On the live board, bust risk sits at 53% and the realistic range is No. 75 to No. 230.
The model graded Jermaine Kelly at pick #231; he was actually picked #222 by HOU. The 4-tier model has him as a likely role.
Headline grade, verdict, and the one-line scout take
We had him No. 231 on draft night, and the league mostly agreed — HOU took him at No. 222. The seasons since are the scoreboard.
The written report — drafted from the fact sheet, every number traceable, honest about what we can't see yet
Which signals drove the grade up or down — in plain English
The analyst room, accuracy-weighted, plus where Draftanomics disagrees with it
Measurables, traits, and college play quality
Bust probability, injury history, character flags, pre-mortem
How well this prospect fits each of the 32 team schemes
Probability he's still on the board at each of your picks
Historical players this prospect resembles
Clips, scout notes, voice memos, written reports
Draftanomics had Jermaine Kelly at #231 on draft night — a Day 3 grade — and HOU took him at No. 222. The model's case leans on his recruiting pedigree. Bust risk projects at 53%, elevated against the position baseline. The room had him #160 pre-draft; our board sat 71 spots lower.
Written from the live model's numbers — projected to round 7, and it updates whenever the model does. See methodology for what drives the take.
Read the two ends and skim the middle: the model has a real track record telling stars and busts apart, but “solid starter vs. rotation guy” is close to a coin flip from public data — that's why those two bars are grayed. For his odds of outright busting, the Bust-risk panel next door is the number to trust.
The bracketed ranges show how far each number could reasonably move — they come from how 23–57 similar players in past drafts actually turned out.
1 public board track this prospect. Per-board breakdown loads with the live board.
Highlight reels indexed from YouTube. Loads only when you press play.
The grade is held down by how many boards rank him top-100, hS recruit rating, and production trajectory.
The grade is the 0-to-1 score the board sorts by — it only means something next to the rest of this class, which is why the rank above it is the number to quote.
Outcome shares from the 4-tier model. The Bust share here is a different model than the headline bust probability — for bust risk, use that number.
Solid points = actual NFL snap share per season. Dashed extension = forecast for upcoming years.
Everything that could go wrong with Jermaine Kelly, gathered in one place — the most serious flags listed first.
Named, concrete signals — every flag cites numbers, comp players, or schedule context.
Based on each team's roster depth at CB / S, age, and recent capital invested.
Closest historical matches among prior CB / S prospects who were drafted, by overall statistical profile. A smaller distance means a closer match.
| Player | Pos | Year | Pick | Snap trajectory | Career arc | Distance |
|---|---|---|---|---|---|---|
| Dexter McDonald | CB | 2015 | #242 | ↑21 pts | Contributor | 1.81 |
| Quinten Rollins | CB | 2015 | #62 | ↑3 pts | Starter | 2.06 |
| Anthony Brown | CB | 2016 | #189 | →1 pts | Starter | 2.32 |
| D'Joun Smith | CB | 2015 | #65 | ↓19 pts | Bust | 2.43 |
| Craig Mager | CB | 2015 | #83 | ↓34 pts | Starter | 2.43 |