Back before MIT introduced analytics to the sports lexicon. Before data warehousing got renamed Big Data and recognized as a tech trend. Even before AWS offered Elastic Map Reduce in the cloud… Kanick was doing number crunching on this question:
What NFL Draft first round selections are most likely to wash-out?
So that was about two years ago.
The context was that going into the 2011, the usual suspects in Cleveland media were pounding the table for drafting ‘a playmaker.’ As though it’s a position. As though the line of scrimmage just takes of itself. As though the coaching and game planning and play calling don’t have an effect on ‘plays’ being ‘made.’ Draft playmakerz. You can’t escape it if you’re a Browns fan. You may hear ‘quarterback driven league; get a QB’ more these days and while both are laughably simplistic, ‘draft playmakers’ is never completely out of most conversations.
Anyhoozles, I suspected that drafting a ‘playmaking’ WR was a dicey proposition and personally, I was fed up with the ongoing problems finding a right tackle and wanted it fixed with a high pick. So I set out to see if we could find trends from past data.
I went through all the first round picks going back to 1985* and tried to determine these things:
- What positions had the highest percentage ‘miss?’
- What positions offer the greatest ‘home run’ potential?
- Do the numbers change markedly when looking at top-10 picks versus overall first rounders?
I am defining ‘wash-out’ as follows: if a player stayed in the league less than five years, he is a wash-out. So, for example, notable bust Tommy Vardell (taken 9th overall by the Browns in 1992) is not a total wash-out because he managed to stay in the NFL eight years. I wanted to use an empirical measurement and thus avoid letting my judgement play a role. (I *did* attempt to rule-out injuries here.. e.g., Jerome McDougle, Bo Jackson, Sean Taylor = not wash-outs. I may have missed some players in doing this.)
Definition for ‘home run’ is wholly touchy-feely and can be the subject of debate. I just looked through the list and asked myself if that player was a stud in the league for an extended time-frame.
Then, as a kludgy way to measure risk/reward, I married the two values in an HR:BUST ratio. This tries to normalize, for example, the risk associated with drafting a RB against the probability of gaining a hero.
Here’s the spreadsheet. The summary page is up front; raw data for each position is found in tabbed worksheets behind it. It’s all sortable. I think I’ve shared it as editable; don’t worry about changing anything, I have a master copy.
The easiest way to work through the numbers is to go position by position. I’ll give a rating for the position then ID a notable bust and homerun. Let’s go.
Interior D-line: safe. 75 first rounders; 10 homeruns; 15 washouts.
Justin Harrell is a notable miss here because he’s Ted Thompson’s miss. The same guy who brought in Raji and maneuvered to get Matthews in 2009 had a notable whiff in 2006… just goes to show no one’s perfect.
As for homeruns… 1995 was a total bag of shit for Browns fans, wasn’t it? In addition to the thing that happened toward the end of the season, 1995 was the Kyle Brady draft which we’ve come to find out included a Belichick erection for Warren Sapp (12th) that went unsatisfied, supposedly due to one Mike Lombardi.
TE/FB: safe. 34 1sts; 3 HRs; 6 washouts.
Not surprisingly, we have a small sample set for TEs. I frankly haven’t heard of any of the washouts, but Junior Miller was taken 7th by the Falcons in 1980 and only stayed in the league for four years.
Tony Gonzalez was taken at 13 in 1997 and I think we can agree that he’s tops. The other homeruns I have are Shockey (14th) and Dallas Clark (24th).
DE: moderate risk. 112 1sts; 13 HRs, 28 washouts.
Courtney Brown anyone? That was easy.
Interesting data point coming: the Jets absolutely nailed it on 2000 with back-to-back homeruns, Shaun Ellis (12th) and John Abraham (13th). But I don’t know how to grade Becht and Pennington who the Jets also took in that first round class.
WR: moderate risk. 96 1sts; 15 HRs, 23 wash.
Charles Rogers (2nd) is tough to top but I think Rae Carruth (27th) provides a cautionary tale on the importance of the interview/background-checking process.
Some good names to choose from for homeruns. You really have to give it to Jerry Rice (19th, 1985) but on a point of personal privilege I want to note Reggie Wayne at 30th in 2001 was some good work.
Interior O-line: safest pick, highest homerun likelihood. 40 1sts; 7 HRs; 6 washouts.
Eugene Chung (13th, 1992) was a disappointment going back to those old Pats teams.
