FanPost

Tarvaris Jackson and Tarvaris Jackson in QB Competition: Injured vs Healthy

133604664_extra_large_medium

This started out as a comment in response (and subsequently exploded to a fanpost) to member Nate Dogg in the thread on Kenneth Arthur's recent entry on QB (roster spot) competition between Tarvaris Jackson and Josh Portis. Much of the discussion the past few days has been centered on how the QB depth chart will pan out. In a response to Tyler Jorgenson claiming Tarvaris was essentially league-average-ish, Nate Dogg wrote the following:

Sounds good, lets look at some advanced stats!
DVOA – 20th
DYAR – 21st
WPA – 35th
EPA – 25th
ANY/A – 23rd
Y/A – 21st
Yards – 20th
TD% – 29th
INT% – 20th
Sack% – 27th
He doesn’t crack the top 20 is a single one of those stats.

There of course has been unending discussion about the way in which we should interpret Jackson's performance last season regarding: the short offseason, completely new teammates (except Sidney Rice, more later), a ridiculously porous offensive line with five players who had never stood together on the same line in a game, a midseason pectoralis tear, the lack of a true WR1 given the injuries to Sidney Rice at the beginning and end of the season, inconsistent performance from Mike Williams (injury, QB play, other?), and an otherwise pretty inexperienced WR group. The discussions often devolve into "I can make more comments than you" arguments on two sides of a fence with a tiny bit of name-calling and trolling (can't say I'm not guilty of any of these).

The biggest problem is that many of these "issues" that Tarvaris had to deal with are non-quantifiable or inextricably linked with one another. How do you separate OL pass blocking from QB play when you want to assign blame for a sack? How much of a role does the failure of the WR group to get open play in a sack? How can you account for BMW's ~66-75% decrease in production year over year (65 to 18 receptions and 751 to 236 yards), when undoubtedly it is some inaccessible combination of regression to the mean for BMW, Tarvaris Jackson's limited (minimal, severe? We can't tell from tape, we don't know how he was coached) vision and aggressiveness, poor pocket protection limiting the amount of time for Mike's route to develop, and coverage schemes that direct TJ's progression away from BMW as the target? It's just not possible.

But, we can potentially quantitate the degree to which the pectoralis tear might have affected his play. Mild pectoralis strains may take 3-6 weeks to heal with complete rest, moderate strains may take a bit longer or require surgery, and severe strains require surgery. We don't have access to Tarvaris' medical records, but we can guess, because the team was talking up the surgery angle from the day of the injury until well after the season ended, that this was either a very bad mild tear or a moderate tear, and we'd expect the healing time to at least be 6 weeks with complete rest and longer for an athlete who takes 2 weeks and then is playing NFL football on week 3.

Ignoring the complicating factors of the OL group, the short offseason, and the injuries and inconsistent performance of the WRs, I performed an analysis of Tarvaris Jackson's sub-par overall performance (described quite aptly by Nate Dogg's stats above) by breaking it up into the season of the "injured Tarvaris" and the season of the "healthy Tarvaris". My choice for dividing the season was arbitrary, but I noted that four of his five worst passer rating games were in the five games following his injury, which were displaced from the injury by the bye week and Charlie's start against the Browns (sorry to bring up that painful memory).

IF his pec was mostly better at eight weeks after the injury, then the five games from week 7-11 would correspond to when his pec was affecting him the most. For this analysis, I considered games 1-5 and games 12-16 as the "healthy Tarvaris" and games 7-11 as the "injured Tarvaris". Look for the breakdown after the jump

For Tarvaris Jackson's stats, I will group games 1-5 and 12-16 and directly compare them to games 7-11 and full season stats. Comparison to other QBs is made based on their complete seasons. Only QBs with over 14 attempts per game are considered.

AY/A was calculated per PFR's formula ( {total yards +20*TD-45*INT}/ATT ), ANY/A was calculated per PFR's formula ( {total yards - sackyardslost* +20*TD-45*INT}/{ATT + sacks} ). *sackyardslost was computed using sacks times average yards lost per sack, as I did not want to spend additional time finding the yardage totals of each sack.

