Consistency is what I would consider to be the biggest factor separating good goalies from very good goalies in the NHL. The two measures that I value most are save percentage and percentage of quality starts (QS). A QS is a game above the league average in save percentage (90.8 currently), or a game with fewer than 20 shots and a save percentage above 88.5. Another useful stat is RBS (really bad start), which is a start resulting in a save percentage under 85.0.
In a lot of cases, goaltenders are the victims or the beneficiaries of the team playing in front of them on a nightly basis, but a quality start takes that into account a little more than wins or goals against average (GAA). For example, 41 saves on 45 shots, resulting in a loss and a GAA of 4.00 (assuming 60 minutes), would be considered a quality start this season as the save percentage is 91.1. Conversely, 17 saves on 20 shots, resulting in a win and a GAA of 3.00, would be a borderline really bad start at 85.0 save percentage.
In my opinion, over time, QS tells the story of consistency fairly well without getting into some other more advanced stats. The current league average for QS percentage is 50.6 in 2019-2020 and 49.8 over the past five seasons. The league average for RBS this season is 17.1 percent currently, while over the past five seasons it is 15.7.
League leaders in QS going back to 2016-2017 were Sergei Bobrovsky at 66.7 percent, Pekka Rinne at 66.1 percent in 2017-2018, and Frederik Andersen at 70 percent in 2018-2019. Here are the current leaders (min 13 GP) in 2019-2020:
Darcy Kuemper is leading with 84 percent QS after his first 25 games, while Thatcher Demko, Tristan Jarry and Jaroslav Halak are all above 70 percent. There is a very good chance that most of these goalies will be below 70 percent by the end of the season, although Kuemper has a good chance to duplicate what Andersen did last year.