Taking a Look at Matt Kalil

As the NFL draft fast approaches, the Vikings have a lot of different directions they could go in trying to reload the roster for the 2015 season.  Months ago Rick Spielman and the team began identifying the team's biggest needs to address this off-season, and while offensive line is surely a part of those needs, it remains to be seen if left tackle is among them.  This study looks at Matt Kalil's performance over his first three seasons as the Vikings starting left tackle using player grades from Pro Football Focus to try to determine what kind of player he is, and if the Vikings should be looking to find an upgrade.

If we flashback to the 2012 draft, history reminds us that Matt Kalil was one of the top prospects and an almost lock to be drafted by the Vikings very early on.  He was a bona-fide top 5 pick, and a "no-brainer" option for the Vikings.  At the time, there were really only three selections the Vikings could have realistically considered with the #3 overall selection: OT Matt Kalil, CB Morris Claiborne or WR Justin Blackmon.  All three players have failed to lived up to high draft expectations.  In fact if you go back and look at the top 10 picks of the 2012 draft, outside of Andrew Luck and Luke Kuechly, it looks pretty terrible.

The goal of this study was to try to figure out if Matt Kalil should be labeled as "good", "average" or "bad" based on his performance grades so far.  Generally speaking, if a player is good, they should perform well on the field regardless of their opponent.  If a player is average, they will more likely perform well against poor opponents, perform average against average opponents and bad against good opponents.  If a player is bad, they will probably perform poorly against just about everybody.  The left tackle position is unique because most of the time they face one opponent, one-on-one for the majority game.  If the left tackle is facing a 4-3 defense, they will line-up against the opposing right defensive end.  If the left tackle is facing a 3-4 defense, then they will line-up against the opposing right outside linebacker.  Obviously, the left tackle will have to take on more than just one player during the game, but since the majority of the snaps are one-on-one against the right side defender, I wanted to focus this study on how Kalil stacks up in those one-on-one situations.  With that in mind, the study considered all 48 games Kalil has started, and each individual game grade from Pro Football Focus was compared against every opponent, considering both that opponent's game grade and their overall grade for that season.  After analyzing all of those grades some very interesting trends began to emerge that helped answer the question: what kind of player is Matt Kalil?

The Study

Starting with Matt Kalil's rookie year in 2012, Pro Football Focus gave him an overall season grade of +12.1.  He was ranked as the 21^st best offensive tackle out of 80 tackles who played at least 25% of their team's snaps (interestingly, he was tied with Phil Loadholt that year).  That suggests that he had a pretty good rookie year.  Most fans and people in the media were very encouraged by his rookie year with some calling it "Pro Bowl Caliber."  As the 4^th overall pick in the draft and a pretty decent showing, expectations were naturally very high for Matt Kalil after the 2012 season.  But, should they have been?

I tracked every right side defender he faced during the 2012 season and averaged together their overall PFF grades and discovered the average defender he faced was graded only -4.6 overall.  In fact, of the sixteen defenders he faced (some division opponents twice obviously), only six of them finished the year with a positive overall grade and only five of them were different opponents (Julius Peppers had a positive grade and was faced twice).  This suggests that perhaps Matt Kalil had a good year primarily because he faced mostly poor competition.

But more than that, I also wanted to see how his per-game grade was affected by the level of competition to see if there was any correlation.  So, I took his per game grades for 2012, and compared them against the overall season grades of each opponent, and then sorted those opponents from worst grade to best grade.  The table below shows Kalil's per-game grade trend against his level of competition:

First off, in reading the graph keep in mind that the red line, which is the final season grade of his opponents, has a much larger range of numbers (from -33.8 to 28.3), because it accounts for 16 games worth of grades all added together, whereas the blue line, Kalil's single game range (-3.1 to 3.9), is much smaller.  So the trend lines will have much different slopes as a result.  At first glance you might interpret it as Kalil's performance being the same regardless of the level of opponent as the blue line hovers close to zero.  But, again the ranges of numbers are so different it's hard to see it on this graph.  If you look closer you can see that there is a downward trend in Kalil's performance relative to the increase of his level of opponent.  It's not a dramatic trend, and it's worth pointing out that Kalil's best graded game came against one of his better opponents.  But, against those six good opponents, Kalil registered a positive grade only twice (33%) whereas against the ten negative graded opponents he notched a positive grade eight times (80%).  So, not only did Kalil face poor competition the majority of the time his rookie year, he also did much better against the poor competition than he did against the good competition.  This would peg him pretty much as an "average" player if my definition at the outset is correct.  He did well against poor competition and poorly against good competition, so would you really call that a Pro Bowl caliber season?

