Two sides

In this past week’s newsletter, which you’ll get if you sign up for a free or paid subscription, I alluded to a disagreement John E. Gibson and I had on the Japan Baseball Weekly Podcast. As with our biggest clashes over the years, it had to do with player valuation.

I was happy to leave it at that until one of our collaborators raised the issue on Twitter, causing us to explain our side of the argument.

Unfortunately, when that started, we both began referencing our takes on the other’s side of the argument, and that didn’t go well.

It became clear I couldn’t summarize John’s point of view to his satisfaction and he may have felt I was being condescending, since I felt his description of my position was simplistic and totally lacking nuance.

You can listen to the podcast, it’s toward the very end, but don’t miss John’s interview with Hiroshima’s Ryan McBroom at the start.

John is more analog than me. He amasses impressions, catalogs them, evaluates abilities. He’s not impervious to digital information, but his visual record of what he sees is vastly more nuanced than mine, especially with hitters.

I am more digital. I see patterns in numbers, and draw inferences from the historical record to see if they can tell me anything. I try to find out why players’ results are what they are. This allows, me, in extreme cases, to make good guesses about how things are going to play out going forward.

We’re both guessing at our interpretations and their meaning. Objective proof is pretty darn scarce, but an advantage of a digital record is that it leaves a trail. The trail may be in code and hard to decipher, but sometimes the signal is strong enough to contribute to actual knowledge of the game.

John has more knowledge about competition and teamwork in sports than I do. I stopped playing team sports after junior high school. He can see red flags where I don’t. He’s careful about what he doesn’t know.

I see the parts of the game as percentages, perhaps because I’ve always been curious about ways to beat the system. Thus it should come as no surprise to anyone that I got interested in counting cards.

One summer vacation while I was at the University of California Santa Cruz, I worked five graveyard shifts a week at a 7-11 to cover my living expenses. On my first night off each week, I would make the four-hour drive up to Lake Tahoe, stalk the $3-minimum black jack tables, and when I either lost my stake or doubled it, would drive back.

My system, playing at tables where hands were dealt from a single deck with only one other player at the most, kept me marginally ahead of the house, doubling my stake more often than I’d lose it.

My big disaster occurred when there were no suitable $3 or $5 tables, so I tried a $20 table where I was the only player. However, I shot past my quit line before I knew it and was up about $1,600 in about 15 minutes. I didn’t stop because I’d just got there.

I ended up leaving with just enough money in my pocket to stop for breakfast at a Denny’s, buy a Chronicle, and read about the downing of Korean Airlines Flight 007 by Russian fighters over the Pacific.

That’s just an aside, but the point is, I’ve been been looking for hacks and good percentage angles since I was a boy, and baseball has a treasure trove of them for those with a sufficient nerd quotient.

John argues rightly that there are lots of things that don’t show up in stat lines that affect the outcome of baseball games. I love that side, too, and contrary to popular belief do not ignore it, but being the person I am, I try not to ascribe too much value to performances that are poorly accounted for.

We can, with some accuracy, measure the contribution of a player’s batting and base-stealing numbers to team runs, and we know the relationship of runs to wins: improve your team’s run scored or runs allowed totals by 10 runs, and that’s worth about a win.

We can also infer that the batter has a larger impact on the result of an at-bat than the pitcher does. We know this because no pitcher strikes out batters anywhere near as often as a batter who strikes out a lot, and no pitcher allows home runs as often as a good power hitter or as rarely as the most powerless slap hitter.

Because the relationship between runs scored and allowed and wins is understood, and because the impact of offensive results on runs scored is understood, and that the batter is the biggest single determiner of the outcome of each plate appearance, one can estimate that the value of the stuff we do know is somewhere around 40 to 45 percent of every win and loss.

Base running is kind of an orphan and without additional information, is assumed to be average when it’s not.

Fielding and pitching, too, are tied together in a knot that is hard to untangle from the perspective of numbers. We don’t know how much better fielding contributes to more aggressive pitching that reduces base runners by allowing fewer walks thus defusing more big innings.

What we do know is that there is only 100 percent of a win or a loss that can be assigned to a team’s players for every game.

Thus everything else that isn’t represented in the offensive record, the pitching, the non-base-stealing running, the fielding at all nine positions, can only add up to around 55-60 percent of the total credit or debit for each win or loss over the course of a season. Go beyond that and you enter the realm of wishful thinking.

