I woke up thinking about Thursday's game. Two days later, I don't feel any better about it. The Mariners had a 7-1 lead and Felix was pitching but Felix fell apart and the Mariners regained the lead but the Mariners still lost and I realize it's just such a vicious cycle. I really needed a distraction.
Then I remembered an Alex Carson tweet that caught my attention.
What's fair is fair: This is opposite of a vulture. Where some relievers steal wins, Farquhar doesn't really deserve a loss if he gets one.— Alex Carson (@AlexCarson) June 21, 2013
Felix didn't give up the lead that night, but he sure didn't do his part to hold it. Danny Farquhar gave up the lead, then Kyle Seager got it back, then Carter Capps imploded and finally Yoervis Medina walked in the go-ahead run. And after all of the back and forth, Farquhar and Capps were credited with blown saves and Capps was tagged with the loss.
Felix gave up seven runs and he got a no decision. Capps gave up one run while he was on the mound and he got the loss.
And then that got me thinking: Why do we still hold on to this archaic mode of thought? Why is this the way baseball doles out wins and losses?
Don't get me wrong -- wins aren't meaningless. Using them to evaluate and compare one pitcher to another is completely daft, but wins do mean something. They are all that matter when wins and losses are tallied up at the end of the season.
And by "we" I mean not the Mariners blogosphere. I know our readers are familiar with the overvaluation of pitcher wins, saves and so forth. But I still have acquaintances that were excited by the thought that Felix could win 15 games this year, and that Wilhelmsen could notch 30 saves. There are baseball fans out there that believe the pitcher tagged with the loss deserves most of the blame.
Someone on the Mariners' roster needs to get the "W" or "L" by their name. That's never going to change. I don't believe it should, either. Is it fair? No. But while baseball thrives on the minutia of the game, someone needs to carry the burden of the win or the loss. It's marketable and mainstream and it's a statistic that has been seemingly ingrained into the baseball fan's thought process.
So how can we change the paradigm?
The first step is to use Win Percentage Added as a benchmark for player contribution towards wins and losses. Lookout Landing already does this -- we include the player that contributes the most both positively and negatively to the game's Win Expectancy with every game chart.
The next step is simple: The player that contributed the most positively to the WE will get the win if the team wins, and the player that contributed the most negatively gets the loss if the team loses.
So let's look at this thought experiment in action. The data on the left is Seattle's Win/Loss record culled from Baseball Reference, and the data on the right is how we would distribute wins/losses if we used WPA as the key metric for distributing wins and losses.
So to state the obvious difference: Now we have position players like Morales credited with three wins and one loss, Montero with two wins, Ryan with a win and so forth. This method acknowledges the player who contributed the most to the victory or defeat within the context of the game.
But this methodology, just like traditional wins and losses, isn't without its flaws.
On June 18, Bonderman and Morales had the same WPA. I couldn't credit both players with a win, since it wouldn't align with the team's overall record. So I made a judgement call: I credited Morales with the win since Bonderman only worked six innings, and Morales' go-ahead hit came late in the game, when leverage is weighed more than in the beginning of the game.
It also still puts the burden of the decision on one player, and WE clearly shows that a victory isn't contingent on one player's contributions.
But now we can expand the responsibility of the win or the loss beyond pitchers. Why was Zunino credited with a loss? Well, it was low-scoring environment, he grounded into a double play in one at-bat, and flied out with runners at first and second in another. Joe Saunders has five wins on the season, but via WPA Win/Loss calculations, he's only really contributed to two of these wins in a major fashion.
Ideally, I'd want this system to mirror shutdowns and meltdowns. As it stands, it's a very crude realignment of the way things are already done, and maybe it will help foster critical thought among non-sabermetric minds. It might not be perfect, but it takes a step towards breaking the archaic cycle.