Wednesday, May 29, 2013

Purdue, home of the QB killer

When I go through my lists of pass rushers, looking for the next potential gem, one peculiarity keeps popping up.  Purdue, despite being a rather mediocre college program, produces a disproportionate number of good pass rushers.  Not Miami.  Not USC.  Not Ohio State.  Purdue.  It makes no sense to me.  They don't really produce many great NFL players at other positions, but this seems to be something they do rather well.  They even have the Den of Defensive Ends listed on their team website, pointing to their primary peculiar accomplishment.

I've already shown the basics of my approach for finding pass rushers here, so I thought I would lay out a list of the rather strong collection of Boilermakers, that keep their school on the national radar.  Some are going to be bigger stars than others, and some might just be situational pass rushers, but they are all interesting to some degree.  These are just the players from 1999 to the present.

Player                      Year           Kangaroo Score        Agility Score        Total        Avg. TFL

Rosevelt Colvin 1999 0.545 0.754 0.556             N/A
Chike Okeafor 1999 0.392 0.707 0.306             N/A
Akin Ayodele  2002 0.496 1.102 0.672             N/A
Shaun Phillips 2004 -0.290 0.893 -0.005 15.75
Rob Ninkovich 2006 0.267 1.013 0.446 13.25
Ray Edwards 2006 1.421 -1.566 0.509 11
Anthony Spencer 2007 -0.215 -0.025 -0.011 17
Cliff Avril 2008 0.287 0.215 0.301 15
Ryan Kerrigan 2011 0.805 0.007 0.632 22.25

As I've said before, I'm usually looking for explosive players (measured by the Kangaroo Score), hopefully with a measure of agility added to the mix.  The Avg. TFL number, is the player's average number of tackles for a loss in their last two years in college.  If a player has a total score over 0.500, and is averaging over 15 TFL, then I consider them a high priority draft prospect.

If there ever was a school that repeatedly produced outliers, who exceeded my expectations, it would be Purdue.  It's not that their players are bad athletically.  It's just that their athletic measurements tend to be in the good to somewhat mundane category.  A few, such as Shaun Phillips or Anthony Spencer, are even well below average, though Phillips was quite agile.  Only Kerrigan, Colvin, Ayodele, and Edwards would have probably caught my computer's attention.  However you want to look at it, I'm past the point of  wanting to bet against players from Purdue.  Almost none of them turn out to be bad.

I've often wondered if they had a head coach, or position coach, who had a particular type of player he recruited.  Or perhaps, they have someone who really does a great job of coaching technique to the players.  Generally, I feel that coaches do very little to change or improve players, but whatever is going on in West Lafayette is intriguing.  The weirdest part about it is that I know they switched head coaches in 2012 and 2009, so it is difficult to attribute it all to one man (though they've only produced one pass rusher, Kerrigan, since the last switch).  Joe Tiller seems like the most obvious person to look to, but he is is 70 now, and probably not looking to dip his toes into the NFL waters at this point.  I kind of like the idea that their is a senior citizen 'meat-head whisperer' out there.  If anyone has any theories, I'd be glad to hear them. 

Maybe it's all just a coincidence.  The sample size still isn't exactly huge.  It only looks peculiar when you compare the results to other top schools that might only produce 2-4 decent pass rushers in the same time span.  All I know is, when draft season rolls around, and I see a defensive end/outside linebacker prospect from Purdue listed, I pay attention, even if the measurements seem a bit off.  Statistics be damned, they're putting something in the water in Indiana.

Sunday, May 26, 2013

The Stat Score: A Rat's Nest of Numbers

Since I refer to the Stat Score rather frequently when it comes to wide receivers, I thought I should lay out how it all works.  Unfortunately, this is something that starts off very simple, and proceeds to get weirder and more questionable as I try to untangle the mess of numbers that I use.  To keep things as clear as possible, I'll try to explain it all in steps, before I get to some of the more deranged issues.

On its most basic level, it revolves around one core idea.  How much of a team's offense was a receiver responsible for?  Well, that is simple to figure out.  If a college team gained 5000 yards, and a receiver in that offense gained 1000 yards, then he was involved in/responsible for 20% of the total yards.  So, why should we care about this percentage?  The reason for this is that raw yardage doesn't tell us very much about how valuable a player was to their team.  For instance, look at this example:

Team's Total Offense                Receiver's Yardage               % of Offense
6000                                             1200                                    20%
5000                                             1000                                    20%
4000                                               900                                   22.5%

As you can see, as the team's total output goes up, the receiver has to produce even more to be as proportionally valuable.  While the receiver who generated only 900 yards initially appears less impressive, he was possibly  more valuable to his team's offense than the player with 1200 yards.  It's easy to get excited about a draft prospect who has gaudy stats, but sometimes we lose track of the context in which the stats are generated.  A player who is generating a larger percentage of his team's offense is probably going to be the focus of the opposing defense.  So, generating significant stats, while being the main focus of your team is more challenging.

As I've mentioned before, this is very similar to Shawn Siegele's Dominator Rating system.  The only real difference is that I base the percentage off of the total offense, and Shawn does it based off of the passing offense.  In the end, it is sort of like the difference between  Dr. Pepper versus Mr. Pibb.  It's debatable which approach is better, and maybe I am wrong to do it this way, but I am too lazy to change my database now.  I think that a team's ability to run the ball should be a part of the score, as it alleviates some of the pressure on the  receivers, so I take that into account.

Beyond this difference of opinion, relating to the running game, there is one other issue.  Instead of giving a score for the player's last year in college, I use their last 2 years.  If one year of production is good, two is better.  This is just my way of weeding out one year wonders.  The average percentage of the offense that a draft prospect is responsible for in his last college year is 17.75%.  In their next to last year, the average is 15.34%.  The actual averages would really be lower, if I considered all receivers, and not just the ones who had historically gotten drafted.

