Colorado Rockies: How to overhaul the pitching win statistic
The Colorado Rockies won their first game of 2022 behind the absolutely stellar pitching of team ace, Germán Márquez. Márquez pitched seven innings of three-hit, one-run baseball, only to have the win taken from him in the eighth by new bullpen addition Alex Colomé.
Colomé only lasted two-thirds of an inning and allowed four hits and a run before being relieved by Carlos Estévez for the final out. Estévez threw two pitches and received the decision for the entire game.
Those two pitches better have been important for the Colorado Rockies. Well, they were.
The Max Muncy fly-out was the third most important play of the game behind Connor Joe’s home run and Mookie Betts bloop single off of Colomé by WPA (Win Probability Added). While Estévez’s work was important, was it truly worth the win? Let’s look at pitching wins and how they can be fixed.
Estévez is a bad example though, so let’s look at perhaps the worst pitching win in Colorado Rockies history to show why a change may be necessary — Bruce Ruffin on June 7, 1994.
Ruffin came into the game in the eighth inning with the Rockies up 8-4 against the visiting New York Mets. When Ruffin recorded out number three of the inning, he had allowed four players to cross home on five hits and a walk, tying the game at 8-8.
The Rockies would rally to score two runs in the bottom of the eighth and then Mike Muñoz proceeded to put three Metropolitans down in the ninth. Bruce Ruffin got the win despite recording an ungodly -0.336 WPA.
For those who don’t know, the reason this is an issue at all is how decision-making for pitchers works. It entirely depends on the offense and is a relic of the time before the designated hitter. Despite being a solely defensive position, a win depends entirely on when the offense takes the lead. Most simply put, a pitching win goes to the pitcher who is listed as the team’s active pitcher when they take the lead, even if they do not take the mound following the lead.
For example, pitcher A allows six runs over two innings, but in the top of the third inning, pitcher A’s team scores seven runs. In the bottom of the third, the team uses pitcher B instead, who throws seven innings of shutout baseball. Pitcher B does not receive a decision but Pitcher A gets the win. That doesn’t seem right, does it? It allows pitcher A to be rewarded with one of the (wrongfully) most important stats for awards season. So how can this be fixed?
I have two fixes for this problem, one simple and one complicated. I feel both would better represent baseball pitching decisions than the system that we have now.
Let’s start with the simple one.
Simple pitching decision fix: Flip it
That’s right, just flip the decision tree, instead of having it be the pitcher who was pitching before the team took the lead, have it be the pitcher to pitch after the team took the lead. This is easily the simplest solution and would fix a lot of the issues that people currently have with the system.
In the examples from before, Daniel Bard, Mike Muñoz, and Player B would have been the ones to get the win for the Colorado Rockies, while Estévez, Ruffin, and Player A would have received no-decisions. The only time that the old style of decision-making should be used is in the ninth inning for walk-off victories. I am sure I am not the only person to propose this, but it is beautiful in its simplicity and fixes some issues that are presented by the current system.
That doesn’t mean that this system is perfect. There are still issues where an undeserving pitcher would get the win.
For example, back to players A and B, but now let’s add in player C.
Player A pitches 2 innings and allows 6 runs. In the top of the 3rd, Player A’s team scores 9 runs, Player B comes in and allows 2 runs over 0.2 innings and is pulled for player C, who plays scoreless baseball for the rest of the game.
Now, using the old method, Player A would get the win. Using this new flipped method, Player B would get the win, but that doesn’t seem to make sense. Player B had an awful outing. The team would’ve probably lost if Player C had not been put in. While immensely rare, this scenario may still cause inappropriate players to receive the win, though at a significantly lower rate.
So what scenario might the wins truly go to the pitcher who deserves them the most? That gets a bit more complicated.
The complicated answer: WPA
An important clutch statistic that has gained prominence recently is Win Probability Added (WPA). This stat calculates how much a player’s performance truly affected a ball game. For example, a strikeout with bases loaded in the ninth inning of a 4-4 ballgame would be significantly more important to the outcome of the game than a pop-out in the eighth in a 10-0 ballgame. The strikeout would produce a much higher WPA than the popout.
So fixing the issue with pitching wins with this is simple: give the win to the pitcher with the highest WPA of a game after their team takes the lead.
For our earlier examples, Daniel Bard, Mike Muñoz, Player B in the third scenario, and Player C in the fourth scenario would all get the wins, as their WPA was the highest of any pitcher after the lead was taken. This would be the most certain way to ensure that the pitcher that deserves the win the most gets the win.
The only wrinkle in this idea is that it would heavily favor closers as they are commonly thrown into high leverage situations and would benefit from having two innings where they could receive a win, whereas most relief pitchers would only have one.
Conclusion
I highly doubt that MLB will change the way wins are awarded anytime soon. It is a flawed system that could easily be improved, especially with the DH in place and the Ohtani Rule in effect. There is no longer a reason to keep the system as it stands. Colorado’s first win of the season showed why that was the case.
Note: Data for this article was provided by Baseball-Reference