Assigning Win Shares
A sane way to credit player contributions to a team's wins
If you’ve read these pages enough you’re likely aware of my battle with the various WAR measures.
It’s really a “me” problem, as everyone else seems fine with naming a metric after something it doesn’t actually measure.
Last year the 88 win Astros finished with 40.5 fWAR (wins above replacement) per Fangraphs. It makes no sense to those of us that are order driven.
I mean, just the thought of typing that makes me cringe a bit.
Really, they should have totaled 7 fWAR right? (or 7.5 since they only played 161 games).
I recognize I’m not going to win this battle, but in an effort to make it make sense to me I developed the “Win Shares” metric, which uses fWAR to calculate each players contributions to wins.
In effect, I’m assigning wins (or portions thereof) to each player.
The Astros have 10 wins. The total win shares I assign should equal that number.
Spoiler alert - it does.
Doing it this way recognizes the value of the fWAR metric, but makes it make sense to me by the total equaling the number of wins a team ACTUALLY has earned.
The Astros currently have 4.8 fWAR and 10 wins, which means it’s taken 0.48 fWAR to win each game on average.
I should note that this metric is net. You can gain and lose win shares just like you gain and lose fWAR over the course of the season.
Meaning, this is a snapshot in time.
The batters have a net of 3.3 wins to date, which makes sense given the struggles at the plate for this team.
The pitchers have accumulated a net of 6.7 wins to date.
The team has netted 10 win shares to date.
Also of note, the team so far has not been as efficient as the 2024 version with a win for every 0.48 fWAR.
Last year’s team registered 0.46 wins per fWAR.
That might not seem like a big deal, and it’s not right now, but as the season develops it could become a factor.
What it means is they are having to work harder for their wins.
I’ll update this periodically during the season.
Expected Wins
Winning the first two over San Diego puts the Astros at .500 on the season, but the slight negative run differential (one run) puts them just below that mark in my formula.
As always, thanks for reading!