Do better teams really draft the best player available?

It is unnecessary to convince the audience of a site called Hogs Haven that the offensive line in general, and offensive tackles in particular, are a necessary condition for success in the NFL. Nearly every consistently successful team in the league has at least a very good offensive line. Of course, many other positions matter, especially quarterback, but a team without good offensive tackles is badly handicapped.

A major point of discussion on Hogs Haven these days concerns whether the General Manager Adam Peters made a mistake by not drafting multiple offensive tackles (OTs) this year. Washington lacks quality OTs on both the right and left sides and this seemed like the ideal year to select at least two prospects.

This year was historic for the quality of the OT class. A record eight OTs were drafted in the first round; no less than 17 were taken in the top 100 picks; and 27 were taken overall. Washington took one, Brandon Coleman, early in the 3^rd round – the 13^th OT taken in the draft. Did the Commanders get the OT help they needed out of this draft?

The discussion raises two different issues. First, is it good policy to steer the team’s draft picks towards positions of need, or is better to simply take the best player available at every pick, regardless of position? This article addresses that issue. A second issue is whether any good options were available when the Commanders picked players at other positions. A follow-up article will examine the latter issue.

Drafting for Need Versus Drafting the Best Player Available (BPA)

Whether a team should draft for value or the best player available probably has been a topic of conversation for as long as Hogs Haven has existed. It’s an eternal dilemma in pro football. Drafting for need means focusing on players that address the team’s most important needs. Drafting the BPA means taking the top-rated player on the board, regardless of position. Often, authors and posters present the position opposing their own in caricature.

Some of the better Hogs Haven past articles on the topic include the following.

• Bill-in-Bangkok offers a full-throated endorsement of BPA
• MattInBrisVegas shows that the probability of actually filling a team’s needs through the draft are low, at least if you expect to get immediate starters through the draft
• Bobby_Gould (formerly KyleSmith4GM) created a multi-part series on draft commandments, which lean toward BPA but make many adjustments for position value.
• Mark Tyler suggests that BPA is often the wrong answer, and argues for BPA at a position of value

My own position is closest to Mark Tyler’s. That is, the team mostly should draft the BPA at a position of need (where “need” considers position value). I have always believed that that’s the way real teams do it.

BPA at a Position of Value: A Practical Stance

Pure positions are easy to express and make for compelling writing but are difficult to use in practice. I won’t exhaust the arguments for and against each pure position, but will offer some reminders of key flaws in each position.

Examples of limitations of BPA:

1. BPA is based on false precision. A player listed as the 60th best each year is not necessarily better or worse than #59 or #61. Rather, there are clusters of players who have roughly equivalent value and who is best among them is unknowable. If the team has a priority need for players at, say, 3 positions, the pool of players available at a given pick will likely include some at positions of need and some who aren’t. Need should be part of the equation for which are selected.
2. BPA proponents can twist themselves in knots in explaining why they need exceptions. Almost all BPA advocates take position value into account, at least when it comes to quarterback. Only one BPA advocate on this board would have entertained taking a RB or WR this year instead of a QB, even if the non-QB was rated higher. But the same logic applies to other positions that have great value (OT, CB, WR).

Examples of limitations of drafting for need:

1. A team’s future needs can only be known imperfectly, due to injuries, trades not yet made, contract disputes, and surprising developmental trajectories of players already on the roster. A team that tries to fill immediate needs through the draft is likely to be disappointed because those needs are ever-changing.
2. Most drafted players do not make an immediate impact unless they are taken at the top of the first round of the draft. That is especially true for positions that typically have a long developmental cycle, including QB, OL, and TE. In general, teams that expect draftees to fill holes by starting from the beginning of their rookie year are going to be disappointed most of the time, unless those players were taken early in the draft.

Because each extreme position is flawed, some middle ground is needed: drafting BPA at positions of value. Great player value trumps position value (as when Johnny Newton fell to Washington this year), but in general, a team should use the draft to address its long-term needs, however imperfectly those may be understood.

Most discussions advocate some position on the BPA / Need spectrum. But only rarely does anyone look at what NFL teams actually do. We consider team behavior next.

Assessing Team Behavior

We examine here whether teams that typically are considered strong drafters and those that are considered weak drafters actually draft for need. There are many different lists of strong and weak drafters. In the first table, I look at eight teams (25% of the league) that almost all raters consider to be strong drafters.

In the second table, I look at six teams that most observers consider to be weak drafters. The weak drafters list was more difficult because of changes at general manager in a number of teams that otherwise would have been appropriate candidates for the list. I removed Arizona, Denver, and Houston, not to mention Washington, from the list for that reason. I left the Panthers on the list because the new GM is inexperienced, they have a meddling owner who often overrules his GM, and because their picks showed all the wisdom of picks in prior years.

I invite any reader who believes that I have missed a team that belongs on either list to do their own analysis and report it in the discussion. I doubt the conclusions of the analysis would be different if one or two teams were to change on either list.

The source of draft needs is Chad Reuter of NFL.com. The lists of needs and their order are solid because team needs are no mystery. For example, who can argue with the list for Washington (last table)?

