February 14, 2017

Nick Mariano examines the advanced baseball metric SIERA as a new way to examine starting pitchers and the production they can be expected to provide fantasy owners.

Welcome back to the fantasy baseball classroom where we’re learning about winning strategies and advanced statistics that help us dominate.

Previous lessons have been focused on hitters, as we’ve explored BABIP as well as hard-hit rates.

Now we turn our attention to pitchers and their SIERA.

Why should I care about SIERA?

You’re probably familiar with ERA (Earned Run Average). Well SIERA takes that to another level as it stands for “Skill-Interactive Earned Run Average”. It was created to quantify how well a pitcher has truly performed based on, well, their skills. That’s rather ambiguous, so let’s wade in a little with a case study.

Scenario A: Pitcher A strikes out the side, notching a 0.00 ERA.

Scenario B: Pitcher B gives up a double, a single, and a walk to load the bases. He then gives up a hard line drive that is hit at his right fielder, who starts a triple play. Pitcher B also gets a 0.00 ERA for this.

Clearly, Pitcher A performed very well while Pitcher B got lucky, yet their ERA says they’re the same. This example is hyperbolic, but Scenario B actually happened last season. The takeaway here is that ERA leaves much to be desired when it comes to analyzing true talent.

What is SIERA?

First, a brief history lesson. Voros McCracken set out hoping to find a way to judge pitchers based on things that were in their control. That meant strikeouts, walks and home runs. Perhaps you’ve heard of FIP (Fielding Independent Pitching) and xFIP (FIP, but with a league-average home run rate allowed assumed). SIERA takes the next step as batted-ball data has become widely available. It incorporates the kind of contact a pitcher allows into the formula, the entirety of which can be viewed here.

SIERA looks at strikeouts and walks, but takes it a step further by positing that a high-strikeout pitcher usually induces weaker contact thanks to his swing-and-miss stuff. With walks, SIERA says a pitcher with better control is less "hurt" by issuing one walk. Most importantly, SIERA takes into account each pitcher’s batted-ball profile. If a pitcher gives up more grounders or more fly balls then SIERA will use that to grade outcomes. These little tweaks may not sound like much, but in this world, even the slightest step towards a more accurate metric makes for a huge advantage.

SIERA is also a park-adjusted metric, accounting for a pitcher calling mile-high Coors Field home versus one who pitches in spacious Oakland Coliseum. ESPN’s Park Factors does well to provide a baseline for which stadiums are hitter or pitcher friendly.

2016 SIERA Leaders

Let’s provide some context for a full season’s worth of data. Here’s your top-10 in SIERA from last season, along with their ERA, for starting pitchers with a minimum of 120 IP:

1. Clayton Kershaw - 2.41 SIERA, 1.69 ERA
2. Jose Fernandez – 2.81 SIERA, 2.86 ERA
3. Noah Syndergaard– 2.95 SIERA, 2.60 ERA
4. Max Scherzer – 3.05 SIERA, 2.96 ERA
5. Stephen Strasburg – 3.18 SIERA, 3.60 ERA
6. Madison Bumgarner - 3.36 SIERA, 2.74 ERA
7. Michael Pineda – 3.40 SIERA, 4.82 ERA
8. Justin Verlander – 3.42 SIERA, 3.04 ERA
9. Chris Sale - 3.43 SIERA, 3.34 ERA
10. Carlos Carrasco - 3.44 SIERA, 3.32 ERA

2015’s ERA leader is absent. Zack Greinke had a 1.66 ERA with a 3.27 SIERA (tied for 14^th) in '15. This led many to see some regression coming in 2016, although no one saw the 4.37 ERA coming. His SIERA ended up at 4.11, as the struggles were earned. He localized most of his damage to a few truly terrible starts, but the overall figure is going to reflect that.

Looking back at 2016 stats may have folks thinking Junior Guerra and his 2.81 ERA is worth a mid-round pick, but that has a mediocre 3.71 FIP behind it, which further shrouds a lousy 4.42 SIERA. Pass.

If someone inspected the SIERA leaderboard at the end of April last season, they'd have confirmed that Noah Syndergaard was good at baseball, as his 1.83 SIERA was No. 1. Some other names looking good at the close of baseball's opening month were Rich Hill (2.57), Rick Porcello (2.72), Aaron Nola (2.77). SIERA also wasn't buying two particularly hot starts: Mat Latos (0.74 ERA, 4.90 SIERA) and Jordan Zimmermann (0.35 ERA, 4.43 SIERA). Meanwhile, Adam Wainwright's struggles (7.25 ERA) were backed by his 5.91 SIERA. See how this can be helpful?

SIERA is a Fantastic Tool, But Isn’t Perfect

Please do not read this as “SIERA is the only stat you’re ever going to need and it is always right!” As always, never rely on any one metric. There are many other stats that help fill out a pitcher’s profile that we will explore over the weeks to come. Might I suggest soft contact generated? This is most relevant in recognizing how Kyle Hendricks (2.13 ERA, 3.70 SIERA) and Tanner Roark (2.83 ERA, 4.32 SIERA) are to be viewed.

Do note that there are types like Sonny Gray who have consistently outpitched their SIERA, and guys like Michael Pineda who have yet to live up to theirs. Pineda posted an amazing 207:53 strikeout-to-walk ratio through 175 2/3 innings last season, so his gorgeous 3.40 SIERA sat behind that unimpressive 4.82 ERA. Healthy control doesn't inherently mean a guy has command of the zone, as Pineda served up more meatballs than the Swedish Chef when he wasn't getting his strikeouts.

This is an evolving science. SIERA will probably be ousted by some metric down the line, just as it was the next step after FIP.

In the end, SIERA is one of the best available tools to analyze pitcher performance. Do they deserve that pristine ERA they’re sporting? Is it likely that their 4.00 ERA is hiding something more? We love targeting talent, so SIERA is very handy. Here’s the leaderboard if you want to poke around. I’m always available for a second opinion as well!