Analytics with Alison: The Deserve to Win O'Meter

We all know that feeling…the one where the team you thought should have won goes home with a loss. It's something that Kraken faithful have dealt with on more than a few nights, especially over a recent stretch of stronger play against teams like San Jose and Boston.
"We got no reward for our effort tonight," Dave Hakstol said after the Sunday loss to the Sharks. "Those are things you have to handle."

But is this just hyperbole when these kinds of things are said? Is it overt optimism?
Thanks to the "Deserve to win O'Meter" at MoneyPuck.com, we can seek to legitimately answer the question "who deserved to win that NHL game?"
Let's dig in.
If you're not already familiar with it, Peter Tanner, the creator of MoneyPuck,

in November of 2019.
You can find the entire write-up
here
, but to summarize, the Meter is a tool that runs 1000 simulations of any given game using expected goals that occurred in a game and applying average goaltending to both teams. There are also other factors involved including which team is home and away. Empty-net goals are excluded.
For any NHL matchup, you can go to the
Games page
at MoneyPuck and after clicking on any game, watch the meter dial fluctuate as it runs through each of its 1000 simulations, reporting the likelihood of each team winning as the simulations progress.
Let's look at the game Hakstol was describing from Sunday. Was he right? Did the Kraken warrant a better fate?

According to MoneyPuck, Seattle DID deserve to win that game.
How does that likelihood fluctuate as each simulation is run? Let's take a look at that, too.

Obviously, this doesn't make up for standings points not gained, but it can make us feel better about what our eyes tell us about what happened and what our gut says should be the outcome of a game.
Lest we think that this means that the Kraken are repeatedly robbed of wins, understand it can go both ways.
Here's the Meter from the Feb. 11 4-3 win in Anaheim.

So yes, there are games that a team should have won but didn't, and there are also "steals" where you take the win and go, before anyone knows any different.
In total, if you go through the Kraken season to date, according to MoneyPuck, Seattle lost 14 times when it should have won; and won six times when it should have lost. That comes out to a net difference of six more wins on the campaign (MoneyPuck doesn't account for overtime or shootout wins).
But before we get too wrapped up in that specific number, it's important to realize that while those missed outcomes did happen, that means 35 games ended with the result the Kraken "deserved." So, more often than not, the Deserve to Win O'Meter does get it right. Even if the margin for a game is sometimes super close.
Here's the Feb. 24 game versus Boston that ended with an overtime loss.

A few things to remember about the Meter. While the major difference it applies in its analysis compared to what happened in the game is goaltending, don't presume that play in net is the sole reason for a certain outcome.
Two of the six "should have lost" for Seattle qualified as a Quality Start for the Kraken and one of those was also a Steal, and seven of the 14 "should have won" were Quality Starts by opponents' goalers and three of those were a steal. Superb or sub-par goaltending isn't going to be the only factor in what the final is on the scoreboard.
At the end of the day, the Deserve to Win O'Meter is just another tool in the toolbox to help inform what we see in a game. It predicts the likelihood of an outcome, but we always leave it up to the dynamic of actual play on the ice to see what the final result will be.