Monday, August 25, 2014

Yoyodyne: The Future Begins Tomorrow!

Everybody can probably remember what it was like to be a small child waiting for Christmas morning.  Time moves much more slowly when you are young, and it slows down even more on Christmas Eve.  Now, imagine what it would be like if you ran downstairs to the Christmas tree and unwrapped the brand new bicycle that you had wished for, only to find that it was completely disassembled.  Your father, a man with no mechanical skills, promises to "get right to it", meaning that it should only take about three months for him to put it all together.  To be fair, my father did lose both of his hands in a bizarre meatloaf related accident, and had to have hams sewn onto the ends of his arms to take their place.  Of course, this also doesn't take into account the additional months of waiting, when your cautious mother ships the assembled bicycle off to the Underwriters Laboratory for safety testing, to ensure that her precious baby boy won't be killed on this mechanical death-trap.

Under these circumstances, you would probably start to have very mixed feelings about Christmas.  A more deprived life in an orphanage might even start to seem like an appealing option.  I don't know.  I guess I was a fairly impatient kid.  Maybe it all balances out in the end, when you get to put your parents in a nursing home (Who's going to be laughing then, Mom and Dad?).

Now, I'm not a child anymore (supposedly), and perhaps I have exaggerated the way in which my parents operate.  Either way, I am older now, and I have hair growing out of my nose, and my expectations for immediate gratification have been somewhat subdued.  I know that the world doesn't move as quickly as I would like, or else we would already have jet-packs (or Japanese robot butts).  You have to learn to be patient, even if waiting can be extremely frustrating at times.

One of the most annoying aspects of speculating about NFL Draft (football Christmas) prospects is the endless waiting.  It's particularly bad when you have a hunch that a late round draft pick might be pretty good, but nobody else really seems to care.  It's even worse when your interests are drawn to some of the less glamorous positions like that of the NFL center.  Try having a serious discussion about NFL centers with a friend, and watch how quickly people begin to fall asleep.  Yet, for some reason, I find this subject endlessly fascinating.  I should I put this...a bit of a dork about such things.  Unfortunately, once the draft is over, all I can really do is sit around and wait...and wait...and wait...just to see if the players I am interested in will ever get an opportunity to play.  Sometimes it takes years for all of this to play out, and even then we sometimes don't get a real chance to judge a player's skills, simply because they may never be given an opportunity.

Normally, I have somewhat mixed feelings about Pro Football Focus, and their methods for grading players.  Sometimes I ignore what they have to say....and sometimes they agree with me and I make sure to mention it here.  Yes, I can be a bit of an inconsistent ass in that way.  Either way, I do find what they are attempting to do quite interesting, even if I sometimes have some differences of opinion with them.  This morning, however, they posted an article that had me howling in delight.

When they posted their grades for the top performers in the 2014 NFL preseason, a stupid smile began to spread across my face.  According to their numbers, the top rated center in the preseason has been Eric Kush of the Chiefs, with a grade of +4.7 cheeseburgers.  Okay, they don't really grade people on a 'cheeseburger scale', but the methods by which they evaluate people can sometimes be a bit mysterious, so I sometimes make up meanings for their numbers.  Either way, Eric Kush scored more cheeseburgers than any other center, at least so far, in the 2014 preseason.

This is particularly funny because PFF has been busy trashing the rest of the Chiefs offensive line, so it doesn't appear as if Kush is likely to be benefiting from an abundance of surrounding talent.  This is all very exciting to me, because Eric Kush was our favorite center prospect from the 2013 draft, although he wasn't selected until the 6th round.  Seeing some signs that things might actually work out for this strange little prospect of ours is sending me into a bit of a tizzy, especially since he is somebody we selected in our ongoing battle versus Ozzie Newsome.  I have to admit that I have the horrible feeling that even mentioning any of this will end up jinxing things for Kush.

Then, things got even weirder...

Among the other centers that PFF had ranked in their top 5, along with established names like Alex Mack and Nick Mangold, was another somewhat peculiar fellow.  Tied for 3rd place on their list, with a score of +3.7 cheeseburgers, was Corey Linsley who the Packers drafted this year in the 5th round.  While there were a handful of people that we considered to be interesting center prospects in the 2014 draft, we finally settled on Linsley as our favorite overall center candidate.  All things considered, I think that projection is starting to look fairly good, considering that we probably spent a whopping 2-3 minutes actually watching him play, and based this hunch almost entirely on his physical measurables.  Our deranged fascination with the short shuttle times of center prospects seems to again be paying some dividends, even if it is a remarkably simple-minded way of doing things.  Yes, Reilly and I do factor in other results besides short shuttle times, when judging centers, but it still remains our favorite score for that position.

Obviously, none of this necessarily means that much.  There's still plenty of time for these players to turn around and make our little projections look incredibly stupid.  For the moment, however, I just want to savor the possibility that maybe...just maybe...we don't have our heads firmly lodged in our buttocks.  In no way am I claiming that these two players are Canton bound, but for the moment, there do appear to be some promising signs of competence.  Even if they were to merely become average/serviceable players, I think that would still be a rather positive outcome, considering how little was invested in them.  My fingers are crossed.

Of course, there is a bit of a catch in all of this.  Eric Kush is still listed as a backup, and I suspect there is a good chance he won't really see the field until 2015, when starting center Rodney Hudson's contract runs out.  I may personally believe that Kush is a much better prospect than Hudson ever was, but that's where PETARD comes back into play.  Hudson was a 2nd round draft pick.  Short of a complete and undeniable failure on his part, the team is likely to give him the starting job, instead of Kush.  Since the same questionable "trust your eyeballs, and not the numbers" type of evaluation goes into assessing NFL players, that also goes into choosing who to draft, it's entirely possible that the Chiefs will squander another year of Kush's career, having him sit on the bench.  How annoyed am I with this situation?  Extremely.

Linsley, on the other hand, appears poised to become a starter this year, even though I don't personally think he is as exciting a prospect as Kush probably is.  Don't get me wrong.  I do think Linsley could be quite good, but I just that think Kush could be really exceptional.  I'm sure Linsley will struggle at times, and there will be ample opportunities to question his abilities, but that is sort of to be expected with all inexperienced offensive linemen.  Still, I generally support the sink or swim approach, of tossing players into games early, and figuring that a few screw-ups will be more educational than they will be harmful.  Hopefully it all works out for the Packers.  Again, I'll have my fingers crossed for a positive outcome here.

Because I haven't said anything negative about the Ravens yet (a requirement), I figured I should tack this on at the end.  In the past year, the Ravens have traded for two veteran centers (Shipley and Zuttah), while starting one of the league's most abysmally poor centers in Gino Gradkowski, in 2013 (who was selected in the 4th round in 2012).  While I suspect Zuttah will inevitably turn out to be an improvement over the performance we saw from Gradkowski (with a fairly pleasing 4.54 second short shuttle time coming from Zuttah), you still have to wonder if all of this scrambling around to address the center position was really necessary or economically sound.  Anytime you sign a veteran player, you are getting somebody who is both older and more costly than what you could have acquired in the draft.  It's still too early to say whether Kush or Linsley will pan out, but if they do, and if the signs that pointed to success were relatively apparent to our Banana 6000 data thresher, then why weren't they apparent to the Ravens?

However this all plays out, I'll just be sitting here waiting.  Like the slogan of Yoyodyne Propulsion Systems says,"The Future Begins Tomorrow", and the NFL has taught me just how true this statement is.  I just find it a bit frustrating how often I see teams with well known areas of weakness spending their preseason games trotting out the same tired players on the field, while giving fewer opportunities to some of the lesser known guys who are buried on their rosters.  The idea of shaking things up, and trying to reevaluate some of the lesser known talents in the league, seems to always take a back seat to simply making decisions based on where someone was drafted.  So I keep waiting...and waiting...and waiting...

Wednesday, August 20, 2014

The PETARD Hall Of The Reasonably Productive

The eventual death of this blog will most likely be a product of how few positive things I generally have to say.  I suspect people are much more interested in hearing good news rather than bad news.  Unfortunately, I come from the school of thought that says "If you don't have something negative to say, don't say anything at all".  It's not that I don't want to be more optimistic regarding the careers of NFL players, it's just that it feels a bit pointless to state the obvious, simply for the sake of being upbeat.  That's why we have Jon Gruden, after all.  We all know that Aaron Rodgers is good.   Jimmy Graham?  Well, he seems to be doing quite well.  What's really left to say about those sorts of players?  I'm much more interested in why certain draft picks fail to meet our expectations, and to what degree such disappointment might be avoidable.

Still, when Reilly and I recently got back from our annual expedition to Bohemian Grove, we were feeling a bit more buoyant than usual.  A weekend of peeing in the woods and sacrificing goats in the Owl Shrine with our nation's overlords just has that sort of effect on us.  So, with a more chipper outlook, I thought we would examine some of the players who have had somewhat surprisingly positive outcomes to their careers, at least relative to where they were drafted.  It's time to give the Ravens credit for their draft picks that seemed to have done well for the team.

