LINK: The Value of Not Doing Stupid Things
I’m stealing this link from Tom Tango, who pointed out Phil Birnbaum’s latest post over at his own blog. However, despite being a recycle for the folks who read both FanGraphs and TangoTiger.com, this one is worth putting here too, since I know there’s a decent amount of you who will read this and wouldn’t have seen it otherwise.
For those who aren’t familiar with Phil Birnbaum, he’s the editor of “By the Numbers”, the newsletter put out by SABR’s Statistical Analysis Committee. He writes about sabermetric stuff at his own blog and has a lot of interesting things to say. Including in his latest post, which is what we’re linking to here. A sample:
Why is this, that you can lose on purpose, but you can’t win on purpose? In this case, it’s deliberate, human-caused. When we invent games of skill, we keep the ones that have an interesting struggle to win. We don’t care whether there’s a struggle to lose, because, who cares? The object is to win.
Or, you can look at it this way. When there’s competition for a goal, it’s hard to win, because you have to beat your opponent, who’s trying just as hard as you. When there’s no competition for a goal — like losing — it’s easy, because nobody is trying to prevent you.
If everyone is trying for X, it’s hard to be the most X. But it’s easy to be the most “not X”.
…
It’s hard to win by intention, fluke, or skill — but it’s easy to lose by intention, fluke, or (lack of) skill.
I think Phil is exactly right here. The kinds of suggestions that we can make with publicly available data are unlikely to create a huge advantage for a team. Our community’s advice basically boils down to “hit the ball hard and try not to let your opponent hit the ball hard”, which, well, yeah. We can go a bit deeper than that, but at the end of the day, the best things sabermetrics have come up with are reasons to not do things.
Our forecasts aren’t so great that we can tell you exactly which free agent you should spend a lot of money on; our forecasts are good enough that we can tell you that Brandon League is not that free agent. Hitting Joey Votto second might be the statistically right thing for the Reds to do, but it matters less that Dusty Baker should hit Joey Votto second and matters a lot more than he shouldn’t hit Zack Cozart there. We can show, with a lot of evidence, that not hitting Cozart second would help the Reds a lot. Hitting Votto there, instead of some other non-Cozart hitter, would only help them a little bit more.
The value of conclusions that have come from statistical analysis seem to lie much more on the side of warnings rather than suggestions.
Anyway, read the whole thing, it’s worth it.
Dave is the Managing Editor of FanGraphs.
This somehow seems to me to be related to falsifiability in the philosophy of science. It’s often not possible to prove a hypothesis is true, but you can prove it false and we therefore assume things are true until shown otherwise.
“Whenever I’m about to do something, I think, ‘would an idiot do that?’ And if they would, I do not do that thing.”
– Dwight Schrute
So, Dwight doesn’t sleep, eat, drink or breathe?
Interesting article. So the Padres intentionally did a stupid thing by choosing Matt Bush first in the draft??
Not sure what this SABR blogger wanted to do. “Elimination of bad decisions” and “figure out what everyone else knows” require many skilled workers. Let me expound here:
The elimination rule requires 1.Knowing why. You don’t make decisions based on blind faith. 2. Methodology. There must be an established system that works first; it’s more of the perfection of the system (“better good decision”) than the reduction of the negative. There are more, but these two professional experience are the most vital.
Now, the figure-others part requires 1. Ability to obtain both published and unpublished information. “Insiders.” That’s not sabermetric anymore. Sabermetric may attempt to obtain more data through public means, but it won’t help. More data mean trash. NEEDED data means gold. 2. Reactive decision making. Not research, not time-consuming, long-term progress, but short-term reward.
Anyway, one last thing: Core competence. Not all organizations are good at all of the above skills. Hell, some might not even have the talents for ANY of the above.
It’s too much to ask, methinks. Stick with what you’re good at, and make the best out of it. You lose some and might win some, but at least can stay–Positive thinking and a happy winner.
*To make a note here, from 1) to 2) for the obtain-info above, an organization might need a bridge here. Not all researchers can do three things at the same time, esp. when the decision making must come fast, while info. pour in furiously.
Fate can play a hand at any time of its choosing rendering statistics eternally vulnerable.