Ray Finkle's Super Secret Custom Perfect Draft Model App

2026

A draft model built for exactly three leagues — two on ESPN, one on Yahoo — and nobody else's.

Password required

The people in my leagues can read. It stays shut to them until the drafts are done.

It is free. If you are not in one of my leagues, ask and I will send you the password. The lock is about my own drafts, not about money.

Draft season Stadium poster for the 2026 draft board: a kicker in a Dolphins jersey kneels under the lights beside a capybara in a headset marked Barnabas Head Coach, with the board running on four screens across the bottom.
The board runs in one browser tab. The capybara is not included.

How it was built

One static page, rebuilt from inputs pinned to a capture date. Ten ranked lists reach a player's row and five of them vote, each re-priced first to the room's own scoring and roster, so the same player carries a different number in each of the three leagues. Six views, no network calls, every mark stored in the browser. The rankings stay behind the password until the drafts are done.

The 2026 draft, in numbers

General 2026 fantasy football analysis — the board's own conclusions are not on this page. Thirteen seasons of real draft prices joined to real box scores, 2,466 drafted player-seasons, plus the 2026 expert consensus, real-money draft prices, and the rooms' own suggested boards. Every rate carries a 95% uncertainty band.

The most veteran board in thirteen years

4Rookies in the top 100 — 2025 had 9
12WRs inside pick 120 who held a 28%+ target share — was 8
17RBs inside pick 120 averaging 18+ touches — was 14
11/36Top-36 names that are new — 2nd-lowest turnover in 12 years
05102014: 6 rookies in the top 100142015: 8 rookies in the top 100152016: 3 rookies in the top 100162017: 5 rookies in the top 100172018: 5 rookies in the top 100182019: 7 rookies in the top 100192020: 8 rookies in the top 100202021: 8 rookies in the top 100212022: 4 rookies in the top 100222023: 6 rookies in the top 100232024: 8 rookies in the top 100242025: 9 rookies in the top 1009252026: 4 rookies in the top 100426Rookies drafted inside the top 100

Four rookies inside the first 100 picks — the fewest since 2014. The unknowns were replaced by proven volume.

2026 expert consensus vs the twelve prior seasons' boards.

Data table
SeasonRookies in top 100
20146
20158
20163
20175
20185
20197
20208
20218
20224
20236
20248
20259
20264
1015201416182022242619.211.9Mean expert rank SD, picks 37-100

The experts stopped arguing. Mean rank disagreement on picks 37–100 is 11.9 — the tightest of thirteen seasons.

2026 captured 7 August with 87 experts; prior seasons early September with 82–237. An August board this tight is understated by the comparison, not inflated.

Data table
SeasonMean expert rank SD, picks 37-100
201417.3
201516.4
201616.9
201717.3
201819.2
201916.9
202015.9
202114.6
202214.6
202314.6
202413.4
202514.6
202612.0
Gone inside pick 31 — nine straight seasons20304014161820222426pick 37Draft pick where the 12th running back leaves the board

The running back dead zone is now permanent. Nine straight seasons the 12th back was gone inside pick 31. The last four, he lasts to 35–37.

Pick where RB12 leaves the board, by season's draft prices.

Data table
SeasonRB12 off the board at pick
201421
201524
201631
201729
201822
201925
202021
202123
202220
202335
202437
202536
202637

What a round actually buys

0%25%50%Round 1: WR 44%, RB 32%44%32%Round 1Round 2: WR 27%, RB 24%27%24%Round 2Round 3: WR 9%, RB 8%9%8%Round 3Chance the pick finishes top-5 at his position

Round 3 is where superstars stop existing. The odds of a top-5 positional finisher drop from 44% to 9% inside three rounds.

12-team PPR, 2013–2025. RB and WR — the only positions drafted in volume that early.

Data table
RoundWRRB
Round 144%32%
Round 227%24%
Round 39%8%
50%60%70%80%90%100%Round-one WR: 88% startable, band 80 to 96, n=57WR88%n=57Round-one RB: 72% startable, band 62 to 80, n=90RB72%n=90

The first receiver is the safest pick in fantasy. 88% of round-one WRs return a startable season, against 72% of backs — and half the bust rate.

