KC Joyner's "The Football Scientist"

[PonyExpress]

I'm considering ordering it, and wondering if anyone out there read his 2005 version and can give a review. Thanks...

 

[primetime]

http://www.thefootballscientist.com/scientific_football/booksample4.pdf

If you're just interested in the Patriots, his sample is the excerpt of the Patriots section. How lucky for us.

I plan on buying the 2006 version.

Some interesting excerpts, including Branch being one of the best WRs in football, Givens not being that great... and this.

"He was the key to the Pats defense last year. When Samuel played at CB, the rest of the defense could
funnel coverage help elsewhere. If Samuel did all of this despite being banged up, imagine how well he
could play if healthy. Heʹs a Pro Bowl candidate this year and I think he has the talent to possibly be the
best CB in football. Part of that is due to this system, but make no mistake about it, this guy is good."

 

[Brady-To-Branch]

I'm considering ordering it, and wondering if anyone out there read his 2005 version and can give a review. Thanks...

I have SF: 2005. It's excellent. However, his sabermetric stats can be overwhelming at first, so read his intros and understand his terminology (it shouldn't take long). Once you do, you'll have a hard time putting it down. I've already ordered SF: 2006.

SF: 2005 focuses on CBs, Safeties, WRs, TEs, and QBs with regards to the passing game.

Here's a few samples from Scientific Football 2005:

Bad decisions - worst things a QB can do

I think this is probably the most important metric when it comes to a team's wins and losses. Bad decisions are simply decisions that a QB makes, usually with the ball, which he shouldn't. They can range from throwing into coverage to throwing the ball while being sacked to staring at receivers. Each time a QB makes a mistake like this, I track it in the bad decision column. In addition to tracking the number of bad passes, I also track the severity of the play on a 1-5 scale, with 1 being the least damaging and 5 being positively suicidal. Bad decisions are tracked by their severity under the metrics, bd1, bd2, bd3, bd4, & bd5.

The level 1 bad decisions are obviously the most common, and the level 5 decisions are saved for those plays that are so bad they defy explanation. If you watched the New England at Miami week 15 Monday night game last season (2004), you might remember the pass Tom Brady threw late in that game. Brady was being sacked, and his body was halfway turned around and near the ground when he just chucked up a pass that was intercepted. That's a prime example of a level 5 bad decision. Brady, who makes bad decisions very infrequently (he was #4 in total bad decision percentage) had no possible reason to throw this pass except to avoid the sack, and in that case he should've taken the sack. Then play came when the Patriots were trying to run the clock out, and gave the Dolphins the chance to come back and win, which they did. All of those factors were taken into account in rating the play a level 5 bad decision.

In addition to measuring the number of each type of bad decision, you'll also find an aggregate metric for the total number of bad decisions. I express this as a percentage of the passes at each depth, and the metric for this is bd%. This metric shows the total number of bad passes made by a QB without taking into account their severity. You'll also find a weighted bad decision total under the metric wtd bd%. The weighted bad decision metric simply adds up the number of bad decisions as weighted by their level (i.e. a level 5 bad decision is worth 5 points, a level 4 bad decision is worth 4 points, etc). I measured the weighted bad decisions as a percentage so that it would give some kind of perspective on how damaging those decisions were.

You'll hear me rail on and on about bad decisions by quarterbacks throughout this book. I can't emphasize enough how much credence coaches give to bad decisions. If a QB makes bad decisions he will almost always be benched regardless of the team's record (see Arizona's team comment among others). I also track this metric for team defense. The column metrics for team defense are identical to the QB metrics.

Lito Sheppard (CB, Eagles)

Lito Sheppard's main stats (2004): 1. Tied for 62nd in completion percentage. 2. 78th in yards allowed. 3. Tied for 63rd in yards per attempt. 4. 80th in deep yards allowed. 5. Tied for 53rd in TDs allowed. Do any of these stats say Pro Bowl player to you?

Lito isn't a terrible CB; he did some things well. He did rank 20th in tight/good coverage percentage. Beyond that, there really isn't a whole lot else. His completion percentages at every level (short, medium, and deep) were 50th or lower, he had the 2nd most deep passes thrown at him, and he was frankly targeted like crazy. Cleveland threw at him all day long, Minnesota went after him almost to the exclusion of anyone else, and Deion Branch won the Super Bowl MVP despite having Lito covering him most of the time.

