Have you ever been sleeping soundly at night only to be woken by a silly question that popped into your head? That just happened to me.
The question was: how often does a goalie steal at least one point for his team? I lay awake for a while thinking about the data that could let me answer the question, figured out how to present the results, and then settled back in for a little more sleep.
Definitions
To steal at least a point, a goalie must have given up at least 2 fewer goals than the expected goals against, his team must have won the game or have gotten it to the shootout, and his team must have scored fewer goals than their expected goals against.
Let’s look at some selected games from the 2023-24 season of the Vancouver Canucks.
In the first game, the Canucks’ goalie didn’t have a “good game”: he gave up 1.97 fewer goals than expected (2.97-1). His team won, as shown by the AW in the AWSO column. AWSO means Any Win (regulation or overtime) or Shootout.
In the second game the Canucks’ goalie had a good game, giving up 2.78 fewer goals than expected. This led to “GG” in the “GoodG” column, but the Canucks lost so it wasn’t a steal.
In the third game the Canucks’ goalie had a good game, and his team won, but it wasn’t a steal because the Canucks didn’t need the goalie to have a good game to get the win. When you score double-digit goals, the play of your goalie in the game does not matter.
Game four is a steal! The Canucks’ goalie had a good game, his team won, and his team wouldn’t have won if the goalie had had a normal game. What could have been a 3-2 loss (based on expected goals against) became a 2-0 win.
Game five is also a steal! The Canucks’ goalie had a good game, his team got to the shootout, and his team wouldn’t have gotten to the shootout if the goalie had had a normal game. Interestingly, this game also counted as a steal for the Minnesota Wild. Both goalies were hot in that game.
Restrictions
I would have loved to name names, but the data (Games report from NaturalStatTrick.com) does not identify the goalies. One must accept the data one has.
Goalie Steals, 2023-24
The columns in this table show: the number of times a team won in regulation or in overtime or got to a shootout (AWSO); the number of times their goalie had a good game (GG); the number of times their goalie had a good game and they won or got to the shootout (WS), and; the number of games the goalie stole at least one point for them (GS). Teams are sorted by GS and then GG.
All teams are shown so the reader can see how their favourite team did compared to the rest of the league.
Florida and Winnipeg had the best defense in the league in 2023-24, each giving up 198 goals, yet they aren’t at the top of the table. That’s because they do a good job of scoring. They had games where the goalie had a good game, but the team would have won had the goalie been average.
San Jose had the worst defense in the league, giving up 326 goals, but are tied for second in the table with 9 goalie steals: this is an unexpected result. The Sharks took advantage of their goalies’ good games. If it wasn’t for those stolen points, the Sharks would have finished even lower in the standings.
Edmonton goalies did not steal a single point for their team. In the eight games in which their goalie played well, the total score was 37-4.
I do not see the performance of Edmonton’s goalies as “lacking”. The Oilers’ goalies happened to have all their good games at the same time as their offense was having a good game. Had the offense had average outputs in those games, the Oilers still would have won them all, but the goalies would have been given credit for several saves (four, by my estimation).
Goalie Steals, 2024-25
The data for this season is partial, including all games up to and including games played on February 22, 2025.
It is reasonable to think that the league leader, when the season has been put to bed, will be one of the top four teams in this table.
The Knights and Kings have greatly benefited from their good goalie games, stealing points in 13 of 15. As a point of contrast, the Jets and the Habitants have had 21 good games from goalies, but only 12 of those games were steals.
Ottawa, with a new goalie in town, has moved up the table and has as many good goalie games (8) and stolen point games (4) as they did last season.
Boston, having lost one of their stud goaltenders to Ottawa, has as many stolen point games as they did last year (5). Boston is unlikely to have as many good goalie games as they did last season (12 last season, 6 so far this season).
T.B. and N.J have 20 good goalie games between them, yet have only 2 stolen games. Their total score in those 20 games was 82-13. Those games were an embarrassment of riches for them, getting both good goaltending and good offense.
Summary
Some articles have taken months to publish. Most articles take a couple of weeks. This article is the fastest one I have ever written.
I had the idea at 2 AM on the morning of February 23, 2025. I got up at 6 AM, had the data analysis done by 7:30 AM, finished my first pass at the article text by 8:30 AM, and had it ready for review by 9:00 AM.
Related Articles
None