2020 NFL Combine: Why It Matters

Now, more than ever, I know NFL Combine results matter. There are definitely examples of “workout warriors” selected too early and failing, but that can be said for any kind of evaluation. Yes, for teams the medicals and interviews matter to a great degree. But we do not receive that information, therefore my focus will be on the numbers generated from this week. 

My goal here is to teach you what I’ve learned over the past few years. To perhaps shift the lens that you watch this week through. And if you aren’t a believer in athletic testing, hopefully your perspective is changed by the end.

As Zach Whitman put it - “Metrics don’t need to be perfect if we do a good job of understanding what they’re saying and what they miss.”

Featured in this post: composite scores, eliminating non-NFL caliber athletes, why teams utilize thresholds, positions most impacted by athleticism and the individual results that best project success.

Composite Scores - Clearer Picture

More or less, the NFL Combine is broadcast as an event to run 40s followed by position drills. In the past there has not been an extensive focus on other results unless a name or number pops out. Why is the 40 more important than the shuttle? 3-cone? Broad jump? And most of all, our understanding of the results is missing one crucial element - weight. In its simplest terms, a 225-pound RB jump of 38-inches is more impressive than a 205-pound back jumping the same distance. Adjust all results for weight.

What if there was a better way? What if we recognized that the forty is just one of seven or eight or nine meaningful results, and a potentially better way of interpreting athleticism is through a composite score which factors in all outcomes along with weight.

Take Nick Chubb for example. On the surface, a 4.52 forty would not grab anyone’s attention. But when coupled with a 4.25 shuttle, 7.09 3-cone, 38.5-inch vert and over 10.5-foot broad jump at 227 pounds? He tests in the 83rd percentile, meaning he is more athletic than 83 percent of all backs entering the league.

So what can we do about it? SPARQ is the approachable composite score I use, and Zach Whitman has years and years, thousands and thousands of scores built up in his database so prospects from each Combine can be compared to their predecessors. Great scores obviously stand out, but it is important to note that an average NFL athlete is not a negative. So don’t immediately knock a prospect for testing in the 48th percentile. 

I love these composite scores. A single number for total athleticism. I can't tell you what percentage of the evaluation it makes up for me. In fact, it might be a sliding scale. An extraordinary score for a high round pick confirms my confidence. And towards the end of the process I geek out when pro day numbers come in, completing profiles, in order to work through the top 15 or so at each position in the hopes of finding someone who has slipped through the cracks. These have hit in the past.

Again, let’s stop clinging on to single forty times when explaining a player’s athleticism. Instead, cite composite scores, which includes all measurements with weight as a factor, as other tests can be equally important.

Eliminate Bad Athletes

The easiest way to implement athletic testing in evaluations is to eliminate non-NFL caliber athletes. Those so distant from “average” that the hit rate is minuscule. From a composite perspective, this means scores two or more standard deviations away from the mean (-2.0 sigma). To better illustrate this, here are the prospects with non-NFL athleticism since 2015 who also attended the Combine.

How many hits do you see out of 70-ish names? Few. And extending that to all prospects, not just Combine invites, expands the cut list to hundreds and hundreds each year. Eliminating these players from draft boards entirely is a quick way to minimize the player pool, theoretically improving the rest of your board, even if a very small percentage do “hit” in a five-year span. The best of this list might be Levi Wallace, an undrafted free agent. Pat Elflein is a three-year starter for the Vikings with mixed results. Ryan Anderson is a serviceable edge rusher at the NFL level. Meanwhile, it could have prevented struggling selections like Jachai Polite and Jamarco Jones.

You might ask, “Where is Orlando Brown?” the Ravens right tackle who produced a woeful workout a few years ago. Thank you for the question, as it leads me to the next point: Pro Day workouts matter! Brown’s testing at Oklahoma thrust him out of non-NFL caliber athlete territory. We all have bad days, players deal with injuries through the draft process and the NFL Combine is a grueling experience. A fusion of fatigue. Therefore allowing players to improve on their testing results in a better environment (pro day) is the correct approach. Prospects know if they bombed an athletic test and know if they can improve. Let’s give them that chance.

