post

Looking at Sample Sizes for Fantasy Baseball

If you’re on Twitter at all, you’ll notice many people taking their victory laps on players through the first week or two. I’m guilty of it. Just look at my tweets about Ketel Marte or Zach Eflin. I have failed to mention Jackie Bradley Jr. or Travis Shaw, two players I was very high on in the offseason (and still am). The point is, there’s more than 90% of the season left to be played, so we need to take a step back. I’ve included two charts below that summarize sample sizes for hitters and pitchers and the “stabilizing point” for many of the metrics. There’s been a lot of research over the years where relative conclusions have been made and I’m not a statistical expert, so I am taking their word for it.


What the stabilizing point means is that a player is likely to repeat his performance if his skills remain exactly the same. It is not a predictor of future performance. A player who changes his approach or gains/loses skills renders the previous sample size data to be somewhat meaningless. However, baseball players are creatures of habit, so you can at least feel comfortable knowing that unless you hear, read, or see a change, this data can be useful going forward. Let’s dive into hitters metrics.

(Note: BIP=Balls in Play; AB=At-Bats; PA=Plate Appearances; BF=Batters Faced)

Stabilization Rate for Hitters - Sample Sizes

MetricNo. PA, AB, or BIPGames Started to Stabilize
Strikeout rate (K%)60 PA~15 games started
Walk rate (BB%)120 PA~30 games started
XBH rate1610 PA:~402 games started
HR rate (HR%)170 PA:~42 games started or 45 FB hit
Batting Average (AVG)910 AB:~260 games started
On Base Percentage (OBP)460 PA~115 games started
Slugging Percentage (SLG)320 AB~100 games started
Isolated Slugging (ISO)160 AB~50 games started
Groundball Rate (GB%)80 BIP~29 games started
Fly Ball Rate (FB%)80 BIP~29 games started
Line Drive Rate (LD%)600 BIP~218 games started
HR per FB Rate (HR/FB)50 FB~61 games started at a 30% FB%
HR per FB Rate (HR/FB)50 FB~45 games started at a 40% FB%
HR per FB Rate (HR/FB)50 FB~36 games started at a50% FB%
BABIP820 BIP~298 games started
Hard Contact% (Fangraphs)80 BIP~29 games starts
Exit Velocity (Statcast)45 BIP~16 games started
Launch Angle (Statcast)50 BIP~18 games started
Hit Distance (Statcast)60 BIP~22 games started
Swing%50 PA~13 games started
Contact Rate (Contact%)100 PA~25 games started
*Where r^2 is 0.5

What we notice is that batting average, BABIP, extra-base hit rate, and line drive rate won’t stabilize within a single season. We can essentially ignore them for in terms of a sample size to date, except of course realizing a player with a .100 BABIP or a .500 BABIP will normalize. What we do see is that swing rate (swing%) and strikeout rate (K%) stabilize at just 50 and 60 plate appearances respectively. We already have some hitters beyond or approaching these numbers. We can see that players like Jose Peraza, Jonathan Schoop, and Adalberto Mondesi have extremely high swing rates (over 60%) and are likely to remain high to rest of the season unless they change their approach. On the other end of the spectrum Ben Zobrist, Andrew McCutchen, and Mookie Betts have well below-average swing rates. Mookie continues to be a more patient hitter than he was in his early 20s and he’s become a much more complete player. Cutch has always been patient and now that he’s leading off for the Phillies, he’s seeing even more pitches which is great!


Now for strikeout rate. Matt Chapman is striking out just 8.6% of the time compared to 23.7% last year. This is huge. Also interesting to note, Cody Bellinger has only struck out just 11.8% of the time after 23.9% in 2018. We aren’t at the stabilizing point just yet and opposing pitcher quality can play a role, so I’m not going to draw any hard conclusions. Even still, it looks like both Bellinger and Chapman have worked on and improved their plate discipline and contact rates. They could both be in for monster years and maybe even in the discussion for MVP. On the other hand, Ramon Laureano, Willy Adames, and Aaron Judge have elevated strikeout rates early on. We knew they would have elevated K rates and I’m confident Aaron Judge will still succeed despite the swing and miss because he’s done it before. But, Laureano and Adames have me concerned. They don’t have Judge’s elite exit velocities and could be at risk of losing playing time especially with their sub-3.0% walk rates.