In the 2013 season Kalil had a worse overall season grade than 2012, clocking in at -6.0, ranked 51^st out 80 offensive tackles who played at least 25% of their team's snaps.  So why did Kalil have such a lower grade compared to 2012?  Well, a common theory is the pneumonia Kalil dealt with in December of 2012 slowed him down and that he entered camp too far underweight, and therefore spent much of the 2013 offseason trying to gain back the weight rather than building extra strength and training for the season.  While he arrived at training camp in 2013 weighing about 311 pounds, a few pounds shy of his 315 pound goal, he was still much heavier than he was at the 2012 Scouting combine when he weighed in at only 306 pounds.  So while it's certainly possible that his bout with pneumonia had an effect on his athletic ability, in terms of his weight he seemed to deal with it just fine.  I don't personally feel it would have made that big of an impact since the illness was in December, and the season was underway 8 months later.  Another theory is that perhaps Matt Kalil dealt with a sophomore slump, a somewhat common occurrence for athletes in their second years.  While it could certainly be one or both of those things to explain Kalil's drop in grade, a more compelling reason as to why he performed more poorly is by looking at his level of competition from 2012 to 2013.

As mentioned above the average season grade of his opponent's during the 2012 season was -4.6.  In 2013 that level of competition jumped up to an average of +1.4.  The total number of "good" opponents (grades above 0.0) increased from 5 different players in 2012 to 7 different players in 2013 (or 6 to 7 total).  In other words, Matt Kalil faced tougher competition in 2013 than he did in 2012, and in my opinion that would have had a much more contributing factor to the decline in his performance.  Again in the chart below, which shows the same data as the chart above just for the 2013 season, we see an almost identical trend downwards as the level of competition increases.

Even more compelling is that against those seven good opponents, Kalil recorded only one game in the positive (14%) whereas against the lesser competition, he notched six good games (67%).  So not only is there another slight downward trend in Kalil's performance as his competition gets better, he has a lot more good games against worse competition.  In other words, Kalil has a similar record of performance as his rookie year, but just faces more difficult competition.  This is almost the very definition of an average NFL player.  I think it's particularly striking that the 2012 and 2013 graphs are very similar.

Last season was generally regarded as Kalil's worst season so far.  He allowed the most sacks and quarter back pressures of his career, and ended up with the worst season grade of his career too.  In 2014, he earned a -29.1 season grade, ranked 81^st out of 84 offensive tackles.  In other words, Matt Kalil was graded as one of the worst offensive tackles in the NFL last year.  There were several popular theories about why this might have happened like having to learn a new offense and scheme, as well as dealing with a knee injury at the beginning of the year.  While it's certainly possible that the knee injury played a role, again, the level of competition makes a more compelling case.

In 2014, the level of competition Matt Kalil faced once again jumped up from the season before, to an average overall grade of +5.2.  That means his level of competition increased by a grade factor of almost +5 each year.  In 2012 the average defender he faced graded -4.6, in 2013 +1.5, and then in 2014 +5.2.  In addition Kalil's grade declines by more than triple that factor each year from +12.1 to -6.0 to -29.1.  The correlation is very strong between Kalil's season grade and the average season grade of his opponents.  Not only that, but the number of opponents Kalil faced in 2014 that ended the season with a positive grade jumped up from seven the year before to a total of twelve.  Against those twelve good opponents, Kalil only registered three games with a grade above zero (25%).  Although, it's worth pointing out that he also didn't manage a single good game against the four bad opponents either.  This change from his first two years is either due to a small sample size, or perhaps the knee really was to blame.  Never-the-less, the 2014 trend lines in the chart below still look very similar to his previous two seasons:

There are a couple of differences in 2014 to notice.  First, Kalil's trend line is much farther below 0 this time around, although the descending trend slope is pretty much the same.  Also, the red trend line crosses the 0 mark much sooner, because there were a lot more highly graded players.  This puts the point of intersection much earlier suggesting that that Kalil had a little bit worse year than the first two, even when taking into account the level of competition.  Again, the fact that he didn't register any games with a grade above zero against his poor competition suggests that he had a below average year in 2014, and merely average years the previous two.  So, perhaps there was some truth to the knee injury impacting him last year.

I think the expectation that he should be an all-pro talent and was destined for many future Pro Bowls was set up by his draft position (#4 overall) and his strong rookie year.  But in looking at this a little more closely, his strong rookie year appears to be more a by-product of facing off against sub-par competition.  Every year he has demonstrated essentially the same relative trend of performing more poorly against good competition and better against weaker competition.  It just so happens that in 2014 he faced an unusually high number of really good defenses, especially when compared to 2012.  To prove this point even more, the table below combines all of the data from his three seasons above to show, essentially, the same trend.  Again, it is sorted by his opponent's overall grade from worst to best, with Kalil's corresponding game grade against each of those opponents:

When you consider his entire NFL career thus far, in 48 games he has faced 24 opponents with an overall season grade greater than zero and 24 opponents with a grade less than zero, split exactly in half.  Against those 24 "good" opponents, Kalil scored a positive grade only six times (25%).  Against the 24 "bad" opponents, Kalil scored a positive grade 14 times (58%).  In other words, Kalil scored a positive grade against bad opponents more than twice as often as he did against good opponents.

