AlNFL19 Posted June 19, 2019 Share Posted June 19, 2019 (edited) I've taken a break but here's another statistical take on this year's rookie class. If you want an explanation, look through one of the other threads linked at the bottom of the OP. I'm too lazy to type it again. What goes into the model: Volume stats like Tackles For Loss and Sacks Efficiency stats like Sacks per Defensive Attempt Combine Testing It provides a projection in terms of average AV in years 3-4 based on a data set from 2006-2015, as well as a % chance to bust (<5.0 AV) or reach Pro Bowl (10+ AV) status. Probabilities of Hitting Thresholds (2006-2015) Projection: >5.00 AV Threshold Reality Projected Avg. Pro Bowl 23.53% 24.48% Hit 82.35% 80.16% Bust 17.65% 19.84% Projection: <5.00 AV Threshold Reality Projected Avg. Pro Bowl 4.17% 3.94% Hit 31.94% 32.46% Bust 68.06% 67.54% Here's the 2019 class rankings: 1. QUINNEN WILLIAMS, ALABAMA Q. Williams, DT, Alabama Statistic Figure Projected AV 6.23 Bust Chance <1.00% PB Chance 33.23% No surprise here. Williams was the best defensive player in the country all year and put up freak numbers at the combine. Williams' projection is third all-time to Aaron Donald and Ndamukong Suh (not bad company). 2. ED OLIVER, HOUSTON E. Oliver, DT, Houston Statistic Figure Projected AV 5.52 Bust Chance 21.15% PB Chance 23.91% Oliver's 2018 season fell short of the astronomical expectations many had of him, but his production was good enough that it spat out a good number when combined with his freak athletic traits, like a 4.73 40 time. 3. JEFFERY SIMMONS, MISSISSIPPI STATE J. Simmons, DT, Mississippi St. Statistic Figure Projected AV 5.01 Bust Chance 36.69% PB Chance 17.22% If I were a Titans fan, I would be thrilled with this projection, and here's why: it's a pretty conservative estimate. Simmons' ACL injury meant he didn't complete much in the way of athletic testing, so the model takes a very conservative path in that regard. Despite that, Simmons' projection is still high. 4. CHRISTIAN WILKINS, CLEMSON C. Wilkins, DT, Clemson Statistic Figure Projected AV 4.99 Bust Chance 37.30% PB Chance 16.96% Wilkins' projection is below the 5.0 hit mark, but he's got a high likelihood of success regardless (62.70% hit chance). His projection is the highest of any since 2006 to fall below the 5.0 mark, at 4.985, to be exact - right between that of Corey Liuget and Trevor Laws. 5. JERRY TILLERY, NOTRE DAME J. Tillery, DT, Notre Dame Statistic Figure Projected AV 4.66 Bust Chance 47.35% PB Chance 12.63% Tillery put up pretty good numbers all around, good enough to have a more-than-half hit chance, including a second-in-the-class 8.0 sacks, tying Quinnen Williams. Tillery's projection suffers, however, from playing for a Notre Dame defense that faced 37.3 pass attempts per game, knocking down his efficiency statistics. 6. DEXTER LAWRENCE, CLEMSON D. Lawrence, DT, Clemson Statistic Figure Projected AV 4.21 Bust Chance 61.06% PB Chance 6.73% Lawrence, serving as more of a traditional nose tackle, didn't rack up the stats at Clemson to give him a great projection. However, he did have some stellar athletic testing. Lawrence's 5.05 40-yard dash doesn't jump off the page in terms of time, but at 342 pounds, it's a very impressive mark. 7. DRE'MONT JONES, OHIO STATE D. Jones, DT, Ohio State Statistic Figure Projected AV 4.04 Bust Chance 66.24% PB Chance 4.50% Viewed as a potential Day 1 pick coming into the season, Jones' stock fell before he declared for the draft despite putting up a rankings-best 8.5 sacks in 2018. Jones wasn't a good performer at the combine, clocking in a 5.12 40 despite weighing just 281 pounds, far and away the lightest of the group. 