NFL Draft Prospects | Breaking down the Draft Sharks Rookie Model
Introducing the Draft Sharks Rookie Model
A lot of people are good at watching college football and “knowing” if a player is good or not.
Is he “bendy”? Does he “have a lot of burst”? Maybe “good vision"?
I am not one of those people. I am good with data. So, let’s talk about some data (and leave the tape breakdowns to the other guys).
One of my first Draft Sharks projects was to build a model combining analytics with film grades to create an all-inclusive approach to rookie prospecting.
Based on research, I knew we would want to break things into three categories:
- production scores
- athletic scores
- film grades
Production Score (0-1)
College production matters
Most (highly) successful players in the NFL had significant production at the collegiate level. There are obviously outliers every year. But as a whole, good-to-great NFL players had very strong college careers.
This Production Score looks at a player's total college output, weighing different factors more heavily by position.
It also adjusts for team competition. For example, if a team has a stud running back like Bijan Robinson, then the other RBs in the backfield (Roschon Johnson) get a little credit for a lack of production. The same is true for wide receivers and to a lesser extent tight ends.
Check out the NFL prospects with the best college production (far right column):
Athleticism Score (0-1)
Being a freak athlete doesn’t hurt
Let's go out on another limb and say that being super athletic is helpful for NFL success. Bold, right?
The truth is everyone playing in the NFL is already a top 0.1% athlete, but we want the best of the best. So, we break down their “athleticism” as a function of how they compare to historical NFL athletes at their respective positions.
It is again threshold based and uses combine numbers as the baseline metrics. If a prospect didn’t participate in the combine, their metrics are estimated using a combination of Pro day or reported values with a correction applied.
Here are the most athletic 2023 NFL prospects (far right column):
You can check out how Anthony Richardson's insane combine and athletic profile stacks up in our updated Dynasty Rookie Rankings.
Prospect Film Scores (0-1)
Rookie film is the missing piece
Grading rookies is hard. Henry Ruggs and Jerry Juedy were graded better than Justin Jefferson heading into the 2020 NFL draft.
It is not easy and specific sources can often have biases or blind spots. So, I created a model to aggregate grades, normalize them, and weigh them based on historical success to generate a final film score.
The goal was to look for outliers and get an overall view of a rookie’s success or lack thereof on film.
Player Grades from Pro Football Focus and Lance Zierlein were given the most weight in the film model.
Additional sources of player grades were included but to a much lesser extent.
Here are the players with the highest film score in the model (far right column):
Agreement
A key metric to analyzing the model outcome is to look at how the production, athleticism, and film grades agree for a prospect. The agreement factor looks at the spread of scores in each category and can be used to identify divisive prospects or prospects with red flags.
It's important to look at the variance in data to see how much confidence we can have in the final output. If the production's there and the athleticism's there and the film grades match, then there's a better chance we have a truly strong prospect.
Or a better chance that everyone was wrong and we can sleep better at night knowing we did everything we could.
Sometimes prospects miss. The goal is to identify red flags or places we should dig in more on certain incoming players.
Disagreement
Disagreement, on the other hand, doesn't necessarily mean you should ignore a player in your rookie draft. Usually there is a narrative or story about why the categories disagree, and a little more investigation is needed.
Jaxon Smith-Njigba (JSN) is a prime example of a good prospect that had a low agreement factor. Our guy Jared dug into the tape and was able to identify why JSN didn't stack up in the production model (injuries) and why the film scores were split.
The agreement factor should be used to screen, not remove rookies from contention.
Here are some of the prospects that have the least agreement (far right column):
Overall the rookie model has been very useful as a baseline to scout these prospects. The functionality to sort incoming rookies by specific categories makes it easy to identify strengths and weaknesses at a glance.
Stay tuned for the next update, which will include defensive players as well.
Check out the full Draft Sharks 2023 Rookie Model Grades here!
Make sure you catch up on the model in action here: