In preparing my yearly NFL season simulations, I typically come across a handful of questions that I want to explore so that I can build stronger projections and feel more confident in the NFL team win over/unders that I want to bet.

In tweaking the early simulations before this year's NFL Draft, I wanted to answer the question of just how much early picks affect team efficiency the following season. Should I be changing things on the fly with each pick? Do we even bother if a team invests in position X? Things like that.

I went into this study with no real expectations so that I don't look at the data in a biased way, though I'm not fully oblivious to my instincts.

Firstly, this is somewhat inspired by a study done by Jim Sannes five years ago that discovered that investing early in offensive linemen had an immediate impact.

Secondly, I know that establishing the run is a recipe for losing in the NFL, so if early running back picks look strong in the study, I'll surely have some caveats.

And thirdly, projecting defensive efficiency year-over-year is tough. If the defensive data returns some ambiguous results, I won't be forcing myself to change my process there.

So, I pulled historical 10 teams that drafted a first-round running back saw the largest shift in median Adjusted NEP per play, and 7 of them improved their offensive efficiency. However, the teams had a median Adjusted NEP per play of just 0.00, which is nothing to be excited about.

Offensive linemen had the second-largest median NEP change as a position group, and 65% of the teams that drafted a lineman in the first round improved the following year.

First-round quarterbacks did lead to an increase in efficiency, typically, but they also were tied to the worst overall offenses of the five positions (with a median of -0.01 Adjusted NEP per play). If you're reading a study like this, you already know that teams that take first-round quarterbacks most likely were bad the prior year and need quarterback help. We should expect the offenses to be weak in general the year prior and have a long way to go from there to get back on par with even the league median rate.

Early receivers might be the most fascinating. They have been part of the best overall offenses the following year (0.07 median Adjusted NEP per play), but that actually coincided with a dip in efficiency year to year. This could speak to a team having success the prior year and be looking to add receivers as a bit of a luxury.

Now, I'll include the first three rounds and then look into some bigger takeaways.

Overall, these results are similar to what we saw with just the first-round picks.

Early-round running backs have been part of some offensive improvements. So have linemen. So have tight ends.

Quarterbacks again are part of the worst median offenses (0.00 Adjusted NEP) with a coin flip chance of improving a team's offense year to year (53%).

Receivers are part of of the best offenses but are not necessarily tied to teams that are improving with their efficiency at a high rate.

What might some of this mean?

Running backs: we know that running the ball is not really the key to winning. Passing efficiency is more vital for securing wins. Teams that take running backs early -- anecdotally, at least -- are probably in a good enough position to draft a running back in the first three rounds because the rest of their roster isn't in shambles. This isn't always the case, but it is more often than not.

Offensive linemen early are taken due to a need, which generally means that a poor line is getting better immediately (so long as the linemen don't bust).

Rookie quarterbacks aren't all hits, and teams that draft them early are already stemming from some weaker offenses.

The receiver thing is interesting because they're also likely viewable as a bit of a luxury pick. Teams with huge question marks -- like at quarterback or offensive line -- are unlikely to spend early capital on high-end receivers.

The overall takeaway here is that investing early in offensive players does actually lead to good odds of improving your team's efficiency year over year. But the results from the larger sample still should have us caution against drastically altering expectations due to a handful of incoming rookies.

Defensive Results

In aiming to predict future-year efficiency, I (and others) have learned that defensive performance is a lot less sticky than offensive performance from year to year.

Side note: that probably inflates the Super Bowl odds and win total projections of teams that had good defenses the prior year (and vice versa for teams that struggled defensively or were just okay).

I bucketed defensive players into just three categories (feel free to miss me with any criticism on that): defensive linemen, linebackers, and secondary.

Here is how first-round investments have impacted defensive performance the following year. (Keep in mind that lower Defensive Adjusted NEP per play scores are desirable, as they indicate how many expected points per play a team allows to opponents.)

Kind of an oof (size large) here.

The median Adjusted Defensive NEP per play scores for teams with first-round defensive investments clings to that NFL average of 0.06 no matter which position is selected, and early-round defensive picks have not necessarily led to significant changes at all -- or even a high probability of improving a unit.

It's a tinge better if we expand this to the first three rounds, at least, which may speak to a concerted effort to improve an ailing defense if multiple early picks are spent there.

We at least see a median improvement in defensive efficiency across all three positions here, but it's minimal, and all of these are league-average units as far as a median goes.

This speaks once again to the volatility of year-to-year defense and, likely, how little one or two defenders can affect an entire defensive unit.

Ultimate Takeaways

It's really tough to get overeager to revamp a team's offensive or defensive projections based solely on how they drafted. If we accept the volatility of how draft picks pan out and look not at singular instances or outliers but instead factor in the overall trends we have over nearly 300 team seasons in this sample, there are some slight takeaways but nothing major.

Defensively, good luck. It's incredibly difficult to project future defensive efficiency, and so it's probably not a terrible idea to scale defenses toward the league average if you're trying to project the upcoming season. I have had the most success doing that rather than projecting outlier teams to remain as outliers.

Offensively, I have to acknowledge that teams that have taken running backs early have improved offensively, but intuitively, these have been teams in good positions overall who can afford to take rushers early. That or the few actual game-changing running backs -- like a Saquon Barkley -- can make a year-over-year difference.

Early quarterbacks haven't shifted team efficiency significantly, and teams that have invested in receivers simultaneously decrease with efficiency but are still strong offensive units. That context should be kept in mind across all positions.

Lastly, we should feel good with Sannes' findings that offensive line investments early are positive for a team's offense for the upcoming year.