How Draft Score Works
One number that grades how well you drafted. The 60-second version is below — read further if you want the math.
Draft Score asks two questions about your draft. Did your picks outscore the average for their slot — and did they outscore the other guys you could’ve taken at the same position?
Both questions matter. A great drafter beats expectations on both. A bad one fails on both. Most managers are somewhere in between.
Three Numbers
For every (position, draft slot) pair, we know what other same-position picks at that slot have averaged across history. A QB at #80 is graded against other QBs near #80 — not against the mixed pool that includes RBs and Ks. So a late QB who scores 18 PPG isn’t magically a steal — he’s just an average late QB.
For each pick, how far above the average drafted player at his position did he score (in standard deviations)? This naturally compresses for tightly-bunched positions like QB — where everyone scores within 5 PPG of each other — and rewards picks that crush spread positions like RB.
We average the two together so you get a single grade. 0 is league-average. Above 0 = above average. Below 0 = below.
2024 in two picks
Both of these picks happened in our 2024 draft. They’re a perfect contrast.
- Season points
- 373
- Average #92 pick
- ~134
- Beat slot by
- +239 pts
- Drafted as QB
- 10th QB taken
- Finished as
- QB1
Crushed both questions. Outscored every QB drafted ahead of him AND blew past what the #92 slot has historically returned. This is what a great pick looks like.
- Season points
- 48
- Average #1 pick
- ~130
- Beat slot by
- −82 pts
- Drafted as RB
- 1st RB taken
- Finished as
- RB40
Achilles injury, played 4 games. Both questions answer 'no.' Note: under our 4-game floor for the slot question, his per-game number doesn't crater the score — but the position rank still hurts.
What the number means
When you see a Draft Score on a page, here’s how to read it.
| Score | What that means | Example |
|---|---|---|
| +1.5 or higher | Elite draft. One of the best of the year. | Carl, career: +1.55 |
| +0.5 to +1.5 | Above-average draft. | Mitch, career: +0.83 |
| −0.5 to +0.5 | Roughly league-average draft. | Dite, career: +0.38 |
| −0.5 to −1.5 | Below-average draft. | Kyle, career: −0.66 |
| −1.5 or lower | Rough draft. The cellar. | Rob, career: −1.18 |
Why we grade position-by-position
QBs score way more fantasy points than running backs. So if we graded a late-round QB against the average pick at his slot (mixed positions), he’d look great just for being a QB — even an average late QB.
The fix: every comparison is position-specific.
- For VoE/Game, your QB at #80 is graded against other QBs picked near #80 — not against everyone picked at #80. Crushing the QB curve at that slot is the bar, and it’s actually hard.
- For PRS, beating a tightly-bunched position (where everyone scores within a narrow band) counts less than beating a spread position. You can be the best of 12 QBs and not actually have outscored your cohort by much in real terms.
Net effect: late-round QBs and Ks no longer get free credit. Drafting an elite RB or WR at the same slot — where positional scarcity is real — gets the recognition it deserves.
Injuries don't ruin your score
Picking a stud who tears his ACL Week 1 isn’t a draft mistake — it’s bad luck. So the slot value side uses points-per-game, not season totals. We also drop picks below 4 games entirely from that side of the math (too small a sample to grade).
The position-rank side still penalizes injured stars to some degree — you can argue durability is part of drafting skill. The two sides balance each other out.
The actual math
We compute two per-pick averages for each manager:
- VoE/Game per pick = sum of (player PPG − slot PPG expectation for that position) ÷ pick count. The expectation is built from same-position picks within ±10 slots, pooled across every year. Skips picks under 4 games.
- PRS per pick= sum of (player PPG − cohort mean) ÷ cohort std-dev, summed across the owner’s picks and divided by pick count. The cohort is the year’s same-position drafted players. Z-scoring within position means beating a tightly-bunched cohort (QB) counts less than beating a spread one (RB). Skips picks under 4 games.
For a given draft year (or career total), we measure how far each manager’s VoE/Game number is above or below the 8-manager average — in standard deviations from the mean. Same for PRS. Then we average the two.
The result: a number where 0 = exactly the league average that year, +1 = one standard deviation above the field (a clear standout), +2 = two standard deviations (rare and elite). Same magnitude in the negative direction.
Career Draft Scores compute the per-pick averages over all years of picks, then z-score the same way against the 8-owner cohort.
What it does not measure
- TradesIf you traded a star away or picked one up mid-season, that doesn't count here. Draft Score grades draft day.
- Waiver wireSame — undrafted scoops and FAAB heroes are a separate skill. (Future page idea.)
- Future seasonsSlot expectations get updated every year as new drafts happen, so old Draft Scores can shift slightly when fresh data lands. Always relative to the league's full draft history.