The reason I'm so upset about this is because of the reasons Sheppard made the Pro Bowl, of which I can count three. First, Sheppard made something of himself at a major college (Florida), so he was already a known commodity when he came into the league. Second, he makes interceptions. Third, he has an easily remembered name (especially once you've heard Chris Berman sing "Lido Shuffle" after one of his interceptions).

He's a name player from his college days, plays on a good team, and makes interceptions, so he gets to the Pro Bowl while his more deserving teammate (Sheldon Brown) stays home. Sure Sheppard makes interceptions, but he is also beaten on passes too damn frequently. I hope he enjoyed his stay in Honolulu, because his reputation will not survive if he keeps performing like he did last year (2004).

 

[PonyExpress]

Thanks so much for your replies. I ordered my copy tonight. Just a thought- I have never warmed to the "FootballOutsiders". I feel as if the "outsiders" almost revel in the obscurity of their numbers and measurements. After reading and reviewing Joyner's work, I feel like he puts their work in need of major revisions. His tape study adds credibility to his metrics and I am sold on his method. I am glad to have another source for real football knowledge than the "outsiders" whose method and presentation never captured my imagination.

 

[Johnny Z]

I have the 2005 and bought the 2006.

As I understand it, he generates his own stats.

He has a stat for how many times a CB was thrown at during a season, and how many receptions he allowed. Some shutdown reputations take a pretty big hit.

The Packers have a CB the fans hate with a passion, Ahmad Carroll. Joyner admits to being a Carroll basher. In his stats Carroll comes of as pretty good. So there are some surprises that way, too.

Deion Branch comes off extremely well.

Stats.com sells a premium stat package to teams and media outlets. It costs several thousands of dollars per subscription.

I'll add, there are a lot of obvious things he does not do. For instance, he makes no allowance for how the score and other factors effect what happens to a CB. Charles Woodson was thrown at just 58 times. Sounds good, but he only played in 12 and 1/2 games. He had 8 coverage penalties, and teams scored decisive points on many of those drives. The Raiders were usually losing and teams were running the ball down their throats. Jason David was a rookie that year who was thrust into starting. The Colts play a lot of roll-up coverage. The safeties were all young and inexperienced. A roll-up rookie CB handing off WRs to an inexperienced safety - it can have its ugly moments. Teams were often down to the Colts and throwing hard to get back into it. Jason David looks bad on paper, but he actually played decent football for a rookie CB. On raw numbers, Woodson looks effective, and David looks terrible. That was not reality. When the packers played the Colts that year they threw on them like demons. When the game was on the line Jason David ripped the ball out of Javonna Walker's hands and that was the kill play of the game.

But at least he presents a lot of raw numbers.

 

[bucky]

I have to admit I've never read his full publication, but I've read his methodology in detail and I totally disagree with it. Most of the "grading" he does is very subjective - such as did the QB make the right read? By definition, science is not subjective. So Joyner is anything but a scientist.

The problem with the whole football sabermatrics thing is that football is such a intricate and team-oriented game. Responsibilities and adjustments are designed ahead of time and made on the fly. There are so many variations based on schemes, alignemnts, calls and reads, that no one person can accurately evaluate each player of every team on every play. Even NFL personnel men have a tough time doing it. And on top of that, Joyner has done this by watching TV tape, as opposed to coach's tape (hopefully, ESPN has helped him to get some coaches tape of the games). The only way he (or anyone else) could accurately make these types of evaluations is if for every snap he knew every offensive and defensive play call and the adjustments that the players were supposed to make.

 

[bucky]

Just to extrapolate on this a little more. The whole premise of sabermetrics is to remove the situational and subjective aspect of statistics and concentrate on the basic pitcher vs. batter battle. This is unique in all of team sports because it's a true 1-1 battle. And the result that sabermetrics focuses on - on-base-pct, slugging-pct, ground/fly ball ratio - are largely independant of ballparks, base runners, etc.

In football it is absolutely and cateogrically impossible to isolate a 1 on 1 battle. What may look like a 1-1 between a WR and a CB is actually not because the CB might have help over the top from a S or underneath from a LB, or the WR may be just a decoy on 1 play, but a primary receiver on another. Or the game's time, score, or down and distance may dictate the way this battle plays out. And while sometimes this can be obvious to the observer, other times it goes completely undetected because of the way the play plays itself out. So an objective/scientific evaluation and comparison of players from different teams without regard for system, teammates, situations, etc. is practically impossible.

