The offense version of this column is here.

My analytics only 2022 NFL Draft Positional Rankings are just that -- the results of what my model spits out at me. I have my personal RB and WR rankings after watching their film. But this process is at least interesting and at best actionable.

My model takes projected draft capital (essentially a film grade), age, college production, team strength, and athleticism to create an NFL projection. I then compare that projection to all drafted players since 2005 at that position to create prospect percentiles. The best prospects will have a 100th percentile score. Bad prospects will have 5th percentile scores. 50th percentile is average obviously. The best prospects in my model often hit the following:

• Projected to go early in the draft

• Are 20 or 21 years old with a bonus for early declares

• Played on elite college teams and were 4- or 5-star recruits

• Were very productive players in their final season

• Were early producers if they didn't go to elite football programs

• Tested well at the Combine

In this column, you'll see two primary metrics: 1) "Adjusted Production" is a combination of the most predictive on-field stats, college team strength, and age, so an older prospect who had more yards at a bad school very well could have a lower adjusted production score compared to a younger prospect on a good team. That's because the latter is more predictive of NFL success. ... 2) "Adjusted SPARQ Athleticism" is an all-encompassing workout score, but unlike other athletic measurables, this one is specific to each position. This is key because the athleticism required at CB is far different than it is at DT. I even have split RB and WR into two different models because small RBs typically are used out in space as pass-catchers and big RBs are asked to run up the middle. Ditto for small and big WRs.

In general, the most predictive stat is draft capital, followed by adjusted production and then adjusted SPARQ athleticism.

2022 DT Prospect Rankings

Jordan Davis (94th percentile) set a new defensive tackle record in adjusted SPARQ athleticism, though we didn't get agility drills from him. Who cares? Davis has 97th percentile speed with 96th percentile weight. At worst, Davis is a disruptive two-down nose tackle. At best, Davis allows defenses to play in two-high shells because he can two-gap and throw centers around the field with ease. His 24th percentile adjusted production is the primary concern here, but 2021 Georgia is one of the rare teams that had enough DL talent to force a true rotation. Anywhere in the 8th to 17th range is fair game to me in this class. Screw positional value. Davis is fun as hell.

DeMarvin Neal (85th percentile) is popping in the model with 97th percentile adjusted production and a 97th percentile three cone. He's way undersized (3rd percentile weight), but he is a disruptive player coming out of the SEC. My hot take is that he's a better prospect than Devonte Wyatt (73rd percentile), who is 24 years old with 28th percentile tackles for loss per game. Wyatt's 93rd percentile adjusted SPARQ athleticism is carrying a lot of weight right now. Undersized 21-year-old DT Perrion Winfrey (78th percentile) and elite athlete Travis Jones (68th percentile) round out the defensive tackle class early on. Everyone else is sub-50th percentile.

2022 EDGE Prospect Rankings

The edge class looks decent, especially in a class that's down in general. There will be three, maybe four, edge rushers that will go top-10.

Aidan Hutchinson (99th percentile) is the headliner after showing 95th percentile adjusted production as a senior. He checked boxes as the Combine with 76th percentile adjusted SPARQ athleticism, but Hutchinson has clear pros and cons. His three cone is in the 99th percentile. His arms are 7th percentile. His weight is 41st percentile. I think he'll be a pretty good, not elite player in the NFL. My model's projected 1st overall draft capital is likely slightly overrating him a bit because in a normal class, Hutchinson gets drafted 7th or so overall.

Like Hutchinson, Kayvon Thibodeaux (97th percentile) is a little undersized (29th percentile weight) but is an above-average athlete (75th percentile). Where they win is completely different. Thibodeaux is an explosive straight-line player (88th percentile forty), while Hutchinson wins with pass-rushing moves and more bend. The advantage in Thibodeaux's profile is that he was a 5-star recruit who balled out as a freshman and then declared as a true junior. There's more developmental upside here.

Travon Walker (97th percentile) might surprise people with this high of a projection. His tackles for loss per game is in the bottom 5th percentile among drafted edge rushers since 2005, but there's a lot of context that my model can also pick up. Georgia's defense was elite, and Walker is a 21-year-old early declare. When in doubt, find SEC early declares, especially ones with 98th percentile adjusted SPARQ athleticism, a 97th percentile forty, and a 94th percentile three cone. Walker will need time to develop, but he has the background and tools of an All Pro pass rusher. In a normal draft, Walker is the flier at 12th overall. In this weak draft, he belongs in the top-five discussion.

Jermaine Johnson (88th percentile), George Karlaftis (86th percentile), and David Ojabo (82nd overall) are the next tier. My model is slightly lower than the market on the first two; Johnson is 23.31 years old with 29th percentile weight and Karlaftis is a 62nd overall adjusted SPARQ athlete. Ojabo is a more interesting prospect. If not for a pre-draft process torn achilles, he'd be close to a 90th percentile prospect. He has 84th percentile adjusted production, is 21.9 years old, and had 81st percentile adjusted SPARQ athleticism. In a down class, I'd be open to taking Ojabo early in the 2nd round, as he still ranks ahead of the following edge rushers even with tanked projected draft capital.

