Who Is the Best Cover Corner?
July 7, 2008
This is the first in a series of pieces on the top cover cornerbacks and secondaries.
Who is the best NFL cover corner? Who has the best coverage unit around? The question is being kicked around a lot these days, with cornerbacks getting free agent contracts that rival those of quarterbacks for sheer dollars and bonuses.
I’ve come across several lists in recent weeks, all going off reputations, some earned, some not. BSR has decided to pursue these questions with the use of the largest database of corner coverage stats around. K.C. Joyner, the author of the Scientific Football books, will soon release Scientific Football 2008, the fourth edition of his series.
Joyner’s forte is breaking down the passing game, from offensive and defensive perspectives. His CB YPA will shortly become, in my opinion, shorthand for measuring a corner’s skills, in the same way that OPS has become the one stat that can best convey a baseball hitter’s skills. It’s easy to calculate and easy to understand. It shows how many yards a corner surrendered on average every time a pass was thrown to his man.
K.C. prints out a YPA ranking every year and in this year’s book has ranked the teams by tandem, calculating the average YPAs for the starting CB duos on each team.
What he has not yet done is compile YPAs over a longer span, say the last three or four years. When he learned I was putting such a spreadsheet together, he kindly forwarded his 2007 YPA rankings, which will not be published for several more weeks, and gave me permission to publish pieces from my calculated averages. BSR readers will therefore be the first to see these numbers. If you are not a regular Joyner reader, I highly recommend his ESPN columns and his books which can be ordered here.
The spread sheet raised as many questions as it answered. Although I have four years of data at my disposal, is it fair to omit a cornerback who has only played two or three years, but at a high level? What about a player who has missed a year due to injury?
In addition, what about players who have solid numbers but show a steady decline. Should I rate them above or below a player who has a similar number but whose averages are improving?
For now, I’ve decided to offer these breakdowns as a starting point for discussion, not as the last word in the debate over which players are best and which units are best. I’m going to offer two individual lists, one showing the best cornerbacks for the last four years and one for the last three years. The reason I’m pushing the data to be comprehensive is the generally inconsistent level of CB play from one year to the next.
It’s fairly routine to see a corner, even a good one, have two stellar years, then drop off a year and then bounce back strong again. For examples, look at the lines for Deshea Townsend and Ronde Barber on the top ten list below.
I’m sure there are many factors for a dropoff. A corner might be playing hurt. He may lose his confidence for a stretch of the season or an entire campaign. His pass rush might be weaker one given year. Whatever the case, there are not many corners in my sample who strung four strong years together. Even the better ones had a so-so year somewhere in the mix.
That said, I believe that a player must display consistency in order to be considered tops at his position. So my first list will, with one exception, include only those players who have posted numbers for the last four years.
| Player-team | 2004 | 2005 | 2006 | 2007 | YPA Avg. |
|---|---|---|---|---|---|
| 1. Springs - Washington |
4.2 | 5.5 | 6.2 | 6.2 | 5.5 |
| 2. Newman - Dallas |
5.8 | 5.8 | 7.1 | 6.1 | 6.2 |
| 4T. Townsend - Pittsburgh |
7.1 | 5.1 | 7.8 | 5.2 | 6.3 |
| 4T. Barber - Tampa Bay |
5.7 | 6.7 | 7.9 | 4.9 | 6.3 |
| 6T. Woodson - Green Bay |
7.0 | * | 5.3 | 7.2 | 6.5 |
| 6T. Lucas - Carolina | 7.9 | 5.8 | 5.2 | 7.2 | 6.5 |
| 7. O’Neal - Cincinnati |
5.8 | 6.6 | 6.8 | 7.0 | 6.6 |
| 8. Smoot - Washington |
7.6 | 7.2 | 6.9 | 5.5 | 6.8 |
| 10T. Jammer - San Diego | 7.8 | 7.5 | 6.2 | 5.9 | 6.9 |
| 10T. Asomugha - Oakland |
7.5 | 6.3 | 6.9 | 7.0 | 6.9 |
(*Charles Woodson did not get a rating in ‘05 because he was playing a rover position in Oakland’s system and did not qualify for the CB ratings.)
Some observations:
– The cornerback positon is incredibly unstable. You don’t think of corner as an attrition position, like running back, but there are only 40 players who had qualifying numbers each of the last four years. That’s just 1.25 per team. This is why cornerbacks who hit the open market make so much money. The demand for solid corners does not come close to meeting demand.
– Raise your hand if you considered Shawn Springs one of the top five corners in the game. I considered him for my top-of-my-head top ten, but he’s been the best YPA guy in the league the past four years.
– Raise both hands if you had Deshea Townsend in your top 20.
– There’s Terence Newman, Cowboys’ fans, parked at #2. His highest ranking in any given year has been 8th, but as you can see he’s been Mr. Steady. His 6.1 last year is even more impressive given his heel injury.
– In the what might have been category, check out Ken Lucas at 6th. The year he hit free agency he was rated higher by the Dallas scouting staff than Anthony Henry. Bill Parcells took the recommendation of Todd Bowles, who had coached Henry in Cleveland and who is now with the Tuna in Miami. Had Parcells pursued Lucas instead — Lucas did cost more — he might still be coaching.
– Notice the big names not on this list: Champ Bailey, Chris McAlister, Nate Clements, Asante Samuel? Inconsistency put them down the list. Bailey had one of the best years Joyner has ever recorded in ‘06, but it’s his only exceptional year the last four. He ranked in the bottom half in ‘04 and ‘05.
– Lastly, YPA tells us a lot, but don’t rush to judgments on guys who are on the list or are not. One fact that nags at me is that gambling, CBs, who Joyner calls “ball-hawks,” often have higher YPAs. Antonio Cromartie, for example, only had a 7.2 YPA last year, which is about league average. Guys like this will give up more intermediate and big throws in exchange for more picks. Conversely, CBs in cover two systems tend to have lower YPAs than they do when they play in more aggresive systems. Should we take the YPAs straight or weight them somewhat depending on a player’s personal style and scheme?
I’ll repeat, take this chart as a point of departure. Discuss.
Next: The top 10 CBs by YPA for the last three years.
Programming Note: I, like our Vice President, will be spending the next few days in an “undisclosed location.” I don’t know if his getaways involve a comely blonde, but mine will. I’ll see you folks again oh, Thursday.
Maybe.
Comments
95 Responses to “Who Is the Best Cover Corner?”
Very interesting stats and list. Thanks for all that work! Everyone will appreciate having real football info to discuss.
