The Mocktopus has returned for the 2013 NFL Draft!
Since it has been a while, here is a quick refresher. The Mocktopus is a series of models that I created to predict the probability that a team on the clock will select a given player in the NFL Draft. Included in the Mocktopus's data set is every first round pick made since 2005. Although much derided, my research suggests that the most accurate predictors of who a player will draft are other mock drafts. The Mocktopus includes several mock drafts, each which is given a weighted score based on its accuracy at predicting the results of past drafts. As an overlay, the Mocktopus also includes takes into account Top 100 value lists from draft websites. Although mock drafts are the best predictors of which prospect will go to which team, mock drafts typically overrate the role of "team need" plays in draft decision-making. Thus, providing this value "check" on mock drafts differentiates the Mocktopus from a "consensus mock" which tallies up mock drafts like votes and slots players accordingly. To take an example of how this works, last year many mock drafts had Mark Barron going to the Dallas Cowboys at number fourteen overall. The Mocktopus, however, recognized that such a valuable player was unlikely to last until number fourteen, and it's instincts proved correct--Barron was drafted seventh overall by the Tampa Bay Buccaneers.
The Mocktopus treats some position groups differently depending on historical trends. For one, the Mocktopus does not include the "value check" for quarterbacks or offensive tackles. In a world where Christian Ponder and Jake Locker are first-round picks, NFL teams tend to abandon value. A bit more surprising is that the Mocktopus has found a similar trend with offensive tackles. Time and again, drafts have shown highly regarded offensive tackles such as Michael Oher, Bryan Bulaga, and Winston Justice slide to the end of the first round, and in Justice's case, out of the first round altogether. Similarly, teams took players such as Sam Baker and Duane Brown about a round earlier than expected. The Mocktopus has a few other factors it considers such as age for quarterbacks and the big play capability of wide receivers on the active roster.
The first version of the Mocktopus was posted on Football Outsiders. You can read the original article here. This year, I am going to be posting the Mocktopus and its updates on this blog. This format will be a little more flexible in that it will allow me to post updates to the Mocktopus as mock drafts change in the final week leading up to the NFL Draft, and if there is time, maybe add a few other posts about some interesting observations about the NFL Draft that are made possible by the Mocktopus.
The first mock of the year from the Mocktopus I call the "Fan Mock." Let me explain what makes the "Fan Mock" unique among mock drafts.
I'm a fan of the Detroit Lions, so when I read a mock draft, I'm naturally most interested in who the author projects for the Detroit Lions. Sure, I might read the dialogues for the first four picks, and maybe skim the rest in a casual fashion, but I honestly care little about who Mel Kiper thinks that the Atlanta Falcons are going to select at number thirty overall.
Now, one of the fun surprises I had when I created the Mocktopus is that it often projects the same player to multiple teams. This is not a rare phenomenon: it occurs multiple times for every single draft that is included within the Mocktopus. For example, for the 2011 NFL Draft, the Mocktopus would have projected Corey Liuget to both the Patriots and the Chargers, Cameron Heyward to both the Jets and the Patriots, and Gabe Carimi to both the Chiefs and the Bears.
This phenomenon occurs because, except in a few instances, there is often only a small probability that the most likely team to draft any given player actually drafts that player. Imagine, for example, that there is a 100% chance that Geno Smith slides to the Cardinals, that the Cardinals have a 50% chance of selecting him, and have a 50% chance of drafting one of five other players. Now imagine that the Bills will have to 50% chance to draft Geno Smith if the Cardinals pass, and have no more than a 15% chance of selecting any other player whether he falls or not. If our goal is to create a mock draft that "gets the most picks right," we would project Geno Smith to both the Cardinals (50% chance) and Bills (25% chance with no other player more than a 15% chance). Indeed, if we think the Jets have a 100% chance to draft Geno Smith if he falls to them, it is likely that we would project Geno Smith to go to three teams.