After I have these scores I convert them, to see how many standard.deviations the result is away from the average, because being able to make bell curves is neat, if perhaps somewhat pointless.  You could really just stick with the basic percentage and still be perfectly fine, but I like to complicate things for myself.  One standard deviation for their final year would be 6.581%, and for their next to last year 6.674%.  So, the math works like this for their final year (next to last year is the same, but obviously putting in a different set of numbers, and the correct corresponding standard deviation):

(Player's % of offense as senior - 17.75)/6.581

The resulting score lets you calculate what percentile the player would be in, or create a bell curve to see how common their results are.  Generally, though distributions can vary from position to position, or subject to subject, one positive standard deviation will place a player near the 70th percentile, and one negative deviation around the 30th percentile.

Now things start to get weird.  Just for the sake of being able to easily peruse a player's numbers I will combine some of their scores to create a sort of overall score.  It's generally a bad idea to rely too much on these overall scores, especially when they start to combine large numbers of smaller scores, as they can hide weird imbalances or irregularities.  It's kind of like judging someone based on their college GPA, instead of looking at their grades in individual classes.  The grades in individual classes will give you a better picture of what is going on.  Maybe someone was average in 6 out of eight classes, excelled in one, and bombed another.   If that one class he bombed was something important (Nuclear Physics for Dummies), and the one he excelled in was something stupid (Basket Weaving 101), that is worth making a note of.  Combining scores into one larger score, just lets me get a quick glimpse of a player, though it can be less accurate sometimes.  I'll still refer to them in my posts for the sake of simplicity, but I'll also try to point out areas of concern when they arise.

Here's one issue/observation that I should mention.  It seems to me that smaller players tend to perform at a higher level in college, at a younger age, while the bigger guys tend to really hit their stride around their junior or senior year.  If you see a player with multiple 1000 yard seasons, he is more often going to be built to a more average size.  This is just mad speculation on my part, but I have to wonder if this has something to do with the bigger guys needing more time to grow into their frame, or reach their physical peak, compared to a 5'10" and 190# player, who might have finished growing by his senior year of high school.  Like I said, mad speculation.  Either way, I combine their 2 years of production in different ways because of this.  For smaller guys (under 210#), I value their final 2 years evenly, in a 50/50 split.  For the bigger guys (over 200#) I value their 2 years with an emphasis on their final year, with a 70/30 split.  This is all a highly debatable and possibly stupid thing to do, but for now, that's what I'm doing.

The last thing I do is to look at their average yards per catch over both years (though I just average their two years here, and don't worry about the two separate averages), and their number of receptions in both years.  This is also highly questionable, but I like having the data.  I'll list here the average result for the 2 years and the standard deviation.  Calculating how many standard deviations away from the average a result is can be done the same as I mentioned earlier.

Yards Per Catch                          Average                       Standard Deviation
Combined 2 year  Avg.                    14.966                                   2.645                         

                                               Avg. # of Receptions        Standard Deviation
Final Year                                        62.06                                      23.6
Next to Last Year                             51.75                                      22.68

I just like to have these scores to get an idea of how big an impact a player has on a per play basis, and what their total volume of plays looks like.  To some extent, a player who does well in YPC will tend to have a lower volume of receptions, and vice versa.  So, a player is unlikely to score well in both.  Extremely explosive deep threats generally have fewer receptions, and reliable high volume underneath guys tend to have a lower YPC.  It's just something I like to look for, but it isn't terribly important, so I usually only let it count for about 10-20% of a player's total Stat Score, depending on their size.  Since the two scores tend to knock each other out, they usually have very little overall impact.  This is just something I sprinkle in there as a garnish.

In the end, I wouldn't get too carried away with trying to divine mystical meanings from these scores.  A lot of stat geeks are trying to find the 'One Stat To Rule Them All', and I'm not convinced that such a thing exists.  I'll use stats that sometimes might appear to be trying to reach this goal, but I don't take them too seriously.  I'm just trying to find an acceptable standard for making comparisons between prospects.  I'm not trying to make any claims of having found the one true answer.  The stats are useful for drawing your eye towards the exceptional versus the mundane, but beyond that you are still going to be better informed when you look at the broader set of data that goes into these stats..  This is an ESPN world though, and people seem to want overly simple answers to relatively complex questions.  With that said, even a method such as this, that is at the very least questionable, one can get a better sense of a player than just raw stats from college.

As time moves on, I will undoubtedly make alterations to my views on certain stats, and will update past numbers, and explain when/why I'm making such changes.  Or, at some point, my eyes will fall out from staring at spreadsheets for too long.

Friday, May 24, 2013

Richard Sherman, and our perception of talent over time

I think we tend to have a rather short term memory when it comes to the players we classify as 'stars'.  Whether we are talking about a highly touted draft prospect, or our team's newest big ticket free agent acquisition, it is easy to buy into the hype.  Despite our general awareness that few of these acquisitions actually pan out, an appreciation for consistency gets downplayed, in comparison to hot streaks. Highlights trump reliability, until the highlights start appearing less frequently.

At one point, you saw the Eagles paying Nnamdi Asomugha $15 million per year.  Then, two years later, he is cut and playing for a different team for $1.7 million.  Did he deserve the initial contract, or the latter one?  Was he the best cornerback in the league, or an overrated bum?  Did your opinion of him sway, or remain steady throughout?

What about Brandon Lloyd?  He produced mediocre to poor results for seven years, then suddenly became relevant in 2010.  Suddenly people were taking him seriously.  People seemed to shift from doubtful, to enthusiastic.  Then they shifted back to doubtful again, in 2013, as he was cut by his sixth team in ten years.  Again, did your opinion ever sway, perhaps more than once?