Like baseball’s OPS, I think this one stat is limiting, and interceptions need to be factored in. They change a game around and are worth a lot of yards on the field mathematically. It would be difficult, since a lower number is good for YPA, while a higher number for interceptions.
Looking at the scale of the numbers, I would suggest something like subtracting .05 off of the YPA for every interception. So for every 2 interceptions, YPA would go down .1 So Newman with his low interception totals would be maybe surpassed by guys with higher YPA’s but who get double-digit interceptions.
Springs had 4 ints. and would have a final score of 6.0 for last year.
Newman aklso only had 4 so would score 5.9 for 2007.
Cromartie led the league last year with 10 interceptions. For arguments sake, say he had a 7.4 YPA. His final score would be 6.9, still behind but less so.
FIRST!?!
Also, in trying to work on that, looking up stats from last year (only), None of the guys on your list show up with many interceptions. As a matter of fact, ints. were low last year, Newman isn’t that far behind. Only 4 guys got 7 or more interceptions!
Maybe a more conservative style of “Break Don’t Bend”, cover 2 defense?
Or just an anomaly.
This is why YPA by itself is not the whole story. No one with a brain thinks Springs is as good as Newman. And Townsend #3?
One point to be made that impacts Newman as well is that he gets thrown at a lot less. One could then look at the number of throws at a CB as part of the equation as well. # of throws, YPA, and completion %= I think you mix all of them together and also factor in yds given up and TDs scored. I frankly think INTs should NOT be part of the measurement at all- maybe only used as a tie breaker.
Say, Jammer is on that list. I remember the year we picked Roy, many people were expecting us to pick Jammer. He seems to be picking it up the past two seasons.
burmafrd, I can’t speak for Springs, but most people, which means fans and some analysts, don’t think Newman as good as the other flashy corners out there. They think of Asante, Cromartie, etc. I remember there was a vote taken on NFL.com on who was the best corner, and people picked Asante and Cromartie over Newman. I think he was last.
Ugh! Lies, Damned Lies, and Statistics.
What could be more misleading?
Champ Bailey, clearly the best CB in the league, gets thrown on twice a game gets one 15 yard out completed would be “ranked” lower than Jacques Reeves, who gives up 15+ passes that average 7 yds. (example, folks).
Like Horse racing, there’s different horses for different courses.
Raf’s last statement about schemes and weighting has a lot more intrigue. A cover two team will never give up the big plays that an aggressive man-to-man group would. They also won’t have the picks.
I’d like to see # of times thrown to, picks (and all turnovers), tackles, and tackles for a loss (incl. sacks) factored in…though I can’t imagine how.
Like GM Jerry, I’d go off manager/scout rankings that can (albeit subjectively) include those other factors in over any single stat.
Real,
YPA by definition factors in number of throws at the CB. Less throws means the value of a big completion has a more meaningful impact on the YPA.
I like the idea of factoring in interceptions, but it would have to be a very small factor. This is one area that does factor in the number of throws at the CB, less throws result in less interception opportunities. Also, interceptions happen for lots of reasons, some of which do not have anything to do with CB coverage skills.
It’s hard to conceptualize a better measure of a CB’s effectiveness than YPA with minimal variables.
Yeah, YPA’s a pretty good stat.
If you could add something to it is the percentage of plays in which the CB makes a PD or an Int. Newman would also rate highly here.
We expect a full report about your vacation with the blonde upon your return.
Rafael,
Ken Lucas is a good football player. Yes, he is better than Henry but do you really think a mild upgrade like that would have made that big of a difference to keep the Tuna around?
I mean sure Lucas would have helped but one man doesn’t win a Super Bowl, nevertheless keep a coach out of retiring.
tuna was done coaching.
i think henry was a good signing who would be a great signing if he could just stay healthy for a season.
I think YPA is a good starting point for evaluation, but it’s not nearly as useful as a baseball stat like OPS. In baseball, you can measure the one-on-one battle between batter and pitcher and give it a precise, repeatable quantifier. Plus, good hitters will all have about the same number of plate appearances, so OPS is a very good comparable measure. (Runs created, or runs created/27 is better because it factors baserunning, but that’s another topic.)
So what would improve on YPA? I think INTs should be factored, and one way to do it would be to subtract the return yardage for the intercepting CB from his yards against base on which YPA is calculated.
Also, if a CB gets an INT and returns it for a TD, that TD ought to be subtracted from any TDs he surrendered in coverage. Of course, TDs surrendered are not in the YPA calculation, so you can’t take them out. But it might be nice if you could find a way to do so.
I’m just thinking off the top of my head here, but here’s some possiblities to improve YPA.
Why not find a way to merge YPA with Yards per Game and Points per Game? Then offset these numbers with yards on INT returns and points scored on INT returns. By adding yards per game and points, you begin to factor in the value of CBs who are so good that other teams avoid throwing at them. You also value the CBs who keep their receivers out of the end zone. I’m not sure how to merge these factors — perhaps there is a formula that could be worked out similar to runs created in baseball.
Of course, in the end, it still wouldn’t be like baseball, because there are very few one-on-one measurables in football. Even a better measure wouldn’t account for double coverage, cover two, speed of the pass rush, abilities of others in the secondary, etc. But you could get to a measurable where the right guys would at least tend to gravitate to the top.
Finally, Rafael, I certainly hope we get to see Adam Jones’ numbers somewhere in this analysis. He doesn’t make the 3-year cut off, so I worry we won’t. But we ought to at least put to rest the nonsense that Anthony Henry is better than he is.
Jones’ numbers were fantastic when he played for Tennessee I believe. And if you add in his return yards off picks, they would look even better. I’ll be disappointed if, with him and Newman in there, plus our pass rush, we don’t have the best secondary in the NFL this year.
As Rafael said, the single statistical analysis is simply a starting point in the discussion. Statistics don’t lie, but when misused they are great for supporting incorrect conclusions.
K.C. Joyner takes on an incredibly tough challenge when he attempts to quantify football the way statisticians have quantified baseball. Baseball is a game of controlled and therefore quantifiable situations. A batter faces the same pitcher / defense that his counter parts. He faces them with the bases empty and no outs and he faces them with two on and two outs. The results can also be quantified, the player was on base, was out, runners advances, scored or they didn’t. It takes some work, but there are a reasonable number of things to track.
