I Hate OPS. Use this instead. A primer…

The other day the following graphic appeared on the screen when Matt Chapman of the As stepped into the box against the Yankees.

A .729 OPS and a 105 OPS+…my head dropped, I rubbed my eyes, looked upward and sighed. I don’t remember in what order.

Let’s start from the beginning. OPS is On Base Percentage plus Slugging percentage.

On base percentage (OBP) is exactly what it sounds like – the percentage of plate appearances that a batter has reached base safely. OBP is more directly correlated to team runs than any other than any other statistic. Runs are the point of baseball, if you were unaware.

Slugging percentage (SLG) is a player’s total bases (1 for a single, 2 for a double, 3 for a triple, 4 for a home run) divided by at bats. It measures a batter’s ability to advance base runners. The ability to advance base runners is damn important, but not quite as important as the ability to get on base.

Enter “OPS”. Let’s add the two together to REALLY measure a player’s contribution to his team’s run production.

Except it doesn’t. I’m hoping that you’ve already spotted a couple of problems here:

First, OBP is a percentage. SLG is not a percentage, it’s a ratio. To be clear, I’m not a mathematical person – I’m average mathematically at best – and adding a percentage to a ratio to come up with one number makes my head spin. The statistic has serious problems already.

Secondly, as noted, OBP is more important than SLG in terms of generating runs. Not only does OPS not represent that, it actually weights SLG more – SLG will be a higher number than OPS, so when they’re added together a larger percentage of it will be SLG.

And something of which you may not be aware, we know that a double is more valuable than a single, but it’s not twice as valuable (a double is 1.8 times more valuable than a single, to be exact). But SLG weights it as twice as valuable.

What we’re left with OPS, is a number that’s a combination of two different computations, which inaccurately weigh the relative importance of two outcomes, with no regard for rather large contextual factors (park effects and era for example).

To use Chapman as an example from above, his OPS would tell you he’s almost exactly a league average hitter. DJ LeMahieu is below a league average hitter according to his OPS, is another example.

The problem is that both are good at getting on base – Chapman’s OBP is slightly better than average, DJ’s is much better than average – but neither are hitting with power this season, so they’re SLG is low. The result is that even through they’re good at the most important aspect of baseball – getting on base -their OPS doesn’t represent that.

Dante Bichette (Dante – the father – not the kid on Toronto who can hit) has a higher career OPS (.835) than Rod Carew (.822). Let that sink in.

So OPS isn’t quite useless, but it’ll certainly do until useless shows up. So what should you pay attention to that’ll accurately reflect a batter’s contributions to his team?

Any stat that ends in a “+” sign, like OPS+ in the picture above. wRC+ and DRC+ work well too. Although they go about getting the final number in slightly different ways, they all have the important stuff in common:

They set league average to 100. So Matt Chapman’s OPS+ of 105 tells us he’s 5% better than an average hitter despite what his OPS says. (An OPS+ of 85 is 15% worse than average, a 125 OPS+ is 25% better than average, etc.)

“+” stats weigh OBP more than SLG, accordingly.

“+” stats are weighted by park factors (because it’s harder to hit in Oakland than Colorado)

“+” stats are weighted by the run scoring environment of the season (because it was harder to generate runs in 1967 than it was in 2000). This makes it easy to compare players in different eras.

When all factors are considered, the “+” stats give us a pretty accurate representation of a batter’s contributions to his team.

Here’s your My Baseball Page take home lesson kids: Ditch OPS.

Oh, in case you were curious, Rod Carew’s career OPS+ was 131. Dante Bichette’s was 107.

Told you so.

Did I miss something? Let me know.