NFL Draft 2014 SPARQ Profiling, Part 8 - Cornerback

If you missed earlier entries in the series, it may be helpful for you to catch up before reading this one.

Introducing rSPARQ

SPARQ Profiling, Part 1: Offensive Line

SPARQ Profiling, Part 2: Tight End

SPARQ Profiling, Part 3: Running Back

SPARQ Profiling, Part 4: Wide Receiver

SPARQ Profiling, Part 5: Defensive Line

SPARQ Profiling, Part 6: LEO

SPARQ Profiling, Part 7: Linebacker

Recap

As a reminder of the basic terminology, here's a little summary of our two metrics.

rSPARQ - a "regression SPARQ" metric, defined by back-calculating NIKE's SPARQ formula and applying regression techniques to get pretty close to the original formulation.

pSPARQ - a "position-adjusted" or "profiled" SPARQ which takes into account a few more parameters, weighting them according to public comments from Seahawks brass and analysis of the existing roster. The main change is using broad jump more than bench press with vertical jump to determine the player's power/explosiveness SPARQ contribution.

Arm Length, Testing Discrepancies

Not to go on yet another lengthy statistical digression, but I found some of the results pretty funny when looking through arm lengths for this post. We might typically view the 40 or shuttle times as volatile, but apparently arm length is as well. Many player arm lengths would vary by a half-inch (and in some cases, close to an inch!) depending on where the test was conducted, with the discrepancy likely arising from a different scout conducting the measurement.

In fact, after I had finished writing this, Field Gulls writer Jared Stanger mentioned that he had come across the same thing in talking to a 2014 prospect. We'll keep the prospect anonymous, but here's what he said about arm length measurements.

"I have had [arm length] measured at 34, 35, and 33 [inches]. The 35 was the proday measurement that [redacted] got from the scouts."

"Arm length is a hard thing to measure one scout told me and it's different with every scout."

This is just a reminder that test results are not absolute. If I was following correct scientific protocol, there's no way I would report the SPARQ values with as many significant figures as I have in all of these posts. I just throw in the extra value after the decimal place (i.e., a 140.1) because it looks better. We should be implicitly placing a confidence interval around each SPARQ. A 140.1 tells us that the player is probably around a 140; saying anything further is implying a greater level of accuracy than is actually available to us.

Getting back to arm length, it might be more accurate to report wingspan or standing reach, as are typically done in the NBA. There's only one obvious way to measure it, so it would likely be a more objective measurement. Unfortunately, it's not widely available in the same way that arm length is.

Keeping all that in mind, make sure to take the arm length values presented here with a grain of salt.

Existing Roster

This is a position group that I've been looking forward to, as Seattle cornerbacks seem to fit into a specific mold. It's somewhat similar to running back, where isolating the potential Seahawks is almost as simple as setting a few parameter boundaries and clicking the "sort" button.

As done in the previous articles, the best way to develop a prototype is to look back on past data. The following table only lists five cornerbacks, and there have certainly been more than five CBs in the PCJS era. The reason for this is twofold: first, arm length is a critical input for Seahawk corners, and we often don't know the arm length for lower-profile players. Second, Walter Thurmond is the highest-drafted by Seattle at the position, but he was injured for both his pro day and the combine. I tried to dig up numbers from his time back at Oregon, but not even data chief Tony Wiltshire could come up with anything.

Because of this, the five players represented include the four corners drafted by Seattle not named Walter Thurmond, with Brandon Browner added in as well. (If you want to see a few more of the camp body names, check out the chart here, 100% more Ron Parker and Marcus Brown!). The average for cornerbacks is a 104 pSPARQ, so adjust expectations accordingly. Richard Sherman's 123 is actually a very good number at corner.

The easy takeaway is that, as we all know, the 5th and 6th rounds are where Seattle likes to draft this position.

As a note, the following table only includes the five corners discussed earlier. I ran the results for all eight corners, and the results actually improved, showing greater athleticism relative to the average. We're not selectively biasing our data set by omitting important names.

We know from previous positions that Seattle likes the broad jump. Bobby Wagner and Christine Michael are examples of high-SPARQ 'Hawks with great broad jumps, but even slightly less-athletic types like Zach Miller have good results for their position. The correlation for this test at corner is almost as high as it was in the linebacker category.

Second, we also know that Seahawk corners are long, and that this probably comes at the potential loss of agility. Perfect athletes who also happen to be tall and long are typically not available in the fifth and sixth rounds, so Seattle seems to allow for only average shuttle performance.

It's worth noting that Byron Maxwell is the SPARQ outlier here. He spoke this past season about having stiff hips when coming out in the draft, and that he addressed the issue with hot yoga. This is a good reminder that a bad test result is not absolute, and that it's easier to improve agility than it is to grow longer arms.