In some cases we can blame injuries or a "bad fit" team-wise, for these fluctuations.  Too often, I suspect we use these excuses to cover up bad player evaluations.  The players are what they are, with whatever limitations they were born with.  It's probably our eagerness or desperation to fill a void on a team,  that lets us see something that really isn't there for more than brief flashes.

Knowing where a player fits in the bell curve of the physically gifted, also has some application after the draft.  When a player starts shooting into the spotlight despite measurables that suggest some physical limitations, it seems reasonable to wonder if this stardom is sustainable.  Is it all just a momentary blip on the radar, that is likely to get a player overpaid, and lead to possible ruination for his team?  Let's take a look at some interesting examples.  As always, scores are given in terms of the number of standard deviations that a player is away from the average results for players in his position group.

Player                                    Ht/Spd Score         Agility Score

Richard Sherman 0.882 -0.334
Champ Bailey 1.272 2.321
Darrelle Revis 0.656 1.176
Antoine Winfield -0.485 0.920
Leon Hall 0.535 1.361
Charles Tillman 0.636 0.599
Patrick Peterson 1.322 1.160
Brandon Flowers -0.870 0.775
Lardarius Webb 0.261 0.581
Devin McCourty 0.436 0.859
Terence Newman 0.338 1.231
Antonio Cromartie 1.082 0.674

I don't think discussions of "shutdown corners" have much value.  Cornerbacks tend to move in and out of the spotlight, the names changing all the time.  For the most part, their periods of celebrated high level play can be traced to the surrounding talent on their defenses.  This is why I don't place a lot of value on corners, in comparison to pass rushers, who I think dictate the outcome of situations more than corners.  Still, a very small collection of corners seem to stick around and remain relevant, regardless of their situation.  Recently, Champ Bailey and Darrelle Revis, would appear to fit in this group, and others may join them in time (of course Bailey is at the end of the road). 

The truly rare agility that most of the top corners have in common seems unlikely to be a coincidence.  Often, their agility score is one full standard deviation above the average for their position group.  It is easy to envision where this sort of agility would come into play for them.  Having excellent speed relative to their height, also seems to be a valuable trait, just as our intuition would have suggested.  So, based on this, what do we make of a player like Richard Sherman, who we are consistently told is one of the top cornerbacks in the game today?

I don't get to see Sherman play very frequently, living on the opposite coast, so my opinions are obviously somewhat suspect.  Still, I have to wonder about the likelihood that he will remain a dominant corner while possessing a rather below average agility score.  From what I can gather, his game is heavily reliant on physical play and jamming receivers at the line, which is a sensible tactic considering his large frame.  Nnamdi Asomugha seemed to benefit from a similar approach before going to Philly, where he appeared to struggle with a different style of defense.  Interestingly, perhaps, Asomugha didn't take part in the short shuttle drill or 3-cone drill, which the agility score is based on.  This normally makes me suspicious that a player is trying to hide an area where they know they might be weaker.  So, is Sherman likely to be similarly limited?  The numbers suggest he might be. Maybe our perception of Sherman will change.  If we were to just compare him to notable cornerbacks that had weaker agility scores, the list would look something like this:

Player                        40 time        Sh. Shuttle Score     3-Cone Score    Total Agility Score
Johnathan Joseph4.31-0.717-0.139-0.428
Tim Jennings4.32-1.198                         N/A            N/A
Brandon Browner4.63-0.511-1.444-0.977
Asante Samuel4.490.174-0.189-0.007
Tramon Williams4.57-0.238-0.189-0.213
Brent Grimes4.57-0.580-1.092-0.836
Mike Jenkins4.38-1.609-1.494-1.552
Cortland Finnegan4.34-1.198-0.189-0.693
Marcus Trufant4.39-1.0600.211-0.424
Rashean Mathis4.45-0.0310.892-0.461
Dre Bly4.510.517-0.942-0.212
Richard Sherman4.54-1.1290.462-0.333
Eric Wright4.36-0.443-0.089-0.266

Yes, there are some noteworthy names on that list.  Many of them were the also the flavor of the month, at one time or another.  Remember when people were talking about Tramon Williams all the time?  Do we elevate Tim Jennings after his 9 interceptions in 2012, or focus on the 6 mediocre years that preceded this explosion?  When we break open the agility score this way, we see that Sherman's 3-Cone Score is actually quite respectable, and that it is his Short Shuttle that is dragging things down.  The most similar players in this respect seem to be Marcus Trufant, Mike Jenkins and Cortland Finnegan.  Generally, I think of the short shuttle score as something that relates to a cornerback's ability to transfer out of his backpedal.  The cornerback prospects that you hear described as having "stiff hips" tend to do poorly on the sort shuttle.  Maybe someone else has a different view on this.  If this is the case, then it would be at this point in the receiver's route that these cornerbacks should be most vulnerable to giving up room to a nimble opponent.  Maybe this isn't an issue for Sherman, or maybe by jamming the receivers, they simply can't blow past him, forcing him to make such a transition very often.

We can also see that out of these 13 somewhat less nimble cornerbacks, 6 of them were running the forty yard dash in the 4.3 range.  That seems like a noteworthy level of speed, and perhaps something which gave them the ability to excel at times, while being weaker on other occasions.  While Tramon Williams, Asante Samuel, and Rashean Mathis ran more pedestrian 40 times, their agility scores also came out much closer to the average result for a cornerback.  Combining the exceptionally speedy guys, and the passably agile, 2/3 of these 13 corners demonstrated at least some measure of physical adequacy.  So, even amongst the seemingly below average, in a rather unscientifically assembled list, a significant portion are showing some degree of promising potential.  While focusing on the speed issue, let's look at how Sherman, and his teammate Brandon Browner, measure up against the average cornerbacks' results, as well as the results for 24 Pro Bowl and All Pro cornerbacks.

