Mark Sanchez: Comparing Performance- Completion Percentage
Time to turn our attention to the most polarizing Jet on the team: Mark Sanchez. Sanchez is clearly at a crossroads in his career. When going over his potential in a prior article I mentioned that he tracked in a very select group of players, but at the bottom end of that group and that we could already write off the ultra high end upside for him. What’s scary is that the real poor part of that group, the Charlie Batchs of the world, that are usually replaced in year 4 seems to be where Mark is headed.
I’ve been fooling around with some charts for QB analysis and wanted to start using them to try to see where Sanchez stands in relationship to his peers. Followers of my writings know I track certain stats for the teams and players and often adjust them into percentages based on opposition. QB is always a very interesting position and one that is fun to look at many different ways. What I want to start off with is a simple look at completion percentage above or below expectations.
What exactly does that mean? Well lets say that Sanchez completes 63% of his passes in a game. Most would say that’s an excellent game for Sanchez who we all know has not been the most accurate passer in the NFL. But what if that comes against the Patriots who are allowing 67% of the passes to be completed against the other 14 QBs they face that season? Not so good. In this case Sanchez’ weekly grade would be -5.97%, meaning he was about 6% worse than the average QB was against New England. Now if we plot that on a graph you will get a large number of peaks and valleys so to smooth that out for the normal deviations what I did was use an 8 period moving average to get a better idea of his true average performance. To get an idea of what this looks like here is a plot of the weekly results in green and the moving average in red.
As you can see there are some spikes and some pretty steep lows, but the overall performance is pretty steady with a slight trend for improvement over the last year and a bit more consistent performances from the young QB. So maybe there is hope. Lets see how he compares with some other QB’s who had the chance to play as rookies over their first 46 games of their career. Note that we are just using the 8 game averages here to smooth things out so what is listed as period one on the following chart actually corresponds to the 8th game the player played. Games in which the player did not attempt at least 10 attempts are taken out of the calculations.
2009 Draft Class: Matt Stafford and Josh Freeman
Stafford is at a little different stage than Sanchez since he missed much of his 2nd season due to injury, but this chart is one of the many reasons why Sanchez right now is considered the worst of the three drafted in 2009. Stafford and Sanchez both essentially started at the same place, but Stafford’s improvement level was off the charts compared to Sanchez’. Freeman started from a higher spot than both of them and, like Stafford, showed some big improvements. When you look at the plots those first 18 weeks or so were virtually identical for all three in terms of the increases and drops. The difference was that Stafford and Freeman’s moves were dramatic reaching around 5% above the average while Sanchez only improved to about 8% below the average. The two continued to improve peaking at about 10% above average while Sanchez regressed to almost rookie year levels over the same time period. Where Stafford lands is anyone’s guess, but Freeman, who did come back to earth, began to rebound and finished his third year on an upswing while Sanchez can’t seem to shake that -5% threshold.
2008 Draft Class: Matt Ryan and Joe Flacco
Back in 2009 I did a pretty extensive look at the handling of Flacco and Ryan compared to that of Sanchez. Both were treated with kids gloves and given every opportunity to play easy mistake free football, especially early in their rookie years. Ryan had more issues in his second season as he dealt with injuries and a league that seemed to catch up a little with him. His low actually corresponds with the timing of Sanchez’ lows and from period 9 thru 25 or 26 the up and down direction of their games are pretty similar. Flacco followed a much different pattern and its one that is more similar to Freeman’s pattern. He fell apart towards the end of his rookie season before showing much improved accuracy thru a good portion of year 2, before crashing in the second half of that season. The big difference between these players and Sanchez is similar to the difference between Sanchez and Freeman in that there was a point around period 26 for Ryan and period 29 for Flacco where there were a great deal of similarities between the directions of where these players were all headed. Ryan continues to improve as did Flacco. Sanchez clearly did not. For what its worth neither Flacco nor Ryan has held any real levels of improvement. Flacco seems to be someone extremely streaky as he fell down to close to 8% below average between period 45 and 50 before rebounding to a 2.3% by his final game, not exactly the performance designed to get you the tier 1 contract he wants. Ryan spent much of the last year fighting to stay around average.
