Model Intelligence · Where upside actually lives

You Don't Have to Draft Risky to Draft Upside

Across the draftable top 150, upside and risk correlate -0.02 — essentially nothing. 26 of the 32 highest-upside players sit in the low-risk tier, and most of them are cheap.

ADY
Auto Draft Years
Aug 17, 2026 · 5 min read
-0.02
correlation between upside and risk scores
26/32
highest-upside players in the low-risk tier
16
of those priced outside the top 50
81.8
upside score, Matthew Stafford — top of the cheap list

The short version

  • Upside and risk correlate -0.02 across the draftable top 150 — essentially independent.
  • 26 of the 32 highest-upside players sit in the low-risk tier.
  • 16 of them are priced outside the top 50.
  • The cheap ceiling list is dominated by stable roles: Matthew Stafford, Dak Prescott, Tucker Kraft, Jared Goff.
  • Independence across the board doesn't make any individual player's risk vanish. Price it per player.

Fantasy managers treat risk and upside like a package deal. Want the big ceiling? Then you have to stomach the big downside. It is one of those beliefs that sounds like finance and therefore feels true.

We checked whether the trade-off exists on the 2026 board. It mostly does not. Upside and risk correlate -0.02 across the draftable top 150 — and if anything the sign points the wrong way for the risk-equals-reward crowd.

The highest-ceiling players on the board are disproportionately low risk. A lot of them are also cheap.

The relationship that isn't there

Across the draftable top 150, the model's upside score and its risk score correlate -0.02. That is not a weak relationship; that is no relationship. Knowing how volatile a player is tells you essentially nothing about how high his ceiling goes.

The clincher is the count. Of the 32 highest-upside players on the board, 26 sit in the low-risk tier. And 16 of those are priced outside the top 50. The best ceilings are not expensive and they are not volatile. They are sitting in the middle rounds being quietly unglamorous.

2026 board · most upside outside the top 50
The cheap ceiling list
Model upside score for players priced outside the top 50 overall.
Auto Draft Years 2026 consensus board. Upside and risk are model scores on a 0–100 scale.

And look at who is on the list

The names carrying the most upside outside the top 50 are not boom-or-bust dart throws. They are Matthew Stafford, Dak Prescott, Tucker Kraft, Jared Goff, Justin Herbert — established quarterbacks and ascending tight ends with stable roles and defined target volume.

That is the opposite of the profile most managers reach for when they decide they need a ceiling. Upside, it turns out, is not bought with volatility. It is bought with a good role at a discounted price — which is a much duller sentence and a much better strategy.

PlayerPositionUpside scoreRisk tierMarket rank
Matthew StaffordQB81.8Low97
Dak PrescottQB79.9Medium78
Tucker KraftTE77Low76
Jared GoffQB77Low111
Justin HerbertQB74.8Low82
Drake MayeQB74.5Low55
Jake FergusonTE74.2Low117
Trevor LawrenceQB71.7Low85

Highest model upside among players priced outside the top 50 overall.

What this means for your draft

Stop paying the volatility tax. When you want to add ceiling, you do not need to gamble on a boom-or-bust rookie or an injury-prone deep threat. You can get a higher ceiling from a low-risk player the market has discounted out of the top 50.

This is most useful at quarterback and tight end. Several of these values live there, and it is exactly where the replacement-value piece in this series already told you to wait. Waiting and getting a higher ceiling is a rare combination — take it.

Your bench can have both. The risk-equals-reward framing insists that a high floor and a high ceiling are opposites. On this board they are frequently the same player. Confirm the role and camp reports on FantasyPros before locking anyone in.

Where I would be careful

Upside and risk are the model's own bands, and a chunk of the board carries a neutral score by construction — so do not over-read small differences between two players a point apart.

Independence across the whole board also does not mean risk is irrelevant for a specific player. A back one injury away from irrelevance is genuinely risky whatever the correlation says. The finding is that you should stop assuming upside requires risk — not that risk is free.

The bottom line

Risk and upside are not a package deal. The best ceilings on the 2026 board belong to low-risk, discounted players outside the top 50 — so add upside by targeting stable value, not by paying a volatility tax you never needed to pay.

Was this useful?

Sources: FantasyPros — expert consensus rankings; Auto Draft Years — 2026 consensus board, model upside and risk scores. Board figures reflect the Auto Draft Years 2026 consensus board as generated on 2026-08-17. For informational and entertainment purposes only.