Showing posts with label Statistics. Show all posts
Showing posts with label Statistics. Show all posts

Wednesday, October 18, 2017

By the numbers

We are knee-deep in the baseball playoffs, which look to be yielding a potentially classic Yankees/Dodgers World Series. The games have been good, if rather long-averaging about half an hour longer than the regular season...in fact no game has been less than three hours (ALCS game two was exactly three hours).

Here, though, I wanted to talk about some regular season questions. We baseball fans like numbers. We argue about which ones matter, when if ever they should take second place to intangibles like the evergreen notion of the "veteran clubhouse presence", and whether "statheads" are universally guys who couldn't play (in my case, yes).




I just wanted to take a quick look at a few correlations between wins (ranging from the Dodgers' 104 to the Giants' and Tigers' 64), and factors like payroll, batter age, pitchers' strikeouts, etc. Nothing too sophisticated here. And yes, correlation is not causation. This is quick and dirty stuff. Anyway...


The highest positive correlation I found was between wins and pitcher strikeouts per nine innings: 0.76. Think teams like Cleveland, the Dodgers, Houston. Baseball is more and more about power arms, and numbers like this one will only encourage the trend.


The highest negative one was between wins and runs against per game, not surprisingly, at -0.86. The correlation for runs scored per game was 0.72.


There is a decent sized positive correlation for pitcher age, at 0.32, and a tiny negative one for batter age. The former figure would seem to contradict the number above on the value of young flamethrowers.


And with modest-payroll teams like the Astros and Diamondbacks making the playoffs, it's not a shock that while payroll is an important factor in success, it isn't necessarily central-the positive correlation here is 0.37

Wednesday, September 27, 2017

The bizarre ritual (anti-ritual?) of anthem protests

The phrase has a strange ring to it-"anthem protests". Who would protest the playing of their own national anthem? The anthem, after all, and the displaying of the flag, and any other civic ritual you could name, aren't some sort of declaration that a nation is flawless.

We celebrate America, or at least acknowledge its worth, as a sort of truism of shared values-that a people, with a given set of civic beliefs, have created and sustained a nation in line with those values. But what happens when those values are no longer widely shared?



I don't intend to wade into all the political controversies that have me convinced we're a nation in name only. I'll just mention that we have a founding document that establishes the rules of the civic game. Increasingly that document, and the laws that operate under it, are seen as just so many words on paper, when one side is unhappy with the results. When such a point is reached, where the consequences of a long-established democratic process are "illegitimate" because the wrong party wins or the wrong policy is instituted, we are in the process of rejecting what made America succeed. The entire system is based on the idea that the results of our political process are by definition legitimate.

The essential orientation of America-as a nation organized towards the promotion of freedom and prosperity-is now rejected by half the nation. Freedom is dicey because people can use it for purposes not tending towards "social justice". Prosperity-the creation of new wealth, with the ultimate purpose of starting and maintaining families-is irrelevant. Redistribution of existing wealth (with the exception of Silicon Valley's vast hordes of same) is paramount. The fact that greater redistribution doesn't yield greater economic equality is irrelevant.


So we have one side wishing to maintain the old order, and another rejecting it entirely. What the latter would establish may still maintain some of the forms of democracy, but they will be toothless relics. Whatever means necessary to get to nirvana will be acceptable.


UPDATE: Miriam has some interesting thoughts on what was intended to be a cost-free exercise in virtue-signalling.

FURTHER UPDATE: A further reason not to watch the NFL-an appalling percentage of these guys are thugs. Here are detailed NFL arrest records. 

ONE MORE UPDATE: "The brontosaurus in the room": Roger L. Simon on the lie behind the protests-the notion that cops are what's wrong with the black community.


AND FINALLY: No matter how you play with the numbers, it is black men who are doing most of the killing of black men.

Thursday, September 21, 2017

Trump tracking-down 4.1%

In the Real Clear Politics average of RV/LV, Trump stands at 42% approval.

This is down 4.1% from his election popular vote percentage, but up from 40% last time I checked this on September 13.


I track this not only because because I'm interested in this figure, but because I track all sorts of stuff-health numbers like blood pressure and resting pulse, twitter tweets per follower, a wide variety of baseball stats, etc. I do love numbers.

Wednesday, September 13, 2017

Trump tracking-down 6.1%

Trump's average approval among registered/likely voters (RCP) is 40%. He got 46.1% of the vote last year.

I've been tracking this and will update it on occasion.


By the way, the full breakdown of the vote totals by David Wasserman is fascinating, at least to politics junkies like myself.

Duke Ellington - "Arabesque Cookie" (Arabian Dance)

It's that time of year again. From Duke's 1960 "Nutcracker" adaptation. I don't think it's a stretch to say ...