8. TRYSTEN HILL, UCF T. Hill, DT, UCF Statistic Figure Projected AV 3.97 Bust Chance 68.37% PB Chance 3.58% Hill was drafted above Jones, but this model thinks he shouldn't have been. Hill's production wasn't very high at UCF (36 total tackles, 3 sacks in his final season), and his athletic testing wasn't really anything special. *There is no projection for Khalen Saunders because he played in the FCS, with which the model isn't really compatible because of the obvious discrepancies statistically between it and the FBS. If you want a more in-depth explanation on how this kind of model works, or are interested in similar things, check out this thread, or the other ones named "(Position) Statistical Projections", which are linked in the thread below: EDIT: I've decided to add the Offensive Tackle projections to this thread because I don't think it merits its own due to the lack of real on-field stats at the position. The offensive tackle projections include: Draft Position Combine Testing Here's the 2019 class projections, without explanations because there isn't much to explain. The model counts Draft Position, 40 Time, 10-Yard Split, and Vertical Jump highly. 1. ANDRE DILLARD, WASHINGTON STATE A. Dillard, OT, Washington State Statistic Figure Projected AV 6.04 Bust Chance 27.42% PB Chance 22.33% 2. TYTUS HOWARD, ALABAMA STATE T. Howard, OT, Alabama State Statistic Figure Projected AV 5.86 Bust Chance 31.58% PB Chance 19.11% 3. KALEB MCGARY, WASHINGTON K. McGary, OT, Washington Statistic Figure Projected AV 5.84 Bust Chance 32.04% PB Chance 18.75% 4. JONAH WILLIAMS, ALABAMA J. Williams, OT, Alabama Statistic Figure Projected AV 5.83 Bust Chance 32.27% PB Chance 18.57% 5. CODY FORD, OKLAHOMA C. Ford, OT, Oklahoma Statistic Figure Projected AV 5.21 Bust Chance 46.61% PB Chance 7.49% 6. GREG LITTLE, OLE MISS G. Little, OT, Mississippi Statistic Figure Projected AV 5.13 Bust Chance 48.46% PB Chance 6.06% 7. DALTON RISNER, KANSAS STATE D. Risner, OT, Kansas State Statistic Figure Projected AV 5.09 Bust Chance 49.38% PB Chance 5.34% 8. MAX SCHARPING, NORTHERN ILLINOIS M. Scharping, OT, NIU Statistic Figure Projected AV 4.97 Bust Chance 52.15% PB Chance 3.20% 9. JAWAAN TAYLOR, FLORIDA J. Taylor, OT, Florida Statistic Figure Projected AV 4.92 Bust Chance 53.31% PB Chance 2.30% 10. TREY PIPKINS, SIOUX FALLS T. Pipkins, OT, Sioux Falls Statistic Figure Projected AV 4.89 Bust Chance 54.00% PB Chance 1.77% 11. CHUMA EDOGA, USC C. Edoga, OT, USC Statistic Figure Projected AV 4.46 Bust Chance 63.94% PB Chance <1.00% 12. BOBBY EVANS, OKLAHOMA B. Evans, OT, Oklahoma Statistic Figure Projected AV 4.44 Bust Chance 64.41% PB Chance <1.00% 13. YODNY CAJUSTE, WEST VIRGINIA Y. Cajuste, OT, West Virginia Statistic Figure Projected AV 4.25 Bust Chance 68.80% PB Chance <1.00% Although with Scarnecchia, Cajuste might be the most likely to outperform his projection. Or not. We'll see. EDIT AGAIN: I've decided to add the Interior Offensive Line projections to this thread as well because I don't think it merits its own due to the lack of real on-field stats at the position. The IOL projections include: Draft Position Combine Testing The most important combine tests for the position by the model include the 3-cone drill, short shuttle, and bench press. Here's the 2019 class projections, without explanations because there isn't much to explain. 1. GARRETT BRADBURY, C, N.C. STATE G. Bradbury, C, N.C. State Statistic Figure Projected AV 6.1 Bust Chance 6.12% PB Chance 25.92% 2. CHRIS LINDSTROM, G, BOSTON COLLEGE C. Lindstrom, G, Boston College Statistic Figure Projected AV 6.08 Bust Chance 6.76% PB Chance 25.58% 3. ELGTON JENKINS, C, MISSISSIPPI STATE E. Jenkins, C, Mississippi State Statistic Figure Projected AV 5.44 Bust Chance 27.25% PB Chance 14.63% 4. ERIK MCCOY, C, TEXAS A&M E. McCoy, C, Texas A&M Statistic Figure Projected AV 4.88 Bust Chance 45.18% PB Chance 5.04% 5. CONNOR MCGOVERN, G, PENN STATE C. McGovern, G, Penn State Statistic Figure Projected AV 4.72 Bust Chance 50.30% PB Chance 2.30% 6. MICHAEL DEITER, G, WISCONSIN M. Deiter, G, Wisconsin Statistic Figure Projected AV 4.67 Bust Chance 51.90% PB Chance 1.45% 7. NATE DAVIS, G, CHARLOTTE N. Davis, G, Charlotte Statistic Figure Projected AV 4.36 Bust Chance 61.82% PB Chance <1.00% 8. HJALTE FROHOLDT, G, ARKANSAS* H. Froholdt, G, Arkansas Statistic Figure Projected AV 4.21 Bust Chance 66.63% PB Chance <1.00% Technically, as a fourth-round pick, Froholdt is ineligible for the model (all of them are built for and on data from Rounds 1-3 draft picks). However, since it was brought up already, I decided to include what his projection would be. Edited July 2, 2019 by AlNFL19 Added Interior Offensive Line Projections Quote Link to comment Share on other sites More sharing options...