Player                                          40 yard                  10 yard             2nd Gear Score
Richard Sherman 4.54 1.56 0.02
Brandon Browner 4.63 1.67 0.04
Cornerback Avg. 4.46 1.53 0.06
Pro Bowl/All Pro Avg. 4.41 1.53 0.13
Not only do they both come in significantly below below the average for Pro Bowl and All Pro corners, but they are below the average for all cornerbacks.  The 2nd Gear Score even makes it difficult to dismiss these 40 times by blaming a bad start to their run, as they seem to be showing only borderline acceleration from the 10 yard mark to the finish.

Like Sherman, Browner is also a rather large cornerback.  While Sherman was taken in the 5th round, Browner went undrafted.  Already this seems unusual for a team to have found two successful corners so late in the draft.  While Sherman ran a 4.54 forty yard dash (below average, but acceptable for someone his size), Browner ran a 4.63 (well below average).  Peculiar irregularities seem to be building upon themselves.  Browner's Ht/Spd Score of 0.657 seems good, but is probably unreliable in this case.  Obviously, the Ht/Spd score becomes a bit skewed when you are looking at a 6'4", 221 pound cornerback.  Sherman's mediocre athletic measurements actually seem spectacular compared to Browner's, yet even Browner managed to become a Pro Bowler   Regardless, it does seem odd to think that this one particular team is having such success drafting corners so late, that conform so poorly to the traditional mold.  This begs the question, are the players truly great, or is the coach/team employing these players in some crafty way that disguises their physical shortcomings?

None of this  is to say that Richard Sherman can't continue to do well in Seattle.  I just mean to suggest that his skills may not be as transferable or universally valuable as Revis' abilities are (assuming Revis recovers from his ACL injury).  I also have to wonder if Sherman's apparent reliance on a more physical style of play can remain as relevant, as officials get more eager to throw penalty flags.  For now, it's only the results that matter, and Sherman seems to be doing a good job.  If he was a free agent though, I would be hesitant to pay the kind of money he will probably be seeking.  Considering that his price tag will probably be in the $10-12 mil./year range, a team could be taking an enormous risk for a player who might be uniquely suited to a very limited number of teams.

Undoubtedly, Sherman will go on to appear in multiple Pro Bowls, and I will look like a moron for continuing to have doubts about him.  He really isn't a terrible athlete, just mediocre in more areas than I would typically expect to see.  If it weren't for his being paired with the similarly enigmatic Browner, I would probably overlook it.  The two of them combined, however, confuse the hell out of me.

Tuesday, May 21, 2013

Ryan Spadola: The Lottery Ticket

I just wanted to throw up a quick post about Ryan Spadola, since the computer likes him so much.  Spadola is an undrafted WR from Lehigh Univ. who was picked up by the Jets, and though I've never seen him play his numbers are interesting enough that I will keep an eye on him.  Here's what his college stats looked like, leaving out his freshman year where he only caught one pass.

Year        Games Played       Rec.     Yards      TDs        YPC         % of offense
2012                 9                  57         851         4           14.9                18.83
2011                 12                96         1614       11          16.8                26.43
2010                 13                78         1130        9           14.5                25.16

While his raw stats are fairly impressive, the deeper numbers are more interesting.  The average draft prospect, at the receiver position, is only responsible for about 17.78% of his teams offense in his last college year, and 15.32% in his next to last year.  Spadola significantly exceeds those numbers, even in his 2012 season where he missed a couple games with mononucleosis.  If he hadn't missed those games, he was on pace for 67 catches, 1005 yards, and 22.25% of his team's total offense in his senior year.  Either way, he significantly surpassed the average players' results for three years in a row.  When a player is responsible for that much of a team's offense, you have to assume that the opposition is going to direct a fair amount of attention towards stopping them.  Spadola seems to have continued to produce, despite being the clear focus of his team's offense. 

Players who repeatedly represent this much of their team's offense are somewhat unusual.  In Spadola's case, the fact that he played at a rather low level of competition is going to be a concern.  So, you have to wonder if he is physically gifted enough to move to the next level.  Here are his combine results:

Ht     Wt    40 time   10 yard    Bench    Vert.   B-J     Short Shut.     3-Cone
6'1"    204      4.4          1.58        15         37      9'11"         4.07              6.72

His weight of 204 pounds means that I could score him as either a 'Big' receiver or a 'Small' receiver, but I think his numbers fit better among the smaller more elusive guys.  This is how he scored in some of the some of the odd measurements that I use.

2nd Gear     Stat Score      Kangaroo Score      Agility Score     Athletic Score
0.18                0.841                  0.019                     0.774                0.384

His 2nd Gear of 0.18 suggests that he might have significantly better deep speed than his 40 time of 4.40 would lead us to believe.  His extremely average Kangaroo Score relative to his very good Agility Score would imply that he is more likely to rely on elusiveness rather than raw explosive power, which is typical of receivers in the smaller group.  Athletically he appears to be similar to someone like Andre Caldwell or Greg Jennings, which inspires confidence that he can compete at a higher level.  His Stat Score just repeats what I said above about his college production being quite a bit better than his average peer, in this case 0.841 standard deviations above average.

While I don't think a player's Total score gives as good a picture of a player as the smaller individual scores, it can be interesting sometimes to see what group the computer lumps a prospect in with.  Here are all of the small receivers (less than 210#) who had Total Scores over 0.500, and 40 times of at least 4.50 seconds.