We'll do this the same way that we did with the players in the previous post, where we examined the players who fell short of our expectations.  As always, we will be using PETARD to define what the average number of games started would be for a player (in their first 4 NFL seasons), depending on where they were selected in the draft.  We will list the number of games that the player actually started, as well as the number of games that PETARD would have projected for them to start.  When we examined the players who failed to meet PETARD's expectations, we were only looking at players who were selected in the first 100 picks of the draft.  This time, when we examine the "successes", we will only be looking at players who were taken with the 90th pick of the draft, or later.  Undrafted players will not be included in this list.  Since teams seem to be a bit overly exuberant in their confidence when it comes to high draft picks, and frequently give them opportunities almost regardless of their actual performance, drawing a line at the 90th pick should help restrict our attention to players for whom the team probably didn't have enormous expectations.  Basically, if these players got on the field, they probably earned their playing time, rather than just being the products of draft hype.

Since we're trying to see if there is a pattern to who succeeds, versus who fails, we'll generally make a note of the player's physical attributes, and perhaps a little bit about their statistical production in college.  Sometimes pieces of this information just aren't available, but we'll do the best that we can with what we have.  As for the physical data, we will generally just refer to a player's Kangaroo Score (our measure of lower body power and explosiveness), and the Agility Score (based on the short shuttle and 3-cone drill).  These scores are given in the form of how many standard deviations that a player is above or below the average result for someone in their position group.  So, the results for a linebacker, for example, shouldn't be directly compared to those of a wide receiver.  In the end, we just want to see if there were signs related to a player's physical traits or statistical production, which would have helped us to differentiate them from some of their less successful peers.

Seeing as how none of these players will have been selected before the 3rd round, it is unlikely that we are going to run across any players of the Calvin Johnson or J.J. Watt type of physical freakishness.  Players of that sort are generally so obvious that they don't get overlooked by NFL teams very often.  Still, even if these players are likely to be of a somewhat lesser caliber, we're just looking to see if the objective facts seemed to point more towards a good argument for their selection.

There is one final note that I should mention.  As I've mentioned before, fullback Le'Ron McClain has generally presented some issues for PETARD.  Because of the nature of the fullback position, it is hard to treat his extraordinarily high PETARD result very seriously.  As things stand, we would be forced to say that no Ravens' selection has exceeded their PETARD projection as much as McClain has, but I think this is more a result of him playing at an unusual position, that has a very different set of expectations placed upon it.  Being selected in the 4th round is about as high as you ever see a fullback taken, and when such a pick is made, the player is probably more comparable to a more highly drafted player, at a more conventional position.  So, for now, we are going to leave Le'Ron McClain out of this discussion.

Now, let's get on to the list...

Ed Mulitalo, G, 129th pick in 1999  Actual # GS:  53   PETARD Proj. GS: 12.9
It's a bit unfortunate that we have to start this list with Edwin Mulitalo, because I really don't have a great explanation for his career.  Since offensive linemen lack traditional stats when they come out of college, we have to rely on their combine performance to a much larger degree.  With a -1.089 Kangaroo Score and a -0.487 Agility Score, I have to admit that there was nothing here that I would have found intriguing at all.  No matter what angle I come from, I would have had a hard time seeing Mulitalo as an interesting prospect, based on the data that was available at the time.

How good was he in reality?  I really can't say.  He did manage to start 128 games over his career, and though he never struck me as anything amazing, that is still quite an accomplishment for a 4th round pick.  Maybe he was just a fairly pedestrian guard, who survived at a position that just draws less attention?  Or, perhaps he's just an oddball exception to my normal rules?  I'm really not sure.  Either way, having him head this list at least shows that we're not trying to unfairly tilt the conversation in a particular direction.  There are always going to be guys like this who exceed my expectations, and it's just a part of life that we have to accept.

Dawan Landry, SS, 146th pick in 2006  Actual # GS:  48   PETARD Proj. GS: 10.89
With Dawan Landry, I think we have an imperfect, but still rather interesting prospect.  That's about the best you can reasonably hope to find in the 5th round of the draft.  His 4.64 second 40-yard dash (which he improved upon slightly at his pro day), along with an Agility Score of -0.588, doesn't suggest a lot of speed or gracefulness.  Still, when it comes to strong safety prospects, I'm usually willing to be a bit more flexible about agility than I would be with cornerbacks or free safeties, and generally have a cutoff in the -0.400 to -0.600 range, which barely lets Landry slip past inspection.  Of course, making such an exception would depend on whether other areas check out okay.  When we look at his exceptional Kangaroo Score of 2.322, we start to see some real reasons to get excited.  Anytime a player has a Kangaroo score that is over two standard deviations higher than his average peer, I pay attention.  That is precisely the sort of lower body power and explosiveness that we might hope to find in an in-the-box hard hitting safety.  But how did he actually perform in college?

        Year         TKL         TFL       Sacks         PBU           Int            FF
2005 76 4.5 0 5 4 0
2004 81 10 3 2 1 1
2003 85 4.5 0 2 2 0

Now, I could easily come up with a list of players that have gaudier college stat sheets, but I think Landry's results would still be viewed by most people as quite solid.  That he produced these results over the course of several seasons, and showed a reliable above average steadiness to his performance, also minimizes any concerns about the flukiness of these sorts of things.  There are few things I hate worse than a "one year wonder".  While Landry might not be a perfectly designed safety, it would have been difficult to argue that he didn't appear to be worth taking a shot at in the 5th round.  With reasonably impressive physical potential, and proven production, I would have to say that his relative success shouldn't be terribly surprising, even if he is perhaps never going to be much more than an 'in the box' type of safety.

Ed Hartwell, LB, 126th pick in 2001   Actual # GS:  46   PETARD Proj. GS: 13.33
With Ed Hartwell, I think we have another fairly interesting prospect.  Athletically he was a bit of a mixed bag of results, though we can't really expect many dominating physical freaks to last until the end of the 4th round.  You gradually have to lower your expectations as the draft rolls on.  His Agility Score of -1.349 is clearly not very good, and might have suggested some potential problems dropping into coverage, but this wouldn't necessarily rule him out as a capable middle linebacker, intended to simply stop opponents from running up the middle.  It's when we look at his Kangaroo Score of -0.025, that we begin to see some of this potential.  Yes, -0.025 might not look like a good result, but since I weigh middle linebackers on the same scale as the larger OLBs and DEs, the results for middle linebackers tend to get pushed down a tad.  As I've said before, I generally view -0.800 to be an average result for the somewhat lighter middle linebacker prospects, with Pro Bowl and All Pro middle linebackers averaging about -0.400.  So, in comparison to those results, Hartwell actually comes out looking quite nice, and well into the range where we find a lot of All Pro/Pro Bowl MLBs, and with a result that is actually about 0.825 standard deviations better than his average peer.  Still, with that said, I probably would have expected him to be a fairly one dimensional player, which he probably was.

When we move past these measurable athletic traits, the picture becomes even more interesting.  While I have been unable to find his college statistics prior to his senior year, that one season of play seems to be worth taking a real look at.

      Year        TKL   Solo TKL       TFL      Sacks         PBU        Int         FF
2000 184 117 12 5 3 1 2

I would have to call that an extremely productive season.  In fact, even if we compared this production to all of the middle linebackers taken in the first round during the past decade, I would say Hartwell appears to measure up quite nicely.  While injuries appear to largely be to blame for the derailing of Hartwell's career (a not unusual case of Exploding Knee Syndrome), that shouldn't diminish what he accomplished early in his first few NFL seasons.  Would I have picked Hartwell in the 4th round?  I'm not sure.  In the end, it would have been a question of whether I felt his exceptional statistical production and lower body power balanced out his poor Agility Score.  I probably wouldn't have bet against him though, and for a mere mid-to-late-round pick, the price was probably right for a guy like Hartwell.

Jason Brown, C,124th pick in 2005   Actual # GS:  45   PETARD Proj. GS: 13.6
Say what you will about Jason Brown's disappearing act once he joined the St. Louis Rams, he still appeared to be quite a good center during his time with the Ravens.  His Kangaroo Score of 0.437 was a fairly good result relative to other offensive linemen, but particularly exceptional when compared to other centers.  When I was recently looking at all of the starting centers from the 2013 season, the average Kangaroo Score was -0.324, so centers do tend to have the least lower body power along the offensive line, at least on average.  Brown was somewhat unusual in this area, at about 0.761 standard deviations above that average starting center's result.  Perhaps more importantly, he had an Agility Score of 0.656, which includes a short shuttle time of 4.52 seconds.  That short shuttle time places him 1.110 standard deviations above the average results for an offensive lineman, and is the number one trait I look for in a starting center prospect.  At least from my perspective, he was precisely the sort of prospect I would expect to see thriving at the center position, and for a period of time, he appeared to do exactly that.  For the price of a mere 4th round pick, the payoff seems to have been quite reasonable, though not terribly surprising when you consider his physical advantages.