In football, the plays are an exercise on organized chaos. Twenty 23 moving parts per play (players + ball), each with their own strength and assignments. When comparing results you can only focus on a limited number of the variables. When analyzing pass coverage; K.C. Joyner chooses route length, quality of opponent, down and distance, percentage of passes completed and average yards per catch among others. To his credit he chooses important factors that are possible to quantify. What is much harder to figure in is scope of each player’s coverage assignment, the overall defensive scheme, the effectiveness of the pass rush on altering the timing or throwing motion of the pass, when a wide out miss ran route and a number of other human elements.
In many areas of the game Joyner has been able to quantify things that were previously subjective. Often his statistics confirm what is already believed about a player and in some cases cause a closer look and give reason to question conventional wisdom. He is the first one to highlight the factors that are not considered in his statistics and avoids using absolutes in most of his conclusions. His approach adds value to the discussion of comparing players and maybe eventually various schemes. In more then one instance his statistical analysis was good enough to highlight something that professional scouts close to the player knew, but fans and the media where a step behind. That said, the area of individual player pass coverage is particularly complex and currently there is not enough statistical data to quantify it enough to make any strong, unobvious determinations about individual players contribution to their teams pass defense.
It’s a work in progress, but until player assignment and the pass rush affect can be factored in he is working with about half the relevant data.
i agree with whoever said that about the ypa not being able to take into account that certain dbs don’t get thrown at often. when they do, the completions weigh more heavily than a db that gets thrown at more.
Trey,
You’ve hit the nail on the head. It’s a team game. Overall team Defense is really the only way to measure the true performance.
A CB that takes away half the field like Deion used to is far more valuable than most…though it adds nothing to his statistics. A great pass rush can mask even the most anaemic players.
The only measure is how a team does against the pass. You can have Deion and Woodson in their prime, but if the other two guys can’t cover, and your system can’t mask them, then what does it really matter?
We had 3/4ths of our secondary go to the Pro Bowl. Yet as a group, they get very little love. A Chicago mediot says their tandem is best.
http://bleacherreport.com/articles/32196-tillman-and-vasher-are-they-the-best-cornerback-tandem-in-the-nfl-right-now
Even the Yankee-homers DMN thinks we’re barely top 5.
http://cowboysblog.dallasnews.com/archives/2008/07/cornerback-rankings.html
Me? There’s no way there’s any group better than what we have. We’re freakin’ 5 or 6 deep at CB and at least 3 deep at Safety. You can dog Roy Willy all you want, but 2/3 of the Pro Bowl voting is players and coaches and he’s been to 5 straight. Reeves & Nate Jones were the problems (along with the injuries) and they still finished top 10 for the year.
“This is why YPA by itself is not the whole story”
That post gets to the heart of the CB equation. YPA is only a piece but if there is no correlation with TDs, first downs, ints then is has to to be mixed with these other parameters. By the way, Burmafrd’s numerartor should be scaled up by %30
Fighter should get the extra %30 too
I typed most of this and then left for an hour. After recent post this is kind of like beating a dead horse, but it was typed so here goes anyway…..
YPA is a relevant as one of many indicators of effectiveness, but it is not even close to being a single statistical measuring rod that OPS is in baseball. It simply ignores too many critical factors, some of which are not possible to track statistically track without inside information that teams are not going to provide.
A really talented corner may be frequently assigned to cover his wide out in the short flat and anything down the sideline or deep middle and in addition he is to help cover the deep middle zone for other receivers when his man goes inside and is picked up by the linebackers. By handling such a large responsibility he allows his the free safety to cheat to help a weaker corner on the other side and he consistently protects his strong safety from getting burned deep by tight ends on seam routes. His YPA may be unexceptional because of the 20 passes he gave up over the course of a season four of them were long ones when he was trying to recover on his man after helping out a teammate.
Likewise, a really weak corner may typically get linebacker help on slants by underneath and safety help over the top can end up with a good YPA and decent interception count. In the meanwhile he is significantly hurting his teams pass coverage and flexibility to blitz. His YPA does not reveal that he consistently gives up a steady diet of nine yard outs on third and eight and he needs so much help that his teammates are often are put in very difficult situations.
Even that does consider that a corner playing in front of a great pass rush is generally asked to cover for 3 to 4 seconds while one playing behind a lousy one may have to cover for 4 to 6 seconds in the same coverage scheme. It also fails to account for the “ball-hawk” factor that Rafael mentioned. Bottom line is I think that YPA has a very limited value when examined in isolation.
Hi all, I’m a long time lurker but first time poster from the UK - this will be my 8th season following the NFL and third following the Cowboys.
I’ve just started reading “The Blind Side” by Michael Lewis and I’m looking to line up my next read - I’d be interested to hear any recommendations from the regular posters on Cowboys themed books and Football books in general.
SmegzOr,
Joyner has a brand new book coming out called, “Blindsided,” which will address the thesis of Lewis’ book and other football myths. It has some good early reviews. I’ll have him discuss it when I return.
Fighter, re: Bailey,
Uh, no. Bailey got about four balls a game last year. 60 total attemps, roughly the same number Newman saw. But his success ratio was only 31%. In his monster ‘06, it was 54%.
The guy’s level of play dropped.
If you want to say he’s the best based on ability you could have a case., But when you go on performance, and for its warts YPA is a solid indicator of performance, Bailey is the kid who gets 1600 on his SATs and parties, figuring he can study an hour the night before and get by on his smarts.
His performance lacks. And he was god awful in ‘04 and ‘05. He was worse than bottom half. He was bottom third. You can claim a lot of external factors played into that, and the Broncos have had a mediocre rush lately, but he’s been healthy.
Terence Newman has ranked in the top 10 twice and the top 15 a third year. In his “down” year he was 31st, which is still top third.
Bailey was first overall in ‘06, bottom half last year and bottom third twice.
He’s not the best corner in the game. They play you to play, not preen.
All these variables have potential problems, except YPA. YPA is as good as it gets. It weighs tangibles and intangibles, different coverage combinations and options, different personnel decisions, game planning by your coordinators and the opponents, and factors it over any time frame and weather.
It’s all about averages and non arguable factors. When you throw in an arguable factor and assign it any weight, you shoot down the indicator.
When a team doesn’t throw at a CB, it indicates your opponent respects you. A good CB normally takes the best receiver which sez your own coaches respect you. Any CB may or may not have help with a safety, or get a LB bump on the LOS. Down and distance, time left in the game, and score, all affect how you play your CBs. How do you assign a value to that? You don’t. You simply use as large a sample as possible, and use as many inarguable numbers as you can.