Startable = WR3-or-better / RB2-or-better finish. Whiskers are 95% bootstrap bands.

Data table
PositionWR3-or-better rate95% bandn
WR88%[80, 96]57
RB72%[62, 80]90
1224366084108RB clears its own replacement level through pick 36RBdies at 37WR clears its own replacement level through pick 60WRdies at 61TE clears its own replacement level through pick 60TEdies at 61QB clears its own replacement level through pick 60QBdies at 61The average pick clears its own replacement level through…

Running back stops paying two rounds before anything else. The average back stops clearing his own replacement level at pick 37. Everything else lasts to 61.

Mean value over replacement by draft slot, 2013–2025. In the 61–84 band every position is under water.

Data table
PositionPositive through pickGoes negative atMean VOR, picks 61-84
RB3637-18.6
WR6061-9.4
TE6061-18.3
QB6061-12.5

The cliffs

100200300400replacement 150−1 to −2: −49 ptsRB×2.30 elite vs replacement1122436100200300400replacement 172−1 to −2: −35 ptsWR×1.97 elite vs replacement1122436100200300400replacement 148−1 to −2: −39 ptsTE×1.64 elite vs replacement1122436100200300400replacement 269−1 to −2: −37 ptsQB×1.35 elite vs replacement1122436

The biggest cliff at every position is first to second. RB1→RB2 is −49 points. The most valuable thing in fantasy is being right about who finishes first, and nobody is.

Average season-end PPR points by positional finish, 2013–2025.

Data table
FinishQBRBWRTE
1390.2382.5365.3272.6
2356.1334.6332.3236.4
3342.8321.2317.2217.1
6306.5271.8285.1186
12268.7221.5247.4147.6
24181.7171.3203.7101.9
3680.2131.6171.570.3
1×1.5×2×2.5×RB: top-3 average 346.1 vs replacement 150.3 — 2.3 timesRB2.30×WR: top-3 average 338.3 vs replacement 171.5 — 2.0 timesWR1.97×TE: top-3 average 242 vs replacement 147.6 — 1.6 timesTE1.64×QB: top-3 average 363 vs replacement 268.7 — 1.4 timesQB1.35×

Why the early rounds are two positions. An elite back is worth 2.3× his replacement; an elite quarterback, 1.35×. That ratio — not any ranking — is the whole argument.

Top-3 seasonal average against the 12-team replacement level (QB12 / RB30 / WR36 / TE12).

Data table
PositionTop-3 averageReplacementMultiple
RB346.1150.32.30×
WR338.3171.51.97×
TE242147.61.64×
QB363268.71.35×

What kind of player actually hits

0%25%50%75%100%WR, picks 1-36Held a 28%+ target share last seasonHeld a 28%+ target share last season: 78.8%, band 71.9 to 85.5, n=6679% · n=66Did notDid not: 61.9%, band 53.0 to 71.8, n=12662% · n=126RB, picks 37-8418+ touches a game last season18+ touches a game last season: 42.9%, band 30.4 to 58.6, n=4243% · n=42Under 13 touches a gameUnder 13 touches a game: 25.9%, band 14.1 to 40.0, n=5826% · n=58WR, picks 37-84Under 15% of team targets vacatedUnder 15% of team targets vacated: 45.0%, band 30.6 to 64.0, n=6045% · n=6045%+ of team targets vacated45%+ of team targets vacated: 26.5%, band 15.6 to 39.4, n=3427% · n=34

Target share held is the marker that works. Early receivers who held 28%+ hit 79%, against 62% without it. Vacated opportunity is the marker that doesn't — the bands overlap, and what signal there is points the wrong way.

Hit = top-24 positional finish, compared inside the same price band. 95% bands, n on each row.