I'm going to skip past Derek Anderson at this point, though he did manage to start more games than PETARD would have expected of him.  Since we're mainly looking at the degree to which athletic measurables could have predicted these outcomes, quarterbacks sort of fall outside of the scope of this examination.  Still, Anderson is the only quarterback that the Ravens have ever drafted, who wound up going to a Pro Bowl, which is sort of a funny bit of (depressing) trivia.

Jeff Mitchell, C, 134th pick in 1997   Actual # GS:  40   PETARD Proj. GS: 12.31
When you go back to 1997, it becomes nearly impossible to find combine results for most players.  A fractured leg in November of his final college season also increases the likelihood that he never would have participated in the NFL Combine or his college's Pro Day.  Since offensive linemen also accumulate no stats, at least of a conventional variety, it is even more difficult to say how I would have viewed Mitchell as a prospect.  Either way, he had a relatively long career, and started a total of 118 games, so by the typical standards of what you expect from a 5th round pick, he was quite productive.  If anybody can fill in some of the blanks here, feel free to let me know.  I have to admit that I would really love to know his short shuttle time, in particular.  I have a peculiar and unhealthy interest in such things. 

Terry Jones, TE, 155th pick in 2002  Actual # GS:  35   PETARD Proj. GS: 9.9
Perhaps this is a somewhat similar situation to what we have with Le'Ron McClain.  On paper, there was nothing about Terry Jones that struck me as exceptional in any way.  When he was on the field for the Ravens, there was similarly very little to get excited about.  He mainly seemed to serve as a blocking tight end, who rarely contributed much in the passing game.  In what was perhaps his most productive season, at least from a statistical standpoint, he accumulated 159 receiving yards on 20 receptions, with 3 touchdowns.  It seems highly probable that his role as a blocking tight end put him in a position somewhat similar to what we find with our Le'Ron McClain fullback dilemma, as a peculiar role player, which may have allowed him to get on the field a bit more than PETARD might have expected. 

Adalius Thomas, DE/OLB, 186th pick in 2000  Actual # GS:  26   PETARD Proj. GS: 6.87
This is probably going to be one of my favorite late round oddball picks that the Ravens made.  From the perspective of pure physical potential, Adalius Thomas was a rather stunning prospect.  At 270 pounds, running a 4.56 second 40-yard dash, while also displaying an extremely good 1.573 Kangaroo Score, Thomas was quite the physical specimen.  His only minor shortcoming was a slightly below average -0.306 Agility Score, though this never stopped the Ravens from dropping him into coverage (though I never felt he excelled in this area).  Thomas clearly had the sort of physical ability that teams shouldn't have ignored, though he still slipped to the 6th round.  Maybe his on-the-field performance in college just didn't live up to the these physical traits?

      Year         TKL        TFL      Sacks         PBU            Int         FF
1999 67 16 9 0 0 1
1998 71 20 12.5 5               ?           ?
1997 65 15 9 10 1           ?

Umm...nope...that doesn't appear to clarify things at all.  His college stats actually appear to be exceptionally strong.  In fact, it only makes his late round status more of a mystery.  If I employed my normal methods for determining where a pass rushing prospect should be selected, as shown in the post on Explosive Pass Rushers, Thomas would actually come out with a 1st round grade.  I suspect most people would view that as a bit of a reach, since Thomas was probably a 'very good' but not necessarily 'great' type of NFL player.  Still, I would say that the computer's view of Thomas' potential was probably a bit more accurate than the view most NFL teams seemed to have of him at the time.  In retrospect, I think he probably performed at a level that would have made a 2nd round pick seem like a fairly reasonable price to pay.

Jermaine Lewis, WR/KR, 153rd pick in 1996  Actual # GS:  27   PETARD Proj. GS: 10.11
What I can say about Jermaine Lewis as a draft prospect, is somewhat limited.  When you go back as far as the 1996 NFL Draft, finding combine data gets to be a bit difficult.  While a good portion of my hunches typically rely on that information, we can still look at his statistical production at Maryland. 

In 1995, his final college season, Lewis had 66 receptions for 937 yards.  That would have been good for 26.8% of his teams total offense, and significantly above the 17.75% mark that we consider to be an average result for a draft prospect.  In his Junior season, Lewis had 45 receptions for 692 yards.  That would have been good for 16.98% of his team's total offense, and slightly above the 15.34% mark that we consider an average result.  Now, normally, I don't mention much about a receiver's sophomore season, but in this case I will.  In Lewis' sophomore year, he had 52 receptions for 957 yards.  This would have been good for 19.69% of his team's total offense, and clearly an above average result.  It should also be noted that there were 3 games in his sophomore year that Lewis didn't even get to start, so his results could have been even higher if we adjusted for this factor.

So, while we may not know how he measured up at the combine, there's really no denying that his actual college performance did merit some serious attention.  I'm not usually a fan of drafting small receivers (Lewis was about 5'7", and 183 pounds) but it's also clearly much easier to justify a move like this later in the draft.  Lewis was primarily known for his skills as a kick/punt returner, and was selected for 2 Pro Bowls, as well as 2 All Pro teams, though he also produced 2,129 career receiving yards and 17 receiving TDs.  All in all, not a bad result for a 5th round receiver, but also not terribly shocking considering the information that we have about how he performed in college.

Tony Pashos, OT, 173rd pick in 2003  Actual # GS:  23   PETARD Proj. GS: 8.07
I don't think that anyone would claim that Tony Pashos was a star, but as I mentioned in the post about the Lobotomy Line, I think we could argue that there is probably minimal difference between Pashos and the more heralded and hyped Michael Oher.  It's one of those unfortunately familiar cases where the perception of the difference between two players seems to hinge upon their draft status (Oher was selected in the 1st round, Pashos in the 5th), regardless of any actual differences in their performance.  Sadly, I can't really say much about how I would have viewed Pashos' athletic ability, as there seems to be practically no data available on this subject.

Arthur Jones, DT, 157th pick in 2010  Actual # GS:  20   PETARD Proj. GS: 9.68
Due to an injury in his final college season, there is no combine information to help clarify some of the questions we might have about Arthur Jones.  If we ignore his final injury plagued college season, and instead look at his performance prior to that point, we see that he had 13 tackles for a loss and 3.5 sacks in 2008.  In 2007, he produced 17.5 tackles for a loss, and 1 sack.  Yes, his sack totals may not have been outrageous, but he was clearly producing a lot of tackles behind the line of scrimmage for an interior linemen, and I put a fair bit of stock in that.  All in all, I would say those results are extremely encouraging signs of a prospect's potential, and might have warranted a fairly high draft pick if we could have filled in some of the blanks about his physical abilities.  Still, without any combine data, I probably would have been afraid to pull the trigger on a player like Jones, even if the limited data we have is fairly exceptional.  In the end, things worked out reasonably well for the Ravens with this pick.  It's just a shame that we'll probably never know where he would have measured up athletically.

Casey Rabach, C,  92nd pick in 2001  Actual # GS:  23   PETARD Proj. GS: 18.55
It's really unfortunate that acquiring combine results prior the the early 2000s is sometimes so difficult to find.  This leaves us with a somewhat incomplete picture of Casey Rabach.  While we are missing several bits of data about Rabach, I am particularly disappointed to not have his short shuttle time.  Nevertheless, the one piece of measurable data we do have is his vertical jump.  This one sliver of information gives him a Vertical Kangaroo Score of 1.284 standard deviations above his offensive linemen peers, which is quite an exceptional result.  Typically, I find that centers have the lowest Kangaroo Scores of all of the offensive linemen, with results that frequently go into into the negative range.  They often seem to make up for this with quickness and agility.  In Rabach's case, it seems safe to say that his lower body power did not appear to be an area of weakness, but instead an area of surprising strength.  While I wish we had more to go on, this alone would be an extremely encouraging start towards any explanation of Rabach's rather solid career, where he would have 118 total career starts.

But wait...the Ravens must have had numerous other late round surprises, right? really.  Answering questions such as this is going to make it rather difficult for me to maintain the illusion of optimistically supporting the beliefs people wish to hold about the Ravens.  For the most part, the rest of the players that the Ravens have selected in the late rounds of the draft have contributed at a level that is rather close to what PETARD would have expected of them.  Just for the sake of argument, let's look at some of the players that people might suspect were much more beneficial to the team, relative to where they were drafted.

Jarret Johnson, DE/OLB, 109th pick in 2003   Actual # GS:  15   PETARD Proj. GS: 15.73
While I've never thought Johnson was a particularly exceptional player, he has turned into a fairly useful OLB/DE, who has his positive moments from time to time.  Still, he only started 15 games in his first four years, when PETARD would have predicted an average outcome of 15.73 games started for a player who was selected with the 109th overall pick.  Whatever the fans may think of Johnson's ability, the team clearly didn't exhibit much enthusiasm about getting him onto the field, at least while he was operating under his rookie contract.  Basically, the team got out of Johnson, exactly what we would typically expect them to get out of any player taken with the 109th pick. 

