CURRENT FORM. CURRENT ROLE.
The players worth your pick.
A ranked list for your Yahoo H2H points league.
THE MODEL’S PROJECTED ORDER
Your draft shortlist.
Current form, role, line support, team strength and advanced analytics—combined using your Yahoo scoring.
Skaters rank by projected season fantasy totals; goalies rank by projected season value with expected starts and team-supported wins. Both apply the 30-game workload filter with established injury exceptions. Detailed stat exploration stays in the separate tabs.
Projected skaters
Projected total points · projected games
| # | SKATER | PROJECTED TOTAL FP | PROJECTED GP |
|---|
Projected goalies
Skill × workload × team support
| # | GOALIE | PROJECTED FP | STARTS / WINS |
|---|
Click any player for the evidence and scoring breakdown. These are the same model’s projections, not a separate subjective ranking. Snapshot ; model forecasts remain uncalibrated.
Skater rankings
Projected rates combine current production, sharply declining historical weights, current non-PP and PP minutes, linemate support, team PP strength, and chance quality. Click a player for the full explanation. Goalies are ranked separately by projected totals and starts. Players below 30 projected games/starts are hidden by default. Established injury-affected players can remain via a labelled exception. Data is a dated snapshot, not a live lineup feed.
PLAN YOUR NEXT PICKUP
Waiver wire · add / drop.
Your player projections × the games they can play in your selected week.
Loading season schedule…
YOUR YAHOO ROSTER
Who could you drop?
AVAILABLE IN YOUR LEAGUE
Who could you add?
Selected drop
Selected add
Weekly model & assumptions
Skaters: projected FP/game × eligible team games. Goalies: projected FP/start × team games × projected season starts / 84. Goalie dates show team games, not confirmed starts. Extra games can outweigh a small per-game advantage; no arbitrary four-game bonus is applied. IR/OUT players score zero in the default unavailable scenario; this is a dated injury snapshot, not a week-specific medical forecast. Select “Assume healthy” to explore a return. Other injury flags remain warnings.
Weekdays follow the NHL schedule’s Eastern date. Only games on or after the effective date count; games already elapsed earlier today are not automatically removed. Match the effective date to your Yahoo waiver processing. Snapshot schedules can change. This is a temporary streaming comparison, not a season-long instruction to drop a star. Actual lineup slots, daily congestion, acquisition limits, waiver priority and Yahoo ownership are not connected. Your roster and available-player lists are entered manually and saved in this browser. Player projections use your current Model & scoring settings.
TEST THE RETURN
Trade analyzer.
Compare packages using the same projections and your saved Yahoo scoring.
League depth
Sets the estimated waiver replacement level. Adjust to your league’s roster depth, including bench slots.
OUTGOING PACKAGE
You send
INCOMING PACKAGE
You receive
How this verdict works
Trade value = max(0, projected fantasy total − replacement FP per game/start × expected games/starts). Skater replacement is the first player beyond your league’s skater capacity, ordered by projected FP/game. Goalie replacement is the first goalie beyond capacity, ordered by season FP; its per-start rate supplies the baseline. Each player’s availability and goalie workload already affect the total. Free roster spots can be filled at replacement level; no second roster-slot bonus is added.
A gap of up to 10% of the two package values’ average is approximately fair; 10–25% is a slight edge; above 25% is a clear model edge. These are editable-league-depth, heuristic comparisons, not calibrated probabilities or market prices. Values use the model’s season window, not a remaining-season schedule. Full rosters, positional needs, actual waiver availability, Yahoo trade rules and lineup limits are not modeled. Injury badges with no return date may understate availability risk. Review individual assumptions by clicking a player.
Team scoring, power plays & roster context
2026–27 · 84-game schedule · dated forecasts and updated win support
Historical columns show goals for / against per game. Forecast inputs are blank by default; green placeholders show model estimates. Enter an outside scoring or win forecast to replace the model. PP assignments use dated Daily Faceoff snapshots. Final team wins include banked wins; goalie scoring uses the remaining-game win chance.
| TEAM | 23–24 GF / GA | 24–25 GF / GA | 25–26 GF / GA | CURRENT PP% | PROJECTED PP GF/60 | PROVIDER PP1 | PROJECTED GF/GP | PROJECTED GA/GP | FINAL TEAM WINS | PUBLISHED POINTS / WIN SUPPORT |
|---|
Current lines & injury audit · all 32 teams
Provider units, not a promise of the next game lineup. OUT/IR conflicts are flagged in player details; no replacement is invented. Seven-defenseman layouts may contain an extra unpaired defender. Names without a model profile use neutral teammate support. Six provider names could not be matched to the current team pool; they do not create new ranked players.
