In 2012, Greg Gabriel postulated an interesting hypothesis that teams use historical draft data to predict how many players at a certain position will get drafted. I did this analysis in 2012 based on the draft board and draft selection of the Packers back then and thought it was a pretty interesting exercise. Basically, a quarterback (regardless of the specific player) is more likely to be drafted in the 1st round than say a kicker is. Extrapolating that further, every draft can expect to see 2-3 quarterbacks drafted in the first round and expect 0 kickers to be drafted in the 1st round. To narrow that down even further, the Packers can expect to see around 2 quarterbacks selected before pick 21 this year and hence if they were interested in drafting a quarterback, they could predict that the 3rd best quarterback will be available for them when they pick (assuming they don’t trade the pick of course).
Naturally, the Packers aren’t likely to pick a quarterback in the 1st round, but this hypothesis can be applied to any position. Below is the number of players picked at their respective positions up to the 21st pick from 2005 (the first year of Ted Thompson’s tenure as the Packers GM) to last year. Also note no punters or kickers have been picked in the top 21 selections so I’ve dropped those positions from the list.
I wouldn’t say the data is all that surprising, quarterbacks, defensive ends (i.e. pass rushers) and wide receivers are the most highly drafted players in the top 21 picks while centers, guards and tight end almost never get drafted in the 1st round. There’s also a very striking decline in the number of running backs drafted in the 1st 21 picks, with last year being only the 2nd time in 9 years that a running back wasn’t selected.
The following list is composed of the top players from their respective positions based on current rankings from CBS Sports’ NFL draft page. One of the biggest caveats is choosing which big board to go off of, I personally like CBS Sport’s because their rankings have been the closest to the actual draft compared to other large media draft rankings. Players names which are italicized are likely to have already been selected by pick 21 and players with their names in brackets meaning that position typically won’t be picked again by the 21st round (for instance, only 1 tight end has been picked higher than 21st in a single draft so the Packers would be breaking the trend a little by drafting a second tight end in the top 21 picks.
- QB: Blake Bortles, Johnny Manziel, Teddy Bridgewater, Derek Carr
- RB: Tre Mason, Carlos Hyde, Bishop Sankey
- WR: Sammy Watkins, Mike Evans, Brandon Cooks, Odell Beckham Jr.
- TE: Eric Ebron, (Jace Amaro)
- OC: Weston Richburg, (Marcus Martin)
- OG: Xavier Su’a-Filo, David Yankey, Gabe Jackson
- OT: Greg Robinson, Jake Matthews, Taylor Lewan, Zack Martin, Cyrus Kouandijo, Morgan Moses
- DT: Aaron Donald, Timmy Jerrigan, Louis Nix, RaShede Hageman
- DE: Jadeveon Clowney, Kony Ealy, Dee Ford, Scott Crichton, (Kareem Martin)
- LB: Khalil Mack, Anthony Barr, CJ Mosely, Ryan Shazier, Kyle Van Noy
- DB: Justin Gilbert, Darqueze Dennard, Hasean Clinton-Dix, Calvin Pryor, Kyle Fuller, Jason Verrett
Keep in mind there is a lot of variability in the draft; it’s highly likely that all the quarterbacks listed will be drafted before the Packers get to pick and its also likely that none of the running backs will be with the continued devaluation of the position as well as the general lack of talent this year at the position.
Overall, I would say it’s a pretty good year to draft a tight end; since 2005, only 1 tight end has been drafted in the 1st 21 picks, so presumably the Packers could pick from anyone outside of Eric Ebron. Wide receiver is also another position where the Packers will have a lot of choice, with any receiver outside of Watkins and Evans likely available for the Packers to choose from. Guard and tackle also would be favorable choices for the Packers as both have a lot of depth, which means some blue chip players should still be available. On the other side, defensive tackle and linebacker are probably two positions where most of the top tier players will already be selected by the time the Packers pick.