Over the course of a season, your weak players will be attacked and the good ones will prevail, in YPA that is. Simple as that.
Trey,
I appreciate your detailed posts. But rather than throw up your hands and say YPA isn’t enough, what’s needed is to see if some of these numbers can be refined and improved upon. Heck, when did Bill James come along in baseball analysis? Seems like Joyner is trying, but he needs help.
For different reasons I agree that Bailey’s play has dropped off. The team pass defense has been below average and he no longer has an exceptional level of responsibility. Denver plays him strait up and he sometimes gets cover two help and he is still giving up plays including TD’s and thrid down conversions.
If Bailey’s not the best corner then who is? Who ever it is they are very close in talent and production to any other top ten corner. There is a nice sized pack of very good, but not great corners in the NFL right now. I love Newman and thought he deserved two of D Hall’s pro-bowl invites; but none of the year 3 to year 9 corners inspire Hall of Fame conversations.
Since Primetime only Champ Bailey and Ty Law have played at an inspiring level for more then one year at a time. Antonio Cromartie had an interesting career start, but we need more time to know how his carrer will play out.
Rha,
We all like things to be simple. The problem with using simple neat mathematical tools to quantify complex subjective phenomena is that it leads to a lot of false conclusions.
The issue is that YPA does not do a consistent job of differencing good corners from the bad one. From a statistical standpoint it is highly vulnerable to outliers and it does not account for many things that are critical to distinguishing good play from bad play. Really terrible play generally show up, but if YPA tells you something that contradicts what is commonly believed by coaches and scouts then it is probably not true.
VAFan,
I applaud Mr. Joyner’s efforts and think that he and football outsiders have made a lot of progress in how football is analyzed. The area of individual pass coverage is more complex then analyzing QB play, pass protection, pass rush etc. In general we can determine with some degree of certainty the assignment of offensive players on most plays. It is easier to compare situations and players for different teams. Pass coverage is a team activity that is covered with some degree of secrecy. It is very difficult, to quantify the impact of individual play.
I do not think YPA + % targeted + INT + penalties yards + total yards given up + 3rd down stops + Touchdowns given up creates a complete picture; but a well thought out score based on all those things would create a decent base line for comparing players.
VAFan,
To your challenge of comming up with an idea, maybe the ansewer is to change to the goal to measure the effectivness of a teams overall pass defense.
It would be much easier to quantify the amount of time the QB had to pass in varius situation and what the results of those passes where in those same situations. Blitz packages could also be factored in as well as the quality of QB play being faced.
If the pass defense as a whole had a meaningful score it could be taken into account when examining individual play. Individual scores would have more meaning in a team context.
After seeing those stats, I’m kind of feeling bad we got Roy instead of Jammer. Jammer seems to be getting better, and he would be perfect right now. I remember he had a good game against Indy and the Patriots last season in the playoffs. On the other hand, Roy was good when we were really bad. Does anyone see Roy improving this season? In addition, Jammer would have gotten the chance to practice against TO, Glenn, and Romo versus Rivers and whoever.
Trey, that’s the point. There are too many variables that are affected by too many variables affected by too many variables. The only options are to not measure at all, or measure only using indisputable factors.
You can’t statistically measure a CBs effectiveness factoring in all these variables accurately, without recording and weighing everything from wind velocity, rainfall or snowfall, temperature, QB quality and injury, WR quality and injuries, down and distance, score, DL quality and injuries, OL quality and injuries,your own defense quality and injuries, all of which would require the same person viewing every game and quantifying everything evenly and without predudice.
Not gunna happen.
Rha,
We agree on the problem that individual performance in pass coverage is too complex to model with a mathmatical statisitic; but we disagree on the solution.
I think we are better off with a subjective model on individual cover play then with a “objective” one that is hopelessly flawed.
Since pass protection is a team based effort, a team based model would be more accurate and useful. There is not much out there on this either and it would be useful to evaluate this aspect of the team game. There are lots of good questions to be explored; what the pass coverage trade offs are between a one gap 3-4 and a two gap 3-4? What are the average benefits (improved time to hit QB, turnovers, etc) of blitzing when compared to the costs (converions, TDs, ect.) for different defensive packages? In what game situations does it clearly benefit the defense to blitz and when it the risk much greater then the benefits? How much did the Cowboys pass defense improve in 2008 after adding Pacman and sitting Roy on the bench on third and 5 or more? To me those are a lot more interesting, useful and quantifiable then if Smoot is really a top 10 corner. (That would be a no)
Say what you will about YPA and the impossibility of measuring a CB’s performance, but I doubt you find this level of intellectual analysis of statistics on many football blogs!
Good info for a thesis, Raf?
No one’s mentioned tackles. What if a CB has 70 tackles and is great in run support??
rha:
Is that “Magnum”?
TO me you have to go in this direction. Take YPA, # of times thrown at, and completion %. That is just ONE collection of data.
THEN you have to factor in pass rush, and QB rating.
AND then you have to factor in the quality of the WRs he is facing.
AFTER you do all that then you add in what kind of coverage the team uses (Cover 2, etc)
THEN after you have done all that you should have a decent idea who is really good and who is not.
There is NO way that Springs is even close to TNew in being a top CB. Just no way. LET ALONE BETTER.
I just had to log in to say congratulations to Drew Henson and his biggest (read: only) fan, TEX!
http://myespn.go.com/blogs/hashmarks/0-8-63/Wedding-bells-for-Drew-Henson.html
Now if only I could figure out if those are tears of joy or heartache streaming from Tex’s eyes…
Trey,
I agree with just about everything you’ve said here. But I wonder, what the hell do teams use to evaluate CB talent? Both for drafting purposes and salary purposes? I would think there would be a great deal of interest from the players and the owners standpoint to be able to measure one CB against another on some reasonable statistical basis and not just on “street cred.”
Plus, not just for salary purposes, but depth chart purposes. Do they just eyeball guys and say — well, he’s better? Or do they reasonably try to evaluate whether that’s true on the field?
You may not have perfect measurables, but why haven’t they taken some of Joyner’s ideas and built on them?
As a baseball fan who loves that game in part because it breaks down so beautifully to statistical analysis and debate, it’s frustrating as hell to try to apply those same concepts to football. It’s maddening. For example, why does the team that gains the most yards in a game often lose? Why do rushing yards seem to count much more than passing yards? What is the value of a turnover? Etc., etc., etc.