Daniel Posted June 19, 2019 Share Posted June 19, 2019 Simmons projecting that high is surprising. I love Simmons, but based on my understanding of your model, I was expecting Wilkins and Oliver to be ahead of him. Quote Link to comment Share on other sites More sharing options...
Danger Posted June 20, 2019 Share Posted June 20, 2019 20 hours ago, Daniel said: Simmons projecting that high is surprising. I love Simmons, but based on my understanding of your model, I was expecting Wilkins and Oliver to be ahead of him. Simmons probably would have been a top 10 pick if healthy. Quote Link to comment Share on other sites More sharing options...
Daniel Posted June 20, 2019 Share Posted June 20, 2019 23 minutes ago, Danger said: Simmons probably would have been a top 10 pick if healthy. You think he'd get taken before Oliver? I had him fourth in the class myself because I was nuts for Christian Wilkins and had him as pretty much equal with Ed Oliver. Quote Link to comment Share on other sites More sharing options...
Danger Posted June 20, 2019 Share Posted June 20, 2019 1 hour ago, Daniel said: You think he'd get taken before Oliver? I had him fourth in the class myself because I was nuts for Christian Wilkins and had him as pretty much equal with Ed Oliver. True. Oliver and Simmons are both top 10 talents, hell Oliver (alongside Montez Sweat) was the steal of the draft IMO. Quote Link to comment Share on other sites More sharing options...
Awsi Dooger Posted June 21, 2019 Share Posted June 21, 2019 Very close to how they came off the board, certainly tighter than other positions especially considering where Simmons would have gone minus the injury. Tillery is somewhat of exception by going ahead of Lawrence. But even that is a gap of 10 picks or so, hardly outlandish. Very good point that Tillery's production number takes a ding based on how many passing attempts Notre Dame faced. Pass defense is my favorite category so I check the related numbers all season. The national champion is almost always within the Top 10 in yards allowed per pass attempt. In fact, you can basically identify the playoff frauds whenever the committee is stupid enough to allow a weak pass defense into the final four. None of the teams who have entered the playoffs allowing 7+ yards per attempt have gotten beyond the first game and some of the results have been ugly. Off the top of my head the examples were Florida State blasted by Oregon, Michigan State embarrassed by Alabama, Oklahoma giving way late to Georgia, and then Oklahoma again falling to Alabama. If I were in that committee room I would be talking about pass defense and nothing else. I would have no trouble laughing at any fellow committee member who was ignoring the significance of pass defense. As a Canes fan one of the tipoffs that our 1983 team was legitimate and could upset Nebraska was that Miami led the nation in yards allowed per pass attempt at 4.7. The beauty of the pass defense category is that it sounds like a pass only category but it also aligns with overall aggressiveness on defense. You cannot hold opponents to low yards per attempt unless you attack the line of scrimmage and take away the cutesy stuff, whether it is pass or run. Clemson was weaker than typical for a national champ at 6.4 YPA allowed. But lots of that was one game, against South Carolina. BTW, Michigan has been at the top or right near the top in YPA allowed during every one of Harbaugh's seasons. He knows what he is doing. It is just a matter of time. One of the great litmus tests right now is opinion of Jim Harbaugh. Anyone who denounces him is the typical simplistic fan who allows the big picture to fly comfortably above his head. Michigan would be moronic to overreact and get rid of him. Sorry for the detour. I can tend to detour. I remember that Houston was also very high all season in pass attempts faced, which likewise cuts into Oliver's efficiency numbers. Houston actually faced more attempts per game than Notre Dame entering the bowl season. But when you draw Army as bowl opponent obviously the pass attempts faced per game average is going to drop. 1 Quote Link to comment Share on other sites More sharing options...
