Goalie Goals Saved Above Expected (GSAx)

Quantifying a goalie's true stopping power.

What It Does

It calculates the total expected goals (xG) a goalie should have allowed based on the location, type, and context of every shot faced. This expected value is then compared to the actual number of goals conceded. A positive result indicates the goalie is performing above expectations, effectively saving goals that an average goalie would not.

Why It Matters

Traditional stats can be misleading; a goalie on a strong defensive team may face easy shots and post a high save percentage. GSAx isolates individual goalie skill from team defense, offering a more stable and predictive metric for a goalie's future performance and impact on a game.

How It Works

The core formula is GSAx = Expected Goals Against (xGA) - Goals Against (GA). The xGA model is built on historical NHL shot data, considering factors like shot distance, angle from center ice, shot type, and whether the shot came off a rebound or a rush. The value is often normalized per 60 minutes of ice time (GSAx/60) for better comparison between goalies with different workloads.

Key Indicators Analyzed

  • GSAx/60
  • Rolling 5-game GSAx
  • High-Danger Save Percentage

Data Sources

  • MoneyPuck
  • Natural Stat Trick
  • Evolving-Hockey

Details

Category
sports
Subcategory
hockey

Try This Pillar

Use the Goalie Goals Saved Above Expected (GSAx) pillar in PillarLab to analyze any prediction market.

Open Lab →