Projected goalie starts analyze which netminder a fantasy hockey platform expects to begin a game on any given night. These projections combine recent performance, matchup difficulty, and roster trends into a single expectation value.
Accurate projections reduce noise in daily lineup decisions, especially in weekly or season-long formats where small edges compound. The sections below define core concepts, compare models, and show how to integrate projections responsibly.
| Metric | Definition | Impact on Projected Start | Typical Data Source |
|---|---|---|---|
| Save Percentage Projection | Estimated SV% against upcoming opponent quality | Higher expected save rate increases point potential | League-wide models, team Corsi, recent GFGA trends |
| Goals Against Average Estimate | Projected GAA based on shot quality and goalie tendencies | Lower GAA improves weekly scoring rank | Regressed to mean, recent even strength performance |
| Game Script & Venue | Home ice, back-to-back status, rest days, playoff seeding stakes | Back-to-backs and low-seed road games often depress volume | Schedule API, team depth chart releases |
| Injury & Lineup Context | Starter health, backup tendencies, defensive partner quality | Rehab starts or thin defensive corps can increase risk | Official injury reports, practice updates |
| Advanced Signatures | Catches preferred side, control of rebounds, blocker/screen usage | Goalie styles create exploitable edges in certain systems | Evolving proprietary tracking models |
Interpreting Projected Start Metrics
Projection outputs often appear as percentages, ranks, or normalized scores. Understanding the scale helps you compare goalies fairly across teams and schedule clusters.
Translating a percentage into expected points requires league context. A .920 SV% may be elite against weak teams but mediocre against top units, altering the implied starting value.
How Projections Are Calculated
Modelers blend small-sample data with regression to the mean, avoiding overreaction to outlier streaks. Inputs include raw stats, expected goals, and quality of competition at five-on-five.
Machine learning layers handle time-series patterns such as day-of-week effects and travel distance, while human analysts adjust for intangibles like locker-room confidence.
Matchup Analysis In Practice
Strong goalies on weak defenses can still post good numbers when facing bottom-six lines, but may struggle in high-danger chances against elite units. Matchup depth matters.
Home ice often boosts save percentage and lowers GAA, while congested schedules compress recovery windows, increasing error risk and lowering volume metrics across the week.
Integrating Projections Into Decisions
Use projected starts to rank your daily or weekly options, then layer in ownership, floor, and volatility. A high ceiling with low floor may suit tournaments, whereas steady rankings fit cash games.
Track alignment between platform projections and your own observations. When forecasts diverge, check practice reports, emergency call-ups, and late scratches before finalizing lineups.
Best Practices For Using Projected Starts
- Check practice updates and injury reports before locking lineups.
- Compare multiple projection sources to identify consensus and outliers.
- Adjust for travel load and rest differential in back-to-back scenarios.
- Balance projection rank with ownership and tournament scoring rules.
- Review model assumptions to understand which metrics weigh most heavily.
FAQ
Reader questions
How often should I refresh projected goalie starts during the season?
Update projections after every regulation game, prior to waiver wire moves, and immediately after significant injury news or practice surprises.
Can projected starts replace watching film entirely?
No, treat projections as a filter; film remains essential for spotting mechanical flaws and subtle trends that models may miss.
Do projected starts work differently in playoff formats compared to the regular season?
Yes, playoff samples shrink, models lean more on recent history and goalie temperament, and single-game variance rises sharply.
Should I ever start a goalie with a poor projected start if my roster is thin?
Only when injury depth is exhausted and the alternative is a guaranteed negative impact; accept the downside as a calculated risk.