Brandon Stokley, WR, 105th pick in 1999   Actual # GS:  11   PETARD Proj. GS: 16.35
Brandon Stokley has always been a bit of an enigma.  Still, many Ravens' fans probably still view grabbing Stokley in the 4th round, as a bit of a steal.  In reality, Stokley would start about 5 fewer games than PETARD would consider to be the average result for a player chosen with the 105th overall pick.  Stokley's only real value seemed to emerge when he went to the Colts, and got to play with Peyton Manning (surprise!).  Prior to that, I think we could argue that he did no more, and perhaps slightly less, than a typical player who the Ravens would have selected at the same point in the draft.

Dennis Pitta, TE, 114th pick in 2010   Actual # GS:  8   PETARD Proj. GS: 14.9
If I haven't made it clear in the past, I really like Dennis Pitta.  Unfortunately, the Ravens seemed to have been less enamored with Pitta than they were with Ed Dickson, as I discussed in an earlier post.  Come hell or high water, the Ravens were going to give the starts to Ed Dickson, no matter how often he dropped the ball.  In the end, this created a situation where Pitta saw the field even less than you would have expected of someone taken with the 114th overall pick.  Ed Dickson, on the other hand, actually started 44 games, when PETARD would have only expected him to start 23.08 games.  This seems like a good time to start asking questions about how much faith you really have in your team's management and decision making abilities, since the only apparent reason for Dickson's edge in this area was probably due to being selected one round ahead of Pitta, in the same draft.  Let the players compete on a level playing field for a starting job, regardless of when they were drafted, you say?  Hahahaaahaaha.  That's hilarious.

Chester Taylor, RB, 207th pick in 2002  Actual # GS:  8   PETARD Proj. GS: 5.1
I've always had a soft spot for Chester Taylor, and think he perhaps got screwed over by landing on teams that had players like Jamal Lewis and Adrian Peterson installed ahead of him on the depth charts.  If given a chance, I think he could have done fairly well as a team's primary running back.  Nonetheless, Petard only projected him to start 5.1 games in his first four years, and he ended up actually starting just 8.  Yes, that was more than the computer projected, but not so much more as to really matter.

Okay, now for some final thoughts on all of this.

Despite the fact that numerous players on this list have incomplete sets of data, we have to do the best that we can with the information that is available to us.  All we can really do is try to weigh their individual pros and cons, and see what sort of picture emerges from all of this.

First of all, amongst the players for whom statistical information is available and relevant, the picture seems quite clear.  For Dawan Landry, Edgerton Hartwell, Adalius Thomas, Jermaine Lewis, and Arthur Jones, their statistical production in college at least matched an average player's results, and frequently exceeded it by a fairly large margin.  Compared to the players in the previous post, who fell short of their PETARD expectation, the differences in this area should be pretty clear.  It seems to make sense to invest in players that ave a history of dominating statistical production, which should come as no real surprise, though team's can be surprisingly erratic in giving this any real consideration.

Secondly, while our data related to these player's physical traits is frequently incomplete, particularly for drafts from the more distant past, the bits of data that we do have almost invariably seems to point to this being a rather important factor too.  Even when only a few bits of data are available, these small glimpses into their potential seem to point towards these players having strong physical gifts.  Particularly for Dawan Landry, Edgerton Hartwell, Jason Brown, Adalius Thomas, and Casey Rabach there did appear to be signs of some physical superiority.  Particularly in the area of lower body power and explosiveness (the dreaded Kangaroo Score), these players seemed to do exceptionally well.  Again, when we look to the players who fell short of their PETARD projections, the differences in this area become a bit more stark.  The players who were disappointment in the eyes of PETARD tended to be a bit below average/horrible when it came to their measured athletic ability.

Does it look like focusing a team's attention on physically superior players, with a history of solid statistical production, might be a good idea?  I certainly think so.  Still, this is not meant to suggest that such players are ever a sure thing.  There will always be players who end up succeeding, where I never would have bet on such an outcome.  Likewise, there will be physical freaks with goods stats, who end up becoming disappointments.  Human beings always have an annoying ability to surprise us in that way (my robot army will solve this problem).  All I'm trying to suggest is that the trend tends to point in a fairly obvious direction.  When a team has multiple draft picks at their disposal, it just seems to make sense to focus them on players who have provided the most objective evidence for why they should succeed.  Over the course of time, such repeated investments do appear to work out better, and put the odds a bit more in your favor.

Friday, August 1, 2014

The PETARD Wall of Shame

I might be the only person to find PETARD a bit entertaining, if admittedly a little simple-minded, but that won't discourage me from delving even deeper into my foolishness.  While PETARD specifically focuses on the efficiency of particular draft day decisions that the Ravens have made, I suspect it could just as easily be applied to any other franchise.  NFL GMs probably aren't beautiful and unique snowflakes.  Outside of the occasional run of good or bad luck that briefly separates them, I would guess that most of their results are generally quite similar.  I'm just looking to kick around some general purpose 'cause and effect' types of ideas here.

Since we have already used PETARD to try to examine the Ravens' year-to-year drafting efficiency over the previous decade, and come to the conclusion that the team might be headed for a bit of a decline, I thought we would narrow our focus a bit.  On an individual basis, players do occasionally fall short of, or exceed, the expectations PETARD might have had for how many games they would end up starting in their first 4 NFL seasons.  Missing the mark with our projections, isn't necessarily a bad thing, or the least bit unexpected.  These peculiar outcomes give us an interesting group of players to examine, and ask the important question of "Why did this player fall outside of the predicted range?".  Was the player exceptional?  Did they suck? 

To start off our attempt to answer these questions, I thought we would just look at all of the Ravens' draft picks who have fallen short of PETARD's expectations.  In specific, we will be looking at players who were selected in the first 100 picks of the NFL Draft.  The 100th pick seemed like a reasonable place to draw the line, since it is common sense to assume that players taken in this range generally have higher expectations of contributing.  We're not terribly interested in beating up the players who fell just 1 or 2 games short of their PETARD projection.  Instead we'll just list all of the players selected from 1996-2010 who fell the furthest from their PETARD projection, in order of the number of games started below our expectation.  For each player, we will list their Actual # of GS (umm, the actual number of games they started in their first 4 seasons), as well as their PETARD Proj. GS (how many games PETARD would have projected them to start in the same span of time).

Later on, in another post, we'll take a look at the players who appear to have exceeded expectations.  You have to eat your vegetables first, or you get no dessert.  Yes, a spoonful of sugar would probably help the medicine go down, but I was really torn between these two conflicting philosophies, so I flipped a coin.  It landed on tails, so we're starting with the bad news.

For each individual player, we will also look at their basic athletic profile, based on their combine/pro day results.  This will usually just consist of a brief look at the Kangaroo Score, and Agility Score, which are shown in the form of how many standard deviations that a player diverged from the average result for an individual in their position group.  Obviously, this means that a wide receiver's results are not directly comparable to those of a defensive linemen.  Keep that little detail in mind while perusing the numbers.  Occasionally, other measurable qualities might be thrown into the mix, as well as some discussion of a player's statistical production in college.  The goal is simply to see whether the warning signs of failure/mediocrity should have been clearly visible to the decision makers on the team.

You'll probably notice that this list will contain very few 1st round picks.  Part of this is because picking a reasonably talented player at the top of the draft really shouldn't be that difficult.  Another part of this apparent lack of 1st round failures is also probably because teams will continue to start below average players who were taken in the 1st round, with almost no concern for how foolish this makes them look.  If a team wishes to stick their fingers in their ears, and say "Nyah, nyah nyah, I can't hear you!", when the reality of a player's performance doesn't match what they imagined it to be, that's fine with me.  PETARD isn't intended to make judgments about such things.  Accounting for a team's stubbornness is beyond my abilities, so we're really just looking at the instances where a team clearly relented, and threw in the towel on a particular player.

Before we get started, I should mention that I did make the decision to exclude Jamal Lewis from this list of shame.  Technically, he probably should have been fairly high on this list, since his 44 games started in his first four seasons, fell short of the PETARD projection by 20 games.  Of course, he missed a full season in 2001 because of a torn ACL.  Without that injury, it's quite easy to imagine that Lewis would have otherwise been within 4 games of his projection.  The PETARD projections are already nearly impossible for top 5 draft picks to meet/exceed, since at that point in the draft PETARD expects every player to start every single game in their first four seasons.  So, I felt it was a reasonable compromise to exclude him from this group of disappointing players.

