" tackle count.

Willis was credited with 174 tackles by NFL stat crews in 2007. Upon further review, the San Francisco coaches gave him credit for 52 more tackles, bumping his "official" count to 226.

How is it possible for stat crews league-wide to miss 52 tackles by Willis? That's a 23 percent increase in his tackle count. That's like the New England offensive staff studying weekly game tapes and uncovering 1,400 passing yards that stat crews missed in Tom Brady's 2007 season.

New England finished fourth in the NFL in defense and was one of only two teams that did not credit a player with 100 tackles. Miami was the other. Indianapolis finished one rung ahead of the Patriots on defense at third – but had a league-high five players with 100-plus tackles.

The Indy defense was on the field for 980 snaps last season. But there were no tackles on 199 of those plays because they were either incomplete passes, interceptions or offensive touchdowns.

That left 781 defensive snaps that involved a tackle – and the Colts' coaching staff awarded a league-high 1,416 tackles. That's an average of 1.81 tackles per play. The Indianapolis staff awarded 891 "solo" tackles in those 781 plays. Figure that one out.

The Tampa Bay Buccaneers, who finished second in the NFL in defense, were even more generous. Their coaches awarded 1,385 tackles in 755 plays, an average of 1.83 tackles per play. The 49ers averaged 1.63 tackles per play.

I surveyed a handful of NFL head coaches a few years back, asking them what they considered a legitimate tackle average. The consensus was 1.44 tackles per play. I still consider that a bit high. I'm of the belief 1.3 tackles per play is about right. Nonetheless, 15 teams were at or below the 1.44 figure in 2007.

The four teams that count the press-box tackle stats as official are Houston, Miami, Oakland and Seattle. All finished in the league's bottom six in tackles per play. Ryan's Houston defense averaged 1.23 tackles per play, and Tatupu's Seattle defense averaged 1.13.

The first time I put together a league-wide tackle chart in 1992, there were three teams that averaged better than two tackles per play. That was ridiculous, not to mention impossible. So there has been slight improvement from one decade to the next.

Here's my solution: The NFL office should hire a retired defensive coach, bring him to New York on Sunday and give him three days to break down the tape for all 16 games. Let him credit all the tackles for all the games. One set of eyes would foster a league-wide consistency. It would also eliminate any and all bias – and padding – by stat crews and or coaching staffs.

I doubt you'd see any more 200-tackle seasons. And I'd start accepting tackle statistics as legit.

Here's the NFL tackle chart for 2007. The first column is the number of defensive snaps by a team. Now subtract the non-tackle plays (incompletions, interceptions and offensive touchdowns) in the second column, producing the number of actual tackle plays (third column). Now divide the number of credited tackles (fourth column) by the number of actual tackle plays (third column). That produces an average number of tackles per play by a team (fifth column).

Team Plays Non-T T-plays Tackles Avg.

Tampa Bay 977 222 755 1,385 1.83

Indianapolis 980 199 781 1,416 1.81

Philadelphia 979 250 729 1,284 1.76

San Diego 1,013 248 765 1,298 1.69

San Francisco 1,078 231 847 1,381 1.63

Baltimore 968 233 735 1,193 1.62

Tennessee 978 252 726 1,173 1.61

Jacksonville 970 256 714 1,145 1.60

Green Bay 994 268 726 1,131 1.55

Detroit 1,087 231 856 1,320 1.54

Chicago 1,036 234 802 1,199 1.49

Washington 1,026 281 745 1,112 1.49

Cincinnati 1,011 227 784 1,167 1.48

Pittsburgh 933 272 661 966 1.46

New Orleans 964 236 728 1,056 1.45

NY Jets 1,016 213 803 1,162 1.45

St. Louis 999 246 753 1,096 1.45

Dallas 1,008 270 738 1,052 1.42

New England 933 242 691 968 1.40

Cleveland 1,066 275 791 1,089 1.37

Denver 992 218 774 1,052 1.36

Minnesota 1,063 260 803 1,091 1.36

Arizona 1,013 248 765 1,036 1.35

NY Giants 984 253 731 970 1.32

Carolina 1,032 235 797 1,046 1.31

Atlanta 1,042 239 803 1,038 1.29

p-Houston 994 225 769 950 1.23

Buffalo 1,047 247 800 961 1.20

Kansas City 980 214 766 917 1.19

p-Miami 983 213 770 912 1.18

p-Oakland 952 219 733 847 1.15

p-Seattle 1,035 266 769 872 1.13

p-Press box statistics considered official.

Agree with him entirely that it's maddening to see so many "phantom" tackles and so much stat-fudging on the IDP side of FF.

 

[BigSteelThrill]

Jessie Tuggle-itis.

