Score away–home after each goal; season totals in parentheses.
Shot quality
Expected goals (xG) rate every unblocked shot by how often shots like it go in, from where it was taken,
the shot type, the strength and what happened just before it. How it works
TOR 49%Deserve to win51% BUF
TOR
BUF
2.81
Expected goals
2.85
3
Goals (no shootout)
5
1.86
Expected goals, 5 on 5
2.10
68
Shot attempts
59
45
Unblocked attempts
45
34
Shots on goal
30
Game notes
What stood out as the game went on, from the play-by-play and the shifts. Time on the bench is
counted in 5-on-5 play outside the penalty box; the note says the player sat, not why. Goals saved are against the
shots' expected goals (xG).
2nd Period
0:20
BUF
No shift
No shift · Kozak, 45:08 on the bench
No shift after 4:56 of the 1st; 3 shifts (1:49) before that
13:10
TOR
Goalie
Goalie · Stolarz −2.2 goals saved
4 goals on 1.78 xG, 20 shots on goal
3rd Period
14:40
BUF
No shift
No shift · Metsa, 10:46 on the bench
No shift after 4:37 of the 3rd; 14 shifts (9:41) before that
Shot map
TOR shoots ←Bigger circle: higher xG; filled: goal→ BUF shoots
About these numbers
xG comes from hockeystats.io's own model, a gradient-boosted tree model trained on every unblocked shot
from the 2021-22 to 2025-26 regular seasons using the NHL's public play-by-play. Each shot's value is the
share of similar shots that became goals. A team's xG is the sum over its shots; "deserve to win" asks how
often the team would outscore its opponent if every shot went in at exactly that rate. Blocked shots count
as attempts but have no xG (the NHL records where they were blocked, not where they were taken), so the map
leaves them out.
An expected-goals take on Dom Luszczyszyn's Game Score: goals and assists, the quality of a
player's own shots, shot quality for and against while they were on the ice at 5v5, penalties, faceoffs and puck plays,
all in goals; goalies by goals saved above expected. From the NHL play-by-play and shift charts.