Player                        Stat Score      Kangaroo         Agility      Athletic          Total    Yrd/GP
Kevin Curtis 2.245 -0.837 1.213 -0.208 1.018 40.7
Lee Evans 1.516 -0.604 1.153 0.194 0.855 50.9
Roddy White 1.093 1.044 0.260 0.536 0.815 68.2
Greg Jennings 1.414 -0.414 0.697 0.086 0.750 68.0
Torrey Smith 0.648 0.891 0.805 0.637 0.643 53.0
Andre Roberts 1.226 -0.480 0.610 0.053 0.640 35.9
Ryan Spadola 0.841 0.053 1.022 0.384 0.613          N/A
Victor Cruz 0.622 1.034 0.073 0.557 0.590 75.1
Emmanuel Sanders 1.144 -0.234 1.109 0.032 0.588 32.3
Justin Blackmon 1.239 0.029 -1.101 -0.071 0.584 54.1
Karsten Bailey 0.405 0.620 0.650 0.760 0.582 4.6
Deion Branch 0.996 -0.762 2.167 0.152 0.574 47.5
Mike Thomas 0.312 0.358 0.433 0.694 0.503 33.3
Golden Tate 1.166 -0.440 -0.932 -0.163 0.501 30.9

So, out of 13 prospects (not counting Spadola), nine players (69.23%) became at least average receivers, by my standard of having reached 35 yards per NFL game played.   A few more like Golden Tate and Emmanuel Sanders could also crack this list at some point , improving the odds further (though I wouldn't count on it being Tate).  Obviously this comes from a rather small sample size, and as I said, I don't like boiling things down to just one score, but it is interesting.  Those are some fairly impressive receivers for him to be lumped in with.

There really isn't much else that I can say about him since I have never had a chance to see him play.  Being blind that way, and considering his level of competition, I wouldn't have considered him as a draft pick.  As an undrafted prospect, however, he is very interesting.  Being picked up by the Jets also puts him in a decent position.  They don't have much talent at the receiver position, and are probably moving away from Santonio Holmes in the near future, so he will probably get a shot.  Again, it is very frustrating to not be able to see any of his college games, but as a lottery ticket type of prospect, I can really see the appeal.  He won't cost the Jets anything, but he does appear to have significant potential.

Monday, May 20, 2013

The 2013 Wide Receivers

This could end up being a future source of embarrassment, but I can live with that.  These are the receivers that the computer feels are the safest picks from the 2013 draft class.  As I've mentioned here, these players are filtered to include only the prospects that had at least average college production (with their stats adjusted to somewhat normalize the differences in offenses), and athletic ability, relative to their peers.  The cutoff for what we will accept as within the average range, is a score no worse than -0.1 standard deviations below average.  This is still close enough to the 50th percentile to not make any real difference.  If I didn't make such minor allowances the pool of prospects would get excessively small.

I'll try to make some comments on a few of them, and might even have future posts to discuss the ones I find most interesting.  This list shouldn't be taken as an endorsement of the players, merely that they fit the profile of a relatively safe pick.  'Big' receivers (over 200#) will also have prospects filtered out with forty yard dash times over 4.6 seconds.  'Small' receivers (under 210#) will be filtered out with forty yard dash times over 4.5 seconds.

For the most part, the Total Score should probably be not be treated seriously, as it can lead to some erroneous conclusions, but I'll leave it in anyway, since I've shown it in the past.  While they are ranked according to their Total Score, it makes more sense to look at how well their Stat Score and Athletic Scores turned out, and how balanced they are. Think of this my computer's WR shopping list.  These guys passed the initial tests to get onto my radar, then I just try to whittle the list a bit more after watching them play.

Big Receivers
Player                             Stat Score               Athletic Score            Total 
Aaron Mellette 1.943 0.457 1.052
Marcus Davis -0.040 1.675 0.989
Charles Johnson 0.652 1.206 0.985
Da'Rick Rogers 0.454 1.052 0.813
DeAndre Hopkins 0.481 0.398 0.431

Small Receivers
Player                             Stat Score              Athletic Score            Total
Ryan Spadola 0.841 0.384 0.613
Quinton Patton 0.799 0.247 0.523
Josh Boyce 0.144 0.688 0.416
Eric Rogers 0.779 -0.069 0.355
Ryan Swope 0.038 0.637 0.338
Markus Wheaton 0.732 -0.061 0.335

I've already discussed Aaron Mellette, so I can skip past his peculiar issues.  Sort of the opposite of Mellette, Marcus Davis is physically rather ideal, but his inconsistent college production makes me a bit nervous.  One other oddball on the list that I really feel I need to make an excuse for is Eric Rogers, who played at Cal Lutheran.  He barely passed the 40 yard dash threshold, coming in precisely at 4.50 seconds, as well as having an athletic score on the fringe of what is acceptable.  I really know very little about him as a player, though I suppose almost everyone else is in the same boat in that sense.  Who watches Cal Lutheran games?  So, Rogers should probably be scratched from the list.  Ryan Spadola is also a bit of a strange pick, coming from Lehigh, but I'll probably get into his story another time.  Spadola could be pretty interesting.  I still have to decide on how much of a deduction should be applied to the Stat Scores for players from Division II and III.

Overall, I would stand behind this list (with some serious reservations about Eric Rogers and Marcus Davis).  Sticking to such an approach should produce a success rate around 65%, though there will be fluctuations from year to year.  That's the way the trend has run, and I'll stick to it for now.  Some added subjective analysis might bump the success rate even higher.  This could be an odd year for the computer's prognostications, since many of these receivers weren't taken very highly, if at all.  In some ways this year reminds me of the computer's view of the 2006 draft.  That was another year where the computer felt the most highly touted prospects weren't as  safe a bet as some of the lesser known players (like Colston, Jennings, Austin and Marshall).  What happened to Chad Jackson and Sinorice Moss?  Hmm.  The computer wasn't a fan of those two.

I have more faith in some of these players, and less in others.  I just have to hope that their ability shines through, when given an opportunity.  When I toss in my own subjective opinion of the players, they get sorted somewhat differently.  The upbeat way of looking at things, is that the computer really only needs to beat the league wide median success rate of 22.5%, and I do think it should exceed that simple mark.  If NFL teams are essentially blindly pulling names out of a hat, then this approach just removes a lot of names from consideration.