Because a mock draft that was interested in maximizing its accuracy at projecting each of the 32 selections in the first round would project the same player to multiple teams, it should follow that nearly every mock draft in America should do so. I have read a number of mock drafts, however, and I can safely say that I have never seen a mock draft do this, even though, in theory, every mock draft should do this. Indeed, all mock drafts will invariably have a pick or two that are forced and are not likely selections for the subject teams. It appears that an unwritten rule of mock drafting is that a mock draft should be plausible, i.e. it is something that could happen on draft day. I'm not convinced what, if anything, internal logic adds to the enterprise of mock drafting. Given that the first round consists of thirty two selections, and that there are at least twenty players who could plausibly be selected at any one of those thirty two selections, the chances that any mock draft will later turn out to be 100% accurate is probably several times less likely than winning the lottery (a much better use, if you ask me, of such amazing, odds-defying luck). It seems that mock drafters do a (admittedly mild) disservice to the fanbases of the teams that have been shoe horned into making improbable selections in the interest of chasing a goal that is a virtual statistical impossibility.
Hence, the Mocktopus's "Fan Mock." The Fan Mock is perhaps the first mock draft ever to say to hell with internal logic and to match each and every one of the NFL's 32 teams with what it views as their most likely draftees. The Fan Mock includes many glorious duplicates. It is a mock draft for the people!
1. Kansas City Chiefs / Luke Joeckel, OT, Texas A&M
After trading for Alex Smith, the mock draft world is largely in agreement that the Kansas City Chiefs are eyeing a tackle. Most have Joeckel, but a few contrarians are suggesting Fisher.
2. Jacksonville Jaguars / Ezekiel Ansah, DE, BYU
The Jacksonville Jaguars have the chance to blow up everybody's mock draft early this year. Ansah, Dion Jordan, Geno Smith, and even Sharif Floyd have been paired with a team that nobody seems to be able to figure out on draft day.
3. Oakland Raiders / Sharif Floyd, DT, Florida
The Mocktopus sees Star Lotulelei as the stronger prospect, but the consensus sending Floyd to the Raiders is so strong it overrides the Mocktopus's typical modus opperendi.
4. Philadelphia Eagles / Geno Smith, QB, West Virginia
The quarterbacks in this class are frustratingly hard to project. There are too many teams in desperate need of a quarterback, and the draftnik community thinks that the best of the bunch, Geno Smith, has a talent level closer to a number twenty pick than a pick in the top five.
5. Detroit Lions / Dee Milliner, CB, Alabama
Millner edges Eric Fisher by the slightest of margins. The non-role that "value" historically plays at the offensive tackle position, the lack of any cornerback prospect that could plausibly go here (there is a non-zero chance that if the Lions wanted a tackle, they could go for Lane Johnson), and the appreciably higher chance of Fisher being gone by this pick pushes the Mocktopus towards Millner.
6. Cleveland Browns / Dee Milliner, CB, Alabama
Our first double! The mocks like the Browns to stay put and draft Milliner or trade down and take the less heralded Xavier Rhodes.
7. Arizona Cardinals / Star Lotulelei, DT, Utah
Star Lotulelei gets the edge over Lane Johnson because of value.
8. Buffalo Bills / Jarvis Jones, OLB, Georgia
This is an odd pick, and it is likely due to the Top 100 lists failing to catch up fast enough to Jones's falling stock. They rate Jones as the best edge pass rusher in the NFL Draft, but after Jones ran a 4.88 forty-yard dash at his Pro Day, it is unlikely that NFL teams will agree. Look for Jones to slide down the Mocktopus as the draft draws near.
9. New York Jets / Dion Jordan, OLB, Oregon
This pick would be the same whether you used the Mocktopus or a "consensus mock."
10. Tennessee Titans / Bjoern Werner, DE, Florida State
Many mocks try to stick Chance Warmack, a highly regarded player who is hard to mock to any one team, on the Titans. The Mocktopus sides with those who favor Bjoern Werner at this slot because mock drafts have historically done a poor job of pricing in the lower positional value of interior linemen, who tend to "slide" on draft day (see DeCastro, David).
11. San Diego Chargers / Lane Johnson, OT, Oklahoma
The Chargers act as almost a floor for Johnson. The mocks see the Chargers trading up for Johnson or breaking his fall if others are leery of burning a top ten pick on an offensive linemen who is not far removed from playing quarterback.
12. Miami Dolphins / Kenny Vacarro, SAF, Texas
This is where projecting this draft gets really messy. The Dolphins would like a tackle, probably, but are unlikely to get one worthy of this selection unless they trade out. As far as team needs, they could go anywhere, so the Mocktopus gives them the best player available that is not an interior lineman.
