2019 MLB Draft Guide: Understanding Sample Size

I’m going to take a page out of the Colton & The Wolfman playbook here and start you off with one of my favorite movie scenes of all time. If you’ve seen Coming to America, then sit back and enjoy. If you haven’t, well you better get to NetFlixing here and grab a comedic education, especially because they're currently in-process of filming the long-awaited sequel. If foul language offends you, cover your tiny, little baby ears or just don't watch this clip.

Now how does this relate to fantasy baseball? Because every time you ask a fantasy expert about a hot or cold start in April, they come back to you with the phrase, “sample size” just as fast as a white guy mentions Rocky Marciano. It’s a knee-jerk reaction.

You: “Hey Howard, what do you think of Michael Conforto hitting just .182 here on April 15^th?”

Me: “Not much. Look at the sample size. Dude’s had 42 plate appearances. You gotta give him some time.”

You can insert any player’s name and any statistic you want, good or bad, and you’ll probably get the same answer, at least up until about mid-May. You may think it’s a cop-out, especially if you’ve got a guy surging like Jed Lowrie last year, but it’s true. You need to give players a relatively longer look before you make assessments. You can get excited about a hot start or kick rocks over a cold one, but making any add/drop or trade decisions at this time is extremely dangerous.

For example…

Player A is batting just .205 with no home runs through the first two weeks of the season. The rest of your team is doing fairly well and you see Player B on waivers batting .385 with four dingers. You drop Player A, add Player B and watch as the calendar flips to May. Player A is now batting .240 with two home runs and Player B is batting .285 and stuck on the same four he had when you added him. Whoops.

How many of you have made this swap? Don’t be embarrassed. Raise your hand. We’ve all been guilty of this. Yes, even people who work in this industry. It’s a novice mistake and everyone learns the hard way. It’s why today we include Draft Guide articles like Patience is Your Best Friend.

And this is why we immediately come back at your questions with the ol’ reliable “sample size” answer.

Think about it. If Player A crushes it for the first half of the season and comes out batting that same .205 to open the second half, your level of concern is minimal. First of all, his first-half stats mask just how bad that .205 looks and your immediate thought is, “well he raked for the first three months of the season, so he’ll snap out of it.” You’re willing to give him a pass because three months’ worth of stats tells you a better story than two weeks does.

Think back to 2017 for a moment. Imagine if Aaron Judge opened the year the way he opened the second half that same season. How many people would have dismissed him as a strikeout-prone failure and dropped him immediately? The guy bats .329 with 30 home runs in the first half and then proceeds to hit .230 in July and .185 in August. Did anyone even think about dropping him in August? Of course not. You have to understand there’s a ton of ebb and flow in this game and players need a certain number of plate appearances or innings pitched to find their center.

How many at-bats? How many innings? I hate to invoke the snowflake metaphor, but every player truly is unique and some click sooner than others. But if you’re looking for a good rule of thumb to keep you from jumping off the ledge every April, here you go:

NOTE: How to read the charts below. If the number listed is 60 that doesn’t mean 60 total plate appearances, that means 60 plate appearances in which that event ended (i.e. the 60 listed for strikeout rate means there needs to be 60 strikeouts in place before you can read too much into the result).

OFFENSE

PITCHING

Again, these are rough estimates, not the gospel.

And if you do find yourself struggling with decisions as to what you should believe and what you shouldn’t, looking at past big-league seasons is a must. Let’s go back to what Jed Lowrie  did to open the season last year, batting .339 with six  homers and 27 RBI through the month of April. It was fun to watch and certainly exciting if you owned him, but looking back at what he once did in the majors and minors, you had to be skeptical. He hit 14 homers all year in 2017 and his career-best was 16 back in 2012. In between those seasons it was middling power, a blah batting average, mediocre on-base numbers and, of course, let's not forget the massive amount of time missed due to incessant injuries. Were we suddenly believing that this soon-to-be 34 year old was finally finding his power stroke? Come on.

For veteran players, a guy with at least three years’ experience tends to give you a better idea as to who he really is as a player. You can look at career averages for BABIP, strikeout and walk rates and ISO to understand if he’s playing over his head or even under. Sure, you’ve got your exceptions -- your late-bloomers -- like Whit Merrifield and Tommy Pham , but overall, three-year averages (another article that will be coming out soon here in the Draft Guide), work as solid sample sizes. Rookies and second-year players obviously don’t have the track record you’d like, but with a full understanding of minor league totals should certainly help you figure some things out.

So the next time you get worked up about a hot start or someone comes looking to trade for your struggling superstar in the month of April, don’t get too upset when the response to your question from me is a simple two-word answer – sample size.