Still, I think some kind of formula considering the factors you listed above (YPA, completion percentage, INTs, 3rd down conversions, etc.) would at least advance the debate Rafael is trying to start — namely which CB would you rather have on your team? And that’s a fun debate to have, even if it’s inherently flawed.
BTW Mr. Vela — I’m still hoping you clear up the nonsense that Henry should keep his starting job over Adam Jones.
VAfan
VaFan,
Teams are gathering and analayzing statistics and they can use them to ansewer some of the questions you highlight, but so far they have not figured out how to determine which DB is going to do well on their team. Currently they have to rely on scouts and it seems to be a very inaccurate science as every year some 4.3 guy gets picked early in the draft and never becomes a decent corner and a few free agents change teams and never produce near as much as the team hoped for when they gave out the huge signing bonus.
Coaches have a huge advantage when it comes to evaluating their own players in coverage. They know what they told the player to do and they know to how well he did waht they asked. They know which guys leave gaps in coverage and require extra help and who makes a metal mistake one in five plays and who makes one every couple of years.
VaFan,
Your question to Trey “How do teams evaluate CB talent?” is extremely valid and gets to the crux of the matter.
Elite CBs are making QB money for a reason…because their so rare.
From what I can surmise, Speed, Ball Awareness/Aggresiveness, and Size seem to be the main criteria.
Speed and size are simply measurables and like the WR position, don’t tell the whole story. But without at least one of them, it’s a very rare individual that can succeed.
The Ball Awareness criteria is not so much a measurable, but rather an observed evaluation of coverage “talent”. Passes Defensed, INTs, and # of attempts are all resultant measures of this talent. YPA may very well be the best measure, but like most statistics, I believe it to be misleading.
Like the QB position, would you rather have a statistical stud like Marino, Favre, or Fouts or an Aikman, whose balanced offense and deep strike game plan never allowed him to put up the same numbers? Is it completion pct (where the West Coast dink/dunk QB excels)? Is it QB rating (where West Coast QBs also excel)? Or is it total offense, where when Aikman was at his prime, the Cowboys were top 3 every year?
And really, does it matter? Does it matter if we have the best, second best, or 10th best CB? Hell, NE hasn’t had a decent CB in quite a while. Yeah, Ty Law was decent while there, but since? I profer that the scheme made the man, not vice versa.
I’ll take our unit every day and twice on Sunday, even if T-New is only middle of the pack each year. The whole average thing skews the ranking, but that’s for another rant.
Real,
Since I got the black lung, I can’t feel my face anymore. I think this is Blue Steel though, the difference in it and Magnum is very subtle as you know.
Before someone points it out, QB rating is probably the closest thing to an equitable measure because it takes in almost all stats. Though it is flawed, it still is a better measure than yards, tds, or completion pct alone.
However, GQ a few years ago wrote an article on the QB rating found here http://www.bluedonut.com/qbrating.htm
One passage talks about the West Coast bias:
“But most influential of all was Bill Walsh, the QB-nurturing 49ers coach. The West Coast Offense he pioneered in the 1980s turned precisely timed, high-percentage short passes into catch-and-run long gains. You couldn’t invent a better scheme to juice QB rating numbers. Because, it turns out, the formula mathematically whacks guys who try to throw long. ”
On it’s true measure:
“Is there a perfect way to enumerate the performance of one man on the field with 21 others? What about perfectly thrown, dropped passes? What about John Elway, who ended his career with a saggy rating of 79.9, even though he had the most career wins by a quarterback and a record 41 fourth-quarter game-saving drives. What about Troy Aikman (81.6), who had Emmitt Smith running behind him, so he didn’t throw a lot of four-yard touchdown passes, “even if he made the pass that got the ball to the four,” says his agent, Leigh Steinberg.”
And lastly, though it’s conclusion is that at least it’s a standard, they talk about the contrary nature of stats in football:
“Football is a team game, a game of drives and momentum. Individual numbers struggle to describe it. Still, numbers are all we have. Statistics are what separate sports from just playing around in the yard. Coach Lombardi said it: if you’re not keeping score, you’re just practicing. So you calculate what you can.”
Fighter,
I think Troy Aikman developed his own QB rating system, which I think he did to address the problems you raise. I can’t recall the details of his method vs the current.
My last to try to get it back to CB.
QBs, when evaluated for the HOF, are not measured by their QB Rating; otherwise, Troy would not be inducted. They are measured by 1. Championships, 2. Wins, 3. Yards…in other words, team success.
CBs are measured by their ability to change a game. INTs are huge, but really, those great ones are measured by their ability to shut down their receiver. Never is YPA used.
If I had to pick a stat, it would be # of catches given up. Like Wade says, T-New simply doesn’t give up many and seldom a deep ball for a TD (1 in the last two years). That has to be tops in the league.
I’m all verklempt…talk amongst yourselves.
With all the factors people are asking to put into rating a CB, I’m thinking to myself, is there any rating used by the NFL that can holistically evaluate a player? We need PhDs to come up with the ratings you guys are asking. The YPA gives a ballpark of how good a CB is.
Obviously YPA by itself does not do the job otherwise Springs would not be on top.
Here is a little question to the oil dealing hombre Tex….
If Stewart isn’t calling the shemes and he’s not really hurting Dallas.. then why in the hell do you have a problem with him?
If he hasn’t had a negative effect on the field then why would you consider him the weak link?
I have no problem with someone being critical but you really don’t give a reason why. You hate the fact that he has the title? What the hell was he supposed to do say no? Generally ALOT of people say no to when defensive coordinating jobs are offered to them. To me it sounds more like your jealous.
And to say that you would go face to face with him is just silly.
Your not man enough to admit when you wrong on here, I would doubt that you would be man enough to get in a grown mans face.
AFB, I know you have taken this on as a personal crusade but you are only encouraging him.I think he and Donny push your buttons ad nauseum and you never fail to respond like an annoyed little kid. I am positive this is the very reason that they both post such inflammatory statements because they know some oversensitive on the board will get all riled up. Do us all a favor and like the rest of us try to ignore the button pushing. I know you feel safe attacking those who spew unpopular opinions but responding only validates their arguments. Is anyone but tex arguing this point about stewart? Just let him have his rant and if no one responds to it it will wither up and die.peace,B
I appreciate your enthusiasm for YPA but how do you differentiate between a CB who faces more West Coast offenseive schemes comapred to a CB who may face more teams that throw the ball further down field.
A CB whose division opponents may all deploy a similar West Coast offensive style would be covering a lot of dinks and dunks that would pull his average down every time he was asked to cover a 5yd pass.
