I get the question every year about this time. It takes some form of the following.

Which players should I target in the second half because they are traditionally ‘second half’ players?

I hate to burst your bubble, actually I don’t but wanted to write that to seem nice, but first/second half splits don’t mean anything. Here’s why.

1 – First/second half end points are completely random. Let me ask you a question. What’s the difference between July 5th and July 15th to a player? Answer is nothing. There’s no difference at all yet one date falls in the first half, one the second.

We all do this, chose random end points. I do this. It’s easy to do.

* Joe is batting .333 since May 25th.

* Frank has a 2.19 ERA his last nine starts.

You can have fun with numbers.

There are natural break points of course, with months and halves being the easiest to turn to. I get it. It’s pleasing to the mind to look at things in those blocks. Truth though, they are just arbitrary end points. There is no practical difference between July 5th and July 15th. Just the truth.

2 – The first/second half split numbers are not always instructive. By that I mean that the splits may tell you a story that just isn’t true. Let’s explore.

Robert hit .270 in the second half in 1901.
Robert hit .322 in the second half in 1902.
Robert hit .270 in the second half in 1903.

Which hitter is Robert in the second half?

You could say he’s a .287 hitter in the second half. That’s what you get when you add all three average numbers together and then divide by three. There are multiple problems with that position though.

First, how could Robert be a .287 second half hitter when only once in three seasons he actually hit .287? In fact, in two of the three years he wasn’t even close to being a .287 hitter.

Second, you’re assuming that he had exactly the same at-bat totals each year. Why? Because that’s how human minds work. However, what if I gave you the following data points. In addition to batting average, here are the hits/at-bat total of Robert each season.

Robert hit .270 in the second half in 1901. He had 68 hits in 252 at-bats.

Robert hit .322 in the second half in 1902. He was injured in the second half and limited to 118 at-bats and 38 hits.

Robert hit .270 in the second half in 1903. He had 71 hits in 263 at-bats.

So what type of hitter was Robert in the second half in 1901-03

Robert produced 177 hits in 633 at-bats. The result is a .280 hitter in the second half.

Not .287 if we divide by three.
Not .270 like he was two of three years.
Not .322 like he hit in 1902.

Robert was a .280 hitter the second half despite the fact that only once in three seasons he actually hit .280.

That’s the main point here.

Numbers can be deceiving when we talk first/second half splits.

 

Another example.

Dexter played from 1970-74. Here are his homer totals in the second half of each of those five seasons.

1970: three homers
1971: six homers
1972: ten homers
1973: four homers
1974: ten homers

So who is Dexter as a second half home run hitter? He hit double digits twice, but the other three seasons he didn’t more than six.

Add up all five numbers and the total is 33 homers. Divide that by five and the result is 6.6 homers a season. That means only twice in five seasons did he reach his 5-year “average.” Is he really a 6.6 homer guy? Again, we don’t have the context of at-bats for each season. Let’s do that exercise again.

1970: three homers in 43 at-bats
1971: six homers in 204 at-bats
1972: ten homers in 218 at-bats
1973: four homers 194 at-bats
1974: ten homers 211 at-bats

Here is the homer rate per season for Dexter.

1970: 14.3 at-bats per homer
1971: 34.0 at-bats per homer
1972: 21.8 at-bats per homer
1973: 48.5 at-bats per homer
1974: 21.1 at-bats per homer

Three times Dexter posted an at-bat per homer rate of less than 22 at-bats per blast. However, twice that number was 34 or higher including one season in which the mark was double the mark he posted in his big three seasons. Moreover, what is the total 5-year rate? Dexter hit 33 homers in 870 at-bats which leads to a 5-year average of a homer every 26.4 at-bats. Note that only three times in five years did he reach that mark and twice he was well above it. So I ask again, who is Dexter as second half homer hitter?

In summation this is what you need to know.

Merely depending upon second half splits to tell you the story of whether a guy is a “good” or “bad” performer is foolish.

The data doesn’t mean much.

Much more important than the splits are a whole host of factors such as the skill of a player. When did that stop to matter? I could even say take a look at the last three weeks as a better option than looking at meaningless split data. What about the health of a player right now? Is there an underlying injury speaking to first half struggles? All of that, not just some first and second half stuff, matters. Even the vaunted second half star Adam LaRoche, the poster boy for second half runs of excellence, the man who had an OPS for his career that was .071 points higher in the second half than his mark in the first half (.838 in the second half), posted second half OPS mark of .732, .788, .702, .789 and .514 in his career.

First and second half splits mean very little in the grand scheme. Don’t be tempted by the party line that suggests the numbers are meaningful. They aren’t.

 

Ray Flowers can be heard Monday through Friday, 7 PM EDT and Friday on SiriusXM Fantasy Sports Radio (Sirius 210, XM 87). You can also hear Ray Sunday nights at 7 PM on the channel talking fantasy sports. Follow Ray’s work at Fantasy Alarm and on Twitter (@baseballguys).