 

[BigSteelThrill]

Just ignore team stats. Problem solved for the most part.

 

[Jene Bramel]

The NFL made a concerted effort in the summer of 2007 to increase the uniformity and consistency in its unofficial tackle and passes defensed stats. It's a work in progress, but improvement was made last season. The coaches count has always been well above an expected number. The coaches give out assisted tackles in bunches, presumably as part of their grading process. The stat crew count has anomalies, as Aaron Rudnicki has noted for years, but it was never as crazy as Gosselin suggests here.

The team tackle stats aren't any more "official" than the stat crew tackle stats.

I was able to get access to the NFL's new set of guidelines last month. Aaron and I will both have articles coming out in the next couple of weeks addressing the statistical anomalies and how things have changed over the past two seasons.

 

[Workhorse]

This is the main reason why I'll never play in an IDP league. It's like playing fantasy gymnastics. Too arbitrary and subjective.

 

[Rozelle]

This is the main reason why I'll never play in an IDP league. It's like playing fantasy gymnastics. Too arbitrary and subjective.

Be careful not to misread this ... for the most part, it's the NFL teams who are jacking the numbers up, their numbers are not used in IDP leagues. The stat crews do a fairly decent job. I'll be honest, there were times, years ago, I would scratch my head and ask "what game were they watchin'"BTW ... thanks for posting this onionsack

 

[Jene Bramel]

This is the main reason why I'll never play in an IDP league. It's like playing fantasy gymnastics. Too arbitrary and subjective.

Team defense is less arbitrary and subjective? The Dolphins, Panthers and Ravens had a high ADP last season. None finished in the top 20 in most scoring systems last year. Who was drafting the Colts and the Vikings in the top five?What's arbitrary and subjective about individual defensive players? I'm always curious about what that means. Here are some common complaints.Bad players end up atop the rankings? Look again. Plenty of All-Pros atop the preseason rankings and end-of-season finishes. The anomalies in unofficial defensive stats account for very little in the year end fantasy rankings.Think the same thing doesn't happen from time to time due to system and opportunity on the offensive side? Jon Kitna is a top quarterback? Kurt Warner (last season's #2 overall QB in the fantasy rankings) leading his team to any playoff appearances? Don't like the fact that shutdown corners don't score well? Well, what about Lawrence Vickers or Lorenzo Neal. Those guys had just as much to do with their offensive output in recent seasons and get less love than the shutdown corners. Unexpected players that come out of nowhere too often on the IDP side? How about Ryan Grant? Kenny Watson? Mike Furrey? Shaun McDonald? Too much turnover from season to season? The team defense position has the fewest year-to-year consistency of any fantasy "position". The linebackers and defensive linemen barely trail the offensive positions in consistency percentage.I'm fine if folks don't care or don't have the time to learn to understand why defensive players aren't any more of a crapshoot than the offensive positions. But I can't let comments like this slide, especially in the IDP Forum.

 

[John Norton]

This is the main reason why I'll never play in an IDP league. It's like playing fantasy gymnastics. Too arbitrary and subjective.

Team defense is less arbitrary and subjective? The Dolphins, Panthers and Ravens had a high ADP last season. None finished in the top 20 in most scoring systems last year. Who was drafting the Colts and the Vikings in the top five?What's arbitrary and subjective about individual defensive players? I'm always curious about what that means. Here are some common complaints.Bad players end up atop the rankings? Look again. Plenty of All-Pros atop the preseason rankings and end-of-season finishes. The anomalies in unofficial defensive stats account for very little in the year end fantasy rankings.Think the same thing doesn't happen from time to time due to system and opportunity on the offensive side? Jon Kitna is a top quarterback? Kurt Warner (last season's #2 overall QB in the fantasy rankings) leading his team to any playoff appearances? Don't like the fact that shutdown corners don't score well? Well, what about Lawrence Vickers or Lorenzo Neal. Those guys had just as much to do with their offensive output in recent seasons and get less love than the shutdown corners. Unexpected players that come out of nowhere too often on the IDP side? How about Ryan Grant? Kenny Watson? Mike Furrey? Shaun McDonald? Too much turnover from season to season? The team defense position has the fewest year-to-year consistency of any fantasy "position". The linebackers and defensive linemen barely trail the offensive positions in consistency percentage.I'm fine if folks don't care or don't have the time to learn to understand why defensive players aren't any more of a crapshoot than the offensive positions. But I can't let comments like this slide, especially in the IDP Forum.

Excellent post Jene. People have tried to make this silly argument for years. Notice the most important thing he said "I'll never play in an IDP league". Obviously he never has or he wouldn't have made this statement. You can lead the horse to water but you can't make it drink