2006 Draft: Vince Young and Jay Cutler
Cutler began in a system designed to make things easy on a QB and had some early success before basically setting his mark in that 0-4% above average range. He has been pretty steady with the exception of his first year in Chicago where he fell apart as the year went on, plummeting down to a low of -7.7% before rebounding in the Martz seasons right back to that 0-4% range. Young started at close to -25% below average as he was wild as a rookie before making a steady climb to become a relatively average passer. He had a similar progression to Sanchez, but showed more improvement and didnt have that big dip around games 26-30 that Sanchez had. His fall over his final 9 games was identical to Sanchez’ and for much of the time he found himself on the bench.
2005 Draft: Alex Smith
Smith continues the trend that we seem to see from the others in that they all seem to have at least one spike that gets them to be above average at some point in time. That said Smith ended up in the same place as Sanchez and has been benched so many times that his small blip to try to climb to average did not occur until his 7th season in the NFL and by the end of the season was back down to a -6.1
2004 Draft: Eli Manning and .Ben Roethlisberger
This is our first real look at two proven successful QB’s who got the nod early in their careers. Roethlisberger was one of the most managed QBs in the game early on and played a very accurate game. As he got more responsibility it began to come down but eventually he rebounded in his third season. Manning is much more interesting because he is someone that we often hear compared with Sanchez. The numbers are very similar between the two, except for that period between 15 and 20 where Manning saw that spike so many others did to get over the hump into average or just above average territory. After that Manning steadied out at -5% similar to Sanchez.
The question becomes when did Manning improve? In year 4 Manning was essentially a disaster. Towards the end of the season he bottomed out at a -17.2% his worst mark since his 10th game as a pro. What salvaged his career was the playoff run that led to a Super Bowl trophy because he was close t being sent out of town at that point. That seemed to give him the confidence he needed as his numbers steadily rose beginning in year 5 where he seemed to settle down into an average or just above average category. He has had a few dips back into the negatives but all in all it would seem his norm is to be slightly above average in this category. With Tebow waiting in the wings Sanchez will likely not get the opportunity to fail the way Manning did in year 4, but I guess you could say that this is the ultimate upside for Sanchez. Here is the graph for the continued development of Big Ben and Manning.
2003 Draft: Byron Leftwich and Kyle Boller
Kyle Boller actually surprised me a little bit because I remember him being so bad that I would have assumed he was worse than Sanchez. In actuality thru 29 games he and Sanchez ran almost identical looking patterns with Boller just being slightly more accurate. Even after that the graphs are similar with the exception that Boller did finally jump into positive territory, though by that point in time Boller was into his 5th NFL season and just hoping to hang onto a spot in the NFL. Leftwich had that second year surge that many others did but came crashing down following a knee sprain in 2004 which just seemed to kill his momentum and he soon found himself struggling to keep a starting job.
2002 Draft: David Carr and Joey Harrington
I have to admit that these are the charts that scared me. On the positive side for Sanchez these are the only 2 QB’s who had real significant periods where they were worse than Sanchez, in particular David Carr. What continues to disappoint is again you have two bad QB’s who somehow were able to at least run a short a period of time where they were above average while Sanchez has never come close to that. The Carr trend level us near identical to Sanchez’ from period 27 to 39. That hovering around -5%, though he did have a small spike, though that was in his 4th season as a pro. Eventually they completely dumbed the offense down for Carr which saw his numbers rise dramatically but they stripped him of almost all chance taking in the process and he quickly lost his job as a starter. That little negative spike that Sanchez had in period 34 is something that affected Harrington as well, though on a much larger scale. It also was evident in the case of Kyle Boller and Vince Young. Harrington ended up one of the historic top of the draft NFL busts.