Just looking at Hutchinson, McDaniel, Faneca, Mangold, and Mankins… all taken after 17th… let’s remember these names when the prospect of drafting interior offensive linemen is disparaged.
DB: safe. 122 (most) 1sts; 17 HRs, 22 washouts.
Even though he went at 29th in 1997, I have to single out Chris Canty as a notable wash-out. Being that this was Pete Carroll’s pick and that Canty was a 5’9″ CB, I wonder if that informs Carroll’s current preference for taller DBs. (I know it does mine.)
Some good names here, but I’d point out that taking a safety can work out: see Polamalu (16th, 2003), Atwater (20th, 1989), Reed (24th, 2002).
/Raises hand in favor of Steve Atwater for HOF. Guy was a beast.
RB: riskiest by far. 93 1sts; 18 HRs, 31 washouts.
Avert your eyes Browns fans. For every Barry Sanders (3rd, 1989) there’s an Alonzo Highsmith (3rd, 1987). I’ll call Tim Biakabutuka (8th, 1996) the worst pick.. seemed dicey even at the time.
As for homeruns, well, you know I won’t pick Emmitt Smith. Steven Jackson (24th, 2004) is the big hitter in the group.
LB: safe, high HR potential. 84 1sts; 14 HRs; 15 wash.
Probably the best bang for your buck is at linebacker. The main thing is to avoid taking Dukies too high: Mike Junkin (5th, 1987). Errr.. yes: Mike Lombardi “.. joined team as Pro Personnel Director in 1987.” Honorable mention: Craig Powell (30th, 1995)… errr… Lombardi was “.. promoted to Director of Player Personnel in 1992..” and was in that position to 1996.
In the caption at the head of this post I suggest Mike Junkin might be the worst pick of anyone since 1985 (my dataset). Here’s why I say that:
- Our data establishes linebacker to be a safe pick with very few wash-outs. But when the linebacker is taken in the top 10, there are only TWO washouts: Trev Alberts and.. that’s right, Junkin. It’s quite an accomplishment to miss here.
- The Junkin pick was secured by trading 1983 Rookie of the Year, 4x Pro-bowler, and young Kanick’s personal favorite: Chip Banks to the Chargers.
- The trade of the pick was initiated by Art Modell.
- The selection of Junkin heralds beginning of Mike Lombardi, Pro Personnel Director.
Add it all together and the Junkin pick was a horrible pick but also points up the way the stench from old Cleveland sports problem can lives on and taunts fans. Who’d think 1987 Mike Junkin would be a current event in 2013… but there it is, Lombardi’s first pick.
I hate him just like you.. really will avoid him on ESPN as much as I can.. but you have to give it up to Ray Lewis (26th, 1996), a complete game changer.
OT: borderline safe, safest top 10 pick. 87 1sts; 14 HRs; 19 washouts.
The only name in the ‘less than five years’ category that I am familiar with is Korey Stringer and he’s not subject to a wash-out tag.
The interesting observation on OTs is that the homeruns are mainly top 10s. But what a list: Pace, Thomas, Samuels, Ogden, W. Jones, L. Brown, Gross, Roaf, Kennedy.
QB: risky. 59 1sts, most top 10s (41); 11 HRs; 16 washouts.
QB has by far the most TOP THREE washouts — JaMarcus, Couch, Leaf, Shuler, Akili, Harrington, VY. I don’t mean ‘busts.’ I mean out-of-the-league in five years WASHOUT. Those are crippling.
At the same time, QB ‘homeruns’ do more for the team than any other position. That’s why there’s more #1 overalls taken and with a list like this — E. Manning, Aikman, P. Manning, Testaverde, Elway — there will continue to be QBs taken #1.
What’s the takeaway, Kanick?
- If you can’t afford to miss in the broader first round: INTERIOR OFFENSIVE LINE.
- If you’re in the top 10 and want to plug a hole with an A-player for the next decade: OFFENSIVE TACKLE.
- Want a playmaker in the first round? LINEBACKER.
- Think you’re smarter than everyone else and want to go against decades of data? RUNNING BACK.
- Drafting #1 overall? QUARTERBACK.
* I looked at data going back to 1980 for tight ends and quarterbacks; to 1983 on interior d-linemen. On TEs, I’m sure I did it to get a larger data set. For QBs, I wanted to have the 83 draft in the dataset. Not sure what I was thinking on the NT-DTs.