Either way, this is a pretty interesting experiment that shows likely what kind of players are likely to be available and not available for the Packers at pick 21. Of course, this hypothesis isn’t all that predictive; for instance, Aaron Rodgers would not made the list back in 2005 nor would Clay Matthews shown up at all.——————
Thomas Hobbes is a staff writer for Jersey Al’s AllGreenBayPackers.com.
23 thoughts on “Packing the Stats: Who can the Packers find at pick 21?”
Thomas, if you don’t knock this stuff off, the draft is going to start making sense to a lot of us.
Seriously, revealing stuff here that isolates a lot of the variables as much as anyone can before a ‘buy’ panic hits and one particular position starts flying off the boards.
Then all bets are off and maybe Ted winds up finding another AR-quality player.
I think its important to know the difference between actual variability in the draft and perceived variability in the draft.
Actual variability would be factoring that different teams have different schemes and hence will rank players different, have different opinions on financial considerations, different opinions about player character and injury history.
Perceived variability would be trying to factor in mock drafts, combine results, pro-days, interviews, general “rumors” spread by the media in general.
I think the 1st set of factors truly makes the draft hard to predict, the second set just obfuscates the draft further.
Picking at #21 but 28 names in italics?
Its based on rankings for a mock draft — these are the 28 guys they *think* *could* go by pick 21.
Which is good news for the Packers because 8 of those guys in italics will still be available when we pick.
I wonder what running the same statistical screen for, say, pick 15 would show. Those would be the guys the packers really have no chance to get (unless they trade up).
I was rather conservative with the methodology, hence why there are 28 names but only 21 spots. Keep in mind each position is independent, I’m not saying that if 3 quarterbacks are picked before 21, the chance of any other position getting picked more of less.
No OC,OG, or RB is going to be selected before 1-21. If C.J. Mosley is not there, I would trade down.
You never know, though these three positions typically have some of the lowest chances of being picked. However, guards are becoming a more popular 1st round draft pick (sort of in contrast with running backs, which is an interesting dichotomy), although this year might not have a name as big as the last couple years.
DB are listed as a block, here, but cover corners likely represent a significant majority of those selected. Part of the problem is guys who aren’t clearly identified in either category, but I would be curious to see how those numbers break down.
I agree with both statements, frankly it’s kind of hard to split players between two positions that have a lot in common. For instance, do you put a player into a position based on what he was drafted for (though the question becomes, how do we know what position they really were drafted for). Or do we place them in the position they ultimately play (though the question becomes how long do we wait and what about truly position versatile players?)
I would argue that linebackers as a whole and defensive backs as a whole run into the problem. Perhaps the worst is 3-4 OLBs and 4-3 DEs because they essentially fill the same role i.e. pass rusher but some most 3-4 OLBs are DEs in college.
One of the elements to any draft that must be considered which wasn’t included in your piece, Thomas, is the depth at each position. By that I mean if a team wants a impact player that matches their need with their 1st pick, they may be forced to pick a certain position because when they pick again they may not get an impact player at that position with their 2nd pick.
The reason I bring this up is because I think this relates to the Packers in this draft. The WR position in this draft is considered to be the deepest and many impact players can be gotten in later rounds. There are a couple of impact Safeties for 1st round picks and about half a dozen others that show signs they could develop into impact players with the right coaching over time. However, there are only a few ILBs considered to be impact players available and if the Packers don’t pick a ILB with their 1st pick when the 2nd pick comes around the ILBs available won’t be much better than what they have now. If the Packers want to upgrade the ILB position in this draft, I think they’ll have to pick either Mosley or Shazier with the 21st pick.
A lot of people are projecting Shazier outside due to his thin lower half. I think he has a lot of talent, but I would think a team running a Tampa 2 type 4-3 would have a lot higher value on him than a team running a 3-4.
Mosely is the clear cut impact MLB in this draft. Most teams in the 20s would probably love to trade down to pick up a pick at the top of the second and an extra third round pick. the first 6 picks or so of the second round have really close grades to the players in the early to late twenties.
Good analysis. I love this stuff.
I should mention that linebackers don’t differentiate between outside and inside, nor do they differentiate between 3-4 and 4-3 mainly for the reason that teams likely have wildly different opinions on who is what.