Some Final Thoughts
Not a very promising scenario is being painted here for Sanchez. One of the things I would like to comment on is the coaching situation that Sanchez has been involved in. I remember hearing a story about how Eli Manning, after being destroyed by the Baltimore Ravens, who were actually mocking him during the game, as a rookie came to his coaches on the train ride home and apologized for the game and wanted to talk about the things he was comfortable with in the offense. He asked that maybe if they worked those things in early it would calm him down and help him get in a little groove so he didn’t implode like that again. The following week he went 16/23 to help at least calm the critics and his own nerves. Now after being able to visualize the development of all these players and seeing these spikes in year 2 or 3 and seeing Sanchez being essentially flat after year 1 I have to wonder if the staff has done him no favors in helping him in this area.
Some guys are not accurate passers but you can create situations that help them be more accurate. 3 step drops with timed routes. Passes they are comfortable with. Scripted plays run over and over in practice. With the exception of a few weeks at the end of his second year where they seemed to do just that, have the Jets ever tried creating those situations for him? None of us are in those rooms where those gameplans are made, but to the untrained eye it seems as if the Jets solution when things go bad is to blame the QB more than help him. Color coded systems to remind him to not make dumb passes. More handoffs. Lamenting that you did not re-sign Brad Smith and force feeding players like Jeremy Kerley and Joe McKnight into the Wildcat roles. Trading for Tim Tebow. That shatters confidence and its not a solution to improve the QB. It’s just a way to hide his faults. There may be a time for that, ala David Carr and Alex Smith in the later years, but when the player is developing it can be damning and Sanchez was in a development phase these last three seasons.
I’m sure I missed a rookie or two that played significantly early on, but I think its safe to say that out of these 14 QB’s Sanchez is either the least accurate or tied for least accurate with Manning. Sanchez showed the least upside of any of them as the only player to never get above the 0% mark. The dip that Sanchez saw in year 2 was a phenomenon only seen by players who had a period above 0% while most others saw their completion percentage rise during that time. Sanchez is most likely never going to be good in this area and if they run an offense anything like they did the last few years he won’t make it out of this season as the starter.
Early next week I post a look at the YPC category and see if Sanchez’ regression has been as bad as I, and many of us, all thought. I think the results might be surprising.
Of games started in college. I think a large majority of the qbs listed above had significantly more college experience.
Sanchez has the least college QB experience of any of them and that is a significant issue. With only 16 college starts under his belt before arriving in the NFL, his NFL rookie year was essentially a redshirt season for him.
Slide those stats back 15 games (his rookie year) and you will see that Sanchez projects quite nicely with your sample group.
When the Jets drafted Sanchez as a junior, they drafted him based upon his projected potential coming out of college ball. Pete Carroll was right - Sanchez needed another year and those 12-14 college games would have done wonders for Sanchez prior to coming to the NFL.
The issue with a young QB is not necessarily his performance but whether or not he is showing improvement. Carr, Harrington, Leftwich, Boller, and most recently Henne never got better and that was the problem. It is becomes even more magnified with players that had 4 solid years of college ball. There simply isn't alot of room for growth left.
Sanchez has alot more field to grow. His stats have steadily improved, though I am sure not as fast as he or we would like. However throw in dissaray on the o-line, a revolving door at WR and Schotty playcalling, I would consider his play to be alot better than everyone gives him credit for.
Does this mean that Sanchez will improve? Maybe or maybe not. But as far as I am concerned, this statistical analysis does not accurately portray the circumstances for this particualar QB.
As the saying goes, there are lies, damned lies and statistics!
And we all know how Eli's 4th year ended.....
I would agree with you that its all about improvement. Thats one correlation that seemed to be in both DVOA and QBR stats if you tracked year 1->2 growth. Its all about improved decision making which I think Sanchez did show, which is shown in some other areas. But for this particular stat Im not sure there was much improvement. The busts you mentioned all showed periods of improvement followed by periods of regression over those same number of games as Mark. Just look at Harrington. It was a steady increase before he crashed and burned. Doesnt mean he improved in other areas, but in this one he did. Sanchez has been relatively flat since game 11 in year 1.
I think in Sanchez' case he was handled poorly but at the same time its hard to make someone more accurate. I think its probably a point the team should concede and instead try to work to his other strengths. It took Manning over 5 years to be just slightly above average, but they never have tried to make him that kind of QB. They just use his skills he is good at to make him a success.
