32 Counter Pass
Footballguy
How does he define "open"? How does he know if the open WR isn't a result of a broken coverage by the DBs?
What he says in his post is that open mean "creating separation, enough for the quarterback to have a reasonable target." Plays where the "receiver is obviously held or interfered with (called or not)" also count as open. If there is broken coverage then that still counts as getting open.How does he define "open"? How does he know if the open WR isn't a result of a broken coverage by the DBs?
RE the edit: your genuine interest comes through pretty clearly and does not come across as an attack. I hope that you do have some doubts about these ratings, because I know that I do.How much weight are you giving route diversity and how much of a NFL success predictor is it?
What are you using for Coleman's 40 and how much of a swing is it if he runs a 4.29 vs a 4.49?
Edit: I think I've posted enough to not have to preference a post like this but just in case... My questions aren't an attack or doubt in anyway. It's either that I'm genuinely interested in your process or that I'm interested in you thoughts/numbers on a player. Keep up the good work.
PFF's top 100 leaderboard includes 17 WRs - the 14 from their earlier list plus Arizona State's Devin Lucien (at 48th overall and WR9), Keyarris Garrett (at 72nd overall and WR14), and UCLA's Jordan Payton (at 91st overall and WR16).Just saw this from Pro Football Focus with their top 14 WRs. Their rankings look to be closer to mine than to the conventional wisdom (e.g., relatively high on Carroo & Mike Thomas, relatively low on Treadwell & Michael Thomas). They put a lot of weight on college performance, just as I do, though they do it by watching tons of tape and grading receivers on each play instead of by crunching numbers from season totals.
The biggest differences between their rankings and mine: They have Daniel Braverman as their WR8 (he is not on my leaderboard), and they don't have Keyarris Garrett, Roger Lewis, Jakeem Grant, or Bryce Treggs on their leaderboard. They are higher than I am on Treadwell (their WR3), and lower on Carroo (their WR4) and especially Fuller (their WR10).
Braverman's production at Western Michigan was slightly below the baseline level that I want to see from prospects, similar to Keyarris Garrett's at Tulsa. Unlike Garrett, Braverman is undersized (though I guess we still don't have accurate weighin information for him) and primarily a slot receiver, which is typically not good for fantasy value (outside of certain situations like New England). We'll see what numbers come out of his pro day this Tuesday. For now I'll stick him on the bottom of my list of WRs worth watching, making extending it to 18.
PFF's writeup sprinkled in a few stats which I have used to update my spreadsheet. The biggest surprise is that Rashard Higgins had a better than average drop rate, rather than worse-than-average as I thought (based on this article on his 2013-14 stats). Higgins didn't have great measureables at the combine, in terms of size or athleticism, but he put up huge numbers in 2014, he gets open according to Harmon's data, he runs good routes according to Roster Watch's grading at the combine, and he apparently catches the ball reliably. I'm thinking of putting another tier break after him in my subjective rankings, separating the top 8 from everyone else.
Could rounding be the cause of Jared Goff's hand size change?Test-Retest Reliability (aka Whose hands grew the most over the past month?)
The Senior Bowl, Shrine Game, and Collegiate Bowl ,,, opportunity to measure test-retest reliability - how consistent are the two sets of measurements with each other?
QB Brandon Allen, ,,, 8.5" hands. At the Combine, ,,, 8.875" hands. He ,, gained 0.375" of hand size. ... Did Allen get an unusually large increase in hand (and arm?) size from his work with a masseuse?
On average, the 80 players gained 0.27" of hand size. So the growth of Allen's hands was far from unique - it was actually fairly close to the typical result.
Those numbers are just the average change - some players had an increase in hand size, some had a decrease. We can find the typical size of the change by taking the standard deviation of the change in hand size, and comparing it to the standard deviation of the players' hand sizes as measured at the Combine. Players' weights by 0.09 standard deviations, height by 0.13 standard deviations (though that is inflated by rounding), arm length by 0.31 standard deviations, and hand size by 0.55 standard deviations. Another way to put it: the correlation between the Senior Bowl (etc.) measure and the Combine measure is 0.85 for hand size.
