Projections, ADP, floor/ceiling and the SLEEPER/BUST calls come from the PredictionMarketsPicks model (2026, CBS ADP), pulled 2026-09-02 — but Proj, Floor, Ceil and VOR are re-scored for Coast2Coast , not standard. PMP stays the forecaster; each player’s stat mix from ESPN supplies the league/standard ratio that converts their number. Base points (yardage, TDs, INTs, fumbles) are exact under your rules. The per-game bonus tiers (80 rec, 100/200 rush, 250/300 pass and the rest) cannot be read off a season total, so how often each threshold is crossed is modelled , not measured — a lognormal on each stat’s per-game mean. Checked against John’s 2025 sheet it lands almost exactly on a mid-volume tight end (22% of games modelled vs 24% actual) but understates a high-volume receiver (48% vs 82%), so elite target hogs are, if anything, still underrated here. Long-TD bonuses (+1/+2) need play-by-play data and are not counted. Kickers are re-scored the same way from ESPN’s FG-distance buckets; D/ST are left on ESPN’s own scale — their tiers differ from yours, so read the order, not the points.
The blue T1–T4 tag is a second opinion, not PMP’s — a one-time transcription of FantasyPros’ consensus tiers (Sep 2026), sitting alongside PMP’s own Sleeper/Bust call rather than replacing it. A player can carry both, one, or neither. Coverage: QB/RB/WR/TE tiers go about 15–19 players deep per position (FantasyPros’ own free cutoff); 22 of 32 team defenses and 17 kickers carry a tier; the other kickers are real players too, just untiered. It won’t update again this season — reload gets you nothing new here. D/ST and Kickers now carry real season projections and ADP from ESPN’s public fantasy API (2026, ESPN’s default league scoring, snapshot 3 Sep 2026): every team’s starting kicker by name, and all 32 defenses. Two caveats: ESPN’s default D/ST scoring is not CBS’s, so treat the order as the useful part more than the exact points; and their ADP is ESPN drafters, not CBS. Ceil/Floor read “—” because ESPN publishes no range. Their VOR is real but sorts after the skill positions by default, and they only enter Cost of waiting in John’s last rounds — a kicker’s edge is illusory when a comparable one sits on waivers every week. ESPN’s skill-position projections are deliberately not used: their default is reception-scored and runs 40–55 points above PMP’s standard for RB/WR/TE while matching on QB, so the two are not the same basis and were never blended. That mismatch is irrelevant to K and D/ST, which score no receptions. Every bye week is real, from the same PMP data as the rest of the board. The + Other button stays for anything genuinely off this board.
The two sources talk to each other in two small ways. Sorting by VOR (the default), players within about 3 VOR points of each other — a genuine toss-up by the numbers, not a real gap — are ordered by FantasyPros’ tier instead; everywhere else VOR alone still decides. And a red ⚠ marks a real conflict: PMP calling BUST while FantasyPros has him in the better half of that position’s tiers, or SLEEPER while FantasyPros has him in the worse half. Agreement (a Sleeper who’s also low-tiered, a Bust who’s also high-tiered) isn’t a conflict and gets no mark — only the cases where the two sources are pulling in opposite directions.
“Lasts?” and “Cost of waiting” are mine, not theirs — a logistic curve around each player’s ADP, spread widening as ADP gets later, giving the chance he survives to John’s next real decision point (the header names the pick). Early in the draft that window is short and nearly everyone reads 95–99%; it sharpens across the 18-pick gaps. On the wheel it looks past his own back-to-back pick — at 10 the question is whether he lasts to 30, not to 11, since nobody picks in between. Cost of waiting values each player at 80% projection and 20% ceiling — the 20% matching the share of this pot that pays on raw point totals — so it leans slightly toward upside; VOR and the board itself stay pure projection. It also docks a candidate 6 points for each of John’s starters beyond the second he’d share a bye week with (bench byes don’t count), looking only at the top four at a position — enough to break a close call toward a clean week, never enough to reach past a genuinely better player. When the recommendation still stacks, the verdict says so. The bye banner and red bye markers count starters , not the whole roster: with 16 players over about nine bye weeks a 3-stack somewhere is near-certain and means nothing; three starters out in the same week costs a game. K and DST are left out of the bye logic entirely: they are streamed, and whoever is on waivers the week of the bye will have changed anyway. It models draft-day noise only: it knows nothing about injuries, camp news, or your leaguemates’ habits. Boom/bust week odds and tier breaks sit behind PMP Pro and are not shown here.
Sync runs through Firestore on the room in the address bar. Anyone with the link can read the board, and anyone who switches to Drafting can write to it — so keep the link between the two of you, and leave it on Watching unless you are the one entering picks.
A green B-number is that player’s bonus share: B31 means about 31% of his projected points come from the per-game bonus tiers. Anything at or above 30% is a player standard rankings systematically underprice, because those points are invisible to them. It is shown only where the share clears 30%.
The orange edge on a row marks a position that still fills one of John’s open starter slots (including FLEX). It recalculates after every pick, so expect most of the board lit up early and only a few rows late.
A red OUT / IR / SUSP tag pulls once, at page load, from ESPN’s public injury report — an unofficial endpoint, not a documented API, so it can go quiet without warning and the tag never updates again this session (reload to refresh it). It only flags the season-ending tier — not Questionable , which is normal in-season noise this early. Matching is by exact name after stripping punctuation and Jr./Sr./suffixes, so an unusual name could slip through unflagged; hover a tag for ESPN’s own note. Treat it as a prompt to double-check, not a verdict.