13. Tampa Bay Buccaneers / Sheldon Richardson, DT, Missouri
A lot of mocks have Star Lotulelei dropping here, but the Mocktopus deems this as somewhat unlikely, given that a player of Lotulelei's caliber is unlikely to slide quite this far.
14. Carolina Panthers / Kenny Vacarro, SAF, Texas
The Panthers could be in a great spot here to grab either Vacarro or Richardson, two good values that fill needs.
15. New Orleans Saints / Barkevious Mingo, OLB, LSU
Mingo is a consensus in the mock drafts, but he is hardly a lock here. Mingo is the favorite to go to the Saints, but only appears there in just over 10% of simulations.
16. St. Louis Rams / Tavon Austin, WR, West Virginia
After just missing out on Justin Blackmon last year, the Rams are likely to go for a wide receiver.
17. Pittsburgh Steelers / Cordarrelle Patterson, WR, Tennessee
The mocks like Jarvis Jones or a receiver to the Steelers, but the Mocktopus thinks Jones is unlikely to fall, given his value. That may changes as the Top 100 lists catch up to the mocks on draft day.
18. Dallas Cowboys / Kenny Vacarro, SAF, Texas
A triple! Vacarro or one of the guards are the favorites for the Cowboys.
19. New York Giants / D.J. Fluker, OT, Alabama
Offensive tackles are going to be a premium in this draft, and the Giants have a solid chance to grab the fourth rated offensive tackle here.
20. Chicago Bears / Johnathan Cooper, OG, North Carolina
Despite being a very highly regarded prospect, Cooper is a guard, and thus has a chance to slide here or out of the Top 20 altogether.
21. Cincinnati Bengals / Alec Ogletree, LB, Georgia
The Bengals have a storied history of betting on prospects with alleged character issues, which makes Ogletree a natural fit here.
22. St. Louis Rams / Cordarrelle Patterson, WR, Tennessee
Do you see what the Mocktopus did here? The Rams are likely to draft a wide receiver, but they could easily go with either Patterson and Austin. Because the Mocktopus's Fan Mock is more interested in correctly pairing players with their ultimate teams than creating a mock that is a plausible whole, the Mocktopus has the flexibility to use the opportunity to hedge its bet on Austin at number sixteen by matching Patterson with the Rams at twenty two.
23. Minnesota Vikings / Alec Ogletree, LB, Georgia
The Vikings are likely interested in a linebacker, and both Ogletree and Te'o could be available here.
24. Indianapolis Colts / Datone Jones, DE, UCLA
The Colts need front seven help and badly. Plenty of big bodies are likely to be available for the Colts at twenty four.
25. Minnesota Vikings / Manti Te'o, LB, Notre Dame
Similar to what it did with the Rams picks, the Mocktopus thinks that the Vikings will likely go linebacker at some point, and wants to corner both likely possibilities.
26. Green Bay Packers / Tyler Eifert, TE, Notre Dame
Tyler Eifert wins or is close to winning a number of small pluralities at the end of the draft.
27. Houston Texans / Tyler Eifert, TE, Notre Dame
The Mocktopus favored a tight end for the Texans last year as well.
28. Denver Broncos / Damontre Moore, DE, Texas A&M
At one point, the mocks had Moore as the favorite to go number two overall. His star has sufficiently fallen that he could slip into the late first or slide out altogether.
29. New England Patriots / Sylvester Williams, DT, North Carolina
The Patriots draft day intentions are famously obtuse, so projecting players to them at this point in the draft is little better than speculation.
30. Atlanta Falcons / Zach Ertz, TE, Stanford
The mocks like the Falcons to take a tight end to ultimately replace Tony Gonzalez. Eifert is much more likely to be gone by the Falcons' pick than Ertz, so Ertz is the best bet here.
31. San Francisco 49ers / Eric Reid, SAF, LSU
After losing Dashon Goldson in free agency, Reid could be an appealing choice to add to a talented 49ers roster.
32. Baltimore Ravens / Keenan Allen, WR, California
Keenan Allen is more of a value pick for the Ravens, as the linebacker group is likely to be considerably picked over by the time the Ravens go on the clock.
No comments:
Post a Comment