On another topic of the coaching what surprised me was when I ran his passing splits for this season. He had the 8th least attempts in the NFL in the short passing category of gimme passes (those that essentially are thrown behind the line) with a higher concentration in those mid tier categories where he begins to fall apart. Some is his fault Im sure but alot of that is on coaching not drilling into his head where to throw.
Bill- I dont put the playoffs in there because it just becomes too difficult for me to pull the stats to quickly populate a spreadsheet so whenever I run these things, unless I mention otherwise, its always just regular season.
If you want to actually overlap the games by eliminating the first season for Mark as a "mulligan" you can start Sanchez at period 16, which would completely eliminate every game from year 1 in his moving average. It still would not change the fact that he never grew above the 0% mark or that its relatively flat, but it would put his finish mark above a few on the list, but the trends themselves wouldnt really change.
BD- I think the one issue with Manning is that his 4th year was really a bust until the playoffs. But he never had anyone breathing down his back. I still dont get why the Jets needed to bring in Tebow, but if Sanchez has the year 4 like Manning he Sanchez wont be the starter in the playoffs.
Thanks for the comprehensive response.
I do understand what you are saying about college start issue, but from my perspective the issue isn't necessary the level of competition but the number of repetitions. College players need those throws as it aids in their development down the road. Essentially they begin to understand what they can and cannot accomplish on the field. Take away 12-14 games and it will impact development. Plus, there really is no comparable statistical model in the NFL to look at in developing a junior QB, which is what the Jets are trying to do.
With Sanchez it translates in a couple of areas. First is the decision making process. The Jets can help by simplifying and running an offense that plays to his strengths. They made that mistake earlier last year and I think it really hurt his self-confidence BUT despite that there was some improvement.
Second, the overall Jets performance took a step back. The o-line and RB play was not as strong as it was the previous two years. There were times last year when Sanchez simply didn't have time to throw the ball.
All young QBs need to be protected - heck Tom Brady is a the best example (not saying Sanchez will be Brady) - as his first 5 years of starting reflected him playing within himself and being carried by a very talented team. It wasn't until 2007, his 8th year in the league, that he became all-world and started lighting things up.
I think both of us would agree that the Jets and us as fans need to accept what we have in Sanchez: a guy who needs a good supporting cast and an offense tailored to his strengths in order to succeed. There is nothing wrong with that and I do believe his is good enough to get the Jets to the SB. Whether that will happen or not is upto him at this point.
You certainly do raise several valid points, and I do agree with every one of them.
Having said that, however, I'm not certain how anyone could devise a meaningful statistical analysis of any position when trying to include all of the aforementioned intangibles.
What percentage of the final analysis should be attributed to a lack of college games? I have been ranting for the past two seasons that Sanchez has been terribly coached. What percentage should poor coaching be attributed to the statistical analysis, or the poor play of the O line?
At the end of the day, I find the statistical analysis an insightful tool when analyzing teams and players, and while I do appreciate the value of all of the aforementioned intangibles, I also know better than to try to bog down that analysis with those intangibles. They are subjective in nature, and accordingly are subject to the bias of the analyzer.
This, of course, will always lend itself to calling the statistical analysis into question and producing arguments over the data on a regular basis.
I do agree with Klecko73, but I also agree with the formula you have established here to yield meaningful analysis.
Thanks for the reply and insightful comments.
I agree that the statistical analysis is a useful tool but I would argue that the analysis needs to be framed in the appropriate context.
I don't have a % that should impact the final analysis - if both of us did, we would work in the FO of an NFL team!!!
None of this should be considered a critcism of Jason's hardwork. I only raise this issue in that without the benefit of context, no analysis is complete. There will always be some level of subjectivity in that evaluation proces but I think in certain situations important factors can be inadvertently left out.
Its typical that the first comment on this thread uses the excuse of not having played enough in college. Sanchez could play till he's 40 and his fans will still bring this up.
Sanchez tracks pretty normal with the others which is scary in some ways because they all saw improvements in percentages as the throws got easier while Marks compl. % had yet to grow.