Educating Twitter: Judging Joey Gallo

I’m warning you in advance, there’s going to be a little bit of math in this one. I’m sorry, but it can’t be helped, and the reason is that the tweet I’m responding to this week butchers some basic baseball math.

The tweet was from an account called “@notgaetti”, whose whole schtick is to throw out insulting commentary about more modern methods of evaluating baseball players. It’s actually pretty entertaining most days. Anyway, here it is:

Lots to unpack here, so I’m going to break this into chunks.

Let’s start with something simple: Don’t use a single embarrassing video clip to judge a player’s entire skill set. If you go to the clip he’s referring to, it’s a pretty bad look for Joey Gallo. No question it was a bad play. But one play doesn’t give Gallo a “low baseball IQ,” and it certainly doesn’t make him a bad baserunner.

You know who was caught stealing more than anyone in the history of baseball? Right, it was Rickey Henderson, the greatest base stealer of all time. I guarantee that I could find one boneheaded steal attempt in the 335 times Rickey was caught stealing during his career, but that wouldn’t be a fair representation of his skill set, would it?

The same is true of Gallo. Overall, Gallo is a pretty good baserunner. He has a 78% career success rate when he tries to steal, which isn’t often, but that’s an above average success rate. He takes extra bases on the basepaths a bit more than league average, too. His career total of baserunning runs is +8, a perfectly fine number that means he’s a touch better than average. He’s simply not a liability as a runner, so taking his worst moment and representing it as if his entire baserunning career is filled with moments like this is sort of a jerk move.

Next, let’s talk about that “he’s a .177 hitter” stuff. I apologize in advance if this is remedial for some folks, I’m not trying to insult anyone’s baseball knowledge. I just want to be sure that everyone is on the same page. For those who know these basics, feel free to skip ahead.

There was a time in baseball when batting average was king, and players were valued for the percentage of times they came to the plate and got a hit. That’s what batting average (abbreviated either as BA or AVG) is: It’s hits (or H) divided by at-bats (or AB), so the calculation is:

H / AB = AVG

The gold standard was .300, meaning the player got a hit in 30% of his at-bats, with a league leader typically having an average of .330 or better. Someone batting .400 hasn’t happened in more than 75 years.

Batting average has always been a flawed statistic, because hits are not of equal value. Any extra base hit (home run, triple, or double) is worth more than a single, because they move runners further along the basepaths. Yet they all count the same in batting average. That leaves lots of room for ambiguity.

Enter slugging percentage (or SLG), which give the hitter credit for every base their hits account for. A single is one, a double is two, etc. The formula for SLG is total bases (TB) divided by at-bats. Or:

TB / AB = SLG

A slugging percentage of .500 or better is the mark of a particularly good hitter, one who collects a lot of extra-base hits. League leaders will typically have a SLG mark of .600 or more. This doesn’t tell us as much about the frequency of a player’s hits, but it tells us way more about the kind of hits he gets.

But we still don’t have a complete picture of the hitter, because there are all sorts of plate appearances (or PAs) that don’t count as “at-bats”. Being hit by a pitch (HBP), for instance, counts as a plate appearance, but not an at-bat. Hitting a sacrifice fly (SF), same thing. The big one is reaching base by walk, or base on balls (BB). None of those kinds of plate appearance are included in either batting average or slugging percentage, but they’re all valuable because their either drove in a run (like a sacrifice fly does) or put a runner on without making an out (like a walk or hit by pitch does.) Those plate appearance are included in the calculation of on-base percentage (or OBP), which is Times on Base divided by Plate Appearances, or (Hits + Walks + Hit by Pitch) divided by (At Bats + Walks + Hit by Pitch + Sacrifice Flies). Abbreviating that, high school algebra style gives us:

(H + BB + HBP) / (AB + BB + HBP + SF) = OBP

I warned you there would be math.

If it helps, just think of on-base percentage as “The percentage his times at bat when he didn’t make an out.” The only limitation you have in baseball is outs. There’s no game clock to stop the contest regardless of the score. In baseball, you get to keep playing as long as you don’t make three outs in the inning, so it’s really, really important NOT to make an out when you’re batting.

A good on-base percentage is .350 or so, meaning “I didn’t an out in 35% of my plate appearances.” League leaders usually have an OBP north of .400.