For similar reasons, related to injury, I will also exclude Sergio Kindle and Dan Cody from this criticism.  Normally, I would have to put them at the head of this list as 2 of the most disappointing draft picks the Ravens ever made (relative to where they were selected), but the injuries they dealt with makes this difficult to assess.  Personally, I think there are good reasons to suspect that they would have failed even without this sort of excuse, but I'm sure some people would find it a bit unfair for us to make judgments about how sensible these picks actually were.  So, being a kind and reasonable sort of guy, I'll leave them out of this.

Let's get the ball rolling, and  move on to the list!

Yamon Figurs, WR, 74th pick in 2007   Actual # of GS:  0   PETARD Proj. GS.: 22.16
Can a player get drafted in the 3rd round, simply because he ran a fast 40-yard dash?  Well, yes, that should be pretty obvious to everyone.  Teams pull this sort of ridiculous stunt all of the time.  Sadly, Figurs' 4.30 second 40-yard dash, while impressive, was probably the only thing he really had going in his favor, and generally 40 times are a bit overvalued (not worthless, just a tad overvalued).  His -1.816 Kangaroo Score would suggest that he had laughable lower body power, though this isn't terribly surprising for a prospect who only weighed 174 pounds.  His Agility Score of 0.197 was basically just about average, and nothing to get terribly excited about, one way or another.  Athletically, he was what he was, a mosquito with speed, and not much else.

The bigger concern would probably be his actual production in college.  In his final college season, he produced 418 receiving yards, on 28 receptions, which was only good for 10.02% of his team's total offense.  That's well short of the 17.75% that I usually consider to be an average result for a draft prospect's final college season.  In the year prior to this, Figurs produced a mere 243 yards, on 14 receptions, for a lowly 6.5% of his team's offense.  This would also clearly fall short of the 15.34% we would consider to be the average for a player's next to last college season. 

In the end, you have an undersized player who wasn't very productive in college, and who's only real claim to fame was being fast in a straight line.  People may claim that Figurs was only expected to contribute as a kick returner (though he failed at this too), but for the 3rd round pick that was used to acquire him, I still would have hoped to see a lot more evidence for making such an investment.  His failure seems utterly predictable.  Some people might say that 3rd round picks frequently don't amount to very much, so what's the big deal?   I would say that if you squander such picks on players of Figur's caliber, such prophesies are sort of self-fulfilling.

Musa Smith, RB, 77th pick in 2003   Actual # of GS:  0   PETARD Proj. GS: 21.5
I feel like I should possibly exclude Smith from my criticisms, since the broken leg he suffered in his second season probably did play a significant role in his disappointing career.  He's largely remembered as being the less famous person to get injured and bring about the NFL's current 'horse tackle' rules (the other player being Terrell Owens).  With a 4.58 second 40-yard dash time, and a rather good Kangaroo Score of 1.317, I would say that he seems like a plausible between the tackle type of power running back.  Unfortunately, his injury makes it difficult to say what might have been possible.

When it comes to his production in college, Smith may not have been Adrian Peterson, but he appears to have been quite competent.  With a career 4.9 yard/carry average, 19 rushing TDs, and 1,324 rushing yards in his final season (a 5.1 yard/carry average in that year), he had a fairly solid if not spectacular resume.  He never contributed much in the passing game, but I think there's still a place for somewhat one dimensional power running backs.  I might not have been ecstatic about Smith as a draft prospect, but he certainly isn't someone I would have laughed at.  Things may not have turned out terribly well for him, but he's probably the lone player on this list that I would have viewed as a legitimate prospect.

Jay Graham, RB, 64th pick in 1997  Actual # of GS:  4   PETARD Proj. GS: 24.56
I have to admit that I've completely forgotten that Jay Graham ever existed, let alone that he was a Ravens' 3rd round pick.  Since I can't find any of the data I would need to say whether this selection was insane, or not, I'll just let it slide for now.  If anybody can dig up Graham's combine numbers, feel free to send them my way.

Paul Kruger, DE/OLB, 57th pick in 2009   Actual # of GS:  7   PETARD Proj. GS: 26.48
Criticizing the 2nd round selection of Paul Kruger, might seem a bit odd.  After all, he did manage to sign a fairly lucrative 5 year/$40.5 million contract with the Browns, when he departed the Ravens in 2013.  The problem is, regardless of what you think of Kruger as a player, the Ravens failed to get much production out of him when he was operating under his rookie contract.  Only getting 7 starts out of a 2nd round pick is not a very efficient use of your team's draft resources, so PETARD views this as a bit of a failure.

To some extent, I think the Ravens may have hindered his development by constantly having him add, and then drop, weight.  They never seemed to be terribly clear as to what position they wanted him to play, whether it was to be a 3-4 OLB or a larger 3-4 DE.  Still, I personally suspect that Kruger will never be anything more than a somewhat below average player, especially in his current role as a 3-4 OLB, where explosive athletic ability is very valuable.  With a Kangaroo Score of -0.368, and an Agility Score of -0.868, he just doesn't have the sort of physical traits that you generally see in the more successful players at his position.  Honestly, his results are quite disappointing.

While his statistical production in college was acceptable, it wasn't exactly mind-blowing.  In his final season at Utah, Kruger produced 61 tackles, 16.5 tackles for a loss, and 7.5 sacks.  Overall, that is a fairly good season.  In the year prior to this, he had 63 tackle, 7.5 tackles for a loss, and 3 sacks.  Those are much less interesting results.  If we weighed him according to the method I use in the post on Explosive Pass Rushers, his average result of 12 tackles for a loss in his final two college seasons would leave the computer unable to give him a draft grade any higher than the 4th round.  When we factor in his limited athletic ability, it only gets worse.  Personally, I doubt if I would have drafted him at all, though that may seem a bit harsh.  We'll see over the next couple of years whether he lives up to the Browns' investment.  I'm obviously betting against it.

Devard Darling, WR, 82nd pick in 2004   Actual # of GS:  1   PETARD Proj. GS: 20.45
Along with Dwan Edwards, who is also on this list, Darling contributed to what the computer feels may have been the worst draft in Ravens' history.  Personally, I think Darling was a borderline tolerable prospect, though I still wouldn't have taken him as high as the 3rd round.  With a 4.52 second 40-yard dash time, a 0.785 Kangaroo Score, and a -1.346 Agility Score, his athletic ability was nothing to write home about, though there are much worse physical specimens.  Normally, you might expect me to get moderately excited about a relatively good Kangaroo Score like Darling produced, but it's still just one ingredient in a fairly complex soup.

So, now we have to consider his statistical production in college.  In his final college season, he was responsible for 16.5% of his team's offense. This is slightly below the 17.75% that we would consider to be an average result, but close enough to not matter too much.  In the year prior to that, he was responsible for 14.58% of his team's offense.  Again, just a bit short of the 15.34% we would view as an average result.

In terms of physical potential, as well as proven statistical production, all the numbers screamed 'this guy is fairly average/unexceptional'.  Compared to many of the other receivers that the Ravens have selected over the years, I suppose it could have been much worse.  I guess I'm trying to put a positive spin on this.  In the end, the computer wouldn't have seen much of a compelling reason to spend a 3rd round pick on a player such as Darling.  His 578 career receiving yards, over 5 NFL seasons, even fell short of my very humble expectations.

David Pittman, CB, 87th pick in 2006   Actual # of GS:  1   PETARD Proj. GS: 19.47
Is your team in the market for a relatively small (5'11" and 182 pounds) cornerback, from a low level of competition (Northwestern Northwestern)?  Well, David Pittman might be your man.  With a 40-time of 4.44 seconds, Pittman had reasonably decent speed, but beyond that, his results were quite a bit less impressive.  His Agility Score of -0.478 would have probably given him the sort of gracefulness I normally associate with strong safeties.  Unfortunately, he didn't have the power of a strong safety, with a Kangaroo Score of -0.568.  In this sense, he was a true dual threat cornerback.  An opponent could simply run him over, or choose to outmaneuver him.  He was last seen struggling to make a roster in the CFL.  What exactly was the team thinking when they drafted him in the 3rd round?  I have no idea.

Patrick Johnson, WR, 42nd pick in 1998   Actual # of GS:  15   PETARD Proj. GS: 31.55
It's really unfortunate that finding data from some of the older drafts is so difficult to hunt down.  I can't really speculate too much about Johnson's physical traits, since I can't find any record of his combine results.  It's always been suggested that he had blazing speed, being a former track star, but outside of those rumors I have no clue.  This obviously makes criticizing Johnson a bit more difficult, since we place a significant value on the measurable data from the combine.