I guess I should mention some of the notable exclusions.  Tavon Austin doesn't make the list, since the computer is seriously biased against players with such small frames.  There is a third category of receiver that he does fit into, which I generally don't discuss.  Instead of just the Big and Small receiver groups, these players would be in the Midget range.  This group covers guys like Johnny Knox, DeSean Jackson, T.Y. Hilton, Jacoby Ford, and a few others.  I'm just not a fan of this type of receiver, though I wouldn't deny that they occasionally do prove valuable.  Their upside is somewhat limited.  The only recent player I can think of, who proved to be a consistent high level performer, is Desean Jackson, and I still prefer the more conventional Jeremy Maclin to him   Still, if taken in the mid-to-late rounds I have no problem with these sorts of guys, as the potential payoff versus the investment makes more sense at that point than it does in the first round.

It's not that the computer couldn't target players in this Midget class; it's just a question of whether we should even try.  The traits for these players are generally easy enough to identify.  They conform to the normal Small receiver standards as far as the agility score goes, but tend to have even more blazing speed, and a truly shocking 2nd gear.  It's just their extremely low body mass and diminutive height that kills them on the Athletic Score.  There are usually one or two guys in a draft class that fit this mold, once players are weeded out for inferior college production, and they tend to leap out at you.  For now, I like to leave this as an unofficial sub-category of receivers, until the sample size improves a bit.  It pays to be wary of the little guys as this scientific demonstration will illustrate.

Cordarelle Patterson gets left out due to insufficient production.  If he becomes a star, that's great, but he presents way too much risk.  Rattle off a list of successful receivers who produced as little as him?  Yup, very short list.  In my opinion, the Vikings went out on a limb with this pick.  Still, He does have some similarities to Percy Harvin, who they recently traded away.  Then again, Harvin was also somebody who the computer thought was too risky, so we'll see what happens.

Justin Hunter gets bumped off for similar reasons as Tavon Austin.  The computer doesn't like guys that are excessively thin framed.  The best comparisons I could come up with for Hunter were A.J. Green and Sidney Rice.  Not bad company, but also not a large pool of comparable players.  I suspect he is closer to the Sidney Rice end of that comparison.  Again, too much risk.  Playing across from Kenny Britt could be a good fit for him though.

Aaron Dobson is mildly of interesting, but his college production was too weak to make the list.  Robert Woods was the opposite, with good (though declining) production, but mediocre athletic ability.  I'm curious to see what happens with Woods since he could be competing with the much more explosive Da'Rick Rogers for playing time.  I'll be betting on the undrafted Rogers, obviously (with my fingers crossed).

Friday, May 17, 2013

A Player's 2nd Gear

None of what I am about to say should be taken too seriously.  It is just something I like to consider when I try to understand just how fast a particular player might be.  Judging how fast a player actually is can be problematic.  On the one hand, there are the timing  instruments and methods that the combine employs, which seem less than ideal.  On the other hand, there is always the issue that a lot of the players were former track stars, and some might just be better at getting out of there crouch, giving them a quicker start.  You can have several players with virtually identical 40 times, yet one still seems to be faster than the others.

Sometimes people talk about a player's "second gear", to describe a guy who really seems to get an extra boost of speed at a certain point.  As odd as it may seem, I think there is probably some truth to this, and believe that combine numbers can somewhat display its existence.  To examine this, we're going to look at a player's 10 and 20 yard splits from the forty yard dash.  Instead of the actual start of the run, we'll pretend that the race starts at the 10 yard mark.  This should eliminate some of the advantages of a player who just had a good start, versus a guy who maybe stumbled out of his stance.

The value of reaching the ten yard split quicker won't be ignored completely.  A player shouldn't be penalized for being quick.  That is a good trait to have, and worth noting.  What we are trying to do, however, is identify which guys might be faster than their 40 time would seem to indicate.  Who, once the initial launch is accounted for, seems to be gaining the most speed as they cover ground?  Or, on the flip side, who is merely quick, but not really showing acceleration past their initial burst off the line?   To do this, we'll look to see how many tenths of a second a players shaves off  from the 10 to the 40 yard mark.  The math behind this questionable idea will be: 10 yard split -(40 time - 3 seconds).  Later on, I'll get into the 20 yard split, since this relates to another weird issue.

I figure that I should also note what the average 10 yard split times are for different positions, since this will form a sort of  baseline for everything that follows.  Amongst the positions where you would expect speed to be an issue, the results are nearly identical.  These are based on 970 players, so the averages should be pretty reliable.

Position                            Avg. Ten Yard Split
Wide Receiver                         1.552 Seconds
Running Back                          1.548 Seconds
Cornerbacks                           1.532 Seconds

What I'm about to say should all be taken with a huge grain of salt.  I generally lump players, somewhat roughly, like this.  If they are shaving off something in the area of 0.08-0.12 seconds, then I feel they probably are moderately faster than their forty times would indicate.  Or, you could say, they might have better long speed than the 40 would indicate.  If a player is shaving off more than 0.12 seconds, then they probably have an exceptional second gear.  Players who are shaving off less than 0.08 seconds, or perhaps even losing time (reflected by a negative score), might be more reliant on quickness, but might have less deep speed than their 40 times would indicate.  Some sensible judgment should be used when considering all of this, and it might all be nonsense.