Duffman57 Posted June 22, 2019 Share Posted June 22, 2019 I'm confused how this model doesn't think that Hill's athletic numbers are anything special..... He's in the 95th and 97th percentile of defensive tackles for broad and vertical jumps, and then is in the 90th percentile for the short shuttle, which are probably the 3 most applicable drills for DL. Here's his RAS..... Quote Link to comment Share on other sites More sharing options...
AlNFL19 Posted June 22, 2019 Author Share Posted June 22, 2019 1 hour ago, Duffman57 said: I'm confused how this model doesn't think that Hill's athletic numbers are anything special..... He's in the 95th and 97th percentile of defensive tackles for broad and vertical jumps, and then is in the 90th percentile for the short shuttle, which are probably the 3 most applicable drills for DL. Here's his RAS..... Sorry, I should have clarified this. The broad jump and vertical jump are actually not included in the model, because they didn’t have a significant correlation to success at the NFL level in the way that other drills did. Specifically, the 40, 10-yard split, 3-cone drill, and short shuttle are all included because they correlate a lot better. The only one of those in which Hill performed very well was the shuttle, which is only included in a formula that measures it, the 10-yard split, and the 3-come together. So of what the model really looks at, his numbers weren’t special. Quote Link to comment Share on other sites More sharing options...
AlNFL19 Posted June 23, 2019 Author Share Posted June 23, 2019 Added Offensive Tackles to this thread. 7 of 11 projection models are now done (remaining: IOL, EDGE, CB, S). The eventual goal is to be able to compile a big board, regardless of position. So that's what's coming in the (hopefully near) future. Quote Link to comment Share on other sites More sharing options...
Ragnarok Posted June 23, 2019 Share Posted June 23, 2019 One guy not on that list that I think surprises is Hjalte Froholdt. He is also learning under Scar and I think is a better prospect than Yodny. Quote Link to comment Share on other sites More sharing options...
AlNFL19 Posted June 23, 2019 Author Share Posted June 23, 2019 33 minutes ago, Ragnarok said: One guy not on that list that I think surprises is Hjalte Froholdt. He is also learning under Scar and I think is a better prospect than Yodny. Sorry, should clarify two things: Froholdt is a guard. Also, these models are built for players drafted in the first three rounds because that’s where the data they’re built on is from. Quote Link to comment Share on other sites More sharing options...
Ragnarok Posted June 23, 2019 Share Posted June 23, 2019 Just now, AlNFL19 said: Sorry, should clarify two things: Froholdt is a guard. Also, these models are built for players drafted in the first three rounds because that’s where the data they’re built on is from. My bad, I saw a few guys that could play inside and out, so I wasn't sure how that was determined. Quote Link to comment Share on other sites More sharing options...
JaguarCrazy2832 Posted July 1, 2019 Share Posted July 1, 2019 Hopefully Taylor plays better than projected but I do love reading these threads. Thanks for posting them 1 Quote Link to comment Share on other sites More sharing options...
AlNFL19 Posted July 2, 2019 Author Share Posted July 2, 2019 Added Interior Offensive Line Projections to this thread. @Ragnarok I included Froholdt's projection, despite the fact that he was drafted in Round 4. His projection is very similar to that of Cajuste. 1 Quote Link to comment Share on other sites More sharing options...
Ozzy Posted July 2, 2019 Share Posted July 2, 2019 (edited) Where are you getting these offensive line numbers, like sacks given up? In terms of statistics finding OL numbers in terms of pass blocking even as simple as sacks given up is pretty hard I would say. Outside of PFF crap which I rarely ever look at much less pay for one can rarely find OL statistics of any kind. Not even in the players bio from the college program more often than not. Then again stats are not everything even for the OL, take Taylor Lewan he gave up a good amount of sacks at Michigan and struggled at times, and the guy is a three time Pro Bowler in the NFL. Edited July 2, 2019 by Ozzy Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.