The numbers in the table are the rounds in which a player was picked by the team. Some teams had multiple picks in some rounds, and some had no picks in certain rounds. The left side lists the priority needs, with highest priority on the far left. The right hand side of the table lists positions not of priority need for which players were drafted by the team.

Table 1

What does the table tell us?

1. The best-drafting teams pick for need. Period. Fully 68% of the 69 picks by these eight teams were at their five positions of priority need. Every one of these teams drafted more players at positions of priority need than at other positions. In other words, they looked for the best available at positions of need.
2. Not only did they focus on need, they also focused on their highest priority needs. Every team drafted to cover its top two needs except for the Packers, who put two picks on every other priority need. Every team put its first round pick, second round pick, or both on at least one of its top two priorities.
3. All best-drafting teams but one put both their first and second round picks on positions of priority need. This especially indicates drafting for need, because the first two rounds are by far the most likely to produce rookie starters.
4. The only team that did not put its first rounder on a position of priority need was Baltimore. Why? A great value at a premium position fell into their laps. The run on offensive players left cornerbacks available who “should” have gone earlier. They took Nate Wiggins with the 30^th pick and doubled down on CB in the fourth round when TJ Tampa was still available. In these cases, value trumped need. Wiggins had a consensus value of 23 and Tampa had a consensus value of 50 (80 picks earlier than they chose him).
5. Although these teams continued to draft players at positions of need throughout the draft, they selected players at other positions (indicating BPA) starting in Round 3. Therefore, it appears that they were not dogmatically going either BPA or need as the basis for their picks after the first two rounds.
6. The best drafting teams often double down on positions of need. There were 16 cases in which a team put two draft picks on a position. In 13 of these cases, those were positions of priority need. Why would they do that? The low yield of draft picks means that it makes sense to take more than one player at a position that the team hopes to fill through the draft. There is a low probability that any one pick will work out.
7. Good teams drafted offensive linemen (tackles and interior linemen) in a historically good year for OL. These eight teams collectively drafted 21 OL. In fact, every team on the list drafted multiple OL, whether the position was a high priority or not. The Chiefs, Steelers, Packers, and Seahawks drafted three. The best drafting teams maintain a pipeline of talent to develop for their OL, and they do not put all their eggs in a basket with one draft pick.

Now consider the teams that are historically poor drafters (See Table 2). What is the pattern for these teams?

1. Bad drafting teams draft for need also. Of their 42 picks, 26 (62%) were at positions of priority need. The percentage is slightly lower than for the best drafting teams, but the difference is too small to be meaningful with such small samples. In other words, nearly all teams in these samples draft mostly for need. However, there was an exception on this list. The Jets (who else) drafted more players that were not at positions of priority need.
2. These teams were less likely to put their top two picks on positions of greatest priority. Two of the six (the Jets and Falcons) did not, and indeed the Falcons were the most criticized team after the draft for taking Michael Penix, a QB, with their first pick.
3. Only two of these teams (Jaguars and Falcons) doubled down on one or more positions of need. The Panthers and Jets, on the other hand, put multiple picks on positions of lesser priority.

Table 2

4. The poor drafting teams did not fill their offensive lineman pipelines. None of these teams drafted multiple offensive linemen. The Giants, Panthers, and Falcons drafted no OL.

5. Subjectively, it seems to me that most of the poor drafting teams made questionable moves. The Jets put fourth- and fifth-round picks on RB, a position of lesser need. The Bears, who only had five picks in the draft, put a fourth rounder on a punter. (Apologies to Mr Party.) The Falcons chose Penix.

Finally, let’s review Washington’s draft (Table 3). Washington has some characteristics of the best drafting teams and others of the worst drafting teams. They put most of their picks on positions of priority need – barely (five of nine on priority needs). Their first round choice was widely applauded, and they put one (and only one) pick on each position of need. However, the Commanders put two second-round picks on positions of lower priority, more like the poor drafters than the best drafters. Johnny Newton looks to have been a very good value (about half a round versus consensus) in the second, but Ben Sinnott at TE was a reach. He will need to show that the reach was a good one. The biggest question in this draft was whether the reach of Sinnott – which may not have been necessary – was worth missing out on an OL, a position of far greater need.

Table 3

Washington’s draft is obviously different from the drafts of the best drafting teams in some respects. Bobby_Gould’s article this week on San Francisco’s draft history – which includes Adam Peters’ contributions – left me concerned. He found that San Francisco usually reached in the draft and its reaches typically did not work out. Peters and his team reached on every pick versus the consensus after Mike Sainristil (data to back up that claim will be in the next article). Will their gambles pay off, or be as misguided as San Francisco’s? Check back in two years.

Conclusion

The data are clear: BPA is not the norm in NFL drafts. NFL teams do not zealously follow the cult of BPA or the cult of drafting for need, but rather they draft the best player available for their priority needs most of the time. This is especially true in the first two rounds. Washington’s path was in some respects more like that of poor drafting teams than top drafting teams this year.