3
A change of 1/8 of an inch is actually an unusually small change between two measurements - there apparently is enough measurement error for the typical change to be larger than that.Could rounding be the cause of Jared Goff's hand size change?
Good question.ZWK said:A change of 1/8 of an inch is actually an unusually small change between two measurements - there apparently is enough measurement error for the typical change to be larger than that.
Hand size is measured to the nearest 1/8 inch. So we know that Goff's hand was measured at 9 0/8 inches at the Combine and at 9 1/8 inches at his pro day.
My current guess is that the Combine hand & arm measurements aren't any more accurate than the measurements at the Senior Bowl and postseason events. (I'm less sure about pro days - who does the measurements there?) The reason for that guess is that players' heights are correlated with their Senior Bowl (etc.) arm length measurements, even after controlling for their Combine arm length measurement. That is what you'd expect to see if the Combine and Senior Bowl (etc.) measurements were all fairly noisy. If the Combine measurements were really accurate (and the only reason that they didn't match the Senior Bowl measurements was because the Senior Bowl measurements were noisy) then you would expect to see that height was uncorrelated with Senior Bowl arm length measurements after control for Combine arm length measurements.
1
Nice listMatt Harmon's big Reception Perception analysis of this year's draft class came out last week, with a bunch of great stats that he compiled from his game charting. He has 21 WRs in his data set, including all of the big names and all of my top 10 (including Roger Lewis and USM's Mike Thomas). On average, he charted 175 routes run for each of the receivers, using the all-22 tape.
The most interesting stat that he tracks, which you can't get anywhere else, is success rate. "Success" here means getting open, and a receiver's "success rate" is the fraction of his routes on which he got open. Harmon breaks down success rate by coverage (man vs. zone, with further breakdowns into press coverage and double teams) and by type of route (screen, slant, curl, nine, corner, ...). You can see his Success Rate Versus Coverage (SRVC) and SRVC By Route tables in his post to look at questions like who struggles to get open versus press coverage, or who does the best job of getting open on nine routes (though you do want to be careful about small sample sizes with some of these).
I crunched some numbers to get an overall Adjusted Success Rate for each WR. This is just the receiver's success rate (the fraction of routes that he got open on) adjusted up if he ran more routes that are hard to get open on (e.g., double coverage, nine routes, corner routes) and adjusted down if he ran more routes that are easy to get open on (e.g., zone coverage, screens). Here are how the 21 WRs came out by that metric:
81.4% Sterling Shepard
79.5% Josh Doctson
77.3% Rashard Higgins
75.6% Corey Coleman
74.3% Laquon Treadwell
74.2% Michael Thomas
73.5% Malcolm Mitchell
72.9% Mike Thomas
69.9% Leonte Carroo
69.6% Kenny Lawler
(69.2% average)
69.1% Keyarris Garrett
67.4% Will Fuller
67.0% Braxton Miller
67.0% Demarcus Robinson
65.4% Tajae Sharpe
62.6% Pharoh Cooper
62.3% De'Runnya Wilson
62.0% Roger Lewis
61.1% Charone Peake
59.7% Aaron Burbidge
58.7% Tyler Boyd
Interesting that the top 6 contains 3 of my top 5 WRs (Coleman, Doctson, Shepard), 2 highly-touted WRs who my numbers have been down on (Laquon Treadwell & OSU's Michael Thomas), and Rashard Higgins. The other 2 WRs in my top 5 (Carroo & Fuller) both come in averageish.
I am planning to incorporate this in my WR ratings, along with some of Harmon's other stats like contested catch rate.
There haven't been many slot receivers with significant fantasy value. Welker, Edelman, Cobb, Landry. Probably not a coincidence that 3 of those 4 have had a HOF QB. I'm probably forgetting some others. Colston, I guess, although he was a very different type of WR. So if a WR looks like he's only a slot guy, I typically downgrade him and put more weight on which team he ends up on.I heard Braverman on a podcast. He believes that he is a slot guy and also believes that he has mastered the college level slot position.
Have you watched him at all? Would the fact that he's strictly a slot guy change anything for his metrics? I'm guessing that his numbers may be better when stacked against other slot only players. Or it may mean more weight is put on certain numbers?
