clock menu more-arrow no yes mobile

Filed under:

Evaluating The Draft: Linebackers

I continue my historical analysis of the NFL draft, showing performance for each position based on round selection, Combine results, and PFF grades.

Kevin Hoffman-USA TODAY Sports

Alright, my fellow Falcoholics. It's time for the latest rendition of my draft evaluation. This time, we're moving away from the more obvious positions of need and tackling linebackers. Injuries aside, it's hard to ignore the issues we've had at the position. Atlanta's linebackers finished fourth worst in missed tackles with 14.7 per linebacker. They rank in the bottom ten in terms of opposing QB rating, including yards gained per snap in coverage and touchdowns allowed. We could always use some improvement in both areas, as well as some spark to our always embarrassing pass-rush.

I had some feedback in the last evaluation article about separating linebackers by inside versus outside, as well as 3-4 defense versus 4-3 defense. Unfortunately, I don't have a large enough sample size when producing results by round or position type. I will tell you that close to 33% of Linebackers in the NFL played a different position in college, whether it was Safety, Defensive End, or Inside versus Outside. That number pops above 45% when separating 3-4 from 4-3 outside linebackers. This implies to me that we shouldn't see any difference between the positions, though I want to revisit this when I have more information available for comparison.

For those of you that haven't read my previous draft evaluation articles, I'll post the links below, but you are in no way obligated to read them. I'll summarize the assumptions here to prevent that:

  • "Significant Snaps" is defined as 300 or more snaps within a season.
  • I used ProFootballFocus.com (PFF) to grab grades on each player, if it exists. For our purposes and to keep it simple, I have adjusted scores to be comparable with other positions, where zero is a neutral grade.
  • Grades are standardized to a per game estimate of 60 snaps.
  • This analysis includes 211 Linebackers drafted between 2007 and 2013.


Linebacker PFF Grades

To evaluate draft pick success, it's important to look not only at performance on the field, but also whether or not a player gets on the field. To measure that, let's look at what percentage of players had significant snaps.

Sm0jwuv_medium

All fourteen first-round picks have contributed in year one. Even in the second round, over 65% see the field. Compare that to defensive end at 70% and 45, respectively; Defensive tackle at 75% and 50%; Offensive Line at 80% in both. Of the fourteen, eight graded positively in the first year, and four of the remaining six had only slightly negative grades (above -0.25 per game). Essentially, that means 85% of the players delivered average performances or better. That same ratio continues into the second year.

Now, let's take a look at the average PFF grade for each round:

Id8i5ml_medium

The rookie year difference between first and second-round picks is minor, and second years favor first-round picks. It also appears as if players in round three generally have the same potential as round two. The drop off in playing time is moderate, but it also shows that the talent has historically been consistently available. All of this implies that college success translates well to the NFL.

Similar to the other positions, I'll differentiate the top ten selections from the rest of the first round:

Fu3u85v_medium

In all, there's not much of a difference with a top 10 selection compared to later in the first. The first year favors the former, but the second year difference is insignificant.

Also, we may get on Weatherspoon for lackluster performance coming out of the first round, but he graded as one of the best Linebackers in the NFL in his only healthy season (2011). It's a shame he can't stay on the field.

This was the primary delay in writing the article. I started making the analysis too complicated. I took a step back and approached the information a different way, only to stumble upon some pretty interesting results. I won't share all the detail...yet! We'll keep the exposure pretty similar to the other articles so that the comparisons remain simple.


WEIGHT

First, let's see if weight has any effect on performance.

Gsfadlp_medium


I notice a distinguished performance cap on players less than 230 pounds. I'm not trying to say that a good linebacker can't fall below that threshold, but the overall distribution indicates player limitations. Of that group, more than 75% recorded negative PFF scores in each category.

Pass rushing linebackers seem to be more effective as their weight increases. However, it shows a coverage trade-off. I imagine that larger players just can't match the speed of their opponents. On the flip side, I believe the smaller linebackers just get overpowered. The intermediate weights usually perform well, possessing a good balance of size and speed to match the receiver.

Against the run, you will notice a distinct boost in performance over 240 pounds.


HEIGHT

Wz2tvfl_medium

I won't go into as much detail here, but I'll list what surprised me. First, 6'3" happens to be the optimal height across the board. Also, taller linebackers grade worse in pass rushing and coverage, thought the sample of those is small (5). Some of these differences are relatively minor, but still pretty interesting. Here's what the overall scatter plot looks like for the heights:

Uqimesw_medium


This shows that the performance potential holds steady at each height. The downside, on the other hand, shows a sizable decrease. I think it's really a little too early to say something definitive about the 6'4" players, but the grouping doesn't show a great start.


EVENTS

I spent a good deal of time grouping similar event performances. I created matrices to compare average scores and the percentage of positive performances. When I cross referenced the two, I found three events that more significantly correlate with NFL success. I used these, as well as height and weight, to create my own predictive score where 100 is neutral, and 5 points represent a moderate change in Combine performance.

The red bars match with PFF scale on the left side; the blue line represents the percentage of players that graded positively within the group.


Utdfoip_medium

The three events: Broad Jump, Vertical Jump, and 3-Cone. Please remember, that doesn't mean other various combinations can't be successful. This combination just produced both the highest average score, as well as the greatest proportion with positive grades.

Using this combination, we have distinguished improvements between each group. Pretty interesting stuff! I'm trying not to go into too much detail here, but, like height and weight, these events correlate a little differently based on pass-rushing, coverage, or run defense. I'll expose some of that in another post, possibly when the Linebacker results start pouring in from the Combine.


Final Thoughts

I won't bore you with a long drawn out conclusion like always. I'd rather try something new. You guys feel free to share your thoughts on the position, both historically and how the Falcons could approach it in the draft. I will at least say that this process has made me even more intrigued with drafting a linebacker. I know that both of our lines are more obvious choices for our first round selection, but a versatile linebacker opposite Biermann would give Nolan's scheme a little more juice. It also may open up more 3-4 schemes within the hybrid defense.

As always, I remind you that I don't intend for this data to be used independently. Rather, its function is to challenge preconceptions and assist in differentiating prospects. In this particular scenario, we can use data to hopefully differentiate top prospects or uncover hidden gems.

I will do my best to fulfill any requests you may have for more information. I've continued to dig more and more into this position, so don't hesitate to ask for something. There's a good chance I have it! I'm always open to criticism to ensure I can relay my information to you in the most effective way.

I'll close by updating the round-by-round position comparisons within the same graph.

Ho5jletl_large

Cvnzkfzl_large