Evaluating Early Season Pitcher Performance

In the first month of the season, fantasy sites and blogs will be littered with your standard buy low and sell high columns. The authors will highlight hitters suffering/benefiting from a low/high BABIP or pitchers whose xFIP (or other expected ERA metric of choice) is dramatically different from their ERAs. I won’t get into why I think these articles are rather useless, but I do want to discuss how to go about evaluating early season pitcher performances. There is a two-step process I use and since I am feeling generous, I would like to share it with you.

I ask myself two questions when analyzing a pitcher during the season:

1. Are his results (ERA and WHIP) stats real?

2. Will he sustain those underlying peripherals?

Are his results (ERA and WHIP) stats real?

The first order of business is to do what the majority of fantasy owners likely do by now, or at least those a step above the newbie level. That is to compare the pitcher’s ERA to his SIERA (my personal expected ERA metric of choice) to determine how real his results have been. It is easy to notice a high BABIP and blindly assume the pitcher’s ERA should drop, but what if that comes with a low HR/FB ratio? It is difficult to do the math in your head and figure out how the luck in one metric offsets the poor fortune in another. So a quick look at SIERA can tell us how all those metrics mix together. Of course, this is FanGraphs, so you all probably do this step already.

It cannot be overstated though how much small sample size can affect a pitcher’s ERA. The actual number early on means so little as to render it almost meaningless. Only a couple of earned runs here and there can make a strong season look mediocre. To illustrate this point, imagine a pitcher who has allowed 16 earned runs over 45 innings (around a month and a half of innings). That results in a strong 3.20 ERA. However, if this same pitcher allowed just 4 more runs (just 1 extra run every 11 innings or so) to total 20, suddenly his ERA shoots up to 4.00 and he doesn’t even look appealing to a shallow league owner! So fantasy owners would be wise not to focus on ERA at all when analyzing a pitcher so early. That is why metrics such as SIERA are even more useful when dealing with small sample sizes.

Will he sustain those underlying peripherals?

This step is just as important as the first one, but I am unsure how many fantasy owners actually perform it. It is simply not enough to determine that a pitcher has been unlucky this year because his SIERA sits at 3.40, but a .330 BABIP and 68% LOB% has inflated his ERA to 4.20. One must also look at how sustainable his underlying peripherals are.

Remember that a pitcher has a greater opportunity to strike hitters out when they face more batters. So all else equal, a high BABIP is going to increase K/9 and reduce SIERA. Of course, it will also increase his BB/9, but the K/9 increase probably has a greater effect on SIERA than the BB/9 increase. So once his BABIP drops back toward league average (or his career mark), the strikeout rate is going to decline and the SIERA will rise. The better luck will help his ERA improve, but it’s not going to drop all the way down to where his SIERA had been.

Next, a pitcher may be getting lucky with strikeouts based on sequencing or an unsustainable rate of foul ball strikes. A check of the pitcher’s SwStk% and how his marks in that metric have historically matched up with his strikeout rate can give you an idea as to whether he will sustain his current K/9. Similarly, looking at the pitcher’s F-Strike% will tell you how sustainable the pitcher’s BB/9 is.

Even if the strikeout rate does match up with the SwStk%, you must then go a step further if the K/9 is dramatically different than past years. A look to see if there have been changes in pitch mix, velocity or a new pitch added to his arsenal is required to help determine how sustainable this new K/9 level is.

The bottom line is that it is not enough to simply compare SIERA with ERA to identify buy low and sell high candidates. You must dig deeper to determine if the components that drive SIERA itself are sustainable. If the components themselves change over the remainder of the season, then that SIERA mark is going to change as well and the pitcher’s ERA cannot be counted on to eventually meet it where it currently stands.

We hoped you liked reading Evaluating Early Season Pitcher Performance by Mike Podhorzer!

Please support FanGraphs by becoming a member. We publish thousands of articles a year, host multiple podcasts, and have an ever growing database of baseball stats.

FanGraphs does not have a paywall. With your membership, we can continue to offer the content you've come to rely on and add to our unique baseball coverage.

Support FanGraphs

Mike Podhorzer is the 2015 Fantasy Sports Writers Association Baseball Writer of the Year. He produces player projections using his own forecasting system and is the author of the eBook Projecting X 2.0: How to Forecast Baseball Player Performance, which teaches you how to project players yourself. His projections helped him win the inaugural 2013 Tout Wars mixed draft league. Follow Mike on Twitter @MikePodhorzer and contact him via email.

newest oldest most voted
Brent Crossman
Brent Crossman

Good post. One thing I’ve noticed in your posts is you reference swinging strike rate as being a predictor for strike out %. I’ve never found that to be any more correlated with future K% than simply using the historical K%. Perhaps there is a study linking the residual between (Expected K% based on SwgStrk% and Actual K%) vs Future Movement in K% but I haven’t seen it, and every time I test it I don’t a meaningful connection.

This is another good article on regression http://www.hardballtimes.com/main/article/it-makes-sense-to-me-i-must-regress/

(K rate and GB rate being the two that get reliable quickly, although still not for a while)