Keep up the excellent work.
My impression watching the LSU game was that there were a lot of carries where Jackson went down relatively easily, and someone like Alex Collins or Jordan Howard could've picked up an extra yard or two. Since Jackson weighs 220, I was hoping to see better. But I do feel like I don't have as much data as I'd like; I haven't found other game videos of his.On Darius Jackson, Eastern Michigan @ LSU is just about as big a mismatch as you could create between two Division I opponents. A lot of questions about the guy (obviously), but that LSU game probably shouldn't be one of them.
FWIW, in terms of build, athleticism and NCAA efficiency he's an elite prospect -- along the lines of Murray, Matthews and Addai, but with more raw athleticism.
The questions are about volume/total production and draft position (which do matter a lot independent of the other stuff). But, like CJ Anderson (not quite elite, but a very good profile), sometimes the guys who check the first three boxes pan out despite one year of production and a low draft pedigree.
My numbers aren't that high on Hooper & McGee, although my numbers aren't that good at picking out TEs. I like Hooper more than McGee. Significantly better workout numbers. I'm tempted to put Higbee as the #2 TE in this class.Hooper is creeping up boards as he should. But I agree that Higbee and Mcgee are flying under the radar, I believe after Hooper those are the two guys to watch.
The experts have him as a borderline 2nd/3rd round pick, on average (according to massraider's spreadsheet from the Tracking the Outliers thread). If he winds up actually being drafted there then he'll move up a bit in my rankings. I put less weight on predicted draft position (relative to my own analysis) than I do on actual draft position.Good stuff ZWK.
In regards to Booker, everything I have heard is that he is still expected to be a 2nd round pick. Is that expectation baked in to your above ranking? Or are you expecting him to be drafted later than that? April 12 in when he will work out IIRC. So perhaps some more information becomes available then.
There isn't one. But if you search for his name, you'll see that it appears 21 times and the median one (11th of 21) is in row 70 of the spreadsheet, which is pick 68. It's similar if you only look at the ones since March 15.Where is the consensus list on massraiders sheet? Is this on page two? That is the only possible combined list I see. On page two Booker is ranked 29th overall.
Nfldraftscout estimated Howard at a 4.57 at the start of the offseason, so that's what I have him at. This article has him at a 4.50 at his pro day, and those sorts of reports tend to err in the direction of a faster time, so that's roughly consistent with the nfldraftscout estimate. He's a downhill power back, good at picking up yards after contact. Has a shot to be a solid starter. Hasn't done much in the passing game. Would be ahead of Prosise in non-ppr.I may have already asked but what's your opinion on Jordan Howard. If you exclude the 2 QBs he's at 13. Did you end up get a 40 time for him? I saw something from April 2015 that says he ran a 4.39
http://www.sportingnews.com/ncaa-football-news/4640997-jordan-howard-injury-uab-transfer-indiana-hoosiers-running-back-freeuab-kevin-wilson-report-big-ten
I guess that was Faust. Core is not in my spreadsheet because he didn't make the top 100 in receiving yards. That's not a good sign. Looking up his numbers, he doesn't have a very promising profile. Tiny market share (15% of rec yds, 11% of rec TDs). 9 25+ yard receptions is below average. One bright spot in terms of production is a pretty good YPT at 10.1. His 2014 stats were a bit worse, on the whole (16% of rec yds, 25% of rec TDs, 4 25+ yd rec, 9.2 YPT). Average athleticism (better than average 40, but bad vertical). Tall (6'2.6"), but on the thin side (25.8 BMI) and with relatively short arms (0.8" shorter than expected for his height).I not sure if it was you or Faust that posted about Tony Pauline liking Ole Miss WR Cody Core. Either way he didn't make your list but I was wondering if you any numbers/opinions on him.
Western Kentucky TE Tyler Higbee is still recovering from his knee injury so he didn't do any drills besides the bench at his pro day. But PFF is pretty glowing about him, as the #2 receiving TE in the draft class. So I'm pretty excited about him as a sleeper TE (which is a valuable sort of prospect in some league types).