Those things other things are hugely valuable, but there is a real ceiling somewhere on how much credit one can assign to them. It’s not a matter of comparing home run numbers, and ignoring what is uncountable as has been suggested.

The spark of our argument was over the relative contributions to wins by Koji Chikamoto and Munetaka Murakami in 2019, when Murakami won the Central League rookie of the year award. Chikamoto was a better than average center fielder, and an excellent base runner. John characterized Murakami as a player who would make you cringe except when he had a bat in his hand, but admitted this week that he was probably extreme in that criticism.

John said Murakami didn’t run the bases well, but Delta Graphs which has somebody counting how often extra bases are taken, rated his base running as a better than average. His evaluation of Murakami’s fielding was equally brutal.

Murakami DID miss a number of foul pops in the first half of the season, but he had decent range, again as measured by people who watched every game. John’s original evaluation of Murakami’s defense was similar to the one Bill James attributed to the New York press’ evaluation of Wally Backman’s fielding with the Mets, that “They’d have you believe he’s been barred from McDonalds because they were tired of picking up all the French fries he’d dropped.”

Better than average defense in centerfield is of course worth far more than average defense at first base. The question is how much more the extra “uncounted for winning plays” are worth in the final summation.

By my evaluation, Murakami produced a little more than five wins with his batting and base stealing, and another 1/3 of a win by being able to play first base. I have Chikamoto contributing three wins with his batting and winning an additional 1.1 games over the course of the season with his fielding.

How much better could Chikamoto be than that evaluation? The impact of his exceptional base running could easily be worth an extra third of a win a season, perhaps even a half a win, but that’s a lot.

The absolute best valuation for an outfielder in my evaluations is 2.43 wins, by Koji Yamamoto in 1975 and Hichori Morimoto in 2007. Chikamoto wasn’t in that range, but if he was 1-2/3 wins instead of 1.1, and we subtracted credit for those wins from his outfield teammates. That would make his season worth.

To get him past Murakami, we need to give him even more credit for his base running, and give Murakami even less for his defense. Neither of these are beyond the bounds of reason, so I would argue that Chikamoto could have had a more valuable season rather than couldn’t, and I’m willing to accept others seeing it as being more valuable. It’s not a huge stretch.

But while we were borderline slagging each other off during the week, I brought up this exchange we had during our Pacific League preview in March that kind of illustrates where we stand.

Jim: “The biggest predictor of how well you’re going to do the next year is not how many games you won, but how well you scored and prevented runs from scoring. As bad as the Hawks were, they were better than anybody else in the Pacific League at scoring and preventing runs.”

John: “You say that, and I agree…that scoring runs and preventing runs help you win games. But you know what helps you win games? Winning games. And they didn’t do it.”

“How can I pick this team to finish fifth? When they were in games they had a chance to win, they didn’t win them.”

OK. The thing about “winning games helps you win games” isn’t really indicative of a reasoned argument, because every team is in games they have a good chance to win but don’t.  I see where John is coming from although I dare not paraphrase him lest he once more belittle my ability to state his point accurately.

We both know that teams can play better than their opponent and still lose, so simply winning is not in itself a predictor of future success. But I think John is talking about more than just winning but about little stuff.

So often, however, people confuse being better with winning. We define champions as the teams that win the most games, because we want to celebrate champions. It’s not contradictory to say champions aren’t necessarily the best, but because it SOUNDS like a contradiction, some people get all huffy about it.

The idea that a 143-game season determines the best team can objectively be proven false, but people say it anyway, since no one wants their favorite team’s championship to be slagged off by those who say they weren’t the best. The two things are not the same, and that pisses people off. Champions don’t have to be the best team, it just helps.

If you want to pick a team to do well next season, go for one that scores a lot of runs, allows few but doesn’t finish high in the standings, like SoftBank in 2021.

Thirty-three teams in Japan since 1950 led their league in win percentage, but not in Pythagorean win percentage. Seven of those “winners” finished first the next year, 10 second, eight in third, two in fourth, three in fifth, three in sixth.

Thirty-three teams led in Pythagorean win percentage but not winning percentage. Sixteen of these finished first the next year, four in second, one third, four in fourth, three in fifth, five in sixth.

There’s something else at play there, too, the law of competitive balance. Teams that do well, are less likely to proactively address potential failings than teams that don’t.

These are historic trends that are knowable, and they are great help when you want to see more than just trees and comprehend the scope of the forest, and what the odds of success are.

Subscribe to weekly newsletter

Leave a Reply

Your email address will not be published.