Instead, I only have a somewhat rough accounting of his statistical production at Oregon to examine.  In his final college season, Johnson had 1072 receiving yards, on 55 receptions (with the mighty Akili Smith!).  That would have been good for about 21.1% of his team's total offense, and above the 17.75% average for his peers.  In the prior year, Johnson had 218 yards, on 14 receptions.  That would mean he was responsible for about 4.32% of his team's total offense, which is well below the 15.34% average we would hope to see.  Those are clearly some wildly differing results.  Was he just a one year wonder, or did he have real potential?  Without more data to complete the picture, it is difficult for me to really say.  Still, I suspect I wouldn't have had the guts to gamble on a player with such inconsistent college production, with something as valuable as a 2nd round pick.  We should never forget, at one point in 2004 Johnson was actually cut by the Ravens to make room for a then 32 year old Kordell Stewart, who was well into his decline.  The horror...the horror...

Dwan Edwards, DT, 51st pick in 2004   Actual # of GS:  14   PETARD Proj. GS: 28.33
Once again, we run into a player from the Ravens dreaded 2004 draft class.  While Edwards has gone on to have a seemingly serviceable journeyman type of career, I don't think anyone would argue that his selection in the 2nd round was a highlight moment for the Ravens' organization.  I'll keep this section short and sweet.  With a Kangaroo Score of -0.038, and an Agility Score that is an amazingly average 0.000, he had just the sort of stunningly mediocre athletic ability that you would expect of an undistinguished journeyman type of player.  It might actually be a challenge to find somebody with more uniformly average physical traits.  If, on the other hand, you are actually seeking to spend a fairly high draft pick on a rather bland player, this might be just the recipe to achieve that sort of outcome.

Oniel Cousins, OG, 99th pick in 2008   Actual # of GS:  5   PETARD Proj. GS: 17.33
Oniel Cousins may be one of the worst offensive linemen I have ever seen.  That's an impressive accomplishment, since the Ravens also used to employ Bennie Anderson.  Comparisons to a turnstile were made for men like these.  Of course, that is merely my worthless and subjective opinion.  When we look at his -0.515 Kangaroo Score, and his -0.348 Agility Score, it starts to become apparent why he might have been such a disappointment to the team.  Poor lower body power, as well as below average agility, don't tend to make for great offensive linemen.  There are players who have gotten by with somewhat similarly weak measurable traits.  It just isn't something I would tend to bet on.  There was no apparent reason for why the Ravens should have selected Cousins in the 3rd round.  Maybe they thought he had moxie.

Tavares Gooden, ILB, 71st pick in 2008   Actual # of GS:  12   PETARD Proj. GS: 22.84
I vividly remember when the Ravens selected Tavares Gooden.  Perhaps because of the Miami connection, and because of his alleged exceptional speed (though his 40-time of 4.62 seconds was actually fairly average for a MLB), he was often referred to as 'Baby Ray', with numerous people suggesting he might be the heir to Ray Lewis.  Of course, this was all nonsense.  It seems to be a common problem in Baltimore, to prematurely claim that every new MLB might be another Ray Lewis.

While there is unfortunately no data related to Gooden's short shuttle and 3-cone drills, we can still pick him apart a fair bit.  As I've already said, his timed speed was fairly average.  When we look at his Kangaroo Score of -0.748, we see that he measures up just about the same as the -0.800 that I usually use as the definition of an average result for a middle linebacker.  This is still somewhat short of the -0.400 result where we start to see most All Pro and Pro Bowl MLBs showing up.  So, there isn't really any evidence of superior physical potential, but how did he do in college?

        Year         TKL         TFL        Sack         PBU            Int            FF
2007 100 3 0 3 1 0
2006 41 5 0 2 0 0
2005 2 0 0 0 0 1
2004 68 5.5 1 3 0 0

These aren't horrific results, but they also aren't the sort of numbers that merit comparison to one of the all time great MLBs.  I view Gooden's stat sheet as basically fairly average, though it should be noted that he did spend time at OLB in his early days at Miami.  Now, comparing Gooden's statistical production in college to what we see coming from the MLBs who have been selected in the 1st round during the past decade, might not seem very fair (yes, just click the link, and be magically transported to even more of my cynicism).  Still, I think it does put his production into perspective.  In the end, you have a fairly average athlete, with fairly average statistical production, who failed to make much of a name for himself when he came to the NFL.  Is this surprising?  Not in the least bit.

Kyle Boller, QB, 19th pick in 2003   Actual # of GS:  34   PETARD Proj. GS: 44.7
I would normally skip past criticizing a QB selection, since my normal methods of critiquing measurable physical traits don't work as well at this position.  Still, I think it is safe to say that selecting a QB in the 1st round, when they have a career college completion percentage of 47.8%, is absolutely hilarious.  In his best season at Cal, Boller produced a whopping 53.4% completion percentage.  Given a long enough stretch of road, and a steady and consistent acceleration to his performance, I suppose the Ravens were envisioning him hitting the 60% mark somewhere around his third contract.  The most interesting aspect of Boller's failure, was that he still came closer to meeting his PETARD projection for games started, than his performance probably merited.  This just goes to show how remarkably stubborn teams are about continuing to put their 1st round picks out on the field, regardless of how they are actually performing.  Take a moment and think of the number of 1st round QBs that you have seen in the past decade that seemed utterly doomed to failure/mediocrity.  Did their team bench them?  Did their team make them actually earn the starting job?  Hmm, probably not.

Adam Terry, OT, 64th pick in 2005   Actual # of GS:  18   PETARD Proj. GS: 24.56
Ozzie Newsome has a fairly obvious infatuation with exceptionally tall offensive tackles.  I suspect he thinks he is going to stumble across another Jonathan Ogden, by taking this approach.  Unfortunately, the 6'8", 330 pound Adam Terry was no Jonathan Ogden.  Despite his height, his arms measured a laughable 32.25" in length, which became even worse when we adjusted this measurement in relation to his height.  His Kangaroo Score of 0.009, suggests merely average lower body power, meaning that his exceptional size/bulk was mostly an illusion of power.  His Agility Score of -0.324, while not horrific, was slightly below average.  Even when we look at other traits that I place somewhat less value on for offensive linemen, like his 40 time of 5.40 seconds, and 10-yards split of 1.85 seconds, we continue to see a pattern that shows no reason to believe that Terry was going to be able to compete at a higher competitive level, against better athletes than he faced at Syracuse.  His eventual failure, appears to have been rather predictable.

Let's wrap this all up...

What sort of picture does all of this present to us?  Well, I feel like I should be very careful in how I go about saying this.  There's probably a good chance that what I'm trying to suggest here could rather easily get misinterpreted.  Despite the fact that I have listed these players in order of how many games started that they feel short of PETARD's expectations, I don't necessarily want this to be seen as an evaluation of the degree to which a player was unable to perform.  This isn't meant to be a Suck-O-Meter, even if PETARD sometimes works fairly well in that capacity.  Since there is no universally accepted and simple measure to evaluate a player's ability, I just view this list as a way to gauge a team's tacit admission that perhaps someone fell short of management's expectations.  The promptness with which a team is willing to give up on a player is somewhat erratic and unreliable.  Basically, we're just looking for players who appeared to under-perform, and are less concerned by the degree to which they may have done so.

Still, since this is just a list of the most outrageously under-performing players, at least in terms of games started, I think certain trends do appear to be quite common, and could explain why these individuals failed to set the world on fire with their play.  While I may be missing bits of data on some of these players, they're physical and statistical traits almost invariably fall outside of the range for what I would consider to be exceptional or even draftable prospects.  Musa Smith may be the one possible exception to that statement.  This may all sound like a "hindisght is 20/20" type of argument, and the numbers I use to make these judgments may seem a bit odd to some people, but I think my criticisms here don't vary too much from how I evaluated most of the 2014 draft prospects.  If nothing else, I'm fairly consistent in my stupidity.  If you are hoping that a player will contribute, especially at a high level, it's probably beneficial to look towards prospects that have some sort of distinguishable signs of excellence.  That's not something that seems to show up very much, if at all, with the players on this list.

None of this is meant to suggest that players with weaker measurable traits, or poor statistical production in college, can't turn out to be quite good.  There are always going to be a handful of players each year that seem to come out of nowhere and surprise the hell out of me.  I just tend to suspect that these aberrations occur a bit less frequently than we are sometimes led to believe.  I don't like to think of the draft as a crap-shoot, but it is probably a game where playing the odds is a factor.  When GMs throw the odds out the window, and just play the game based on what their gut is telling them...well...they might have a gambling problem.

Tuesday, July 15, 2014

What is a petard anyway?

It seems that I've always had an incorrect understanding of what this word really means.  As it turns out, a petard is basically a bomb made for blowing up gates or doorways.  A soldier would approach a fortification with their petard in hand, light the fuse, and then run like hell.  According to the mighty Wikipedia, "Petard comes from the Middle French peter, to break wind, from pet expulsion of intestinal gas".  How awesome is that?  We basically have a fancy way of referring to a fart bomb.

Unfortunately, things didn't always go according to plan, and we are probably most familiar with this word from Shakespeare's Hamlet, where he referred to the possibility of an individual being "hoisted with his own petard", or launched into the air by a bomb of one's own making, at least in a figurative sense.  Sometimes things just don't go the way you expected them to, and the results can be a bit unpleasant.