Running Backs
Player                                    40 time                    10 yd split                   Time gained/lost
DeMarco Murray 4.37 1.52 0.15
Stevan Ridley 4.65 1.60 -0.05
Ryan Mathews 4.37 1.49 0.12
CJ Spiller 4.27 1.48 0.21
Chris Johnson 4.24 1.40 0.16
Ray Rice 4.42 1.47 0.05
BenJarvus Green-Ellis 4.55 1.55 0.00
Adrian Peterson 4.40 1.53 0.13
Marshawn Lynch 4.46 1.53 0.07
Matt Forte 4.44 1.49 0.05
Ahmad Bradshaw 4.55 1.59 0.04

Player                                    40 time                    10 yd split                   Time gained/lost
Champ Bailey 4.28 1.48 0.20
Johnathan Joseph 4.31 1.53 0.22
Antonio Cromartie 4.47 1.58 0.11
Brandon Flowers 4.55 1.47 -0.08
Patrick Peterson 4.31 1.49 0.18
Terrence Newman 4.37 1.56 0.19
Richard Sherman 4.54 1.56 0.02
Lardarius Webb 4.35 1.49 0.14
Chris McAllister 4.53 1.58 0.05
Antoine Winfield 4.41 1.53 0.12
Carlos Rogers 4.44 1.55 0.11

Wide Receivers
Player                                    40 time                    10 yd split                   Time gained/lost
Torry Holt 4.44 1.57 0.13
Steve Smith 4.41 1.51 0.10
Chad Johnson 4.57 1.56 -0.01
Deion Branch 4.47 1.51 0.04
Lee Evans 4.39 1.56 0.17
Vincent Jackson 4.46 1.57 0.11
Greg Jennings 4.42 1.57 0.15
Marques Colston 4.50 1.60 0.10
Calvin Johnson 4.35 1.52 0.17
Mike Wallace 4.28 1.43 0.15
Darrius Heyward-Bey 4.25 1.44 0.19
Percy Harvin 4.39 1.47 0.08
Brandon LaFell 4.58 1.55 -0.03

None of this is meant to be a judgment on a player's overall ability.  It is just a way to look at their acceleration.  These brief lists were just somewhat randomly put together to show different types of players.  A good example of this second gear might be Torry Holt.  While his 40 time is good, but not shocking, at 4.44 seconds, his 10 yard split was slightly below average at 1.57 seconds.  After ten yards though, he really seemed to take off, shaving 0.13 seconds.  Perhaps he just had a bad start to his forty, I don't know.  That 0.13 seconds of shaved time tells me that he may have even better deep speed than his 40 time indicates.  Based on how Torry looked during his career, I would say that fits with the speed he displayed.

Or, we can take Deion Branch, for example.  He has a respectable 40 time of 4.47, but is only shaving off 0.04 seconds after the 10 yard split.  When we look at his 10 yard split time, however, we see that it is somewhat above average (1.51 versus the 1.55 second average).  That would seem to fit with what we see from him.  He is more quick than fast.  C.J. Spiller has a very good 10 yard split of 1.48 (versus the 1.54 average), yet is still shaving an additional 0.21 seconds from that point on.  He would appear to be exceptionally quick, and have and remarkable acceleration even past the 10 yard split. 

Some players might not show exceptional deep speed.  That doesn't mean they can't be useful.  Brandon Flowers for example is quite quick and agile (which I don't show here), a useful trait for a cornerback.  At the running back position Ahmad Bradshaw is, similar to Flowers, quite capable of being elusive.  Benjarvus Green-Ellis or Richard Sherman might not appear to be exceptionally quick or fast, but that doesn't mean they can't play a game based on raw power.  I'm just trying to examine one aspect of a player's combine data here, not the whole package.  Overall, I think this tends to fit what I end up seeing on the field.

One other weird issue arises when we dig deeper, and additionally look at a player's 20 yard splits.  Normally, there is a correlation between a player's vertical jump and their speed, but not always.  The idea that one sort of explosiveness, common to athletes with quick twitch muscles, would carry over into other areas, probably isn't surprising, but there are exceptions.  First let me show you a typical breakdown of a "Big" wide receiver (players over 200 pounds).

Player                                10 yard split                20 yard split                     40 time
Calvin Johnson 1.52 2.53 4.35
Vincent Jackson 1.57 2.63 4.46

In this case, Calvin and Vincent are both shaving off a good amount of time relative to their 10 yard split (0.17 and 0.11 seconds respectively).  However, from their 10 yard split to their 20 yard split they are actually adding rather than shaving time (0.01 and 0.06 seconds respectively).  This is extremely common among large receivers, and I suspect it has something to do with overcoming the inertia of their massive bodies, to get them up to speed.  On the flip side of this, we have something odd that occurs almost exclusively with very small receivers (though it could obviously occur at other positions too). 

Player                             10 yd split          20 yd split          40 time         10yd to 40 yd diff.
Tavon Austin 1.50 2.49 4.28 0.22
DeSean Jackson 1.53 2.52 4.35 0.18
Jeremy Ebert 1.54 2.53 4.38 0.16
Aldrick Robinson 1.55 2.46 4.35 0.20
Emmanuel Sanders 1.49 2.46 4.40 0.09
Titus Young 1.58 2.54 4.43 0.15
Jacoby Ford 1.46 2.44 4.22 0.24

All of these players show an excellent second gear, ranging from 0.09-0.24 seconds shaved off their 10 yard split.  Oddly, they also seem to be shaving off time by their 20 yard split, ranging from 0.01-0.09 seconds.  This might not seem like much, but it seems to mean that they are constantly accelerating all the way through the 40 yard dash, which is bizarre.  You are probably only going to see this with smaller players, in the 190# or lighter class.  This would also fit with my view of the role of mass and inertia, in relationship to acceleration, though that is all just a wild hunch.  These players also to tend to have rather weak vertical jumps (though there are occasional exceptions).  There is an article here, that kind of gets into some of the kinesiology that might explain this odd group, and the muscle development issues behind them.  Just to be clear, I feel kinesiology is just a step away from wearing crystals and magnetic healing bracelets, so I'm not sure how much stock I would put into this.

So, yup, that's something I have been giving some thought to.  Maybe I'm nuts, and I will change my views on this later. 