It really has more to do with boredom then anything.
Until training camp comes along that is.
And I really never saw it as attacking someone stoproyce.
I guess asking someone to give a valid reason or asking someone not to tell you their political beliefs is attacking?
Yeah, all right.
Someone said something a few posts ago about Chris Simms coming to Dallas, and Simms has openly admited that he would love to go to Dallas.
JJ says that he wouldn’t find the need for another backup QB, but do you really want 40 year old brad johnson coming in for a few games?
i feel you man ,i do understand. Its just getting tired . Its a pointless argument, he has no reasoning you know this, and its hardly worth to your effort thats all.
Point Taken.
Anyways, I know this is a little off subject, but would anyone want to be Green Bay front management right now? If I was Aaron Rodgers I’d have a hard time concentrating on football.
I think there was an article this morning (Dallas morning news or the other dallas paper)about the team actually looking to Simms as a possible replacement to johnson. I’d tend to prefer Culpepper as he has actually accomplished something besides having internal organs blown up mid game. But i bet he’d want way more money than Simms. i am just not sold on Simms. But i do find it encouraging that this is being discussed at the ranch.If romo went down we’d be pretty mediocre with Johnson.
Especially if they let him go you know he’d end up in minnesotta or chicago. Awwwww poor packers.
I didn’t read all the posts but I agree that this stat alone is not a great way to rank CBs. Besides the fact that a shut down corner only gets thrown at a limited number of times a game, the tempo of a game or an offense would have an effect on the ypa. If our team is passing for 300 yards a game and scoring on average 4 td’s a game, the opponent will also be game planning to keep up with the high powered offense they are facing. So the game becomes more of a shootout and defenses start making sure the opponent doesn’t get a cheap score. They keep everything in front of them and probably give up the 5 to 7 yards a play. In addition, if you are way up on someone they are going to do the same thing.
Stop, I bet that Green Bay would somehow work it in that they will let him go or trade him, but that he will only go/or can go to someone outside of their division.
The problem with Culpepper, I think he doesn’t realize that he’s still going to just be a backup QB. I personally would like him as well, he at least has an arm.
I do think as far as corners go… that Champ Bailey gets a free pass quite a bit.
Does he have more INTS usually?
Yep.
But I don’t remember off the top of my head, a time when Newman has been burned straight down the field for a TD.
Maybe a crossing route, or maybe a missed tackle but not just straight up burned.
I know of two times that Bailey was torched in the past 4 years. Once Chad Johnson ran right by him and last year when Jennings scooted right past him.
And it wasn’t because he was going for a big play either, they both ran right past the guy, i don’t even know if I remember a double move being made on him.
Figther who did Newman give that long TD pass too?
I would think total yards against a CB would have to be some how incorporated into a calculation that included YPA.
Cash do you mean receiving yards?
Because that is included.
If your talking rushing yard at their side, then thats a totally different thing. That really doesn’t measure their coverage skills.
The truth of the matter is that there isn’t any more lockdown corners.
Part of that is because the rules have changed to help offenses out alot more, but even without those rule changes, there isn’t anyone even close to Deion Sanders.
My bad by the way that was Dre Bly that got burned by jennings.
Oh wait a minute, I wasn’t completely wrong.
JAMES JONES, a rookie burned Champ Bailey for a TD in that game.
Not Simms, he couldn’t even take the job from Applewhite.
Champ, I saw him get burn on the first play in OT by Jennings for a TD last year, it was not impressive.
Give me Newman anyday.
Jones…sorry
Thats ok, counter. I thought the same thing. He got burned by Jones in regulation, and Jennings burned bly in OT.
If I’m not mistaken, T-New gave it up to Burress in the first ballgame.
Anyone?
YES YES YES, A real football debate. Thanks Raf why don’t you just ask us to figure out which came first the chicken or the egg? (chicken). This is just what we needed! Thanks again Raf and K.C. for the advance stats. I believe there is no BEST corner, there is however a couple of real good ones. Kind of like trying to choose between Carrie U. and Jessica S. Both not bad!! The best scenario would be to have both at the same time– T New. and A Jones shuttin them down!!!!!!! Can’t wait.
This year?
He was hurt in the first Giants game. Remember Jaques Reeves was covering Plaxico.
You’re right AFB. Maybe the 2nd game in NY? T-New was out for the first game. I remember the play, don’t remember the game.
The point is that T-New has given up nada in two years.
Now, the other point is what CB group (as we’ve come to find out the hard way…you need 3) would you rather have?
With Pacman as the third (I’ll get to why in a sec) there is not a better group, when healthy. Our fourth is a first round pick, or three other very promising prospects (Scandrick, Alan Ball, Evan Oglesby) for 5th, 6th, etc.
Pacman, alone, adds to every ST return unit. I mean, Hester & Deion-type impact. That’s why he doesn’t need to start in the base. He’s got other duties.
Then, the dime might truly be scary with Henry repacing Roy Willy.
I wouldn’t trade our group for any other unit in the NFL. Potential for all time great units of all time.
Trey,
Very Impressive.
Not much to say til camp starts
Donny,
Thanks for the props. I am not sure how impesssive it is as it is always easier to evaluate someone elses work then it is to propose something yourself.
Rafael,
Thanks for all your work here.
I do have a question for the next time you get to talk with Mr. Joyner. Has he considered the merits of evaluating pass defense on a team by team basis as opposed to a player by player basis. It seems that this is one area that players depend on the scheme and teammates as much or more then thier individual talents. Please see post 27 for a few ideas behind the question.
Trey:
Just like you said, to have a firm grasp on actual performance there’re a lot of factors to be taken into account in coverage when it’s about CBs…
Now think about what needs to be taken into account when we’re talking about overall defensive level of play…. YPA for CBs looks like a really close to actual performance benchmark if you think it through, because there’re a lot of intangibles like this: The best CBs will see the best WRs here and there, they won’t see them in an every game fashion. The best Defenses won’t see the best Offenses, either. How do you measure the difference, the changing level of play and the rivals that they aren’t facing? I would call that nearly impossible.
Chandus:
Very good point. Thats one of the reasons why I think that Asante Samuel AND Champ Bailey are overrated.
And for that matter I’ll throw in Al Harris who is constantly praised by people.
Bailey’s division isn’t one loaded with great wr’s OR good qb’s. Same really with Samuel.
Harris got a probowl nod, but woodson is way better. Garret abused Harris by putting Owens into motion.