With that sort of unexpected result in mind, I made a chart that we named PETARD.  This was intended to show what the Ravens might expect from one of their draft picks, when it comes to the percentage of games started in their first four years in the NFL, depending on when the player was selected.  While individual results will vary, based on numerous factors such as actual ability (though this probably matters less than you would hope), I just wanted to get a sense of what the average result for a player might be.  With this, we can get some rough idea as to what the Ravens' management is expecting from a player, and how close a player came to meeting these expectations.

While many people may try to grade a team's draft class, based on their own subjective opinions, the degree to which we should pay attention to any of this jibber-jabber is highly debatable.  Like many other people, I have my own views on these things, but there's no reason for anyone to take the opinion of a dilettante of NFL statistics, such as myself, too seriously.  In the end, it is really just a question of whether the players appear to be meeting the team's expectations, which I want to examine.

So, what I decided to do is to calculate the total number of games we should expect a player to start in their first four years, and find the cumulative total for each draft class, and then see how close the entire class actually came to this projected average result.  Whether the player is arguably any good, really doesn't matter to me.  If the team wants to start a player, despite that player performing horrendously, is entirely up to them.  Take Gino Gradkowski, for example.  According to PETARD, Gradkowski would have been projected to start 17.5 games in his first 4 years, based on where he was drafted.  So far, he has started 16 games, performing quite horribly according to most people's accounts.  If the Ravens choose to continue giving him starts, that is up to them, though it seems unlikely that this will happen.

If, in a given year, PETARD projects that a draft class should generate a hypothetical 140 total starts over a four year span, we are simply seeing how close that class came to meeting this expectation.  If a draft class accumulates the 140 projected games started, PETARD would give this a result of 100% efficiency.  This wouldn't necessarily mean that a draft class was good, it merely means that a particular draft class fully reached the average expectations the team might have had for it.  The total number of projected starts will obviously vary from year to year.

The number of draft picks that a team has in a given year, and how high those draft picks happen to be, really makes very little difference.  A team could just as easily have one draft pick in the 7th round, and still be judged to be quite respectable in the eyes of PETARD.  For example, if this solitary and imaginary late round prospect was chosen with the 238th pick, we would only expect this player to start 4.35% of all games within their first four years, or just 2.78 games in total.  If this imaginary player started this insignificant number of games, PETARD would still view this draft class as having met expectations, and give it a grade of 100%.  If a draft class produced more starts than expected, the grade could easily rise above the 100% mark.

Now, since the Ravens are frequently viewed as a team that does quite well in the draft, I wanted to see how they have done over the course of time in meeting what appear to be their own expectations.  Since much of their reputation seems to stem from their notable successes from the late '90s to the mid 2000s, that was the period of time I chose to target for all of this.  The chart below shows how PETARD would evaluate their performance from the years 1999-2011.  The goals isn't to judge the Ravens, but instead, to let them judge themselves.  Basically, this is sort of like sending the Ravens' management to a Montessori school.  We can just sit back, and let the team hoist themselves on their own petard.

For the most part, you can just ignore the trendlines that are on the chart, as they are just there for a later discussion, which I may or may not end up pursuing.

Before we get much further into this, there is a minor adjustment I think we should make.  When we initially came up with our calculations for PETARD, we excluded a small handful of kickers, punters and fullbacks, since the expectations are clearly quite different for people who play at these positions.  In the above chart, you will see that the Ravens' result for the 2007 draft class is a rather insane 142%, which would suggest that they performed 42% above what we would consider to be the average expectation.  Unfortunately, a lot of that is driven by the inclusion of fullback Le'Ron McClain, who was selected in the 4th round with the 137th overall pick, which is about as high as you would normally expect a fullback to ever be taken.  Normally, we would only expect a player to start about 18.6% of all games in their first four years, if selected at that position, while McClain actually started 84.375% of all games.

Since the inclusion of such a strange player, at such an arguably peculiar position,  produces rather disturbing results, I thought I would also include an alternative chart for this same time period, that excludes Le'Ron McClain from the equation.  I think this might present a more sensible picture of the situation, but you can make up your own mind about this issue.

The Ravens have had their ups and downs, just like any team, but it does seem to be quite clear that from 1999 to 2007 they almost always managed to draft players who succeeded in meeting their expectations, at least in a cumulative sense.  This doesn't mean that I fully endorse all of the decisions the team made during this period of time, but they certainly appear to have been at their best during this period.  The only major blip on their radar was their horrific performance during the 2004 draft, where their selections produced practically nothing while under their rookie contracts.  On the other hand, in 6 out of these 9 years their draft classes fully met expectations, and frequently exceed them by a rather good margin.  The average result during this period would have been about 103.5%.

Sadly, once we start looking at the years after 2007, things seem to have hit a bump in the road.  From that point forward, the team seems to have consistently fallen a bit short of the mark, with their best results only managing to measure up with what might have been viewed as a rather disappointing result in their better days, with an average outcome of around 84.8%, during the 2008-2011 period.

Technically, I really shouldn't be including the 2011 draft class in any of this, as they haven't completed their fourth year yet.  Still, it's fairly simple after three years to project things forward a bit, and see that they too will probably fall a bit short of expectations.  While their result for now is projected to be about 90%, I actually suspect it might end up being perhaps 1-2% higher.  A lot of this projected grade hinges on Jimmy Smith and Torrey Smith starting 32 combined games in the upcoming 2014 season, and injuries can obviously have a huge impact on this.  According to PETARD, the 8 players that the Ravens selected in 2011 were projected to generate 127.6 cumulative games started.  So far, after their first 3 years in the league, they have produced 83 games started.  With just one year remaining, that leaves them 44.6 games below expectation.  Now, if Jimmy Smith and Torrey Smith both start all 16 games this season, that reduces the 2011 draft class' debt to 12.6 games that would still need to be picked up, or 2.1 games started per player, for the six other prospects that were taken in 2011.

That would seem like a fairly easy goal to achieve, but it probably isn't.  Tandon Doss and Anthony Allen aren't even on the team anymore, and seem unlikely to start for their current teams (which Ozzie would still get credit for).  Tyrod Taylor, as a backup quarterback, also would appear unlikely to see any action.  This increases the debt that each of the 3 remaining players would have to make up for from 2.1 games started/player, to 4.2 games started/player.  Considering that this debt falls on the shoulders of Jah Reid, Chykie Brown and Pernell McPhee, I think it is safe to say that the goal will most likely not be reached.

When we consider the 2012 draft class, things don't seem to improve very much.  So far, the bulk of the heavy lifting being done by the 2012 class is coming from Courtney Upshaw and Kelechi Osemele.  While Upshaw has started 68.75% of his games so far, his shortcomings seem to have been a likely factor in the team's decision to sign the veteran pass rusher Elvis Dumervil to split snaps with him.  It's not exactly a ringing endorsement of your top pick when a team is making those sorts of moves after a player has been in the league for just one year.  Still, it's way too early to assign a grade to this class.  Nonetheless, I have some concerns about this group.

I realize that some people won't like the idea of judging a draft class based on how many "starts" it produces relative to a weird idea of "expectation".  All I can really say is that I prefer to keep things fairly simple.  At some point, a team has to find/produce starting caliber players, or else they are likely to run into problems down the road.  There's plenty of room to debate certain aspects of this.  Sometimes a player manages to emerge a bit later in their career, even if they weren't a starter in their first few seasons.  The problem with that is a player's long term contributions probably shouldn't be a deciding factor in evaluating a draft pick.  One of the primary values of a draft pick is the relative affordability of their rookie contracts.  The draft isn't really about acquiring talent.  It is about acquiring cheap talent.  If they're not getting on the field fairly early, a team really isn't getting much value out of their investment.  It also means that the team will have to compensate for this by signing/starting veteran players, who are more likely to be expensive, while also limiting their ability to evaluate whether the youngsters have anything to offer.

People might also point to the Ravens' fairly successful run during the past 5 seasons, where they have frequently managed to make it to the playoffs.  Of course we could argue that a lot of the foundation of those successful teams was probably built during the drafts of the late '90s and early 2000s, with selection like Ray Lewis, Ed Reed, Jonathan Ogden, Marshal Yanda, Haloti Ngata, Terrell Suggs, etc.  As these players get older and retire (which some have already done), does it appear that the team has really found replacements of a similar caliber?  Or, will the team start to decline, as the draftees of the past 7 years are forced to become the new faces of the franchise?  I may end up posting something a bit later that relates to this, but much like the popular opinion you hear on this subject, I do think it generally takes about 3-4 years for the full effects of a draft to be felt, whether they are positive or negative.  If the Ravens are going to go into a decline, I wouldn't be surprised if the consequences of recent drafts only start to show up right about now.  Perhaps this is already occurring based on their relatively poor performance in 2013.

