ADP Value Tiers: Where Drafters Systematically Overpay (and Where to Pounce)

Average draft position is the closest thing fantasy football has to a betting line. It tells you what the market is paying for every player. Like any market, it has structural inefficiencies — places where drafters consistently misprice, year after year, even when the analytics community has been pointing it out for half a decade. Our Shark Snip projection model breaks ADP into four tiers and identifies where each one systematically overpays or underpays. The model lives on the same projection stack we expose in /tinker, so you can recompute every delta against your own assumptions before draft night.

Why ADP misprices things

Three forces push ADP off equilibrium:

1. Recency bias. Last year's top-5 RB gets drafted as if he is a guaranteed top-5 again. The hit rate is closer to 30%.
2. Name brand. A 30-year-old former star at QB gets drafted three rounds before a more productive but unfamiliar starter, every time.
3. Highlight memory. A WR who had a viral playoff catch goes a round earlier than identical players who didn't. This is real and measurable.

None of these factors show up in projections, which is why models routinely produce different rankings than ADP. The art is knowing which differences are real and which are noise. A Ja'Marr Chase or Justin Jefferson tier usually deserves its price; a name-brand RB2 being pushed up because he used to be an RB1 is a very different bet. The same kind of consensus-versus-projection gap shows up in our betting models on /picks, where the room anchors to a stale number and the model rebuilds from current usage every week.

The four ADP zones

Zone 1 — Picks 1–24 (Rounds 1–2): Mostly efficient

The top of the draft is the most carefully studied stretch in fantasy. Bad calls get punished publicly. Consensus gets very close to model projections here, and our model tends to agree with ADP within half a tier on the top 18 picks. Where the model disagrees, it is usually about RB age or injury risk that consensus has not yet repriced. Historically, players drafted top-24 hit a top-24 finish about 50% of the time, which is the best rate of any zone.

The 2026 names that anchor Zone 1 are familiar: Bijan Robinson and Jahmyr Gibbs at the top of the RB board, CeeDee Lamb and Ja'Marr Chase at the top of the WR board, Puka Nacua and Justin Jefferson rounding out the elite tier. Bijan goes around ADP 2 overall, Gibbs around ADP 4, Lamb at 5, Chase at 6, Jefferson at 7, Nacua at 8. The model agrees on the names; the order can flip half a tier depending on whether you weight pass-game involvement or pure touch volume.

Strategy: don't try to be clever. Take the projections and ADP at face value. The edge in Zone 1 is structural — securing two top-12 finishers is more valuable than chasing a Round 6 sleeper. If you pick from positions 9–12, the model usually prefers Brock Bowers or Amon-Ra St. Brown to the late-Round-1 RBs going at the same price, because the TE premium and target volume profiles are sturdier than a third-tier bell-cow.

Zone 2 — Picks 25–60 (Rounds 3–5): Where overpay lives

This is the most inefficient zone in fantasy drafts. It is where consensus chases name-brand veterans, last year's top-12 RBs who slipped slightly, and the "second WR on a great offense" archetype. Think Eagles or Dolphins WR2 pricing: the player can be excellent and still be over-drafted if the room treats him like a locked WR1. Our backtest data on Zone 2 is rough on consensus:

• Round 3 RBs hit a top-24 finish roughly 35% of the time.
• Round 4 WRs labeled "WR2 on great offense" hit WR2 finishes under 30% of the time.
• Most of the value lost goes to Zone 4 sleepers (see below) who outscore them at a fraction of the cost.

The 2026-specific traps the model is flagging in Zone 2 are the third RB on a Round-2-RB team (Tyrone Tracy or Bucky Irving second-year valuation drift), the WR3 on Houston going behind Nico Collins as if the targets cascade evenly, and any 29+ veteran whose route share dropped 6+ points in the second half of 2025. Run those names through the /desk trade-analyzer before draft night and you can usually find the same projected output in Round 6 at a third of the cost.

Strategy: This is where the model-vs-consensus delta matters most. Use the model's ROS rankings to flag every player whose projected rank is more than a full tier below their ADP and avoid them. The RB regression piece covers exactly which Zone 2 backs to fade, and the model vs consensus deep-dive walks through how to read a tier-gap correctly instead of treating every delta as a green light.

