NHL advanced statistics can make player pages and team trends far easier to understand, but only when each metric is used for the question it actually answers. This guide explains Corsi, Fenwick, expected goals, high-danger chances, PDO, and related measures in plain language, then provides a repeatable workflow for comparing players without letting one hot streak or short sample distort the picture.
Overview
Traditional hockey statistics tell you what happened: goals, assists, shots, saves, and points. Advanced statistics add context by describing how a team created those results and how sustainable they may be. A player can score on a small number of chances, for example, while contributing positively in possession and chance creation. Another player can have a strong point total while spending most of his shifts defending.
No single number provides a complete player evaluation. Shot-attempt metrics are useful for estimating territorial play, expected-goals models add shot quality, and shooting, save, and PDO figures help identify finishing or goaltending results. The most reliable interpretation comes from combining several categories with deployment, teammates, score effects, role, and enough games to reduce noise.
Definitions can vary by data provider. “High-danger” locations, expected-goal formulas, blocked-shot treatment, and score-adjustment methods are not necessarily identical across websites. Before comparing two player pages, confirm that the figures use the same season, game state, sample type, and provider.
What to track
Use the following groups to organize an NHL advanced stats dashboard rather than treating every available number as equally important.
| Category | Metric | Plain-language meaning | Best use |
|---|---|---|---|
| Possession and volume | Corsi | All recorded shot attempts for a team while a player is on the ice, including shots on goal, missed shots, and blocked shots. | Assess territorial pressure and share of shot attempts. |
| Possession and volume | Fenwick | Unblocked shot attempts: shots on goal plus missed shots. | Compare unblocked attacking volume while reducing the effect of blocked shots. |
| Possession and volume | CF% or FF% | The team’s share of Corsi or Fenwick attempts when the player is on the ice: for divided by for plus against. | Compare on-ice attempt share, especially with similar game-state filters. |
| Shot quality | Expected goals, or xG | An estimate of the probability that a shot becomes a goal, based on factors such as location and shot type. The exact model varies. | Measure the quality and quantity of chances created or allowed. |
| Shot quality | xGF% | The team’s share of expected goals with a player on the ice. | Evaluate whether the player’s team is generating better chances than its opponent. |
| Shot quality | High-danger chances | Attempts classified as coming from especially dangerous areas or situations under a provider’s definition. | Track access to premium scoring areas, but verify the provider’s criteria. |
| Finishing and goaltending | PDO | Usually on-ice shooting percentage plus on-ice save percentage, commonly expressed around a 100-point scale. | Spot unusually high or low combined finishing and goaltending results. |
| Individual production | Individual points, shots, and individual expected goals | Measures a player’s direct scoring output, shot generation, and expected scoring contribution. | Separate personal attacking involvement from team results. |
Corsi and Fenwick
Corsi is a volume measure, not a direct count of possession time and not proof that every attempt was dangerous. A defenseman may post a strong Corsi share by moving the puck out efficiently, while a team may collect attempts from the perimeter without creating many goals. Fenwick removes blocked attempts, which can make it useful when examining shot generation and defensive pressure, but it remains a volume measure rather than a quality measure.
Percentages are generally easier to compare than raw totals because they account for the opponent’s attempts. Still, on-ice results reflect teammates, opponents, zone starts, score state, and coaching decisions. A player used primarily in offensive-zone situations may look different from one assigned difficult defensive shifts.
Expected goals and high-danger chances
Expected goals translate shot characteristics into estimated scoring probability. If a team records ten chances worth 0.10 expected goals each, its total expected goals is 1.0, regardless of whether it scores zero, one, or several actual goals. That makes xG useful for separating chance quality from short-term finishing.
xGF% compares expected goals for and against, so it can answer a practical question: is the team creating the better overall chance profile with this player on the ice? High-danger chance counts provide a more intuitive companion, but they should not replace xG. Two chances can both be labeled high-danger while differing in angle, pressure, rebound conditions, or other model inputs.
PDO and the limits of small samples
PDO combines on-ice shooting percentage and on-ice save percentage. It is often used as a descriptive signal for unusually favorable or unfavorable results. A high PDO may reflect excellent finishing and goaltending, but it can also indicate a short-term run that is difficult to maintain. A low PDO may point to poor luck, weak finishing, goaltending problems, or a genuinely poor defensive environment.
PDO is not a quality grade and should never be used alone to label a player lucky. Check shot volume, expected goals, individual shooting, goalie usage, and the number of minutes behind the result.
Cadence and checkpoints
Review advanced statistics at three different speeds:
- After a game: Use the box score and video to identify changes in line combinations, ice time, role, or injury status. Avoid firm conclusions from one result.
- Every two to four weeks: Compare recent and season-long Corsi, xGF%, shot rates, individual expected goals, and special-teams usage. This is a useful cadence for fantasy decisions and following a team’s tactical shift.
- Monthly or quarterly: Recheck definitions, provider methodology, player roles, and larger samples. This is the better checkpoint for evaluating whether a trend has become meaningful.
For each player, record the sample size, position, average ice time, strength state, line and defense partner, offensive-zone start rate if available, and whether the team was leading or trailing. Separate five-on-five numbers from power-play production. A player’s overall points may be driven by special teams, while his five-on-five chance profile tells a different story.
When reviewing lineup changes, pair the numbers with the NHL injury report and lineup impact guide. Injuries can change a player’s teammates, workload, and defensive assignments before the underlying numbers have time to stabilize.
How to interpret changes
Use a simple comparison workflow:
- Define the question. Are you asking who drives play, creates scoring chances, finishes efficiently, or helps a particular lineup?
- Match the context. Compare the same season, game state, score-adjustment setting, position, and minimum-minute threshold.
- Start with process. Look first at shot attempts, expected goals, individual shots, and chance involvement. These describe what the player and team are creating.
- Then examine results. Add goals, assists, shooting percentage, on-ice save percentage, and PDO to see where outcomes differ from process.
- Check usage and teammates. Review zone starts, competition, line combinations, special-teams role, and deployment before attributing every difference to skill.
- Compare recent and full-season views. A recent improvement is more persuasive when it appears in both chance quality and volume, not just goals.
Consider a forward with a rising goal total but flat individual shot and expected-goal rates. That may be a finishing surge rather than a major change in chance creation. Conversely, a player with stable points but improving xGF% and high-danger involvement may be contributing more than the traditional box score suggests.
For defensemen, do not treat Corsi or xGF% as isolated proof of defensive ability. Puck-moving skill, penalty differential, retrievals, matchup difficulty, and the quality of chances allowed all matter. For goalies, team on-ice save percentage is not the same as an individual save percentage, so use goalie-specific measures when evaluating the position.
When to revisit
Return to this guide whenever a player changes teams, moves lines, returns from injury, takes on a new special-teams role, or produces a sharp change in results. Recheck the comparison after roughly a month of new usage rather than reacting to the first few games. For fantasy analysis, pair the statistical review with the weekly waiver-wire process, especially when a player’s ice time and opportunity change before his points do.
Update the article before each NHL season and whenever a major statistical provider changes its definitions or presentation. Confirm how expected goals and high-danger chances are calculated, whether blocked shots are included, and which filters are available. Replace illustrative examples when the underlying data or terminology changes.
For a practical monthly routine, choose three comparable players, save their five-on-five Corsi or Fenwick share, xGF%, individual shot rate, individual expected-goal rate, points, shooting percentage, and deployment notes. Mark what changed, identify whether process or results moved first, and revisit the comparison after the next checkpoint. That habit turns advanced statistics from a collection of unfamiliar labels into a useful tool for reading NHL news, player trends, and team performance with greater care.