There is one other factor in all of this that I can't dismiss.  There is always the possibility, no matter how unlikely I personally find it to be, that a team can simply shift its philosophy towards giving playing time to young players, which could affect their PETARD results.  So, what happened in 2008, that might explain the team's apparent decline in draft efficiency?  The obvious answer might be the hiring of John Harbaugh, who happened to arrive in 2008, right when PETARD seems to suggest things really went downhill.  Whether Harbaugh is hesitant to start young players, or whether the young players have failed to meet his expectations, is a difficult question to answer.  Either way, the result is the same.  It all ends up amounting to a reduced lack of value coming out of the team's draft picks.

While none of this can truly say how the Ravens have performed in comparison to other teams, I do think it points to the possibility that whatever magic or luck guided the team in the past may have begun to fizzle.  Perhaps, much like the way that fans refer to Ozzie Newsome as The Wizard of Oz, he is really just a mere mortal hiding behind a curtain, madly pulling levers in a desperate attempt to keep the citizens of the Emerald City content.  If I only I had a brain, I might be able to answer these questions a bit better, but for now, I remain pessimistic about the team's immediate future.

Tuesday, July 1, 2014

Let's Make Broad Generalizations!

This is only going to cause me trouble.

I'm about to make what is probably a huge mistake, by diving into a subject that I find interesting, but which I suspect is only going to cause me a fair bit of annoyance.  Fortunately, I can be a bit of an idiot, and am not very good at avoiding foolish behavior.  So, here's the question that's currently rattling around in my head.

What should we realistically expect out of a draft pick?  Or, to what degree does being picked highly influence the path of a player's career, for better or for worse?

I realize that a number of people have tinkered around with this sort of question.  A lot of the attention nowadays seems to be directed towards running regressions that compare CarAV(or Approximate Value) to where a player was selected in the draft.  The results of some of these examinations can be interesting....but I just can't fully jump aboard the CarAV train.  I'm not saying that CarAV is a bad statistic, it's just that I have some qualms about committing to it at this point in time.  As these sorts of geeky stats go, I just think it might be trying to do a bit too much.

I think my main problem with CarAV is that many people seem to be using it as a measure of a player's performance or contribution to a team.  Some people seem to be using CarAV as a qualitative measure, when I think it is probably more of a quantitative statistic.  It does a fine job of saying whether a player is getting on the field, but suspect it doesn't say much about what someone does once they are allowed to play.  I'm just not sure we are at the point where we have a "one ring to rule them all" statistic, to easily sum up such a potentially complex question as to whether a player is "good" or not.  Instead, I think CarAV might be more of a 'perceived performance' or 'perceived contribution' estimate, which is a very different sort of idea.  It might tell us who the coaches have some confidence in, but not whether the coaches are right to feel this way.

A player who is on a particularly effective unit (offensive line, defensive line or whatever) probably has a clear advantage in building up their CarAV results.  It tends to be a very end result oriented statistic.  If your team/unit does well, you get graded well, even if you might have been the weakest link in the chain.  Getting selected for Pro Bowls and All Pro teams also has a significant effect on CarAV, and those selections can often be popularity contests that might favor high draft picks over a comparable player who was drafted later.  The single biggest factor in CarAV is probably a player's ability to simply get on the field in the first place, where high draft picks will always have the edge, almost regardless of their actual performance.  I just feel that CarAV might be weighted with some confirmation bias regarding high draft picks.

Over the first 6 years of his career, Evan Mathis' average annual Approximate Value result was a 2.  That's not a very good result, in the CarAV world.  Opportunity clearly plays a significant role in all of this, as he (a 3rd round pick) was only asked to start in 22 out of a potential 96 (22.91%) games during this time period.  In the last 3 years, when he was finally made a regular starter, his average annual Approximate Value was 8.66.  On the other hand, the generally mediocre Michael Oher, who was selected in the 1st round, had an annual average Approximate Value result of 8 in his first five seasons, and his results never dipped below 7 in any year.  Of course, Oher started every single one of the 80 games from the day he was drafted, despite wavering between an average to below average level of play.

Now, some people might argue that perhaps Evan Mathis improved over time, and I'm sure that his coaches would love to take credit for this improvement, but I just can't buy into that idea.  I think there were some fairly solid reasons to suspect that Mathis was always the more gifted player, as you can see in the post on the Lobotomy Line.  Does anybody believe that Michael Oher is a comparable/superior talent to Evan Mathis?  Or, is it just more likely that the challenge of exceeding/altering people's expectations of a player is a more difficult proposal when a team has invested less in a particular player in the first place?  Hope is a flame that constantly burns bright when it comes to high draft picks.  Hell, there are people still waiting for a Tim Tebow resurgence.

Now, I realize that this is just one peculiar example of a possible shortcoming in CarAV, and criticizing CarAV really isn't my goal.  CarAV has its uses.  All I'm trying to say is that there are some issues which cause me to have concerns about whether the best players are consistently being put on the field.  I have little doubt that the majority of high draft picks are reasonably talented, but assuming that the talent is there simply because they are high draft picks is a very different question.

Instead of looking at 'perceived performance' or hype, relative to where a player is selected in the draft, Reilly proposed that we ignore the talent/quality of performance issue altogether.  He was just curious about the degree to which a player's draft position relates to a team's willingness to give someone a starting role on the team.  More specifically, he suggested that we chart the percentage of potential games started during a player's first four years in the league.  A four year period was chosen since, with the exception of some 1st round draft picks, that is the typical length of a rookie contract, which is part of the real issue we're eventually going to try to figure out.  So, without further jibber-jabber, this is basically what the results look like.

We're going logarithmic, like an ALPS Blue Velvet potentiometer!

This chart is based on the Ravens' draft picks from 1996-2010.  Beyond being my home team, and the organization I am most familiar with, they really are a great team to run these experiments on.  Having one GM, Ozzie Newsome,  who has overseen such an extensive period of time running the team, eliminates some of the fickle fluctuations that might come with a more volatile organization.  For better or worse, their behavior should be relatively consistent, at least compared to teams with more turnover.

During this period of time 118 players were selected by the Ravens, though we did remove five of them from the discussion.  Since we are only doing this to look at the team's tendencies, behavior and biases, Reilly removed Sergio Kindle and Dan Cody from the list, because they were relatively high draft picks who never played due to injuries that were factors before a single game had even been played.  Remember, we're not trying to judge the team's performance or luck in the draft (not yet, but we will later), just their tendency to give "starts" to higher draft picks.  If a player was never healthy enough to play, the team never gets to make much of a decision in those cases.  We also removed any kickers, punters (Dave Zastudil)or fullbacks (Le'Ron McClain and Ovie Mughelli) who were selected before the 5th round.  These positions generally don't get credited with "starts" in the first place, so they just create a bit of chaos if they are included.  In reality, only Le'Ron McClain would have had a significant effect if he had been allowed to remain in this list, at least compared to Zastudil and Mughelli.  We also obviously can't include results from more recent drafts, because the four year time period for those players hasn't run out yet, though they will eventually be included.

The data could be tightened up and manipulated a little to produce a better R^2 value, but really, 0.5697 is rather adequate for our humble purposes.  I prefer simplicity over tidiness and perfection, since a better data fit could spark criticisms of nudging the results too much.  We're not trying to predict what percentage of potential games started that a player will have in their first 4 seasons.  Instead, we just want to figure out what an average result might be, relative to where a player was selected.  Why we're curious about this, is something we'll get into at a later point in time.  For now, we're just going to refer to the data from this chart as PETARD (because it makes me giggle like a little girl).  Eventually, we'll have to turn this into an amusing acronym to justify naming it this.  It should also be noted that once we get above the 5th pick in the first round, it becomes impossible to meet or exceed expectations, so we actually capped everybody at an average expectation of 100% of games started in their first four years, regardless of how insane that may be.

Like I said though, this isn't really about judging whether a player is good or not.  There are clearly numerous data points, particularly for mid-round players, that wildly diverge from the trendline, so individual outcomes can be fairly unpredictable.  Despite that, it seems reasonable to suspect that players who exceeded these average expectation for where they were selected were possibly doing something right, and vice versa, for players who failed to meet expectations.  You could say that we are simply looking to identify who the outliers are.  In the future we will explore the stories behind these players who produced surprising and unexpected results, in an attempt to better understand them.   I fully realize that many people aren't going to like the idea of using "percentage of potential games started" as a metric for making judgments (we're still working on the Moxie-Meter), can't please everyone.  At the end of the day, I do think being a "starter" is more valuable than not being a starter, so we're not going to over-think this.  This is just meant to be a simple and reasonably objective way of gauging things, and some shortcomings are naturally going to exist in this process.

Really, there's probably not much in this post that should matter to anybody, but some of the issues I've been considering will refer back to this.  It was just easier to throw this out there in advance, rather than trying to fit it in later.  I think the more interesting discussion will come somewhere further down the road.