I performed a weighted average of passer rating to minimize the effect of outliers with few numbers of passes, specifically the Philly game ( SUM [ att(g1)*rat(g1) + att(g2)*rat(g2)...att(gN)*rat(gN) ] / SUM [ att(g1) + att(g2)...+ att(gN) ] ); the weighted average returned a season passer rating of 79.3 which is incredibly close to the PFR passer rating of 79.2, suggesting this was a valid calculation.

I do not have access to EPA, WPA, DVOA, or DYAR on a game by game basis and cannot do analysis on those advanced metrics -- more later. All games and season stats obtained from pro-football-reference.com, except sack counts which were generated by my inexact method of looking at game logs on espn.go.com.

Total Tarvaris

WeightedPasserRating: 79.3/79.2, 21st
Comp%: 60.2%, 17th
AY/A: 6.2, T 21st
ANY/A: 5.1, 24th
TD/INT: 1.08, 25th
TD%: 3.1, 29th
INT%: 2.9, 20th

Overall, Total Tarvaris was around 22nd if you take the average ranking of these stats, which is the maybe the worst pseudo-stat there is, but it supports a few regulars' contentions that TJ is better than around 10 other starters in the league, right in line with DVOA and DYAR. WPA and EPA rate Tarvaris more harshly, though the summation of poor components throughout the offense will negatively affect both. I think these numbers are an accurate assessment of his overall season.

Injured Tarvaris

WeightedPasserRating: 64.7, 34th
Comp%: 55, 28th
AY/A: 5.3, T 31st
ANY/A: 4.4, T 32nd
TD/INT: 0.43, 34th
TD%: 2.0, 33rd
INT%: 4.6, 32nd

Overall, Injured Tarvaris is almost the worst QB in the league with an average here of 32, essentially as ineffective as Blaine Gabbert. There's not much to say, except that I am fairly confident that we have three QBs on the roster that are better than Gabbert and we should all be incredibly thankful to John and Pete that we don't have to sit through that kind of QB performance again.

Healthy Tarvaris

WeightedPasserRating: 86.6, 13th
Comp%: 62.9, T 7th
AY/A: 6.6, 18th
ANY/A: 5.4, 22nd
TD/INT: 1.83, 11th
TD%: 3.7, T 23rd
INT%: 2.0, T 5th with Tom Brady

Healthy Tarvaris is a bit of an eye-opener. He's still awfully weak in TD%, but if you are relying on Marshawn to get you 12 TDs a season and you bring in Robert Turbin to add a few more, it's okay to have a slightly lower TD% rate. The average of these stats falls around 14, which is somewhat surprising even to me, your local non-hater and open-minded assessor. More appropriately, there is a plateau of INT% around 2.0-2.2% which is virtually indistinguishable, so maybe all of the 5-11 guys tie and its more like a 8 rating. Still, healthy Tarvaris performs just about average.

If we prorate Healthy Tarvaris to 14.5 games, he comes out about at about 3000 yards, 16 touchdowns and 9 interceptions, or 3000/16/9, and for 16 games he comes out to 3400/18/10. That second line looks like the kind of QB that a Carroll team with a strong run offense and a top-5 defense could take to the playoffs and expect to be very competitive. I'd prefer 4000/25/12, but Playoffs are Playoffs.

Don't believe me? Look at this comparison of Healthy Tarvaris vs Alex Smith:

Passer rating: 86.6 vs 90.7, 13th vs 9th
Comp%: 62.9 vs 61.3, 7th vs 12th
AY/A: 6.6 vs 7.3, 18th vs 11th
TD%: 3.7 vs 3.6, 23rd vs 22nd
INT%: 2.0 vs 1.1, 5th vs 1st

Don't get me wrong, I don't idealize Smith as the gold-standard of playoff QBs, but they were only one of two muffed punts away from beating the eventual Super Bowl champs and advancing to the big game themselves, with Mr. Unspectacular himself at the helm.

