#KilltheWin

Updated! Updates in italics.

Believe me, I realize that, including this one, my last two pieces have involved killing. It’s just that some of these anachronistic stats serve no purpose any longer. So, why beat around the bush? Get rid of ’em. You’re out!

This is the second of a series of three articles i’m going to be writing in which I will make the case for “killing” a certain baseball stat. As you may already know, today I #killthewin.

First, let’s break down the Win stat to its core. What is a “Win”? According to MLB.COM:

A pitcher receives a win when he is the pitcher of record when his team takes the lead for good — with a couple rare exceptions. First, a starting pitcher must pitch at least five innings (in a traditional game of nine innings or longer) to qualify for the win. If he does not, the official scorer awards the win to the most effective relief pitcher.

So, in order for a starting pitcher to obtain a win, he must have pitched “at least five innings” and have the lead? If he doesn’t, the official scorer just hands the win over to the most “effective” relief pitcher?

Effective?

Go ahead, you try to define effective in this case. I’m waiting.

Also, a starting pitcher has to give you five innings to qualify for a win, but a reliever isn’t held to the same standard? What type of BS is this?

That’s problem numero uno with this stat, it’s unclear, it’s vague, and it measures absolutely nothing.

In order to tackle this problem, we have to ask ourselves one very simple question:

What does the win stat tell me about a pitcher?

Example #1: Last Friday, Masahiro Tanaka had pitched the minimum five innings, allowing one run, when in the top of the sixth he strained both hamstrings running home from third to score the tying run. He was replaced by Jonathan Holder who pitched a scoreless sixth, followed by Chad Green who pitched scoreless seventh, and then in the top of the eighth, a 2-run home run by Brett Gardner gave the Yankees the lead. Dellin Betances follows up by pitching a scoreless eighth to preserve the lead for Aroldis Chapman who closed the doors in the ninth.

Chad Green was awarded the win that night. That’s no slight on Green, he’s become one of the most effective relief pitchers in baseball, but how does that make you feel? If you’re like most people, you’ve resigned to the fact that that’s just the way things go. BUT does it really make sense that Greene is awarded the win only because after he pitched a scoreless inning, the Yankees offense scored a run? Greene has zero control over the situation. Shouldn’t the win be awarded to Gardner for hitting the 2-run home run that resulted in the Yankees obtaining the lead? Or should it go to Gleyber Torres who scored the game winning run on the Gardner home run?

You see, there is the other problem. A pitcher’s win-loss record is contingent upon his offense, a part of the game that pitchers often can’t control.

In this case, what the win stat told me about Green’s performance was that the Yankee offense scored the game winning run after Green pithed a single scoreless inning.

In other words, it told me nothing about Green’s performance.

Example #2: This season, Jacob deGrom leads the NL in ERA with 1.55. The next guy, Max Scherzer has a 2.00 ERA. However, deGrom’s record stands at 4-2, while Scherzer has a 10-2 record.

Let’s take a 10-game sample size.

In the last 10 games deGrom and Scherzer have pitched, these were the results:

Screen Shot 2018-06-14 at 4.30.03 PM.png

deGrom has allowed 1 or less runs in nine out of his last 10 starts. Scherzer? Four out of 10.

deGrom has pitched seven or more innings in nine out of his last 10 starts. Scherzer? Five out of 10.

deGrom’s team has scored two or more runs in four out of last 10 starts. Scherzer? Nine out of 10.

In early MLB polls, Max Scherzer leads the vote for Cy Young. In second place? Jacob deGrom. The only possible explanation for this is that Max Scherzer has more wins than deGrom does. So, I ask again, how does that make you feel? Is Max Scherzer giving his team a better chance to win games than deGrom is? Or is it simply that deGrom’s offense is failing him? If that’s the case, should we punish deGrom with a loss or a no-decision because of that?

In this case, what the win stat tells me about deGrom is… nothing.

Example #3: The average CY Young Winner post 1967, when a winner was awarded in each league, has 20 wins. The only time a pitcher with a sub .500 record won the award was in 2003. That pitcher was closing pitcher Eric Gagne. Only two pitchers have won the award with a .500 W-L Record: Rick Sutcliffe in 1984 and Bruce Sutter in 1979. The remaining 100 pitchers finished with a better than .500 record.

Here’s the average numbers for all winning pitchers since 1967:

Screen Shot 2018-06-15 at 9.17.43 AM
Check out my spreadsheet for more info.

More than any other stat, voters will rely on the win stat to award a pitcher the most prestigious pitching award.

Just look at the 2016 AL CY Young voting. Rick Porcello edged out Justin Verlander. Why?  Because he had more wins. Just look at the numbers:

Screen Shot 2018-06-15 at 9.33.40 AM

Their numbers are pretty much identical, except that Verlander beat Porcello in most categories. Verlander had a lower ERA, more IP, a lower WHIP, allowed less Hits per Nine Innings (H9), and struck out more batters per nine (SO9). Porcello edged Verlander out in FIP, which I detail further below, walked less batters per nin (BB9), allowed slightly less home runs per nine (HR9), and of course had more wins.

Ask yourself, which of the stats in the table above describes each pitcher best? I bet the win stat was the least informative one.


What good am I, if I just go around killing stats? There has to be an alternative, right?

Not quite. At least, not a single alternative.

One good measure of a pitcher’s performance is Walks + Hits per Innings Pitched (WHIP). WHIP measures how well a pitcher keeps runners off base. Keeping runners off typically results in less runs. However, it doesn’t account for when a batter reaches on an error (#killtheE), fielder’s choice, or hit by pitch (HBP).

WHIP = (Walks + Hits) / Innings

Top 5 in WHIP

  1. Justin Verlander: 0.76
  2. Corey Kluber: 0.83
  3. Max Scherzer: 0.85
  4. Gerrit Cole: 0.88
  5. Aaron Nola: 0.93

Another good measure is Fielding Independent Pitching (FIP). An improvement of ERA, FIP attempts to measure events a pitcher has the most control over, like strikeouts, walks, HBP, and home runs. It completely eliminates what happens when a ball is put in play since a pitcher cannot control what his fielders do behind him. The problem with this stat is that it doesn’t take into account pitchers that have a tendency to create weak contact.

FIP = (13 x HR + 3(BB + HBP) – 2 x K / IP) + FIP constant

Top 5 in FIP

  1. Max Scherzer: 1.88
  2. Jacob deGrom: 2.00
  3. Luis Severino: 2.24
  4. Justin Verlander: 2.28
  5. Trevor Bauer: 2.33

There’s also Skill-Interactive ERA (SIERA). An improvement on FIP, SIERA also measures only what a pitcher can control including batted ball tendencies. So, since a pitcher’s FIP doesn’t tell the entire story, SIERA attempts to fix that. A formula isn’t included simply because it’s too complicated. Trust me.

Top 5 in SIERA

  1. Max Scherzer: 2.20
  2. Gerrit Cole: 2.51
  3. Chris Sale: 2.70
  4. Jacob deGrom: 2.75
  5. Corey Kluber: 2.77

No stat is perfect, but some are better than others. The win stat isn’t even in the conversation. It’s guilty of being the one stat attributed to a pitcher that says absolutely nothing about him. Therefore, I hereby sentence the win to fast and swift death.

May it rest in peace.

Leave a Reply