Hacker News Reader

Comment by warcher | Hacker News Reader

They didn't say that. Here's the quote I believe you skimmed and misinterpreted.

"About 1,300 former players have died since the B.U. group began examining brains. So even if every one of the other 1,200 players had tested negative — which even the heartiest skeptics would agree could not possibly be the case — the minimum C.T.E. prevalence would be close to 9 percent, vastly higher than in the general population."

They're being honest about the selection bias, but they're saying the absolute minimum threshold is 9%. That's high number one, and it's almost certainly much higher than nine percent because they only tested a few of the brains. 1300 is every former pro player that died. Not every one they tested.

View on HN

As I said before, that's in the 10th paragraph of the article. The first line people read is:

> A neuropathologist has examined the brains of 111 N.F.L. players — and 110 were found to have C.T.E.

It's incredibly disingenuous to lead with that. It makes readers think that the rate of CTE is almost 100%. Readers have to get through nine more paragraphs (including profiles of two NFL players) before the truth shows up. Yes, NFL players have absurdly high rates of CTE compared to the general population. Nobody in this thread is saying otherwise. But disingenuity makes your side easier to dismiss and discredit, even if it's right. That's why I criticized the article.

110/111 testing positive for CTE is insanely high, even with selection bias. This is a big, big deal, and should be treated like a big, big deal. This is not hyperbolic in the slightest. If a chemical plant was suspected of causing cancer, and you got those test results, same experimental bias and all, it would be shut down instantly.

Boyoga!

If I picked a sample of 5 cherries and I made sure they were all rotten, it sort of reduces the amount of statistical inference you can do about the true distribution of the cherries to zero. Your use of "even with selection bias" doesn't really make sense here.

If you picked 9% of all cherries in whatever way you like, and found 98% of them rotten it still means that at least 8.8% of cherries are rotten. While not even close to 98% it's still a huge number in regard to CTE.

To say nothing of the many potential causes for potential cognitive decline in later life. I would expect the vast majority of cases of dementia in life to not include CTE.

Except for NFL players.

I'm honestly puzzled at the people coming out of the woodwork to troll this post. What is driving these people and their abandonment of reason?

Reply on news.ycombinator.com