Zone 3 — Picks 61–120 (Rounds 6–10): Tier breaks dominate

The middle rounds are where tier construction beats individual player evaluation. Whether you draft Player A at pick 80 or Player B at pick 88 matters less than whether you have built around the structural needs of your league (PPR, TE premium, superflex, etc.). The model's edge in Zone 3 is mostly about tier alignment — knowing which 6–8 player groupings are roughly interchangeable and prioritizing the player whose schedule, role, or upside profile fits your team build. In full PPR, that might push a target-earning slot WR like a healthy Chris Olave above a lower-volume deep threat even if the highlight reels disagree.

Zone 3 is also the cleanest place to deploy the "second-bite TE" strategy. Brock Bowers and Trey McBride anchor the top of the TE board in Zone 1; if you punt that tier, Zone 3 is where Sam LaPorta, T.J. Hockenson, and David Njoku usually live, and the points-per-game gap between TE6 and TE12 is roughly two PPR per week — small enough that streaming wins. The TE premium framing changes the math, which is why our TE premium piece reranks the position from scratch.

Strategy: Stop trying to crown individual sleepers in Zone 3. Take the player at the top of the active tier whose ROS schedule fits your playoff window. Nine times out of ten, that beats chasing a specific name. If you cannot decide, open the /desk start-sit tool with both candidates side by side — the projection-delta visual usually settles the argument in under thirty seconds.

Zone 4 — Picks 121+ (Round 11+): Where sleepers live

This is the highest-variance zone, and also where the model produces its most consistent ADP-beating value. The reason: drafters get bored and start picking on vibes. Our model's projections here are not magic — they just consistently identify three player archetypes that consensus misses:

• Year-2 WRs with rising target shares — the breakout candidates we covered in target share vs air yards. The 2026 model has at least four Zone-4 WRs projected for a tier-jump based on second-half-of-2025 route participation alone.
• Backup RBs on teams whose starter has injury history — the contingent value math is brutal in our favor, especially for clear No. 2 backs with Kyren Williams-style role-jump paths. Pair this with the FAAB framework in our waiver guide and you have a year-long edge, not just a draft-night one.
• Streamable QBs on offenses with rising pace — see why streaming QBs beats drafting elite. The Bo Nix and Caleb Williams archetypes pay you back the Zone-2 cost of taking a QB early.

Strategy: hammer Zone 4 picks aggressively. The hit rate on individual late picks is low, but the cost is so cheap that any single hit pays back five misses. Historically, about 1 in 6 Round 12+ picks finishes inside the top-24 at their position — and those are the ones that win leagues. If you are also running DFS lineups in-season, the /dfs lineup builder on /dfs often surfaces the exact same Zone-4 archetypes a few weeks before redraft consensus catches up, which is a useful tell.

Position-specific notes

QB

Round 3–6 QB ADP is almost always a trap. The streaming math (covered in our QB streaming piece) means you can replicate elite-QB production with two Round 11–13 picks for a fraction of the draft capital. The model fades the early-QB tier in standard formats. The exception is Josh Allen and Jalen Hurts in superflex — rushing upside is genuinely scarce and the positional crunch in superflex changes the math entirely. In single-QB leagues, take Bo Nix or Drake Maye in Round 12 and use the saved capital on a Zone-3 WR.

TE

Top-3 TEs (Brock Bowers, Sam LaPorta, Trey McBride) are usually fairly priced in TE premium. Anything in the TE5–TE10 range is overpaid in our model — those tier-2 TEs get drafted on name, but the production gap to a streamed TE is small. We cover this fully in the TE premium strategy piece. In standard PPR, the entire top tier collapses — only Bowers and McBride hold a real Zone-1 grade, and even they sometimes lose to the third-RB or third-WR in your build.

RB and WR

RBs are overpaid in Zones 1–2 and underpaid in Zone 4. WRs are roughly correct in Zone 1 and underpaid in Zones 2–4. This is the structural argument for "robust RB" in high-stakes drafts and "Zero RB" in PPR — both are defensible, both exploit a real ADP inefficiency. The RB1 hit-rate backtest shows why the early-round RB tax needs to clear a higher bar. The 2026 WR pool is one of the deepest of the decade — Ja'Marr Chase, Justin Jefferson, CeeDee Lamb, Puka Nacua, Amon-Ra St. Brown, Drake London, Garrett Wilson, Nico Collins, Malik Nabers, Marvin Harrison Jr. all carry plausible WR1 cases, which means even Round 4 has a real shot at top-12 production.