 

[Johnny Z]

I don't see a stat for bad reads. He has a stat for bad QB decisions, and he smooths those across 5 levels of severity.

He is pretty conservative in assessing those. The highest level of severity is level 5, and he only counts 10 examples of a QB making that bad a mistake in the entire season.

CBs getting toasted is another. He is very conservative there as well.

As for game film, he uses what is available. I would bet that he spends way more time with film than the premium stat services do. They miss all sorts of stuff - way too much to be bragging about accuracy.

On the scientist aspect - obviously, it can never be science.

 

[gomezcat]

Anything which allows us to look at the game in a new way, one which challenges us, is a good thing. I like the idea of looking at CBs as he does, in particular. I remember reading about this guy's book on SI.com (I think it was Dr Z, but I'm not sure) and the reviewer was pretty impressed.

 

[Pats726]

Anything which allows us to look at the game in a new way, one which challenges us, is a good thing. I like the idea of looking at CBs as he does, in particular. I remember reading about this guy's book on SI.com (I think it was Dr Z, but I'm not sure) and the reviewer was pretty impressed.

Does his metrics change from year to year??? Does he look at different things one year as opposed to past years??

 

[Johnny Z]

Does his metrics change from year to year??? Does he look at different things one year as opposed to past years??

I don't know how long he has been doing it, but I think the 2005 book (2004 season) is the oldest one he has for sale.

I think he's likely to add a lot of new things in the 2006 edition.

Some of what he is doing is for the fantasy market, which ain't me.

 

[Pats726]

I;m not eaither in the fantasy thing..which I feel is about stats only and little about team..which is what football is more about..and a LOT diifferent!!! Is 05 available?? MAY have to find...

 

[Brady-To-Branch]

I don't know how long he has been doing it, but I think the 2005 book (2004 season) is the oldest one he has for sale.

I think he's likely to add a lot of new things in the 2006 edition.

Some of what he is doing is for the fantasy market, which ain't me.

Sceintific Football has nothing to do with fantasy. Joyner does has books/guides that are strictly for fantasy, though.

SF '05, Joyner only focuses on the passing game. SF '06 he'll be using metrics for the running game as well. Also on SF06, Joyner says he'll tighten up some of his metrics. On of them is YCPA, total yards per catchable attempt, which he ignores passes deflected at the LOS and passes that are throwaways. Joyner noted that, according to the NFL stats, these attempts are credited to the nearest receiver and he felt that was ridiculous. He had an article on ESPN Insider where he showed the difference between yards per catch and YCPA (minumum 40 receptions) Here are just the raw stats from that article...

http://insider.espn.go.com/nfl/insider/columns/story?columnist=joyner_kc&id=2441030

2005 Yards per catch leaders
1. Ashley Lelie -- 18.5
2. Terry Glenn -- 18.2
3. Santana Moss -- 17.7
4. Randy Moss -- 16.8
5. Ernest Wilford -- 16.6
6. Justin McCareins -- 16.6
7. Eddie Kennison -- 16.2
8. Terrell Owens -- 16.2
9. Plaxico Burress -- 16.0
10. Joey Galloway -- 15.5
11. Lee Evans -- 15.5
12. Brandon Lloyd -- 15.3
13. Roy Williams -- 15.3
14. Steve Smith -- 15.2
15. Chad Johnson -- 14.8

2005 YCPA leaders
1. Santana Moss -- 11.4
2. Steve Smith -- 10.7
3. Eddie Kennison -- 10.7
4. Ernest Wilford -- 10.0
5. Terry Glenn -- 9.7
6. Torry Holt -- 9.6
7. Chad Johnson -- 9.6
8. Eric Parker -- 9.6
9. Donald Driver -- 9.4
10. Ashley Lelie -- 9.3
11. Joey Galloway -- 9.3
12. Marvin Harrison -- 9.2
13. Hines Ward -- 9.2
14. Joe Jurevicius -- 9.1
15. Kevin Curtis -- 9.1

 

[PatsFanSince74]

I'm considering ordering it, and wondering if anyone out there read his 2005 version and can give a review. Thanks...

Joyner's stuff for 2005 was amazing and it's a good "spend," even more so, I would imagine, if you're into Fantasy Football (which I'm not). However, an even better read is "The Hidden Game of Football" by Bob Carroll.

 

[Johnny Z]

Th book is not directed at the fantasy market, but his website has stuff for fantasy football.