Simplifying the Complex
It’s not easy to look at all of the metrics coming out of the combine and of pro days and understand the athleticism of the prospect, especially when comparing to other prospects. There are eight basic metrics run and two places they run them (combine and pro day) so there are up to 16 metrics to look at for each prospect. Then you have to compare these scores to hundreds of other prospects and thousands of other players that came before them. It’s darn near impossible at face value. Of course, if Thomas Edison thought like that we’d probably still be living by candlelight!
So we need to take this complex set of scores and information and make it simple. the way we do that is by combining certain scores and narrowing down the field into four basic groupings, speed, agility, power and strength. This allows us to look at 4 total metrics from the 16 available, which is much more convenient. Then we can add this “quad” of scores together to get a total score which represents the prospects total athleticism based on his metric results. Still, even if we combine these scores to four from 16, how do we compare them across prospects and history?
For example, everyone compares 40-yard dash times across players, especially of the same position. People are wowed by an RB who can run a sub 4.45 forty time and all of a sudden that prospect is “rising”. Comparing forty times is irrelevant though if you don’t consider the weight of that prospect. It’s a whole lot more impressive to run a sub 4.45 forty at 225 lbs. than it is if you’re 195 lbs. In order to account for this a concept called speed score was created to account for weight in comparing forty times. This is a much better way to approach looking at forty times and comparing them across players. Even so, we still have to find a way to “normalize” the speed score and all of these metrics so that we can compare them across prospects and across past players. We do this in statistics by creating what’s called a z-score, which is comparing the metric to the average score and then dividing by the standard deviation. You can learn more about z-score here if you’d like. By normalizing each metric we’re able to put it on the same scale as the other metrics which allows us to compare metrics against each other, as well as prospects scores across history and across positions. Very powerful stuff indeed. Below is a more detailed explanation of these concepts that gets us to what we call the Quad Score.
At DFF we utilize what we like to call the Quad Score. The Quad Score looks at the 8 metrics that NFL scouts evaluate for NFL prospects, including height, weight, 40-time, bench press, vertical leap, broad jump, shuttle run and 3-cone drill.
Of course, for a variety of reasons not everyone does each drill, so my method is to always take the metric from the combine (since everyone is together in the same environment using the same testing equipment) unless it is not available, then I go to the pro-day metric. If a drill was not performed by a prospect at either the combine or the pro day, that portion of the Quad Score is made to zero, so it has a neutral impact on the Quad Score results, it doesn’t hurt the prospect and it doesn’t help them in the scoring.
Once I have scores for all 8 metrics, I find the z-score for each metric by position. Z-score is a statistical measurement of a score’s relationship to the mean in a group of scores. What this means is that each score is normalized so that a positive score is greater than the mean and a negative score is less than the mean. So, when we compare across measures by z-score, the higher the number, the better the score and vice versa. This makes it very easy to compare athleticism across athletes both in the current class and in past classes. If your Quad Score is greater than zero you are more athletic than the average athlete and when your Quad Score is less than zero you are less athletic than the average athlete.
When the z-score for each metric is created I do one more thing, I calculate the speed score for each player and create the z-score for speed score. Speed score helps put 40-time into perspective by incorporating weight into the equation. Once we have all of these speed scores and z-scores calculated I pull them together in what I like to call Quad Score. The Quad Score, are four combinations of z-scores that add up to a sum total score and measure what parts of that sum score are deficient vs. proficient. The four combinations are:
Speed = z-score of the speed score (which again, also incorporates weight)
Agility = z-score of the shuttle run and 3 cone drill
Leg Power = z-score of the vertical leap and broad jump
Arm Strength = z-score of the bench press
The Quad Score does not incorporate height, which is the only metric of the 8 that is not incorporated into the Quad Score. Once all of this is calculated, add up the quad scores for a total Quad Score. The higher the total score, the more athletic the prospect of course, though that does not tell the whole story. by segmenting the metrics into four distinct types we’re able to see what each prospect excels at and what may be a weakness. In other words, two prospects may end up with a Quad Score of 1.00 but they could reach that score in very different ways. If you have, for example, two RBs with a Quad Score of 1.00, one may have speed and agility and the other may have leg power and arm strength. Both scored a 1.00 overall, but they are very different prospects and only by looking at the quad scores can we see it.
Now, you have to understand what the Quad Score can’t tell us. Especially when you’re looking at combine and pro day information. At the combine and pro days we measure prospects athleticism. Athleticism doesn’t translate to the field when it’s not combined with football intelligence, field vision and spacial awareness on the part of said prospect. It also won’t translate when you don’t have scheme fit, coaching fit and an available starting spot on the depth chart. Therefore, in our rankings you will see low AS Scores higher on the list and vice versa because we recognize this is only one measure of evaluation, though one that projects much of the potential of an NFL prospect.
Deeper Understanding of Quad Score
Using z-scores allows us to “normalize” our data to compare across many attributes and past prospects. It also allows us to easily identify exceptional athleticism and vice versa. In statistics if you are above two standard deviations from the average you are in the 95th percentile of a normal distribution. If you are above three standard deviations you are in the 99th percentile. By using z-scores in Quad Score we set one standard deviation equal to 1, -1 and two standard deviations to 2,-2 and three standard deviations to 3,-3. This means that if your Quad Score is above 3 or below -3, you are in the top or bottom 1% of athleticism, respectively. The same goes for each of the individual scores. So it becomes very easy to understand scores when you are looking at them. If a prospect is between 1 and -1 on any score, they are slightly above or below average in that score. If a prospect scores 2.27 in speed they are in the 95th percentile of speed per weight. Once you understand this concept, comparing prospects across metrics and history becomes very easy. Of course, as noted above, this is only one piece of the puzzle as athleticism has to be paired with other intangibles for a prospect to make it in the NFL. What Quad Score does do is demystify the pre-draft evaluation process and brings order and simplification to the understanding of the athleticism and potential of rookies as you get ready to draft them onto your dynasty teams.