Your Yahoo H2H points scoring
These are your saved league weights. Verify them against your Yahoo league settings.
Projection controls
Smaller current-season prior = more influence from this season. Role estimates use current and recent minutes independently of historical ability. A partial absence changes projected totals; skater per-game rankings remain rate based. Workload estimates use current participation and historical NHL appearances, without assigning new call-ups a 75-game default. The 30-game filter keeps established injured players when a reviewed injury note or your injury assessment supports a healthy-role workload of 30 or more. Unexplained absences are not automatically classified as injuries.
Projection imports
Optional player JSON replaces the bundled player pool. Imported season totals bypass role/statistical estimation. Skaters and goalies remain separate.
[{"id":"1","name":"Player","team":"EDM","pos":"C/LW","gp":82,"G":30,"A":50,"PM":10,"PPP":25,"SOG":250,"BLK":40}]Goalie fields: W, GA, SV, SHO. Up to 10,000 unique IDs. Optional HIT/xG.
Freshness & coverage
A recency system, not a reputation ranking
Historical ability weights: 50% last season, 35% two seasons ago, 15% three seasons ago. Missing seasons are renormalized. This season’s production gets its own sample-adjusted weight: GP / (GP + current-season prior). By default, 16 games gives 57% influence to current non-PP performance.
Current deployment matters separately. Provider PP membership receives 75% of the PP role estimate (72% of team PP time for PP1, 25% for PP2), with 25% from observed usage. Without a current assignment, usage remains the healthy-role fallback. Observed PP minutes use 85% recent usage and 15% season usage after at least three recent appearances. Current non-PP minutes largely replace historical minutes. Inferred PP1 means the five highest recent usage rates with at least 0.5 PP min/game; PP2 means remaining usage of at least 0.3; No PP means no regular assignment. Losing the power play or moving down the lineup can therefore reduce value before a full season of low totals accumulates.
Skater projection
Non-PP and PP goals/assists are projected separately per minute. Goals regress toward xG and finishing history. Shot attempts and shot-on-goal accuracy support shot forecasts. On-ice expected chances support an assist estimate. Actual line xGF/60, teammate production and measured NHL EDGE offensive-zone puck time provide small bounded adjustments. Power-play team scoring and expected-goal rates, plus the recent PP1 roster, inform the PP environment.
Individual high-danger attempts, shot attempts, xG, created xG, scoring chances, primary assists and on-ice shares are also shown for inspection. They are not all independent ranking bonuses. Offensive-zone puck time is measured while the player is on ice; zone starts are a separate deployment statistic. Missing EDGE data stays unavailable.
Goalies stay separate
Goalie model v2: current-form weight = current minutes / (current minutes + 900), separate from 50/35/15 historical season rates. Regress GSAx toward zero; blend 70% GSAx skill, 20% danger-band skill, 10% league-centered flurry residual. Goals against combine 65% skill-adjusted xGA with 35% current team GA (55% season / 45% last 10). Current depth-chart starters receive a 65% raw role floor; backups a 35% ceiling. Sample-adjusted current starting shares blend with historical shares, then the team’s finite 84 starts are allocated with a 65-start default cap. Depth order is evidence of a role, not a guaranteed start.
Goalies never enter the skater FP/game list. Their default order uses projected season fantasy totals, team-supported wins and a finite starts budget. Current skill and starting share update the GSAx, danger-band and flurry model. A short run of high per-appearance scoring does not grant starter volume.
What remains uncertain
This is an uncalibrated statistical model, not a backtested forecast. Daily Faceoff units and injury badges cover all 32 teams, dated November 3–4. Current combinations use matching shared-unit analytics only; a newly assembled line gets neutral shared xG. IR/OUT players have no active linemates. Badges with unknown return dates do not imply a recovery date. Rosters and stats are snapshots. Team win forecasts blend 60% equally averaged external points-derived probability, 20% internal scoring/xG probability and 20% regressed last-10 win rate. OTL conversion uses current OTL plus 20 prior games at 11%. These coefficients are heuristic. External forecasts are equally averaged from dated hockeystats.io, daily.hockey and Forepitch snapshots. Colorado’s November 4 values are 109, 96 and 105 points (average 103.3). Goalie model v2 blends recent form separately, enforces current depth roles, and uses season plus last-10 team goals allowed alongside xGA. No Yahoo connection, live refresh, schedule optimization or guaranteed draft outcome.
Sources: Daily Faceoff · hockeystats.io · Colorado forecast comparison · NHL Stats · NHL EDGE · MoneyPuck.com. Personal non-commercial use; MoneyPuck data credited throughout.