How to use this with the rest of the toolkit

Reading this post in isolation is not the play. The cornerstone workflow is: open the model on the rankings page, sort by ADP delta, and pull the top fades and top values into a draft board. Then run the riskier names through the simulator on /workshop to see how their range of outcomes interacts with your roster. Finally, save your build to /desk so the in-season tools can re-rank the same players against waiver-wire pickups every Tuesday.

This is also where the trade-analyzer math piece becomes load-bearing — once the season starts, every Zone-2 fade you nailed on draft night becomes a trade target for the manager who took him. The cleaner your draft-night reasoning, the easier those mid-season trades are to negotiate.

The 2026-specific calls the model is pushing hardest

A few names the model has out of step with current ADP, just so this is not abstract:

• Brock Bowers, TE1 overall in TE premium — the route participation rate and red-zone target share both clear the Travis Kelce 2018 bar. Worth a late-Round-1 pick if your league is TE premium.
• De'Von Achane, top-12 RB — pass-catching role plus a high-tempo offense gives him the floor most Round-3 RBs lack.
• Drake London, WR1 upside — the second-half-2025 target share with the new coordinator was already at 30%, and the model treats that as the new baseline rather than the prior-year aggregate.
• Bo Nix, QB1 in Zone 4 — rushing usage plus offensive scheme push his ceiling into the Round-5 QB range at a Round-12 price.

The model refreshes these weekly. Anything that moves more than half a tier between Tuesday refreshes is auto-tagged in /tinker as a "watch this" delta, which is how you keep ADP from going stale on you mid-July.

Build your own ADP delta tool

The four-zone framework is more useful when you can see the actual delta on your own draft board. The Shark Snip workshop ships with an ADP brick — fed by current consensus from Underdog, Sleeper, and home-league snapshots — alongside the same player feature store our projection model uses. Wire ADP into one input and a model projection into the other, and the workshop returns the delta per player sorted by zone. Tune the zone thresholds to your league’s scoring (PPR moves the WR2 line two rounds earlier than half-PPR) and the resulting fade/buy list is genuinely league-specific rather than a generic ADP cheat sheet.

To go deeper on any individual player, click open this projection in the lab and the contributing features behind the delta show up in a single panel. The lab is where you can stress-test whether a Zone 2 fade actually holds — drop opportunity inputs back to the player’s career median, refit, and see whether the projection still beats ADP. If it does not, the delta was efficiency-driven and the fade is real.

If you would rather start from a finished model, public ADP delta sets live in the Shark Snip marketplace sorted by realized accuracy over the prior three seasons. Forking one is one click; once forked, the same zone thresholds and feature weights become tunable to your league. The cross-link worth bookmarking is our sibling RB regression piece — Zone 2 is where regression fades pay off most consistently, and that framework spells out exactly which backs to flag before draft day. The public leaderboards then close the loop: they show which Zone-4 sleeper bets actually paid off this season, so you can calibrate your aggression for next year’s drafts on hard data, not vibes.

Bottom line

ADP is a market, and like every market, it misprices the same things every year: name brand, recency bias, and middle-round veterans. Take Zone 1 at face value, fade Zone 2 hard, exploit tier alignment in Zone 3, and hammer Zone 4 aggressively. Do that consistently and you will outdraft your league before the regular season starts.

Open the Shark Snip draft kit for the model's full ADP-vs-projection delta sorted by zone, with the contributing features for every overvalued and undervalued name. Pair it with /tinker if you want to fork the projection and run your own assumptions before draft night.

Research note board

Use this draft-room board before moving a player up or down. It keeps projection, price, and format separate.

Educational analysis only, not a bet recommendation. Check current lines, injuries, rules, contest terms, and local regulations before acting.

DFS projected ROI vs ownership %

Projected GPP ROI multiplier vs projected ownership across simulated lineups. Sub-10% leverage plays compound when they hit; chalk plays cap your upside even when the projection is dead-on.

Prop OVER hit rate vs line distance from median

Empirical hit rate of OVER bets as the prop line moves away from the player projection median, measured in standard deviations. A line set 1sd below the median hits ~84% of the time — but books price the juice to match.