Thresholds and Their Purpose

Some of the most important measurements have already been recorded prior to prospects touching the field in Lucas Oil Stadium. Heights, weights, hand size, arm length and wingspans can all be vital for this reason: thresholds.

My perception of minimums and thresholds changed after reading this piece. If it needed to be funneled into a single line, one stands out: “Big picture wise, you want to play with the odds, not against the odds.” In this case, the odds mean siding with prospects who possess the measurements that are successful in a specific scheme deployed by the team. So narrowing the group of prospects down theoretically could improve evaluations.

Example: the Seattle Seahawks at cornerback. The last seven outside corners Seattle drafted all possess arms 32-inches or longer. How can this impact their evaluation process? At the Senior Bowl, of the nine or so prospects on the roster who were listed at corner, four had arms 32-inches or longer (or 31 7/8). So, the Seahawks (among other teams) go from focusing on nine outside CB prospects down to four, theoretically improving the evaluations of that group with more time spent. Others who project to the slot will be evaluated separately, but you get my point. In fact, the Seahawks believe in this method so much that they even look at safeties and project transitions to corner, see Tre Flowers.

Other teams don’t take it as far as to eliminate prospects completely but link certain tests with specific positions. Like the 3-cone drill for Patriots’ corners.

Will this mean some teams miss on quality players who do not fit within the parameters? Absolutely, but these decision-makers are banking on good process to win in the end.

What History Tells Us

These next two sections are singular testing results that best project future success for certain positions. That can be quite a task, finding a single athletic test that has the highest hit rate among the top performers, but two fit the bill. I know career starts might not be the best way to exhibit success, but it does show a combination of trust and longevity teams have in a player.

First is the 20-yard shuttle for offensive linemen.

These are the top Combine testers from 2010 to 2019. In that span, over 300 OL prospects completed a 20-yard shuttle. The top 23 are listed above. As you can see, 20 of the 23 were drafted, and those drafted players went on to start 80+% of their career games. Taking it one step further, the eight Day 3 OL started 85.47% of their career games and includes gems like Jason Kelce, Matt Paradis and Charles Leno.

It is not a leap to say that if an offensive lineman at this year’s Combine hits that 4.47 short shuttle, he should be a target player. This type of success is extremely impressive. Again, that result places them in the top percentiles of their position, but it also results in success.

The pass-rushing group is not as clean, but it does show some level of prediction.

Below are the top 3-cone times for EDGE and DL that were drafted in the fourth round or earlier since 2007.
 

19 prospects hit the Combine 3-cone threshold of 6.89 seconds that went on to be selected in the top four rounds (four others hit that mark but were drafted in rounds 5-7 or were not drafted at all).

Cliff Avril and Clay Matthews just missed with a 6.90. Anthony Barr, who now plays off the ball, registered a 6.82 a few years ago. Again, both of these are only including NFL Combine participants. Obviously all are not “hits,” but the rate of success (of varying degrees based on expectations) in comparison to other positions is high.

Let me reiterate that I am a believer in composite scores. The goal here was to find singular athletic tests that best predict success, and the OL shuttle is undeniable for the top percentile.

So How Do We Use This Information?

Not to confuse you, but the application of athletic testing is not identical across all positions. Consistently I have seen athletic testing improve my evaluations along the offensive and defensive lines, both for early and late-round picks. Let’s hypothesize why; trench players frequently feature in one of the few one on one matchups on an NFL field. At the basic level, wouldn’t an athletic advantage be a leg up? This extends to day three/UDFA offensive and defensive linemen: prospects who are incomplete but offer an athletic base to develop with coaching. 

Outstanding athletic profiles (when matched with intriguing tape) led me directly to Charles Leno, Shaq Mason, Grady Jarrett, David Onyemata, Javon Hargrave and Maxx Crosby, among round three and later picks.

Conversely, I do not get discouraged when top-ranked running backs do not test like top athletes. In fact, they often test like average to below-average athletes and that’s okay! So much goes into playing the position: vision, balance, aggression just to name a few. It’s great when top backs also test like top athletes: Saquon Barkley and Christian McCaffrey for example, but it isn’t a necessity. I have found that throwing darts on undrafted uber-athletic ball carriers is highly fruitful. Recent examples include Matt Breida, Mike Boone and Raheem Mostert, all testing above the 90th percentile.