How this board works, and where the numbers come from
"The next man up" does not get worse as the draft
goes on, and this board does not pretend it does. In a 12-team league with 9
forward slots, the 108th-best forward is always the guy you could settle for. That
does not change because other people drafted. What does change is how fast each
position dries up, and the scarcity card shows what it costs you to wait one more
round at each one. That is the number that should move a pick.
Every range bar is drawn on one shared scale, so bar
length and position are comparable between rows, so a wider bar really is a wider range.
Each bar runs from the low to the high we would not be surprised by, and
eight players in ten should finish inside it. The marker is the projection.
Statisticians call those ends the 10th and 90th percentiles.
My league. Who counts as "the next man up" depends
on your league, not ours. Teams times roster slots decides which player you could get
for nothing, so in a 10-team league that is the 90th-best forward rather than the
108th, and everyone above him is worth a little less by comparison. The
cost-of-waiting figures above use
one round = one pass through your league. Daily-starts drives the schedule-collision
warning, which only appears once you hold 3+ skaters from one club in a tight format
(measured: the startable-games spread is 10 / 25 / 32 games at 12 / 10 / 9 daily
slots, so it is a footnote in loose leagues and a real cost in tight ones).
PO column. Games in the final four full fantasy weeks
(Mar 08 – Apr 04), the playoff window, with off-night count. Measured range across teams:
11–15 games and 0–6 off-nights, and only +0.26 correlated with season schedule value,
so it is information the season multiplier does not carry. Championships are won in
that window.
Projections as of 2026-08-21.
Data. Play-by-play, ice time and situational splits from the NHL public stats API. Schedule from the published 2026-27 NHL calendar. No MoneyPuck inputs: this build drops the shot-quality and expected-goals features (skaters) and the goals-saved-above-expected features (goalies), so the page owes no third-party data attribution.