MASTER NOTES: Unlucky pitchers

In assessing and projecting player performance, a core principle at is the importance of understanding and managing luck.

We know players will get lucky or unlucky—that’s what makes fantasy baseball unpredictable, which makes it interesting and fun.

And of all the counting stats in fantasy baseball, the one most prone to luck is pitcher wins.

Some stat sites include stats that seem to address luck in pitcher wins, but the stats depend too much on “Quality Starts.” A 4.50 ERA ain’t quality enough for me.

Instead, BHQ has a better metric: Pure Quality Starts. PQS is a more precise measure of pitcher performance, looking at every start by pitcher skills—counting hits, walks, strikeouts, and home runs, as well as innings.

PQS scores go from a 0 to 5, with 0 and 1 classed as “Disasters,” while 4 and 5 classed as “Dominant.”

I gave every pitcher “luck points” for every start, based on this table:

 0   +3   --
 1   +2   --
 2   +1   --
 3   --   -1
 4   --   -2
 5   --   -3

Then I added them up.

And guess what? It turns out that where luck affects wins, it's all bad luck.

It’s really rare for pitchers to get lucky wins. In 2013, pitchers lucked into wins in only 3% of games with PQS scores of 0, 1 or 2.

But it was really common to be unlucky. The average Luck Score was -11.5, and 38% of starts qualified as “unlucky” to some degree, including 11% of starts where the pitcher got PQS-5, as good as it gets, but didn’t get the win.

I also looked at individual pitchers’ total Luck Scores, and sorted them from the luckiest to the unluckiest. Nobody turned out to be especially lucky, in line with the overall results.

But there was a whole pile of bad-luck starters, including 28 who had Luck Scores of -30 or worse.

Let’s look at the top (or bottom) 5, including ties.

The unluckiest pitcher in 2013 was Cole Hamels of the Phillies. Hamels’ bad luck started when I bought him at the Tout Wars-Mixed auction for $25. And it got worse. He had a Luck Score of -49. His horrendous luck included nine winless PQS-5 starts and nine more winless PQS-4s.

Stephen Strasburg was next at -44, in a disappointing year for Washington.

Then came the real surprise. Yu Darvish had 13 wins and a great year—and he was unlucky! His Luck Score was -42, thanks to 17 winless PQS-4 and PQS-5 starts, including an midseason streak of six in a row.

So think about this: If Darvish had won even half of his unlucky starts, he’d have had 21 or 22 wins, adding $8 to his already impressive value, equaling Max Scherzer, and making the Cy Young discussion a lot more heated.

Chris Sale of the White Sox came next at -41 and then the Cubs’ Jeff Samardzija at -38, tied with Matt Harvey, Matt Cain and Madison Bumgarner.

OK, so how do we use this information? First, we want to draft skills, not outcomes, especially in stats like wins, which depend on so much more than the individual pitcher’s performance. Pitchers with high PQS averages are better pitchers.

Second, we have to think about why pitchers don’t win excellent starts. The primary culprits are poor run support and bad bullpens, so ask if those situations have changed for unlucky pitchers. If so, they might be worth the extra buck. If not, let someone else take the risk.

That could be bad news for Sale and Samardzija, since the Chicago teams won’t be notably better. The Sox added Juan Abreu, but they traded their closer. The Cubs have done little in the offseason.

But Darvish looks even more attractive for 2014. The Rangers have strengthened their run-scoring with Prince Fielder and Shin-Soo Choo, although at the expense of some defense, and their bullpen is something of an embarrassment of riches.

It’s an important fact in our game that we can’t change luck. But we can manage it, and one way to do that could be looking for bad luck in pitcher wins.


Click here to subscribe

  For more information about the terms used in this article, see our Glossary Primer.