I’m not saying that any CB gets to face great WR’s year in and year out, but the AFC West, and AFC East might have some of the worst combinations of QB/WR duos except for the patriots.
I’m not saying they aren’t good, I’m just saying that the constant ass kissing of Bailey isn’t warranted. The guy got ran down by Ben Watson in a playoff game. Not that Watson’s slow, but come on.
Football Outsiders have the concept of DPAR for teams, OL, DL, QBs, RBs, WRs, but not CBs.
Basically, they adjust performance based on situation (a 5 yard pass on 1st & 10 is worth more than if on 3rd & 7), opponent, and weights them vs. other replacement value. It’s complex and they keep adjusting the formula, but they at least try.
I guess my point is that the consensus best CBs based on scouts, personnel guys, and mediots don’t match the YPA rankings. SI did an article in ‘06 that supports my Champ Bailey argument at http://sportsillustrated.cnn.com/2006/writers/jeffri_chadiha/11/21/shutdown.cbs/index.html
Football.com has their rankings here http://www.football.com/top10/top_10_cornerbacks_in_the_league.html
Twominutewarning.com has an interesting rating for team pass defense at http://www.twominutewarning.com/cb2007.htm
USA Today polled 9 national writers for their Top 5 rankings. Champ Bailey had the biggest margin of any position ranked. In fact, he received every first place vote. http://www.usatoday.com/sports/football/nfl/2008-06-22-top-five-cornerbacks_N.htm
So either every mediot is stupid (a possibility) or there is something inherently flawed with YPA.
if we want to go in the direction of most over-rated corner, harris has my vote.
#2: d.hall
Chandus:
There are clearer metrics for measuring offense units pass performance. I like Aikmans QB numbers for QB play alone, plus there are also clear team performance measurables such as 3rd down conversion rates from particular distances. These could be used to provide some comparative factor for the qualifying the quality of the offenses each defense faced. I think Joyners and others work create the basis by which to baseline team pass offensive effectiveness. It would help if pass protection could be isolated in terms of time to throw could be separated from the results of pass attempts when given time to throw.
Applying statistical analysis to any free form activity is challenging, but there are a lot less components to consider when you are evaluating units verse each individual player.
Pass defense could be broken down in to pass rush unit and pass coverage unit.
The pass rush can be measured by the amount of time the QB had before throwing the ball (measurable), how often the QB was hurried (currently tracked stat), how many men rushed (measurable), how often the QB was hit (currently tracked), how often he was sacked (measurable), how much time passed before the sack, etc. These numbers can be placed in context (down, distance, game situation) and compared to how other pass rushes compare against the same teams.
With the pass rush quantified the situation the pass coverage unit is in can be quantified as well. For each situation the result of the pass defense can be measured. Pass completion of X yards, no YAC, pass completion of X yards, Y YAC, pass deflection, pass dropped, pass off target, pass thrown away, pass intercepted. The statistics could be complied and analyzed in multiple ways.
It would take real statistical analysis and a lot of hard work to come out with formulas that consistently produce understandable, meaningful numbers. The hardest part of any data collection effort would be determining the amount of before each pass play resulted in a pass attempt or a sack.
This kind of data would also be helpful in evaluating individual QB strengths in weaknesses in handling differing coverage package and blitz packages.
Trey,
Again, all good ideas. But why don’t there seem to be guys out there tracking these measurables and then devising formulas that work, besides KC Joyner I mean. In baseball, there are hundreds or thousands of those guys. Seems like football is still in the relative dark ages analytically.
VAfan
In baseball the league tracks and publishes a lot of the statistics. In the areas where the league was not tracking the needed data lends itself to an objective, repeatable information gathering process. The fan statistician biggest challenge was to do the data analysis.
In football KC Joyner has to watch film of hundreds of games to collect statistics that are not being collected by the league. If they are being collected by teams they sure are not being published. Supposedly San Francisco was doing a lot of special data collecting and data analysis during the Bill Walsh days. I do not know what they found was helpful to the bottom line, but either way they did not seem inclined to share what they found.
The largest obstacle to the growth of a baseball like stat-nerd community for football is the lack of meaningful data collection beyond what is interesting at the ESPN sound-bite level. Even if the NFL had access to the data, I doubt they would publish it as I doubt the marking guys believe that an open, effective effort of demystifying football is in the best interest of the league.
That is why KC Joyner truly ambitious data collection efforts are impressive and relevant. At this point he and his ilk are fighting an uphill battle. That said, I think Rafael is right in thinking that his approach will grow in popularity as it matures. It will also build on itself as more and more people get interested in quantifying the game.
Just read that article from Chaditha in Nov 2006. No mention at all of Tnew but he says Hall could be a shut down. Now I KNOW never to read anything that MORON ever writes again.
Random thought …
I’d rather have a bottle in front of me than a frontal lobotomy …
We now return you to reality …
Fighter,
Thanks for the links. The two minute warning link was interesting. The statistical based rankings actually line up with common perception of who has good corner play.
In that USA today article, the word that jumps out the most is interceptions. In fact, that’s the first thing they mention about Champ Bailey. I’d rather have a CB who doesn’t give up TDs or long passes than a a CB who gets a few interceptions a season.
I can’t believe Al Harris is on that list. Is it his long hair that got him in? TO burned Al. Al’s interception against TO was pure dumb luck. TO also burned Cromartie in the Pro Bowl.
As Raf mentioned, the YPA is just a starting point. If we are going to say that the YPA is not an active reflection of a CB, we might as well say that the average yards and total yards of a RB are not accurate, the sacks of a LB are not accurate, etc. YPA is average, and I’ll take it as a good ballpark metric.
Fighter,
Some of the voters making the rankings on the USA site were idiots.
(The current list includes rookies and is not a lifetime accomplishment list) Among current corners Newman was ranked ninth, four places behind Trufant. Junior Seau recieves votes as one of the top 5 LB in the game today. Roy Williams recieved a vote as a top 5 saftey.
On the all time list Ronnie Lott recieved 5 vote points for the best CORNERBACK of all time. Mel Blount is listed as the best cornerback of all time, he is not even the best Steeler CB. Everson Walls does not belong in a best of time discussion either. Jackie Smith, Ben Coates and Raymond Chester recieved votes as best all time TE.
Lot’s of votes being cast by people who should not get a vote.
