Rethinking the NFL Combine and Athleticism
A big challenge in NFL player evaluation is the assessment of athletic ability. Most people agree that athleticism is not accurately reflected in single observed event (e.g., 40 yard dash time), but rather it is a complex concept informing many different aspects of a player’s performance. This is why teams use multiple sources of player data when attempting to quantify “overall athleticism”. One widely available public source of player athleticism data is from the NFL combine (and Pro-Days). NFL teams use various formal and informal methods in reviewing a player’s performance from these exercises to come up with an overall player evaluation. Exciting metrics, such as the SPARQ, that attempt to quantify athleticism have recently become more accessible to the general public. However, it is unclear how exactly these tools were developed and whether or not they produce valid and reliable scores that accurately reflect overall athleticism.
Here at VPG, we specialize in optimizing how assessment data are processed, presented, and used for decision making. In the coming weeks, we will showcase a new set of proprietary techniques for grading player athleticism, called the Athletic Aptitude Battery™ (A2B™). The A2B™ was developed using the latest psychometric and statistical techniques, but we will keep the nerd-speak to a minimum. Let’s take a look at what the A2B™ is and its value for player assessment.
What does the A2B™ look like?
The A2B™ has provides three components for player evaluations: (1) Player Types, (2) Domain Scores, and (3) NFL Potential Metrics.
Teams are often looking for a particular “type” of player. For example, a team might have a need at wide receiver, but really they are looking for a specific kind of receiver such as a possession guy or a receiver that can both play the slot and return punts. While both these player types are technically receivers, they possess very different athletic profiles. The A2B™ Player Types address this need by categorizing players into specific player types using custom VPG algorithms (aka math-magic hand waving). A2B™ Player Types can be used to quickly identify players that have the general athletic profile to address their team’s needs. From there, it can useful to examine the detailed Domain Scores to get a more refined understanding of each player’s athletic profile.
Have you ever tried to draw accurate conclusions from just eye-balling players’ data on numerous events and for many of those events players won’t even have data? While it can be fun for an interested fan of football, it is clearly not an efficient or objective way for a team to review hundreds of NFL prospects every year. The A2B™ Domain Scores transform various combine data results into easy-to-understand athletic metrics directly relevant to football. Each player gets a unique set of scores on several important athletic areas such as overall athleticism, speed, strength, and so on. The full range of scores are useful because they quantify different components of athleticism that are relevant to the player evaluation process. For example, if a team is looking for a possession receiver, they are probably willing take a chance on a slower, less athletic receiver who has decent size, but excels at explosiveness and strength compared to other similar prospects. Using the A2B™ Player Type followed by a closer examination of the A2B™ domain scores provides a quick and easy way to sort through prospects while minimizing the interpretational burden.
NFL Potential Metrics
Teams want to know how athleticism translates to actual NFL performance. The A2B™ NFL Potential Metrics fulfill this requirement. Using advanced statistical methods, we rigorously examined how the A2B™ Domain Scores and Player Types relate to actual NFL performance. From this work, the A2B™ NFL Potential Metrics were born. These position-specific metrics provide a simple, single number summary of how well a player’s athletic profile projects to future NFL performance.
Does the A2B™ actually work?
Simply put, yes. We examined the relations between the A2B™ and numerous NFL outcomes and the A2B™ significantly predicted every outcome that we examined. Specifically, the A2B™ identified players that were more likely to make the NFL, be Pro Bowlers, and have longer NFL careers. The A2B™ was also predictive of position-specific outcomes of current NFL players. For example, when looking at wide receivers, as A2B™ Domain and NFL Potential Metric scores went up, so did number of receptions, receiving touchdowns, and receiving yards.
What’s up next?
We spent a lot time this post talking about the A2B™. In the next entry, we will actually show the A2B™ in action. We’ll start with the wide receivers…