3 ms·
This is a beautiful and well written article, but I find the argumentation rather weak. Let's roll up the reasoning backwards: The article claims that changes
by weinzierl 9y ago
This is a beautiful and well written article, but I find the argumentation rather weak. Let's roll up the reasoning backwards:
The article claims that changes to three factors (weight, bounciness and air resistance) could have caused a 25 percent increase in the number of home runs[1]. The increase in the number of home runs could be from the possibly altered balls flying farther [2].
The factors they looked into are weight, bounciness and air resistance and their contributions break down as follows:
- 5 feet, less air resistant, according to previous research [3]
- 3 feet, increased speed around 0.6 mph, according to previous analysis of bounciness [4]
- 6 inches, decreased weight by 0.5 g of the ball, which their analysis did not show. [5]
Never mind that 5 feet 6 inches is a little less than 8.6 feet.
When it comes to bring the argument home the article either refers to other sources instead of linking the claims together or it suggests a connection it fails to support.
All the lengthy talk about the composition of the core is interesting and I enjoyed reading about it. It could have been used to support the argument about the overall weight, except the article didn't show a significant change in overall weight. It could have been used to support the argument about bounciness, but it fails to draw that conclusion while referring to previous research in a hand-wavy way.
I think they should just have skipped the CT scans, thermogravimetric analysis and the scanning electron microscope analysis. For the saved money they should have bought more balls.
They could have weighed them precisely and measured their bounciness.
This wouldn't have made such a beautiful article but it would have probably made a much stronger argument.
--------
[1] "According to Nathan’s calculations, this would lead to a more than 25 percent increase in the number of home runs. Asked whether these changes in combination could have significantly affected the home run rate, MLB declined to comment."
[2] "Combine all these factors together — a lighter, more compact baseball with tighter seams and more bounce — and the ball could fly as much as 8.6 feet farther."
[3] "According to a previous analysis performed by The Ringer, that increase in bounciness alone would add around 0.6 mph to the speed of the ball [..]"
[4] "[..] previous research at FiveThirtyEight showed that they also became less air resistant. The decrease in drag is probably a result of a smaller, slicker baseball with lower seams. The change in air resistance could add an additional 5 feet to the travel distance of a fly ball."
[5] "The overall weight of the balls also dropped by an average of about a 0.5 grams between groups, but, unlike with the cores, this difference was not statistically significant.x
The ball as a whole weighs much more than the core alone, and there was more variation in the weight of the full baseballs than in the weight of the cores, both of which meant that the bar for statistically significant variations in weight was higher for the whole baseball than for just the core."
- nkurz 9y agoCan you unpack their explanation of [5]? It rings false to me, because I think it's based on a false assumption of the distribution of baseball weights. I think their assumptions are that both the weights of the total ball and weight of the core follow normal distributions, but after showing that the cores do not, they continue to assume that the weight of the whole balls does. This feels like one of the classic critiques of frequentist statistics, where effect size gets conflated with sample size. Worse, shouldn't we presume that there is a (strong) quality control step, which eliminates the tails of the distribution for overly heavy and overly light balls, but perhaps not for heavy and light cores? Which means that our assumptions that both the cores and balls follow the same normal distribution for ball weight is wrong from the start, and thus the notion of "statistically significant" shouldn't depend on total weight of the ball. That is to say, I think the evidence is just as strong that the cores are .5 g lighter as that the balls are .5 g lighter, but it's just a quirk of p-value and our (known-to-be-false0 assumptions of normal distribution that make one conclusion "significant" and the other "not significant". They did find that the balls were .5 g lighter, right?