Pro Football Reference Blog

Among his many thoughts in his weekly Monday Morning Quarterback article, Peter King shared some on Eric Berry, the star safety from the University at Tennessee, who is projected as a top ten pick, and on many boards a top five pick, in this year's draft:

If I were an NFL team drafting high, I'd be very careful evaluating Eric Berry.

The Tennessee safety, obviously, is a rare prospect. But the history of safeties in terms of longevity and greatness at the top of the draft is very shaky.

The nature of the position is smallish people throwing themselves around like linebackers, and that doesn't lend itself to long careers. The three best safeties to be drafted in the past decade -- Ed Reed, Troy Polamalu and Bob Sanders -- have missed 78 games due to injury in their 21 combined NFL seasons.

Berry looks like a top-10 pick, but the team that takes him is going to be picking against history. Of the five top-10 safeties this decade, none has had franchise-player impact: Roy Williams (Dallas, eighth overall, 2002), Sean Taylor (Washington, fifth overall, 2004), Michael Huff (Oakland, seventh, 2006), Donte Whitner (Buffalo, eighth, 2006), LaRon Landry (Washington, sixth, 2007). Taylor might have had franchise-player impact if he had not been gunned down three-and-a-half years into his career. But overall, the position justifies the caution lots of teams are taking with it.

. . .

I'm not saying Berry won't be a great player. Maybe he'll be Ed Reed. Maybe he'll know when to dish out the big hit and when to steer a player instead of seek and destroy. But the odds of him being great for a long time -- as opposed to the physical longevity of a tackle or defensive lineman or quarterback not subject to as many high-speed collisions -- are pretty long, based on history.

I had several instant reactions to that segment, including:

1) Well, if someone is talking about history, we might just be able to look into that;

2) It would suck if Eric Berry had a career like Troy Polamalu, Ed Reed and Bob Sanders . . . I mean, who would want 9 pro bowls, 5 first team all-pros, and 3 second team all-pros, and numerous playoff appearances during the combined first five years of their respective careers. Seriously, who would want to settle for that while they were still playing under their rookie contracts? If God gave you some insider info that Eric Berry's first five years were going to be a cross between Troy Polamalu, Ed Reed and Bob Sanders, but he was unfortunately mum on everyone else, and you were a team sitting with pick number five for a team that had horrendous safety play the year before, don't you jump at that opportunity? I don't know how Eric Berry's career will turn out, but I would be primarily concerned with how he will do while he is playing under the initial contract, and not what might happen eight years later.

3) Even just looking at the five recent top ten safeties cited, it is clearly a case of lack of comparison to how other positions perform and selective memory. Roy Williams is a liability now, but he made 6 pro bowls and 1 first team all-pro. As we will see, that's pretty good for an early first round pick. Sean Taylor made two pro bowls in his first four seasons and looked like a star; I don't think we should anticipate the tragic end to his life and project that as impacting how highly drafted safeties turn out. Whitner, Landry and Huff have each started three seasons. Others can chime on Whitner, Landry and Huff, but I wouldn't rule out Whitner or Landry making a pro bowl at some point; Huff looks like the most disappointing but he could turn it around. Those three guys are young and have been starters and we certainly shouldn't make a historical comparison to other guys who are still quite active and potentially entering their primes.

4) On the injury thing, how different is that from other positions? I think it is an interesting question--though one that should be more of an issue when evaluating contract extensions and second contracts. Do star safeties tend to fall apart more frequently than other positions, once they hit the mid to late twenties?

I'll try to take a look at that last injury issue in the future. For now, I want to get back to the history of drafting safeties in the first round. We can always have a recency versus relevancy debate, but I'm not going to look at guys who have played for three seasons and are still active. I used the draft database to find all players in the 1978-2002 drafts--this coincides on the front end with the changes in passing rules and the sixteen game schedule, and ends with a draft where players, if active, should be at age 30 or older. I then sorted each position by number of players who made a pro bowl, and number of players who had a career AV of 50 or greater. Here are the results for all first round picks during that span, sorted by position, and percentage of pro bowlers.

Okay, so this is just a quick look, and is not sorted by draft position within the first round, pro bowls are not a perfect measure of value, etc., etc. We see, though, that safeties grade out pretty well. What about early first round picks, though? Berry is going to go high in the draft, not just in the first round. Here is the same chart, except for top ten picks only, during the same time period.

So, yes, you're reading that correctly. The history of safeties in terms of longevity and greatness the top of the draft is, in fact, not very shaky at all compared to other positions--it is the opposite of shaky. Seven of the nine safeties drafted in the top ten between 1978 and 2002 made a pro bowl. Six of the nine had a career AV higher than 50 (and Bennie Blades finished at 49). The median top ten safety had a career like Eric Turner. Rickey Dixon is the only one who started less than three seasons and could be considered a true bust.

Only nine were drafted in this time period, so we have a small sample size. GM's don't want to draft safeties early. They want to take the next Bruce Smith at defensive end, which is why we see that position rank second-highest in terms of most picks in the first round, but lowest in terms of pro bowl percentage. That's the position where they are willing to reach and to view with rose-colored glasses. They don't reach for safeties. You have to be elite to overcome the bias at safety against using the early pick, much like what we may be seeing with Berry. Contrary to King, I think history (granted, a very small number from history) shows that Berry is a relatively safe pick at the top of the draft to turn out to be at least a pretty good starter during his rookie contract.