First, off, how did you get started with the RAS development and the launch of RAS.football?
I started it much earlier but it officially launched in 2013 – the Le’Veon Bell draft. That year he ran a 4.6 40 and the entire argument against him was that he wasn’t fast – that he was unathletic. They kept throwing that term around – “unathletic” – and it drove me crazy the way they threw that term around. 4.6 was not bad for a player his size! All the other testing showed he was a good athlete. All those terms – quick, explosive – they didn’t mean anything if there was no context. That’s when I decided to create the RAS model.
So what is RAS exactly?
It’s a player’s relative athletic score. We rate all the metrics with a 0-10 score so they are easy to understand. You don’t misunderstand a score of 0-10. The same with the green, yellow and red colors. It’s easy to explain. It’s meant to be a simple way to compare players by position across a database of over 26,000 players’ scores.
The database compiles, for each player since 1987, the tests from their draft years and their percentage place by individual metric of their scores within the history of the position since 1987.
We collect each players’ metrics across at least six tests – height, weight, and other scores and create an average score across each metric and for the player in total compared to others in their position group in the database. This gives you a good idea of their athleticism relative to others who played their position.
How do you work with positions that have changed over the years as the NFL has changed what it looks and asks for from players and positions?
It’s a good point. Right now we go by all the traditional NFL alignments. The NFL doesn’t break out nosetackles or nickel backs, for example. Splitting those players out from their traditional groups is labor intensive and also would have a retroactive effect on other players’ scores, since every score is an average influenced by what’s in the database.
Do NFL teams use RAS scores? Who does use it most?
Its nuanced – most NFL teams don’t use RAS as they have their own analytics databases and departments and have access to much more data. It is relatively close to what teams like the Eagles, Packers and Colts for example use.
It’s more used by the media and fans and agents. But its interesting – kids who were in high school in 2013 now use it to get their own scores and ratings. And I do talk to scouts sometimes who like it more for trends – to see how many players of a certain size or speed have been drafted. They like the size of the database I have to pull from.
Have you seen teams like the Steelers leaning more into using athletic scores as a way to build rosters?
For many years the Steelers had one of the lower emphasis on athletic scores. Up until a couple of years ago they were below a 7.0 RAS average per player. But over the past couple of years there has been a notable shift. They are now eighth in the NFL in average RAS score per player – a 7.6 average. It’s not elite but it’s very high – especially for a preseason roster. That’s not traditionally at all how they built their roster.
No coincidence it coincides with a new front office?
It’s only been a couple of years so it’s hard to call it a trend yet. But yes, teams with new general managers – they tend to like to take less risks. So taking players with higher athletic scores is less risky versus swinging for the fences with players with lower scores. It’s a much safer approach for them.
Are there positions you’ve noticed where teams look more adamantly for higher RAS scores in comparison to other positions?
You see it a lot at tight end. Athleticism is a bigger part in what teams have looked for at that position. Teams don’t draft players high at tight end if they don’t have high RAS scores.
We don’t split out nosetackles from defensive tackles but when we do take a look at them individually we see that the drafted ones have lower scores than those pass rushing defensive tackles which isn’t a surprise. They don’t need to have the same athleticism – they need mostly to be big!
It is interesting though because even in positions like quarterback that aren’t all about athleticism, the RAS scores still correlate with who teams take. Quarterback athletic scores matter but athleticism isn’t the main aspect of a good quarterback. Yet the data tells us that teams still love highly athletic quarterbacks. But’s important to keep in mind it’s indicative of how teams like to roster build, but not necessarily success.
Is that the case with positions like center and inside linebacker which would seem more cerebral?
Center is another example. It’s a tougher position to evaluate because so much of it is cerebral. You have to understand blocking assignments and what the defense is doing. A lot of what makes them successful is North of the neck.
Inside linebackers – that is more difficult to measure because I think teams look for different things from linebackers depending on the team. So success may not be a Pro Bowl for that team – success may not be what we traditionally would think of as success. Some teams just are satisfied with a solid player that knows where to be at all times and does what he’s asked to do.
Which leads me to my next question – how do we tie in things like roster building using RAS scores with actual player/team success? Can we for certain correlate the two?
That’s the next step. I’m gathering data from Pro Football Reference – it’s a great site. They have all of the player stats – receiving and rushing yards, yards after the catch, tackles and interceptions and sacks… I just need to get that integrated into the database and then we’ll be able to look at those trends. What RAS scores – by individual metric and overall score – tend to lead to the most success. Do players with the best 40 times get the most receiving yards, etc.? We’ll be able to do that by player, position, team, conference…
Is that what’s on the horizon next for RAS.football?
We still plan to keep what we have now on RAS.football free to users. But I do plan to add more tools and allow people to dig deeper than the player cards you can use now to do player comparisons with. Those will be behind a paywall to help me cover my costs for the site and work as it gets bigger and more labor intensive.
You’ll be able soon to dig by a lot more factors and we’ll create tools that back out weight and height to see how players compare while being agnostic on those elements. You’ll also be able to do comparisons by height and weight across specific individual metrics like speed to see how players compare that way. You’ll just be able to mine the data more deeply and see what RAS metrics actually lead to on-field success- with more tools.
I think there will be more fantasy football implications as well. We’ll be able to show you what metrics have traditionally led to success for players in their rookie seasons. This will give fantasy football players more information to make decisions.
Lastly, what do you say to folks who say these data are taking away from the emphasis on what you actually see players doing on the field?
People still have to watch film. I think it’s a trap when people talk about these things in such an extreme way that they just ignore the numbers. No team does that. Every team has an analytics department and uses data. Some don’t weigh data as much as others but they all look at the numbers.