A few other metrics to look at soon are the Statcast metrics. Depending on how much contact a player is making, somewhere between 16 and 20 games played those metric stabilize. The aforementioned Bellinger and Christian Yelich both averaging nearly 96 MPH on balls in play. Yoan Moncada is another player who has had a hot start and seems to be putting up elite metrics to back it up. Don’t worry about Ronald Acuna, he’s striking the ball well and things should turn around for him quickly (He smoked a home run last night). I have some concern for Charlie Blackmon, Adam Jones, and Jurickson Profar. They are near the bottom of the list in terms of exit velocity, so keep an eye on them moving forward.

Stabilization Rates for Pitching Metrics - Sample Size

MetricNo. BF, AB, BIPGames Started or IP to Stab
Strikeout Rate (K%)70 BF~17.1 IP
Walk Rate (BB%)170 BF~41.5 IP
Hit By Pitch (HBP%)640 BF~156 IP
XBH Rate1450 BF~353 IP
HR Rate (HR%)1320 BF~322 IP
Batting Average (AVG)630 BF~153.7 IP
On Base Percentage (OBP)540 BF~131.7 IP
Slugging Percentage (SLG)550 AB~134.1 IP
Isolated Slugging (ISO)630 AB~153.7 IP
Groundball Rate (GB%)70 BIP~24.5 IP
Fly Ball Rate (FB%)70 BIP~24.5 IP
Line Drive Rate (LD%)650 BIP~158.5 IP
HR per FB Rate (HR/FB)400 FB~484.8 IP at 30% FB Rate
HR per FB Rate (HR/FB)400 FB~363.6 IP at 40% FB Rate
HR per FB Rate (HR/FB)400 FB~290.9 IP at 50% FB Rate
Soft Contact% (Fangraphs)325 BIP~118.2 IP
Exit Velocity (Statcast)175 BIP~63.6 IP
Launch Angle (Statcast)45 BIP~16.4 IP
Hit Distance *Statcast)100 BIP~36.4 IP
*Where r^2 is 0.5

Well, it looks like home run rates will not stabilize within a single season.

Let’s look at a couple of pitchers who have thrown enough innings to warrant a “stable” strikeout rate. Of course, Max Scherzer has another great strikeout rate sporting a 35.9% K rate through his first 19 innings. He looks primed to finish inside the top three SPs once again this year. Jose Berrios has a 27.6% strikeout rate through his first three starts after 25.4% K rate last year. It will be interesting to see if he can improve on his strikeout rate for the second straight year. I’m concerned about Miles Mikolas who did not post high strikeout totals last year but his K rate through 16 innings this year is just 11.3%. Keep an eye on him, he may not be shallow mixed league worthy going forward.


Ground ball and fly ball rates also stabilize relatively quickly. Take Justin Verlander, who succeeds by throwing high heat up in the zone. He gives up a lot of weak fly balls and popups with this strategy. So far through his first three starts, his ground ball rate sits at 45.8% after a groundball rate of just 29.1% last year. That’s a stark difference. We are still at least one start away from the stabilizing point but he’s throwing his fastball less often (12% less) and introducing a changeup. Given the pitch mix change, the increased groundball rate is starting to make some sense.

Launch angle is the Statcast metric that stabilizes the quickest. There’s Marco Estrada with a launch angle of an incredible 34.5 degrees! This is what he does as his career LA is 21.6 degrees but he typically induces a ton of popups. Currently, his popup rate is an insane 30%. That won’t last but he might be a decent streamer when he pitches at home in that big ballpark. Scherzer’s launch angle is just 11.3 degrees after 20.3 degrees last year. He’s another pitcher who succeeds similarly to Verlander by throwing high-heat. However, there has not been any changes with his current pitch mix. It’s something to monitor but it could mean fewer home runs. Scherzer has averaged 25.8 home runs per year the last four seasons.


Follow me @FreezeStats. Check out my work at FantasyPros and Pitcher List.


Photo Credit: by Steve Slowinski
FanGraphs

Data Sources:

Sources

by Jonah Pemstein
September 19, 2016
Fangraphs
Russell A. Carleton
Baseball Prospectus
by Eric Seidman
May 22, 2009
FanGraphs

Pizza Cutter