AAV rankings
Two numbers per prospect, because a second contract is two bets. The odds column is the chance he earns a second NFL deal at all, priced at draft time — about 87 percent of first-rounders do, about 22 percent of seventh-rounders. The dollar figure is the deal's average annual value — the AAV in the page title — what it pays per season if it happens, in today's cap dollars, built from every real second contract signed by the 2015–2023 classes. We show them separately rather than blending them into one expected value: nobody signs the average of $60 million and nothing. The order does multiply them, though — a $17 million projection you collect one time in five is not worth more than a $17 million projection you collect almost surely, so the rank is odds × dollars even though the columns stay separate. For classes already through the league, compare the projection with what the player actually signed — both translated to the current cap, so a 2015 deal and a 2022 deal sit on the same scale.
We don’t price this position.
Second contracts at quarterback swing so far from the model’s projection — the typical miss is wider than an entire rookie deal — that any single dollar figure would be theater. Every other surface on the site shows a dash here for the same reason, and so does this one. Pick another group above.
How these numbers are made. Two models, both built from real second contracts (as logged by Over the Cap) signed by the 2015–2023 draft classes. The odds column asks one question — does a drafted player earn a second NFL deal? — using only what is knowable on draft night: projected slot, bust risk, size, position. Tested on draft classes it never saw, it lands within a few points of the real round-by-round rates. The dollar column projects what the deal pays as a slice of that year's salary cap, then converts to today's cap — so it answers “what does the deal look like if he earns one,” not a blended expectation. Prospects who haven't played an NFL snap are projected on draft-night information alone; prospects who never cracked the consensus board get the rookie-scale figure and no odds, because the model has nothing honest to say about them. See methodology for how closely the projected odds track what actually happened.