Also, its hard to say that players in a certain range have consensus similar grades since every team is different with different needs, wants and philosophies. I would more say that some teams are likely going to want to get back into the 1st round if they think their guy is still there and there are going to be teams that will want to move back and gather more ammo for later on.
“One of the elements to any draft that must be considered which wasn’t included in your piece, Thomas, is the depth at each position.”
Very likely this averages out over time.
Correct, but I think JH9 was referencing positional depth in this specific draft.
In any case, I would be surprised if TT chose a WR in the 1st. The coaches have been able to develop a lot of talent from later rounds in that position.
A real physical specimen at ILB, not so much (Desmond Bishop notwithstanding).
That is the idea, however, you are still talking about very low numbers for each position as a whole, even with 9 years of data. I didn’t want to add more years because I wanted the article to be in context of what Ted Thompson has seen plus the further you go back the more different the NFL was, I’d expect you’d see a skewing of running backs, guards and linebackers that probably would make the analysis even messier.
I think the problem you immediately face is what do you define as “depth”? Arguably, teams have different opinions on different players and one team might consider a certain position deep while another may not. For all we know, the Packers may feel like there are a ton of safeties worthy of the 21st pick but not many wide receivers. The second issue is that hopefully, by analyzing enough years, the actual depth of draft will be normalized. For instance we can say that 2005 was a perceived deep class of wide receivers with almost 2 STD more wide outs being picked than average. However, having nearly 20% of all the top picks being wide receivers is not all that common and the rest of the years will balance that out (like for instance in 2008 when no wide receivers were picked in the 1st round at all).
Lies, Damn Dirty Lies , and Stats.
I think the crux of that quote is supporting weak arguments with even weaker statistics. I’ve tried to make my argument as neutral as possible because I know this analysis is not all that precise (though I would wager its more precise than the majority of mock drafts). Either way, the draft is just too unpredictable to really figure out with any sort of accuracy, we just got to get as clear a picture as we can.
Don’t pay too much mind to the nitpickers, Thomas. As a numbers guy, I love seeing this kind of analysis. Keep it up!
I’m hoping two or three more offensive than defensive players are taken in the 1st twenty. Then we will still have a chance at some blue chip players on defense.
This only works if you feel that there are an equal number of blue chip players on offense and defense. If there are more blue chip offensive players, then having more picked before the Packers’ selection doesn’t give them any more choice of defensive players.
The draft is more predictable than most people think if you are just predicting who will be drafted in R1. Most good mocks would hit on 29-32 out of 32.
The most overlooked aspect of drafting is defensive i.e., who is drafting BEHIND you and what are they looking for. For example, say a top TE is heading towards pick 21 but you know that there are three teams behind you (next 5-10 picks) who would pick that TE with a high probability. That tells you the odds are good one will trade around you to grab him. The same be said for teams in front of you. If 3 of the 6 teams in front of you figure to have a high prob of taking a top TE, then odds are you will have to trade up to get that TE. TT has never strategically traded up in R1 to make sure he got his guy. He frequently strategically trades down i.e., he is indifferent between several players so he trades down knowing that there is a high prob he will get one of the guys he is indifferent among. Too bad he doesn’t understand that this concept is equally at play in both directions.
I’d argue that Clay Matthews fits the description of Ted Thompson strategically trading up to get the player he wants in the first round.
I guess the big caveat with your idea is that teams know exactly what players are valued by which teams. I’d argue that it’s relatively easy to predict the first couple picks but as the draft goes on, the more the draft doesn’t “go as predicted” and the harder it is to gauge what team is going to pick what player. In your example, how does Thompson really know whether the teams in front of him or behind him are interested in taking a tight end? Having a good tight end doesn’t preclude you from taking another one, nor does not having a great one significantly hinder your chances of being a good football team. Sure, Thompson might have a rough idea of who is interested in picking a tight end, but he doesn’t know with much certainly one way or another, hence he lets the draft fall the way that it does as opposed to trying to get ahead of the process.
Comments are closed.