If we compare 2011 Healthy Tarvaris against the reportedly "never during his career displayed any ability to improve Tarvaris", 2011 healthy version is better by a long shot with passer rating up 9 points (+11% relative improvement), comp% up 3.5% (+6% relative improvement), AY/A up 0.7 yards (+12% relative improvement), TD/INT ratio jumping to 1.83 from 1.09 (+69% relative improvement), TD% up 0.1% (+3% relative improvement), and INT% dropping from 3.3 to 2.0 (+39% improvement).

When healthy, Tarvaris played better than he ever has in his career, and showed significant performance improvement over prior years.

One might argue that "my" stats aren't the "most advanced or sophisticated stats"; DVOA, DYAR, WPA, and EPA may provide a more statistically accurate assessment of how a player being on the field affects the team's chances of winning. I don't have access/capability to these stats on a game-by-game basis. But they are all dependent to some extent on completion percentage, TD%, INT%, and stats parallel to ANY/A and AY/A, and all of the more sophisticated advanced stats would show the same directional effects (of uncertain magnitude) of the healthy vs injured breakdown and the healthy 2011 vs career comparison.

Of course, healthy Tarvaris had a significant disadvantage I hinted at in the introduction. Sidney Rice was on the field with our "healthy" version in only 2.5 games (out of 9.5). Injured Tarvaris had the advantage of Rice being available for around 85-90% of his 151 pass attempts. In contrast, Healthy Tarvaris only had Rice available for a mere 30% of his 299 passing attempts. It's not unreasonable to expect that a fully healed Rice, or even just a more experienced WR group behind Rice will help to boost Tarvaris' production.

Sack rate is at least somewhat dependent on OL, and our OL was less than stellar last season, ranking 24th in pass blocking according to FootballOutsiders.com (surprisingly similar to TJ's average rating among the advanced stats). Let's look at the first 4.5 games when the OL was atrocious vs the last 10 games where it "started to come around".

QB Sack rate for all games: 8.5%, 27th
QB Sack rate games 1-5: 9.8% would have ranked 32nd
QB Sack rate games 7-16: 7.9% still ranked 27th

Interestingly, his sack rate when Sidney Rice was healthy was even lower, although I am sure this is subject to multiple confounders (games 2-5+7-11): 7.3%, 24th.

Tarvaris does hold onto the ball too long. But Healthy Tarvaris did have one of the lowest INT rates...

Statistically, an INT is equivalent to about -45 yards of offense, and a sack is equivalent to about -12.5 yards of offense (7 yards lost and + additional loss of 5.5 Y/P league wide average), so 3.7 sacks is the equivalent of 1 INT in terms of negative impact for the team. It's pretty clear that Carroll would rather give up 3 sacks per game instead of one interception, see 15:25 where Carroll explains that when his college teams won the turnover battle, they were 53 and 0, and in the pros his teams are 10 and 2.

It's probably incredibly unfair to solely blame Tarvaris for his caution throwing the ball because that is what the coach demands of all his QBs (though Tarvaris could throw a few of 'em away). This might hold true for Jim Harbaugh as well, as Alex Smith posted a sack rate of 9.0%, good bad for 29th in the league, even worse than TJ. More telling perhaps is that of the six players with the lowest INT%, only Brady (11th) doesn't rate among the 15 worst in Sack Rate (Romo 20th, Rodgers 21st, TJ 27th, Smith 29th, Bradford 30th).

POTENTIAL CONFOUNDERS

1) Changing quality of the Offensive Line over the course of the season. Likely, based on observation, there was a change in the efficiency of pass-blocking over the season, and it's just a total guess, but I don't think we'd be far off if we estimated that the OL ranked around 30th in the first third of the season, 24th in the second third, and 18th in the final third. Since, Tarvaris' healthy attempts are split 157 with the 30th ranked OL (in his first 4.5 games) and 142 with the 18th ranked OL (last 5 games), I don't think the OL performance would bias the comparison of "healthy" vs "injured" much in favor of healthy Tarvaris. But I'll admit, I certainly don't have the means to quantitate it in any way.