Data table
GroupMarkerHit rate95% bandn
WR, picks 1-36Held a 28%+ target share last season78.8%[71.9, 85.5]66
WR, picks 1-36Did not61.9%[53.0, 71.8]126
RB, picks 37-8418+ touches a game last season42.9%[30.4, 58.6]42
RB, picks 37-84Under 13 touches a game25.9%[14.1, 40.0]58
WR, picks 37-84Under 15% of team targets vacated45.0%[30.6, 64.0]60
WR, picks 37-8445%+ of team targets vacated26.5%[15.6, 39.4]34
0.20.40.6TE weekly coefficient of variation 0.57TE0.57WR weekly coefficient of variation 0.53WR0.53RB weekly coefficient of variation 0.52RB0.52QB weekly coefficient of variation 0.38QB0.38

Tight end is the most volatile thing you start — and quarterback the least, at nearly half the swing.

Weekly coefficient of variation, 2013–2025 starter-level seasons.

Data table
PositionWeekly CV
TE0.565
WR0.528
RB0.524
QB0.382
250/250

ESPN hands two leagues with different rules the identical board. All 250 ranked players, same rank, in a full-PPR / 4-point room and a half-PPR / 6-point room. Confirmed twice, from separate captures: 5 August (258/258) and 8 August (250/250). Its auto-draft runs on that list — in one room it drafts to the wrong scoring, in the other it prices quarterbacks at the wrong rate. That is why you never let ESPN auto-draft for you.

-48-36-24-120+12+24+36+48← host has him earlier than the moneylater →same pickESPN: one dot per player, top-150 skill players. Mean distance from the money market 12.0 picks.ESPNmean 12.0 picks off43% a full round · 14% two roundsYahoo: one dot per player, top-150 skill players. Mean distance from the money market 5.0 picks.Yahoomean 5.0 picks off10% a full round · 1% two roundsa receiver, 51 earlya receiver, 50 late

ESPN sits 2.4× further from the money than Yahoo. Mean gap 12.0 picks against 5.0; 43% of its top 150 a full round off the real-money market, 14% two full rounds. Yahoo: 10% and 1%.

One dot per top-150 skill player: host list rank minus real-money draft price, in picks. Shaded zones: within one round, within two.

Data table
ListPlayersMean gap≥1 round off≥2 rounds off
ESPN13812.0 picks43%14%
Yahoo1435.0 picks10%1%
-12-8-40+4← picks earlier than the moneyTEESPN ranks TEs 12.7 picks earlier than the money on average12.7 earlyYahoo ranks TEs 8.4 picks earlier than the money on average8.4 earlyRBESPN ranks RBs 3.6 picks later than the money on averageYahoo ranks RBs 0.0 picks later than the money on averageWRESPN ranks WRs 1.3 picks later than the money on averageYahoo ranks WRs 2.5 picks later than the money on averageQBESPN ranks QBs 0.4 picks later than the money on averageYahoo ranks QBs 0.2 picks earlier than the money on averageESPNYahoo

The one bias both platforms share: tight ends go early — ESPN by 12.7 picks on average, Yahoo by 8.4.

Mean signed gap to the money market by position.

Data table
PositionESPN tiltYahoo tilt
TE-12.7 picks-8.4 picks
RB3.6 picks0.0 picks
WR1.3 picks2.5 picks
QB0.4 picks-0.2 picks
51 picks early

ESPN's single worst call on a receiver, against where real money drafts him.

50 picks late

Its worst miss in the other direction — also a receiver. Yahoo's worst in either direction: 26.

No names on this page — the people in my leagues can read.

The rules move less than you think. The roster moves more.

2803203604004404-pt passing TD: QB1 361, QB12 289, spread 72QB1 · 361QB12 · 2894-pt passing TDspread 726-pt passing TD: QB1 416, QB12 344, spread 72QB1 · 416QB12 · 3446-pt passing TDspread 72every QB +55

Six points a passing TD is the most over-rated setting in fantasy. It adds ~55 points to every quarterback and leaves the QB1-to-QB12 gap exactly where it was: 72 points. It changes which quarterback, barely. It does not change when you take one.

2026 projected stat lines rescored under both rules.