Arnold Ebiketie (75th percentile) is projected well ahead of Boye Mafe (62nd percentile) despite being projected to go slightly later in the draft. His 82nd percentile TFLs dwarf Mafe's 27th percentile. My model even likes 21-year-old Drake Jackson (71st percentile prospect), the highly productive DeAngelo Malone (69th percentile), bendy Cameron Thomas (64th percentile), and high-end athlete Nik Bonitto (63rd percentile) more than Mafe.

2022 LB Prospect Rankings

There are six off-ball LBs who project between 89th-94th percentile prospects, and all six come from Power 5 schools. Leo Chenal (94th percentile) is arguably receiving the least buzz of them, but he's the prospect who pops the most analytically. He's an early-declare from Wisconsin with 90th percentile tackles per game on a good Wisconsin defense, and then he had 98th percentile adjusted SPARQ athleticism with an 85th percentile forty on 90th percentile weight. What am I missing here? He's 21.5 years old.

Nakobe Dean (93rd percentile) was a staple to one of the best college defenses of all time and is only 21.4 years old, but he only had 10th percentile tackles per game, opted not to test pre-draft, and has bottom 8th percentile weight. He might be situation specific. His teammate Quay Walker (91st percentile) has the requisite 57th percentile size and showcased 82nd percentile adjusted SPARQ athleticism. He just wasn't an early declare like Dean.

Devin Lloyd (91st percentile) is getting love from the NFL/college insiders, but there are red flags with his profile. He's 23.6 years old, only had 61st percentile tackles per game, and is a 50th percentile adjusted SPARQ athlete. Lloyd looks like a serviceable starter, but the elite LBs typically offer more as prospects. Alabama's Christian Harris (90th percentile) and Penn State's Brandon Smith (89th percentile) are more than two years younger and are better athletes. Smith in particular has 96th percentile adjusted SPARQ athleticism.

As for Day 2 depth, LSU's Damone Clark (83rd percentile) is the analytics darling if his medicals check out. Georgia's Channing Tindall (81st percentile) is undersized but has 94th percentile speed. Wyoming's Chad Muma (76th percentile) had 92nd percentile tackles per game with a 61st percentile forty. And Montana State's Troy Andersen (74th percentile) is a 100th percentile athlete at the position.

2022 CB Prospect Rankings

It's a class of two at the very top of CB rankings. Sauce Gardner (98th percentile) checks all boxes on the field and comes attached with 99th percentile height and length for the position. His speed sits at 71st percentile, leaving him with a translatable press-man skillset in the NFL. Derek Stingley Jr. (96th percentile) had generation production as a true freshman in the SEC but has battled injuries since. He showcased 62nd percentile speed, arguably playing on a bad foot, at his pro day. Stingley's biggest concerns are his sub-31 inch arms and this injury, but betting on 20.8 year old SEC prospects is generally smart.

Clemson's Andrew Booth Jr. (91st percentile), Washington's Trent McDuffie (86th percentile), and Florida's Kaiir Elam (80th percentile) are the other potential 1st rounders. Booth didn't test due to injury, but he had 78th percentile adjusted production as a 21.6 year old coming from a top-end program. McDuffie isn't as liked by my model because he's much smaller (32nd percentile height with sub-30 inch arms), had worse production (34th percentile), and played on a worse team. He must have outlier instincts to be a stud in the NFL. Elam checks more analytical boxes. He has 91st percentile height, 80th percentile speed, and hasn't turned 21 years old yet.

UTSA's Tariq Woolen (74th percentile) is a boom-bust Day 2 prospect. He is a 100th percentile adjusted SPARQ athlete because he has 100th percentile height and 99th percentile speed, but he's a 23-year-old small school prospect with sub-par agilities. Roger McCreary (68th percentile) is the opposite, as he comes from Auburn with 12th percentile adjusted SPARQ athleticism. Washington's Kyler Gordon (66th percentile) is somewhere in between. He has 59th percentile adjusted SPARQ athleticism with 55th percentile height. Gordon had 23rd percentile speed with elite lateral abilities -- a 93rd percentile three cone.

Sam Houston State's Zyon McCollum (44th percentile) is a complete wild card with 99th percentile adjusted SPARQ athleticism.

2022 S Prospect Rankings

Kyle Hamilton isn't a generation safety prospect, but he's a damn good one. The Notre Dame product is a 21-year-old early declare with 89th percentile weight. Hamilton can play in the box, cover tight ends, and has the instincts to play deep safety. His 25th percentile speed was disappointing but of all NFL positions, forty time means the least for safety prospects. This position is about reading offenses. That's hard to measure on my spreadsheets.

Michigan's Daxton Hill (92nd percentile) and Georgia's Lewis Cine (85th percentile) are the other potential first rounders, but they look way different than Hamilton at 5th and 22nd percentile weight respectively. They both play more physically than their size suggests, however, and they have 93rd and 95th percentile speed. Hill gets the slight edge in my model due to age, but both look like above-average starters.

Baylor's Jalen Pitre (72nd percentile), Maryland's Nick Cross (66th percentile), and Penn State's Jaquan Birkser (64th percentile) round out the top of the class. Of them, Cross is the youngest (20.6 years old), the most athletic (95th percentile adjusted SPARQ athleticism), and the biggest (69th percentile weight). He also nudged them in tackles per game.

The offense version of this column is here.