2) WR performance over the course of the season. Doug Baldwin and Golden Tate both showed some growth in the second half of the season. As mentioned above, Rice was not available for most of healthy Tarvaris' games, breaking down to present about half the time Healthy was playing in the first half and none of the time Healthy was playing in the second half. It's basically an unknown, but it would not be surprising if the complete absence of Rice in the 2nd half balanced out the improvement in the other WRs.

3) Strength of schedule. Could Tarvaris have faced easier opponents when he was healthy than he faced when injured? In games 1-5 and 12-16, Seahawks opponents had a cumulative win-loss record of 91 and 69 (approximate average of 9-7), and an average ranking of 13.2 for team defense based on PPG allowed. In games 7-11, Seahawks opponents had a cumulative win-loss record of 36 and 44 (approximate average of 7-9), and an average ranking of 15 for team defense based on PPG allowed. Healthy Tarvaris faced better opposition based on both win-loss record and opponent scoring defense than injured Tarvaris faced, so "healthy" also received no help from the schedule.

SUMMARY
So yeah, Nate Dogg is right. Tarvaris, on the whole, performed below average last season, but still better than 8-12 other QBs. When you remove (admittedly somewhat arbitrarily) his play while injured, a healthy-ish Tarvaris performed quite close to league average, under otherwise trying circumstances, and also moderately improved his performance compared to his career stats.

In addition, there was slight bias against Healthy Tarvaris in terms of Rice's availability, schedule strength, opponent scoring defense, and a below-average pass-blocking line for the entire season. Yes, these simple stats don't tell the whole story; they don't change our memory of what we perceived (correctly or incorrectly) or the frustration we all felt at another year of offensive mediocrity. Maybe you will still argue that he just doesn't have good game sense. I'd counter that there might be more sense than we know, because we just don't know. Sometimes, physical tools - the ability to run left and chuck the ball 42 yards downfield with your recently torn pectoralis and still hit your WR in stride - can help you win games. To deny the possibility would be irrational.

Ultimately, I want the Seahawks to have better QB play than we have seen the last four years. After delving into this, I'm not convinced Tarvaris wouldn't perform as the 10th to 15th best QB this season, especially with improved OL, TE, and WR play. Frankly, he is already damn close to being that good average, and I believe in the theory of potentiation of performance by cumulative talent in team sports like soccer, basketball, and football.

I am not wishing, hoping, or expecting Tarvaris to win the competition, it's just clear to me that it is not a foregone conclusion that he isn't the best QB on the roster. Personally, I will likely be happier if Flynn wins the competition, because that will mean we have at least 2 QBs whose floors are league average. On another note, to answer the question about Tarvaris being worthy of a roster spot, it's hard for me to fathom cutting a near league-average starting QB when the rest of your roster has 2 games of NFL experience.

Obviously the rest of the offseason and preseason will tell us far more than this analysis, but what do you think? Do you think there is validity in breaking up the season this way? Does it make you reconsider whether Tarvaris still has room for growth? Does it make you reconsider what his value would be on a roster as a second string QB? Does it make you consider his chances in the QB competition in a different light? Are your opinions, positive or negative, about Tarvaris so fixed that even looking at this analysis makes absolutely no difference to your way of thinking?

Note: In the interest of avoiding the diversions that have hijacked the other recent threads, I would ask that all commenters refrain from personal attacks on this author, other commenters, or Tarvaris Jackson. I think this is a genuinely interesting analysis, please keep comments targetted at the topics in the post. If things do get derailed, I'd request that the admin please redirect the conversation appropriately.

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

Join Field Gulls

You must be a member of Field Gulls to participate.

We have our own Community Guidelines at Field Gulls. You should read them.

Join Field Gulls

You must be a member of Field Gulls to participate.

We have our own Community Guidelines at Field Gulls. You should read them.

Spinner

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9341_tracker