This next table presents the same information, but I have ordered it differently to show the same trend a different way.  This time I ordered Kalil's games in order of worst to best grade and correlated that with his opponent's overall season grades.  This time you can see that as Kalil's per-game grade increases, the level of competition decreases.  Like the graph above, the point of intersection is nearly in the middle.

Now some may have a problem with the way I re-arranged his game grades as it doesn't show his performance in real-time.  So the table below shows his entire career arch, in chronological order.  Notice...the same trend is still there and if anything it is even more convincing as it shows the overall descent of his career grades relative to the overall ascend of his level of competition.

Now, because the overall season grade is such a larger number range than the single game grade, the two tables below consider only the games each opponent played against Kalil, which means the trend lines are much more even.  These graphs are admittedly not as helpful, because they don't take into account the opponent's complete skill-set against a wide variety of NFL talent.  In other words it's not as convincing a number about what type of player the opponent is, because it only considers how well those opponent's did against Matt Kalil.  And naturally since they are faced off one-on-one most of the time, if one player had a good game, it stands to reason that the other probably had a bad one.  The first table is sorted by opponent games from worst to best, and the second is sorted by Kalil's single game grades from worst to best.  Both show a strong correlation that to me, simply confirms that this study might hold some water.  We would want to see this kind of correlation on an individual per-game basis if these one-on-one matchups were to matter.

Conclusion

So what does this all mean?  Well to me, I think it shows that Kalil is probably not as good as we thought or expected he would be after his rookie year.  While the illness, knee injury and sophomore slump may have all contributed to his poor performance in the following seasons to some degree, I personally would not put much stock in any of it, because his trend lines are so very similar in each of his first three seasons.  In other words, Kalil seems to perform the same way against good competition and bad competition every year.  It just so happens that the level of competition got harder and harder each year.  This ultimately suggests to me that Matt Kalil is merely an average starting level player in the NFL and that his performance has actually been pretty consistent.  It's too bad that we had to spend the #4 overall pick on what is essentially a replacement level player, although starting left tackles are admittedly difficult to find.  I believe ultimately it was because of the positional value, and the relative safe-ness of Matt Kalil as a prospect, that he was ranked and drafted so highly.  I don't fault the Vikings for making the pick, because we had a huge gaping hole at left tackle heading into that draft, and who wouldn't take a safe starting NFL caliber player, even an "average" one at a major position of need?

What To Do With Matt Kalil

So do I think we need to replace Matt Kalil?  Not necessarily, although I do think bringing in competition to push him (just like any position) is a good idea.  While the 2015 NFL schedule won't be released until tomorrow, we do know who their opponents are.  The table below shows all 16 teams Matt Kalil will have to face next season, along with the overall 2014 season grade of the defensive end or outside linebacker he will have to face this upcoming season too.

Team	Player	2014 PFF Grade
Chicago	DRE Jared Allen	+5.4
Chicago	DRE Jared Allen	+5.4
Detroit	DRE Ziggy Ansah	+17.2
Detroit	DRE Ziggy Ansah	+17.2
Green Bay	ROLB Julius Peppers	+20.2
Green Bay	ROLB Julius Peppers	+20.2
Seattle	DRE Michael Bennett	+34.8
St. Louis	DRE Robert Quinn	+14.4
Kansas City	ROLB Tamba Hali	+11.2
San Diego	ROLB Melvin Ingram	-1.6
New York Giants	DRE Jason Pierre-Paul	+16.9
Arizona	ROLB Sam Acho	-0.9
San Francisco	ROLB Aldon Smith	+5.8
Denver	DRE DeMarcus Ware	+4.4
Oakland	DRE C.J. Wilson	+0.2
Atlanta	DRE Osi Umenyiora	+1.9
Average Grade		+10.8

I hate to be the bearer of bad news, but things do not look good for Matt Kalil in 2015.  If the 2014 season was his most difficult yet of his young career, then the 2015 season is stacking up to be even worse.  There are only two names on the list above with a 2014 grade below zero and the average defender he will face in 2015 had a +10.8 grade in 2014, nearly 5 more points better than last year!  It's a scary thought that Kalil will, yet again, have to face an average defender that is graded 5 points better than the year before.   So based on how Kalil has fared against good players over the past three years, if the trend continues we could be looking at an overall grade of -45 or worse for him!

I take one look at that slate of opponents and wonder if there is ANY left tackle in the NFL that could string together a solid season against a list that contains essentially the most elite levels of edge defenders in the NFL, let alone an average, replacement level player like Matt Kalil.  So at the end of the day, 2015 a "contract year" for Matt Kalil as it is not expected that the Vikings will pick up Kalil's fifth year option before May 2nd.  The 2015 season will be his toughest test to date and should be the perfect barometer for contract negotiations.  If Kalil passes the test, he will eventually cash in a huge payday.  If he doesn't, he may end up being just an average NFL starter who failed to meet the lofty expectations of being drafted 4^th overall.  Only time will tell of course.