Data table
SettingQB1QB12Spread
4-pt passing TD36128972
6-pt passing TD41634472
ESPN, 4-point TD, full PPR — 2 RB + 1 FLEXESPN, 4-point TD, full PPR slot 1: RBRBESPN, 4-point TD, full PPR slot 2: RBRBESPN, 4-point TD, full PPR slot 3: RBRBESPN, 4-point TD, full PPR slot 4: RBRBESPN, 4-point TD, full PPR slot 5: RBRBESPN, 4-point TD, full PPR slot 6: RBRBESPN, 4-point TD, full PPR slot 7: RBRBESPN, 4-point TD, full PPR slot 8: WRWRESPN, 4-point TD, full PPR slot 9: WRWRESPN, 4-point TD, full PPR slot 10: WRWRESPN, 4-point TD, full PPR slot 11: WRWRESPN, 4-point TD, full PPR slot 12: TETEESPN, 6-point TD, funky rules — Half PPR · 1 RB + 2 FLEXESPN, 6-point TD, funky rules slot 1: RBRBESPN, 6-point TD, funky rules slot 2: RBRBESPN, 6-point TD, funky rules slot 3: RBRBESPN, 6-point TD, funky rules slot 4: RBRBESPN, 6-point TD, funky rules slot 5: WRWRESPN, 6-point TD, funky rules slot 6: WRWRESPN, 6-point TD, funky rules slot 7: WRWRESPN, 6-point TD, funky rules slot 8: WRWRESPN, 6-point TD, funky rules slot 9: WRWRESPN, 6-point TD, funky rules slot 10: WRWRESPN, 6-point TD, funky rules slot 11: WRWRESPN, 6-point TD, funky rules slot 12: WRWR

What actually moves a board is the roster. Half PPR costs a top-24 receiver 44 points — then a 1-RB / 2-FLEX roster pushes receivers back past the scoring hit. Seven backs in the top twelve becomes four.

First 12 picks by value over each room's own replacement level. No generic list knows this.

Data table
RoomRBWRTEQB
ESPN, 4-point TD, full PPR7410
ESPN, 6-point TD, funky rules4800
0−25−50Half PPR costs a top-24 WR 44.2 pointsWR−44.2 ptsHalf PPR costs a top-24 TE 34.1 pointsTE−34.1 ptsHalf PPR costs a top-24 RB 23.3 pointsRB−23.3 pts

What half PPR takes from a top-24 season, by position.

Full-PPR points minus half-PPR points, 2026 projections, top 24 at each position.

Data table
PositionPoints lost, top 24
WR−44.2
TE−34.1
RB−23.3

Which kind of list wins the argument

0.20.30.40.52014: experts 0.415, market 0.39720142016: experts 0.234, market 0.22320162017: experts 0.378, market 0.32320172018: experts 0.397, market 0.34320182019: experts 0.443, market 0.42820192020: experts 0.431, market 0.35020202021: experts 0.475, market 0.43520212022: experts 0.481, market 0.48820222024: experts 0.512, market 0.4262024expertsmarket

Expert consensus beat market price in 8 of 9 clean seasons. Mean rank correlation with season-end points: 0.418 against 0.379.

The market here is free-site mock ADP — no historical real-money archive exists to test. Evidence that a well-built expert board is the better spine, not that price is worthless.

Data table
SeasonExpertsMarket price
20140.4150.397
20160.2340.223
20170.3780.323
20180.3970.343
20190.4430.428
20200.4310.350
20210.4750.435
20220.4810.488
20240.5120.426
0102030picks 1-24picks 25-60picks 61-108picks 109-180RBQBTEWR 17.0 · rank range 91WRMean expert rank SD by pick band, 2026

Where the experts are still fighting in 2026: receiver in rounds 6–9. Rank SD 17.0, a 91-place spread — the same band where receivers still pay and the backs have already died. The widest single disagreement in the top 100 runs from pick 51 to pick 241.

Mean expert rank SD by pick band, 7 August consensus.

Data table
PicksWRRBQBTE
1-244.05.1——
25-609.17.810.5—
61-10817.012.111.712.4
109-18028.122.518.824.1

Read before quoting

  • The 2026 expert capture is 7 August with 87 experts; historical seasons are early September with 82–237. That timing difference cuts against the "experts stopped arguing" finding, not for it.
  • Prices are the 4–11 August market and will move before the drafts.
  • Every number on this page traces to public data — draft prices, box scores, published consensus. The board's own rankings and conclusions are not here, on purpose.