PuckCast Edge
Every pick is free. The Edge is the rest.
Betting edges, stake sizing, props priced against the line, fantasy tiers and the graded ledger. $14.99 a month in season, or $99 for the year.
PuckCast Edge
Betting edges, stake sizing, props priced against the line, fantasy tiers and the graded ledger. $14.99 a month in season, or $99 for the year.
Not all shots are created equal. A one-timer from the slot and a wrister from the point look the same on a shot counter, but one scores at a far higher rate than the other. High-danger chances measure what actually matters: where the shot comes from.
TL;DR
A high-danger chance is a shot attempt from the wedge in front of the crease: PuckCast's pipeline flags it as within 25 feet of the goal and within 45 degrees of a straight line to the net. Shots from there score at a far higher rate than shots from the perimeter. Tracking them tells you far more about a team's offensive ability than raw shot counts ever could.
The wedge in front of the crease: within 25 feet of the goal and within 45 degrees of a straight line to the net. This is where goals happen. One-timers, deflections, and rebounds from this zone score at the highest rate on the ice. Teams that consistently generate HD chances win more games.
This drives scoring
Top of the circles and mid-range areas. These shots score occasionally, especially with traffic or on the power play. Volume matters here — a steady stream of medium-danger looks can add up, but they rarely win games on their own.
Volume matters here
Perimeter shots from the point, behind the goal line, or the neutral zone. These inflate shot totals but rarely beat NHL goalies. A team that racks up 40 shots but mostly from low-danger areas isn't really dominating — they're just shooting from bad angles.
Mostly noise
PuckCast's own high-danger flag (used as a team feature in game predictions) is a single cutoff — within 25 feet of the goal and within 45 degrees of a straight line to it (src/nhl_prediction/native_ingest.py). The per-player Finishing Map uses a different, tighter geometry for its own "slot" and "crease" zones — a box roughly 11 to 31 feet off the goal line and within 11 feet of the centerline for the slot, and within 11 feet of the goal line for the crease (scripts/generate_shotmaps.py). They are two separate definitions built for two separate jobs, not the same box relabelled.
Raw shot totals (Corsi) tell you who has the puck more. That's useful, but it misses a critical piece: shot quality. A team can dominate Corsi while firing 35 shots from the perimeter and still lose to a team that generated 20 shots, 12 of them from the slot.
High-danger chance differential is one of the strongest predictors of future team success. Teams that consistently out-generate their opponents in HD chances tend to climb the standings over time, even if their record looks mediocre early on. The goals catch up to the process.
Think of HD chances as the bridge between Corsi (quantity) and xG (quality-weighted probability). They tell you where shots are coming from without the full complexity of an expected goals model.
A simple zone-based classification. Every shot is tagged as high, medium, or low danger based on where it was taken. Easy to understand, easy to track. The tradeoff: it treats all high-danger shots the same, whether it's a breakaway or a contested wrister from the edge of the slot.
Simplified shot quality
A probability model that assigns a specific scoring likelihood to every shot. It factors in location, shot type, traffic, rebound status, game state, and more. More precise than zone-based danger ratings, but harder to calculate and explain.
Precise shot quality
Both metrics tell you about shot quality. HD chances are the quick read; xG is the deep dive. If you only have time to check one number, HD chance differential is the faster read. xG fills in the rest.
In the model
PuckCast's own pipeline flags a shot as high-danger when it is taken from within 25 feet of the center of the goal AND within a 45-degree angle of a straight line to the net — a wedge in front of the crease (src/nhl_prediction/native_ingest.py). Other public trackers (Natural Stat Trick, Evolving Hockey, MoneyPuck) draw a similar 'home plate' box with slightly different edges. Whatever the exact boundary, the concept is the same: shots from the middle of the ice, close to the net, score at a far higher rate than shots from the perimeter.
The league leaders shift year to year, but teams with strong cycle games, net-front presence, and cross-ice passing tend to dominate HD chance generation. Check PuckCast's power rankings for current team shot quality ratings — teams near the top consistently create more high-danger looks than the league average.
They serve different purposes. Corsi measures total shot volume (all attempts, including blocks and misses) and is a solid proxy for puck possession. High-danger chances measure shot quality. In practice, HD chances are more predictive of scoring than raw Corsi, but the best models (including xG) combine both quantity and quality. PuckCast uses both in its prediction model.