None of these should be viewed as a stand-alone stats for evaluating hitters. The more information we have, the more complete picture of a hitter we get. You’ll often see batting average, on-base percentage and slugging percentage grouped together for a hitter, like this:

.300/.400/.500

That’s sometimes called the hitter’s “slash line”, and it tells us a whole lot more than any one of them alone.

None of this means that batting average isn’t important. It is. You want to know how often a player gets a hit, because it helps paint the picture of that hitter. But it’s an incomplete picture if you stop there. Without knowing the kind of hits the player is getting, or what happens in all of his other plate appearances, you can’t fully judge him.

Back to the tweet about Gallo. Yes, he’s hitting in the .170s this year, and is a career .197 hitter as of this writing. League average this year is .248 as of now, so he’s well below that. But his career OBP is .323, which is roughly league average because he walks a lot (almost 15% of the time, compared to a league average of 8.4%), and his career slugging percentage is .465, almost 40 points higher than league average, because a lot of his hits are for extra bases. So dismissing him as simply “.177 hitter Joey Gallo” is either short-sighted intentionally misleading.

Credit: AP /Abbie Parr

Next, the tweet refers to Gallo’s “OPS+ of 100.” Here’s what that means.

In recent years, folks have taken a player’s on-base percentage and slugging percentage and added them together as sort of a quick-and-dirty shorthand view of him as a hitter. That’s where you get OPS:

On-base Plus Slugging = OPS

So, if a good OBP is .350, and a good SLG is .450, then a good OPS is .350 + .450 = .800. League average would be a bit lower than that, something like .317 + .413 = .730. A league leader will be north of .900, or even 1.000.

Back to Gallo. His slash line this year is .173/.300/.430. As noted, the batting average is well below league average. The OBP is slightly below average, and the SLG is slightly above. His OPS is .730. League average is .731, so it would appear he’s producing at a roughly league-average level.

That’s where the “OPS+” comes into play. As you likely know, some ballparks are harder or easier on hitters. They have shorter fences, or deeper outfield walls, or big/small amounts of foul territory, or wind currents that impact balls in the air, and so on. Because of that, an OPS of .730 produced mostly in a home park that’s HARD to hit in is more impressive than a .730 OPS produced mostly in a home park that’s EASY to hit in.

In short, not all OPS marks are created equally, so adjustments are made. I won’t get into the more complicated math involved, but the basics are that players’ OPS marks are adjusted to account for the ease/difficulty of their home park, then converted to a scale where 100 is exactly league average. Anything above 100 is good (think “he’s producing at 105% of the average), while anything below 100 is bad (“he’s producing at 95% of the average).

Joey Gallo’s home park in Minnesota is somewhat hitter-friendly, so while his OPS of .730 is .001 below league average, his OPS+ converts to 97, which is a 3% below league average. (It’s dropped slightly from the 100 mark he was at when the tweet was written.)

As shown, Gallo’s overall batting numbers are pretty much league-average, or just a touch below. His baserunning numbers overall are also pretty much league-average. He’s done this in a ballpark that’s hitter-friendly, which means he’s starting with an advantage and is still only producing average numbers. If we look at the Win Above Replacement, or WAR, stat (which is explained here in far better detail than I can provide), American League hitters are producing roughly 1.0 WAR for every 300 or so plate appearances (309.5 to be exact). Gallo, as of right this moment, has 327 plate appearances, so he should have produced 1.0 WAR or so offensively. Instead, he’s produced 0.4 offensive WAR. Roughly half a win isn’t a huge miss across 300+ plate appearances, but it is still a miss compared to where an average offensive player would be.

In other words, our tweeter was right to mock the notion that Gallo is an average offensive player this year. Gallo IS below average. But the tweeter is absolutely wrong to reach that conclusion based on just his batting average and one boneheaded play as a runner. The real reason is that Gallo is making outs more than he should, especially given the friendly hitting environment in his home park, and he’s no longer far enough above average as a slugger or baserunner to compensate for those extra outs he’s making.

Basically, our tweeter just demonstrated why your math teacher always wanted you to show your work. Sometimes we get the right answer, but we get there by dumb luck instead of by actually knowing what we’re talking about.