Sample Sizes and Starts Like Story’s

Trevor Story hit seven home runs in his first six games of 2016. Right now Max Scherzer has one of the highest ERAs in baseball. As a Tiger fan, I hear complaints over the struggles of Miguel Cabrera, Justin Upton and Francisco Rodriguez. How much can we trust April stats though? Will these guys continue to struggle? Is it crazy that we are even worried as fans? Should you keep them in fantasy baseball? The answer to all of those is maybe, but usually no.

Also, don’t go out and drop Max Scherzer just yet.

At the beginning of the season it’s easier to point out players’ struggles because there’s no padding to the stats. Miguel Cabrera had a 0-for-19 stretch in April of his 2012 Triple Crown season. At the end of the season nobody remembered that. That’s for two reasons. The first reason is he didn’t end up being bad for the whole season, in fact the struggle was an outlier for Miggy. The second reason is that in the end he had amazing stats, and it was hard to remember the bad stretch because of what happened.

Essentially, I’m talking about sample sizes, and this isn’t a new topic. In fact, it’s been covered many times by sites like FanGraphs. I just want to put my own spin on sample size and exactly when you should panic over your players.

In a sports statistics class that I took during my senior year in high school we went by the rule that generally speaking for a stat to stabilize you need a sample of at least 30. In baseball, there’s 162 games though, so that sample needs to be bigger. FanGraphs supplies us with a list of when a statistic becomes stabilized, or could be taken as a real trend. A sample for hitters from FanGraphs (they have them for pitching stats too):

60 PA: Strikeout rate

120 PA: Walk rate

240 PA: HBP rate

290 PA: Single rate

1610 PA: XBH rate

170 PA: HR rate

This means that Trevor Story’s home run rate has not stabilized yet, as he hasn’t even had half the plate appearances needed for it to stabilize. So what do you go by? You should go by past trends and statistics. Maybe if the player was getting older you could think that you should slightly adjust for age.

In Story’s case, there are no major league numbers or trends so you could look at his minor league stats and figure that they could be similar in the majors too. Matt Duffy hit 12 homers in his first major league season after failing to hit double digits in the minors and in college he struggled to hit for any power at all.

Sometimes young players develop or mature their skills, or even discover a new skill and you have to treat it differently, but when a player like Story starts off hot, you also shouldn’t expect them to keep up the pace.

In 2006, Chris Shelton hit nine home runs in his first 13 games. He hit three in his last 102 games. So do I expect Story to keep it up? No, not at this rate, but reading scouting reports and looking and analyzing minor league stats and armed with the knowledge that he plays half of his games at Coors Field, he could realistically approach 30 homers this year and hit between 20-30 in a normal season. We will have to see what his rate is when it stabilizes. He will go through a cold streak that cancels out the hot streak, and when making choices, it’s important to value sample size.

For a more experienced guy like Miguel Cabrera or Justin Upton, it’s easier to surmise that they will obviously have better parts to their season. Miggy is getting older and coming off two injury-laden seasons, but he’s still a great hitter and hasn’t hit under .300 since 2008 when he still managed a .292 batting average.

Upton has struck out at a rate of 41.7 percent this season, but in his career that rate is usually closer to 25-to-26 percent, so I will take that large sample over his first 80 PAs for a new city and new league.

Francisco Rodriguez has gotten off to shaky start, and he was risky as an acquisition as he has been aging, but I’m not ready to give up on him yet. I’d give it a little longer. There is reason to be more concerned about him than say a Justin Upton because Upton is 28 and coming off many strong seasons where K-Rod is in his mid-30s. Rodriguez could be like what Joaquin Beniot, who was about the same age, was for the Tigers, or he could be what Joe Nathan was. Nothing in the way he has pitched has been overly concerning yet, though, so I would think that he will rebound.

When dealing with sample size it’s important to value the larger trend over the small one, know the context of the situation, and not panic or celebrate too early. That’s because most of the time, the trend will even out close to what the player has done in their career. You don’t have to be a mathematician to figure that out, although being one does really help you understand these sample sizes.

So don’t cut Max Scherzer or say the David Price contract already is a bust, because the sample is so small and some days guys just don’t have their best stuff or they go into slumps, and you have to value their career numbers until the stabilized trend says so or you can really spot a difference in how the player is playing.

Leave a Reply