Context Revisited

A while back I addressed the fact that a ton of people who analyze the game of baseball, mostly those who work for big/fancy corporations, either (A) have no idea what they are talking about, or (B) think that you are so stupid that you couldn’t possibly understand the nuances of analysis so they just continue to spew the same, slightly meaningful crap, as everyone else. Alas, it really isn’t that difficult to place things in the proper context – all you need to do is to apply a little elbow grease. After I re-reading that previous piece entitled Some People Never Learn, I thought I would expand upon the idea of context with a few more concrete examples of why it is so vital to do place whatever you are analyzing on some continuum to root it in the real world.

Which player would you prefer to have on your team?

Carl Yastrzemski who hit .301 in 1968 or Glenn Wright who hit .321 in 1930?

If that was all the information you had at your disposal you would clearly answer Wright since he hit .020 points higher, but where is the context? There is none. We could add the context of opportunity, i.e. at-bats, and try to derive an answer. Alas, that context still says the option to choose would be Wright given that he hit .020 points higher in only seven fewer at-bats than the 539 that Yaz racked up in ‘68. However, another level of context would be to discuss the era, or the level of competition that they faced in their respective season. We could talk about how there were no players of color playing in 1930, or the fact that teams traveled by bus and train and not airplanes, but those things aren’t really quantifiable since we would be dealing in the realm of supposition. So, why don’t we just compare each batter’s batting average directly to those who were also playing at the same time.

Yaz hit .301 in ‘68, the Year of the Pitcher, and the AL batted just .230.
Wright hit .321 in ‘30, and that year the NL batted, hold on to your hats, .303!

To arrive at how each player did in relation to their competition, you simply divide the players batting average by the league average to come up with an adjusted total.

Yaz: .301 / .230 = 1.31
Wright: .321 / .303 = 1.06

So what this context shows is that despite batting .020 points higher according to their raw batting averages, Wright’s batting average was only 6% better than the league average in 1930 whereas Yaz’s piddly .301 average, which led the AL by the way, was a full 31% better than the league average. Given that context, which batter would you have wanted for your hypothetical team?

And that’s the trick of context. Numbers are just numbers without context, and far too often people try to compare players of different era’sby merely looking at raw numbers without any deference being paid to context. Here are two concrete examples

(1) Shoeless Joe Jackson hit all of 54 home runs in his career. Willie Mays hit 660 home runs in his career. So Mays was the better power hitter, right? Not so fast. Partly because of huge disadvantage in batting average (.356 to .302), Mays finished his career with an OPS that was just .001 point higher than Jackson at .941. In addition, if we place each batter’s effort in context by comparing his effort against the league during the years that he played, we find out that Mays’ adjusted OPS says he was 56% better than his level of competition. As for Jackson, playing in the so-called dead-ball era, his adjusted OPS was 70% better than his competition. Ergo, according to this context, Jackson was actually a better offensive weapon as measured by OPS, and in fact by a rather large percentage, over Mays.

(2) Let’s line up home run king Barry Bonds with home run king Babe Ruth. In terms of raw numbers Barry Bonds out-homered Ruth 762 to 714. However, if we use context, who was the better home run hitter?

In 1919 Ruth hit 29 home runs. No one else in the AL hit more than 10.
In 1920 Ruth hit 54 home runs. No one else hit more than 19.
In 1921 Ruth hit 59 home runs. No one else hit more than 24.

I could do on, but the point is pretty obvious. Ruth wasn’t just winning the home run title, he was more than doubling his next closest competitor in some years. For all the home runs he hit, Bonds only lead the league in home runs twice (Ruth did 12 times), and in those two seasons Bonds led the league by a mere six home runs (46 to 40) and nine (73 to 64).

So the next time someone just starts throwing raw numbers out there when they are making a Hall of Fame argument take a moment, adjust things with some context, and come up with an answer to the question or comparison that actually makes some logical sense – even if the majority of sports writers view that concept as being as foreign as the ability to fly to the sun like Icarus.

By Ray Flowers

Tags: Babe Ruth, Barry Bonds, Carl Yastrzemski, Glen Wright, Shoeless Joe Jackson, Willie Mays