PutzZach Greenwell@zach_greenwell
Former#WKU TE Tyler Higbee was arrested Sunday in BG and charged with 2nd-degree assault, 2nd-degree evading police, alcohol intoxication.
Darius Jackson's game against Toledo is now also online. I had a similar impression that he wasn't using his size. He seems kind of hesitant in traffic, instead of driving forward with power to pick up a few extra yards (or wiggling forward, as some RBs do).My impression watching the LSU game was that there were a lot of carries where [Darius] Jackson went down relatively easily, and someone like Alex Collins or Jordan Howard could've picked up an extra yard or two. Since Jackson weighs 220, I was hoping to see better. But I do feel like I don't have as much data as I'd like; I haven't found other game videos of his.
I think my numbers basically agree with yours. I think is NCAA efficiency is only averageish, not elite. But averageish production plus his build & athleticism makes for a very promising prospect.
Bump Hooper again, jesus it never ceases to amaze me this close to draft day.Putz
Not sure if this is the type of thing worth trying to take into account in future years, but in addition to Core and Treadwell, Ole Miss had a pair of WRs who are relatively highly regarded NFL prospects for next year (Adebojeyo and Stringfellow). They also have a future NFL TE who is a tweener in the mold of Jordan Reed (Engram) who is arguably the top prospect at the position for next year. Beyond those top 5 pass catchers, even 5th WR Markell Pack and backup DaMarkus Lodge are also potential NFL guys.I guess that was Faust. Core is not in my spreadsheet because he didn't make the top 100 in receiving yards. That's not a good sign. Looking up his numbers, he doesn't have a very promising profile. Tiny market share (15% of rec yds, 11% of rec TDs). 9 25+ yard receptions is below average. One bright spot in terms of production is a pretty good YPT at 10.1. His 2014 stats were a bit worse, on the whole (16% of rec yds, 25% of rec TDs, 4 25+ yd rec, 9.2 YPT). Average athleticism (better than average 40, but bad vertical). Tall (6'2.6"), but on the thin side (25.8 BMI) and with relatively short arms (0.8" shorter than expected for his height).
There is something to that. In an ideal setup, I would have a way to adjust for that. I don't see an easy way to incorporate it into my formula, though.Not sure if this is the type of thing worth trying to take into account in future years, but in addition to Core and Treadwell, Ole Miss had a pair of WRs who are relatively highly regarded NFL prospects for next year (Adebojeyo and Stringfellow). They also have a future NFL TE who is a tweener in the mold of Jordan Reed (Engram) who is arguably the top prospect at the position for next year. Beyond those top 5 pass catchers, even 5th WR Markell Pack and backup DaMarkus Lodge are also potential NFL guys.
Mentioned this before, but my big hangup on market share is that it doesn't account for how talented the other players on the same roster are. Ole Miss had a really impressive stable of pass catchers last year and I wonder if that skews some of these numbers and causes a guy like Core to look worse than he would if he played as a lesser program without any other NFL caliber receiving talent.
Thanks!Does drop rate/target info go back a few years? Interested in AB, ODB, Dez, Juilo to see their numbers in comparison.
BTW...interesting reads. You are definitely working...thank you.
I didn't include targets in those spreadsheets. I do have them somewhere (along with various other stats which went into creating those spreadsheets). In fact you could calculate target totals based on games, yds/g, and yds/tg (which are all in the spreadsheet). Also, drop rate is not calculated based on targets - the formula is drops/(rec + drops).How did I miss that spreadsheet. Whoa dude.
Anyway, I see drop rate stuff but am I missing the column with target totals? I see yards per target, but not target total.
This is fantastic. Thanks for sharing.Thanks!
Data on college drop rate is scattershot, and the earliest that I have is from the 2012 college season. Beckham had a 6.5% drop rate (according to Greg Peshek) which is pretty good; I don't know about the others. Average is about 7.4% for the prospects who are good enough to have their stats reported. Some other players are in my WR stats spreadsheet. The only WRs who I have at 10% or higher before this year are Justin Hunter, Markus Wheaton, Marqise Lee, Martavis Bryant, Sammie Coates, Breshad Perriman, and Vince Mayle. Apparently the Steelers don't mind a high drop rate.