OT What I hate about statistics.

“Lies, darned lies and statistics.”

A quote often attributed to, and popularized by Mark Twain. But apparently coined by the British writer Sir Charles Dilke in 1891.

And of course “garbage in, garbage out”. And, yes, measuring “garbage” can be statistically valid and reliable.

Aleph, all true. Your fourth question is the most challenging of all: most people don't even see their own biases when they are looking at data. I forget what it's called, something like the judicial paradox: That no judge is ever impartial, no matter how hard they try.

Same with statistics. Confirmation bias is one thing--and it's an easy trap to fall into, even easier in these days of massive data available to everyone. But the deeper search for data that supports one's belief--while ignoring data that does not--is the biggest joke of all among researchers. I've even seen it on posters in laboratories: All data that does not fit the hypothesis is to be discarded.

Aleph, I would argue that mathematics is not a science. It's a symbolic language that transcends all other languages.

When I was in grad school (education) we used this book in our stats class:

https://en.wikipedia.org/wiki/How_to_Lie_with_Statistics

Statistics never lie and all liars use statistics. The best thing I brought out of statistics class.

Except in weather, if they say (in KS) there's 50% chance of rain, that really means 100%. 40% is 50%, 30% is probably not, and 10% is not a chance in heck.

I agree that Stats don't lie. Only those with an agenda, IE politicians.

Yes, we all lie. If we were always absolutely honest, we would be at each other's throats. I guess if we didn't lie, we would always have to do social distancing. 🙂

I love certain stats. Career .297 hitter. 399 HR. 3007 hits. 18 time All Star. RIP #6 Al Kaline.

Growing up in Chicago, Al Kaline was one of my favorite players also. Not sure why but he represented the game well, was a clutch player, and feared by everyone when he came to the plate.

halbert said...
In what other activity is a 30% success rate considered great--and 40% unheard of?

How about these baseball numbers (guess the player and year):

304 innings pitched. 1.12 ERA. 24W-10L, 268 SO, 62 BB, 11HR. Lost 1 game when the opposing pitcher threw a no-hitter. Extra points if you who THAT pitcher was.

If a tie-breaker is needed after halbert's question:

Who played for the New York Rangers, the Brooklyn Dodgers, and the New York Knicks in a single season, and in what position?

Baseball, well sports the only place where stats warm my heart.
Living in KS, my all time favorite was George Brett.
The summer of 1980 was so memorable for more than the heat wave. Here's an excerpt from Wikipedia.

"All these impressive statistics were just a prelude to 1980, when Brett won the American League MVP and batted .390, a modern record for a third baseman. Brett's batting average was at or above .400 as late in the season as September 19, and the country closely followed his quest to bat .400 for an entire season, a feat which has not been accomplished since Ted Williams in 1941.

Brett's 1980 batting average of .390 is second only to Tony Gwynn's 1994 average of .394 (Gwynn played in 110 games and had 419 at-bats in the strike-shortened season, compared to Brett's 449 at bats in 1980) for the highest single season batting average since 1941. Brett also recorded 118 runs batted in, while appearing in just 117 games; it was the first instance of a player averaging one RBI per game (in more than 100 games) since Walt Dropo thirty seasons prior. He led the American League in both slugging and on-base percentage.

Brett started out slowly, hitting only .259 in April. In May, he hit .329 to get his season average to .301. In June, the 27-year-old third baseman hit .472 (17–36) to raise his season average to .337, but played his last game for a month on June 10, not returning to the lineup until after the All-Star Break on July 10.

In July, after being off for a month, he played in 21 games and hit .494 (42–85), raising his season average to .390. Brett started a 30-game hitting streak on July 18, which lasted until he went 0–3 on August 19 (the following night he went 3-for-3). During those 30 games, Brett hit .467 (57–122). His high mark for the season came a week later, when Brett's batting average was at .407 on August 26, after he went 5-for-5 on a Tuesday night in Milwaukee. He batted .430 for the month of August (30 games), and his season average was at .403 with five weeks to go. For the three hot months of June, July, and August 1980, George Brett played in 60 American League games and hit .459 (111–242), most of it after a return from a month long injury. For these 60 games he had 69 RBIs and 14 home runs.

Brett missed another 10 days in early September and hit just .290 for the month. His average was at .400 as late as September 19, but he then had a 4 for 27 slump, and the average dipped to .384 on September 27, with a week to play. For the final week, Brett went 10-for-19, which included going 2 for 4 in the final regular season game on October 4. His season average ended up at .390 (175 hits in 449 at-bats = .389755), and he averaged more than one RBI per game. Brett led the league in both on-base percentage (.454) and slugging percentage (.664) on his way to capturing 17 of 28 possible first-place votes in the MVP race. Since Al Simmons also batted .390 in 1931 for the Philadelphia Athletics, the only higher averages subsequent to 1931 were by Ted Williams of the Red Sox (.406 in 1941) and Tony Gwynn of the San Diego Padres (.394 in the strike-shortened 1994 season). "

He still is. Super Max in Colorado I believe.