The Draft Research Project: Of Starters and Urgency

Previously, I looked at a pool of roughly two thousand players across eight drafts to look for trends in how teams invested their draft capital, what results they saw, and which positions were the most and least urgent in terms of finding potential impact players–players who at least earned a Pro Bowl and played in a minimum of 40 games.

One of the surprises (to me, at least) was the preeminence of the defensive line in terms of urgency, the somewhat middling position of quarterback, and the notable humbling of wide receivers. However, there is more to the story. Teams cannot be composed entirely of impact players, and with an average of 65 regular starters available for 32 teams in each draft, just finding those starters has to be the goal. The table below offers the simple information on how many reliable starters at each position were available in a typical draft on Day Two and then on Day Three, with the ratio noted as a falloff margin.

The falloff ratio for wide receivers and edge rushers should scream to even casual fans that even if a team feels it can wait until the second day to pick one of those players up, they absolutely can’t wait until the third. However, after that settles in, even this should not pass the smell test, because this does not look anything like how NFL teams draft, and it doesn’t match conventional understanding of urgency. The reason? Teams probably don’t just need one starter at wide receiver, they need three (but they do only need one starting running back).

In Context

Here is a list of the number of regular starters found in the average draft across the eight years under study, divided by the number of players at that position a team would need for a “typical” starting set (assuming 11 personnel and nickel formations, the most common across the league).

This resembles the conventional draft priority shown by teams to a remarkable degree. “Clearly,” it seems to say, “teams are looking for starters and hoping for more.” However, as I composed the tables for defense, one issue stood out, and that was the problem with corners. True, across the entire league, roughly 60% of the time, teams play out of nickel, and dime packages take up another 14% of snaps. Football Outsiders reports that in 2021, the Buffalo Bills played nickel 92% of the time (and dime another 3%), with the Dallas Cowboys even more aggressively playing with at least five defensive backs on 98% of defensive snaps. The team most likely to use base in 2021 was actually Seattle–which still had at least five defensive backs on the field 61% of the time. Indeed, for that same year, the Cowboys’ Pro Football Reference data has five different defensive backs combining for 75 total starts (an average of 15 each) with various other players filling in the gaps. Thus, corners should be an urgent need. But...

Limitations of the Model

There are some limitations, however. For that same year, the Bills only had 34 starts at corner listed under Pro Football Reference, and Seattle had 43 starts listed for their corners and defensive backs outside of the 34 credited to their starting safeties–enough for them to “start” nickel or more half the time. In other words, even though the true base formation in the NFL is nickel, not all tracking data reflects this.

As a simple crosscheck, I created a pool of “contributors”, defined as all players who ever earned a Pro Bowl, started in 40 games, or simply played in 60 games. I wanted to make sure that I was not missing players who were on the field but not showing up in my tables. Whereas corners make up 11.6% of all draft selections, corners make up 11.7% of contributors. In other words, there is not a hidden pool of corners making reliable contributions and just not being credited as starters. However, I made one last check to be certain that I was not missing something.

There were another 30 corners who played in at least 60 games without recording 40 starts, or 12.6% of the total group selected. For a frame of reference, there were 28 defensive linemen who met the same definitions, or 12.7% of the total group selected. In fact, that tracks with the draft-wide trend of there being 254 players who were not reliable starters but who were regular contributors out of the 2004 non-specialists (12.7%). Once again, there does not seem to be a hidden body of pseudo-starters out there for corners. These are the players teams have to work with, and the rate of finding such players shows that there is a need to find these players as readily as possible.

Conclusions

Ultimately, the one data point I kept coming back to was that while like many fans I tend to focus on “stars”, finding stars in the draft is a bonus. The simple reality is that even finding players who can provide regular contributions is difficult. Only 694 of the drafted players (35%) even appeared in 60 games across the first 80-82 games available to them. Drafting a player who sticks is in fact the victory. Here is one last look at the numbers, sorted by the percentage of contributors gone by the end of the first round, with the percentage of contributors left on the last day noted as well.

The low total number of guards makes a certain amount of sense given how many of the “tackles” taken as contributors actually end up playing guard. However, it should be clear from all of these tables that teams need to prioritize the trenches, and then they need to either shift to cornerbacks or wide receivers. However, this is where the trend rates for finding impact players might be worth consideration. With only 7 of the 24 Pro Bowl-level receivers going in the first round but 14 of the 23 Pro Bowl-level corners being taken then, the tie-breaker would seem to go to the urgency of cornerback as both regular contributors and impact players at that position are more likely to be gone by the end of the first round (or the middle of the second).

Interestingly, there is one person who at least publicly agrees with these conclusions, and that’s Ryan Poles. Speaking about how to improve the Chicago Bears, he said “I always go back to the premium positions...We’re always going to look at pass rushers, we’re going to look at offensive linemen, corners.”

If he is true to his word, the data backs him up.