Many people also seem to forget the years when Champ played in the NFC EAST. HE faced Joey Galloway twice a year and I clearly remember JG getting the better of him. Thats with Quincy Throwing the ball!!!!! Dont Forget T.O. burning him for 90+ yards(are you kidding me?) Anyway they say he was nursing a hamstring injury and wasnt 100%..however I remember Newman being dinged up and still catching flak for a poor game against the skins…same rules should apply right?
ym,
YPA does not have the same relevance or credibility as the other statisitics you compare it to. Most of those other statisitics cannot be debated as to what the number should be. Two different people calculating the YPA on the same corner will come up with two different YPA numbers. Its a number based on an interpertation if that reciever was the corners responsibility at the point in the play where the ball was thrown. If one of the evaluators was the coach who knew what the corner was assigned to do the number would be widely different then an outsiders assesment. What Joyner is doing is cool, but YPA is not currently a scientific number.
There is no such debate on average rushing yards for a RB or sacks for a LB.
Raul,
I always heard it as “I would rather have a free bottle in front of me than a pre-frontal lobotomy”
Well, one the things I’m noticing is that the level of performance of opposing WRs and offenses should be taken into account when evaluating a CB. Basically, a lot of factors beside the number ball a WR has caught and how many yards the CB has given up. A RB has 1500 yards and a 5.0 average? What defenses has he played against? A LB or DE has 22.5 sacks in a season? What offsensive lines has he played against?
Hey remember the “hey we might ought to look at getting matt jones if he gets cut” talk?
He just got caught cutting what police believe is cocaine in his truck with a credit card.
Probably can put that thought away now.
Sean:
Yeah, I’ve heard it both ways, but never did take to the “pre-frontal” part… What’s that — the medication or the cutting of the scalp?
Trey:
That’s were I believe that you’re wrong…
“There is no such debate on average rushing yards for a RB or sacks for a LB.”
In the first scenario let’s talk about an Offense that in it’s schedule has the 3 or 4 of the Top 5 Run Defenses in the League in that year, average per carry becomes questionable if you know that there’re teams that faced only one of those Ds?
In the 2nd scenario, it’s a shame that Ellis, Ware and Spencer didn’t had to face the Eagles in the Giants game where Winston Justice decided not to show up.
There isn’t anything close to being a reliable benchmark in Football, becuase in the end those use stats and stats can be misleading.
Oh, and BTW, if we think it through, then what Shawn Springs has done in Washington is a lot more impressive….
Their pass rush is and was poor, everyone knows that both of the Safeties could be burned (including when Taylor was on the field, RIP), his partners have been Carlos Rogers (toast) and Fred Smoot…. And last year they faced some of the most powerful Offenses in the League, including our team, twice.
I’ve always respected that guy, he’s a really really good CB, one who’s VERY underrated.
Chandus,
You are failing to distinguish the difference between subjective interpretations on the meaning of a given statistic and having to make a subjective interpretation in order to derive the statistic in the first place.
The meaning and importance of a given statistics is generally subject to debate. There may be debate about how good a running back is who got 5.0 yard average. People have argued that Barry Sanders 5.0 was worth less then an Emmitt Smith 4.5 because Smith did not lose yards and provided his team consistent results allowing for a sounder game plan. In baseball a 300 batting average in the National league is considered better then the same average in the American. Those are opinions; they may maybe well grounded, but they are opinions none the less.
There is not a debate on what each baseball player BPA. BPA is factually derived by dividing the number of hits by the number of qualified plate appearances. Discussion or debate are not needed to determine which running back carried the ball on a given play, or how many yards the league credited the team for the play. Rushing average like batting average can be established in fact.
A cover player’s YPA is measured based on how the yards given up on plays that they were responsible to be covering the guy the ball was thrown to. Who was responsible for the coverage is very difficult to objectively quantify and is very much subject to debate. The reviewer has to make a personal call when deciding if the open man on a 25 yard seam route was the free safety responsibility or was it the left corners. Unless the film evaluator knows the players assignment then it will be “wrong” a good amount of the time. Any two evaluators will come up with two sets of numbers.
In addition to the subjective nature of how the statistic is measured, the data it produces lends itself to conclusions that are counter-intuitive and highly questionable. Not only do the best corners fail to consistently end up ranked ahead of corners who are known to be vulnerable; it highlights an inconsistency in almost all the player’s performance that is widely believed not to exist.
Teams do not pay corners 10 million a year because they think the performance they get is relatively random and in any given season he may be out performed by 10 – 15 corners that make less then a million dollars a year. It is much more likely that the radical jumps in each corners YPA from year to year has more to do with weakness in the statistic value in measuring performance then the reached conclusion that almost all of the leagues corners are horribly inconsistent.
Trey,
I believe sports statistics are designed to quantify performance to some level, for analysis and measurement. The weight you assign to that statistic is a direct function of the quality and undeniablity of the data, and the formula used for the analysis.
Some things are easier to quantify because the value of the data and the formula to analize that data, is either universally accepted or otherwise undeniable.
As your club considers a player, the stats are simply one measurement of the player. Another is the scouting, which involves, unfortuately, subjective observation. There is no perfect statistic, but bringing in subject criteria as you suggest shoots the goose. The only subjective data should be identified as such, and is the scouting.
The number and weight of variables required to statistically guage a CB is different than other positions. So you take what you can get, YPA in this case, and let (hopefully) unpredjudiced subjective analysis (scouting) take it from there.
Your ideas are okay, but they are all subject, IMO. There are simply too many variables affected by too many variables to carry anything viable forward.
Rha,
Any ideas I threw out on an analyzing pass coverage are simply suggestions of approaches could be explored. They are all subject and not well thought out at this point; but the approach did include a lot of aspects that can be objectively measured. My idea of analyzing pass defense on a unit basis may in the end, prove to be as difficult as the individual player performance. To determine if a given approach is viable and if its outputs align themselves with reality requires the effort of collecting sample data and analyzing it. That is a lot of hard work, more then I am willing to do. That’s why I asked the question of if KC Joyner has considered a unit based approach; he is a real statistician who is already doing the hard work of collecting and analyzing the data.
I strongly believe that it is possible to reduce the number of variables that need to be factored into a statistic and still have it provide useful insights that are objective. That is why I value Joyner’s contribution to analyzing football. He is not coming up with some esoteric score; he is providing new data that for the most part is objective.
The outputs of a unit based approach would also be useful for a slightly different set of questions then what KC Joyner’s work on individual contribution to pass coverage. The question is would a unit based approach to evaluating pass coverage provide meaningful answerers to questions that Mr. Joyner or his readership would be interested in knowing.
Raul,
Anything new? Where is Raf?