Wednesday, May 15, 2013

Paul Worrilow: Neglected Hen

Using a statistical approach to finding middle linebackers can be a bit sketchy.  With other positions you can kind of look at a players physical traits, and see how they will compare to his opponent.  How does a particular wide receiver compare physically to the average cornerback?  How does this pass rusher compare to the average offensive tackle?  Over time, and enough plays, a player with a physical advantage should produce results.

Middle linebackers, as well as safeties, are exceedingly peculiar in this way.  They don't necessarily have one guy they will be matched up against.  On one play an MLB could be chasing a tight end, on the next a running back or wide receiver.  It's also not necessarily predetermined who the match up will be, but likely something that won't be known until the ball is snapped.  So, decision making comes into the equation.  That is something subjective, which everyone has to judge for themselves.

Still, certain baselines physical traits do exist, as well as a few somewhat reliable trends.  For one, the majority of MLBs who achieve high levels of success are running sub 4.7 second forties, and often into the 4.5 range.  Secondly, a player's agility score (based on the short shuttle and 3-cone drill) does appear to have a significant relationship to their ability to drop into coverage.  Thirdly, just like OLBs and defensive ends, a good Kangaroo Score has a positive correlation to their likelihood of being a good blitzer.  Nothing is set in stone, these are just the way things tend to go on average.

Being too rigid about what you are looking for here can be a problem.  Not many MLBs are going to be great at everything.  Some are good run stoppers.  Some are good in coverage.  Some are good blitzers.  Some run stoppers are only good against inside runs, while some are better at getting to the sidelines.  It's a mad world!  The only ones worth paying serious attention to, in my opinion, are the guys who can pretty much do it all, but they almost always get drafted before the end of the second round.

Despite being undrafted, and being a Blue Hen, Paul Worrilow actually compares rather favorably to some of the better MLBs in the league.  He's been picked up as an undrafted free agent by the Falcons, and despite the odds, I'm wagering (my ego, not my money) that people will start hearing his name in the next couple years.  Here's how he compares to some of his peers.

Player                               40 time         Bench           Kangaroo Score       Agility Score
Paul Worrilow 4.59 30 -0.302 2.480
Derrick Johnson 4.52              N/A 0.070 1.223
Brian Urlacher 4.59 27 0.458 1.051
Brian Cushing 4.64 30 -0.219 1.133
Patrick Willis 4.51 22 0.217 -0.282
DeMeco Ryans 4.65 23 0.463 0.569
Lawrence Timmons 4.66 25 -0.481 0.702
Lavonte David 4.57 19 -0.547 0.285
Luke Kuechly 4.58 27 0.303 1.256
Sean Lee 4.72 24 -0.219 1.203
Jonathan Vilma 4.61 23 -0.376 1.516
Paul Posluszny 4.70 22 -0.406 0.996

If the Kangaroo scores seem a bit lower in general, this is because these players are graded with all the OLBs and defensive ends who throw off the curve for the other MLBs.  The average result for middle linebackers is probably closer to -0.800 (for now, use that as the baseline for judging these Kangaroo Scores), so all of these players would be above average for their position.  The average Kangaroo Score for Pro Bowl or All Pro MLBs would be -0.362.  I realize that this makes things look a bit odd, compared to a baseline of 0, but I'm sure you will survive.  Scores are shown as the number of standard deviations above or below average that a player is for his position group.

What you will hopefully notice about Paul Worrilow is his shocking agility score that is 2.480 standard deviations above average for his peer group.  He might have the highest agility score of any linebacker currently in the NFL.  The closest current NFL players I can find are A.J. Hawk (1.891) or Von Miller (1.846).  Worrilow also had one of the better 40 times in the 2013 linebacker class, as well as an excellent bench press of 30 repetitions. 

So, what did he actually accomplish in his time at Delaware?
              Tackles       TFL       Sacks       Pass Def.      Int.      FF       Fumbles Rec.
2012          107           9             3                2               -          1            -
2011           97           11             1               2               1          -            1
2010          113          9.5            2               3               -          2            2
2009           60            4             0                1               -          -            2

Now, I'm not claiming that he is going to be the next Brian Urlacher or Ray Lewis.  I'm just saying that he is a very athletic guy, who showed steady and solid production in college.  The most interesting stat to me is his number of tackles for a loss.  This is always one of my favorite stats, since I think it really says something about a player's aggressiveness to go after the play, rather than just letting it come to him.  In this regard, and many others, Worrilow does as well or better than players like Arthur Brown, Manti Teo, or Kevin Minter who were all taken well ahead of him (since everybody was taken ahead of him).

Still, you might wonder if he actually looks like he can play...

He might disappear, never to be heard from again.  In a league where so few middle linebackers are actually worth very much, I would find it to be a shame if Paul Worrilow didn't get a real shot.  Teams seem more interested in recycling a mediocre talent like Rolando McClain.  So how does Paul compare to the overrated McClain, who is being given yet another chance despite accomplishing nothing beyond being a former first round pick?

Player                            40 time       Bench                 Kangaroo Score           Agility Score
Rolando McClain             4.68               24                         -0.564                           0.074

Yup, Rolando was pretty much just average across the board, but somehow people are surprised that he failed.  Paul pretty much crushes him.  However it all plays out, it's always fun to bet on a somewhat odd prospect like Worrilow, even if he does come from an a place that is just a puffed up sandbar.

Oh, one other thing I wanted to make a note of is the annual "he's an undersized linebacker" nonsense that the commentators spout about anyone under 6' 4" and 250 pounds.  The average height and weight, of the 17 Pro Bowl and All Pro middle linebackers that I have numbers for, is 6' 1.2" and 239 pounds (Worrilow is 6'2" and 238#).  People used to always say that Ray Lewis was undersized, when it seems he was very much the typical build for an MLB.  So, maybe, people should shut up about this stuff.