Speaking of longer arms, no Seahawk corner acquired through the draft has had arms shorter than 32", with the caveat that Walter Thurmond does not have a test on file. This means that finding a corner prototype is actually easier than doing the same for other positions, because we have more parameters to use in refining our search. More on this later.

Historical Context

To give an idea of what a cornerback score looks like, below is a table which lists a group of players from the last 10-15 years. There's no intentional selection bias, and there's varying levels of performance among the sample group, i.e. some players are certainly better than others.

Fairly striking is that these players have some freakish times in the short shuttle. While Seattle tends to sacrifice agility for length, most star corners of the last 15 years have had elite quickness.

2014 Draft

The same note applies as before: values that are bolded, italicized, and underlined are assumed. If most of the necessary variables were present, I assumed that the player's 40 would be in the same percentile as his 10-split, 3-cone to shuttle, etc. This means that if a player doesn't have a 10-yard split, his speed index is only influenced by his 40, and things function similarly for other categories.

Again, disclaimer: rSPARQ and pSPARQ average out to be (roughly) the same. The average CB is at a 103 rSPARQ and 104 pSPARQ.

All data provided for the SPARQ study is courtesy of Tony Wiltshire.

There's a little more of the "estimated" variety for cornerbacks than we've seen for other positions. One reason is that a lot of corners elect to not bench press, but I can't quite rationalize why so many are deficient in other areas. Most estimates are fairly minor, as we have data that shows how 3-cone applies to shuttle, 40-yard to 10-yard, etc. Our estimates will likely end up being very close, so the large amount of italicized text in the right column shouldn't be too worrisome.

Prototype Refinement

Rather than just list a few individual players, we have a very specific prototype we're looking at and we can try for one more layer of refinement . All we need to do is limit the sample to players with above-average SPARQ and arm lengths greater than 32". This method will catch some players who will likely be gone before we draft a corner, but it should also show some potential prospects who fit right into the mold and might not require too much draft capital.

Now, these tables are not necessarily exhaustive, because we don't have test scores for everyone. Arm length is often not publicly available. However, it likely contains most qualified players.

I also cheated a little and reduced the arm length requirement to anything over 31-1/2". There's too much inherent variance in arm length to make a firm cut-off at 32", and it's quite possible that the Seahawks wouldn't disqualify someone for being at 31-3/4" anyway. We also don't know what their scouts have measured for any given prospect's arm length measurement,

I did not restrict the sample by height, because we're also in the market for a nickel corner, and my intuition is that we care less about height there. As we don't have data for Thurmond and he's the only true nickel back that we've drafted, we're left with guesswork. Length is probably still important, and I would guess that agility matters much more without a sideline to help cover the receiver. Basically, we're looking a little shorter and quicker, as you'd expect.

Projected round accuracy may vary. It's just ballpark.

Projected Rds 1-3, Arm Length > 31-1/2"

Projected Rds 4-FA, , Arms > 32

Projected Rds 4-FA, Arms > 31-1/2

Highlights

The tables above really cover everything, so I'll just briefly mention a few names of particular interest here.

Mohammed Seisay, CB, Nebraska, 116 rSPARQ, 130 pSPARQ

With potential Seattle fits Stanley-Jean Baptiste and Keith McGill projected to go early in the draft, I've tried to keep my eyes more on those projected to go later on in the weekend. Seisay's the prototypical Seattle corner that's available in the fifth round. He has a more famous teammate (Tharold Simon, Luke Willson, etc.) that might be slightly decreasing his visibility among the draft community. From a purely numerical basis, he's a tremendous athlete with elite length and matches the same criteria as the five PCJS corners discussed earlier.

33-1/2" arms belong on an offensive lineman.

Seattle was present at his pro day.

Shaq Richardson, CB, Arizona, 107 rSPARQ, 113 pSPARQ

There's a bit of Byron Maxwell in Richardson's numbers. The only thing keeping him from being an elite SPARQ guy is a bafflingly poor shuttle time which has very little correlation to the rest of his numbers. He just seems generally undervalued considering collegiate performance and obvious size and athleticism. He's one of my favorite corner prospects, but we'd probably need to enroll him at Byron's hot yoga studio.

Rashaad Reynolds, CB, Oregon State, 127 rSPARQ, 123 pSPARQ

The only real strike against Reynolds is arm length, where he comes in at 31-5/8". As we discussed earlier, this may not be an exact measurement, and we don't have the exact length requirements for a Seattle nickel back anyway. Reynolds was one of Jared's draft gems back in November.

Beyond the arm length, he's an elite athlete who's excellent in all categories. I watch most Oregon State games, and Reynolds is a very capable player who I'd love to see us grab in the fourth round. I'm thinking that he might be selected more toward the early-to-mid fourth, though, which would place him out of our range. There's always the possibility of trading into the late third or early fourth to grab him.

As always, I'll be around in the comments and would be glad to answer any questions you may have about SPARQ or general methodology. I'm anticipating that we'll be looking at our final position group, safety, on Monday.