The point is to show that with 95% for every 1 person saying "no" you have 19 people saying "yes". With 99% for every 1 person saying "no" you have 99 people saying "yes". The common denominator is one "no".
What youre struggling with is the counterintuitive nature of applied statistics vs pure math, and this is the point TFA was trying to make.
> You aren’t getting 80 more signups as suggested
TFA isnt saying you "get" more, but just illustrating how different 95 and 99% actually are. Its restating the potato paradox, linked elsewhere in the thread
Genuinely curious aside—you used TFA twice. It means “the fucking article” here, right? When you use TFA, your otherwise helpful comment reads as if it’s angry, exasperated, or some otherwise negative feeling that stands in opposition to the rest of your comment—which reads as a genuine attempt to be helpful and explain. Why do use TFA?
I think this is one of those things where the meaning has evolved over time, at least as someone who's been on Slashdot since the early '00s but only joined HN a few months ago. Originally, you'd only see "TFA" as part of RTFA, generally with the assumption that the person you're replying to had not read the article. ("If you'd RTFA, ..."; "Maybe you should RTFA.")
But "TFA", although it derives from "RTFA", never seem to had the same negative connotations. It's just that sometimes you want to refer to the original article in question, but "original article" is long to type and/or ambigious. (Do you mean the news article from the NYT, or the scientific paper the NYT article is reporting on?) And "TA" is too short for people to clearly know what you're talking about (And did you mean "teaching assistant"?) "TFA" is short and unambigious: it always means the article linked to from the main page.
Long story short: Although etymology would suggest that "TFA mentions this" is as aggressive as "Maybe you should RTFA", in actual developed usage, they're very much not the same.
> Long story short: Although etymology would suggest that "TFA mentions this" is as aggressive as "Maybe you should RTFA", in actual developed usage, they're very much not the same.
I did assume the best—hence pointing out twice how helpful the comment was. I was merely curious why the author of the comment used TFA, and if they meant something else than the typical meaning of the acronym. Unless you personally know the contents of the author’s mind, I don’t believe you can answer for their meaning with much authority. You seem to have misunderstood my intent with asking (or do you generally like reminding long-time HN users about rules?). I’m not offended by the usage of TFA or assuming the worst of the commenter. I am curious about the juxtaposition of a helpful tone alongside a rather well-established acronym. Either way, thanks. I’ll hope the author confirms that was their actual meaning.
I wasnt aware "TFA" could be interpreted with a negative connotation, although that seems obvious in retrospect. Just trying to participate in the HN community, and also because its less typing : )
As a matter of writing style, repeating acronyms are invisible to the reader, whereas repeating words are annoying and remove value. This may be an opinion I gained from my military service, which was acronym-heavy.
Except that the potato paradox doesn’t apply here. What the business cares about is total signups and revenue, not the variation in rate of signups. The potato paradox only applies if you’re talking about the rate of misses, which, while interesting, is secondary or tertiary in importance.
An 80% decrease in missed signups only causes about a ~5% increase in revenue. That’s an important point if the team that produced that 5% revenue increase costs a large amount of money to run. At close to a million a year for the team, that’s only going to be worth it for some.
And that was the whole point of the article: this optimization usually isn’t worth the cost.
In the context of A/B Testing, if you've decided to stop the test at "95% significance" then you'll stop at the 19 yes, 1 no spot (or maybe 38:2). If you're testing for 99% significance, you won't stop unless you get to 99:1 (or more likely, you'll need to test way past that because "no" number 2 will arrive before "yes" number 99, so you'll need to test to 198:2 or 297:3 or further.
This is why you should never stop an A/B test once you've "hit your statistical significance". Always choose the number of tests you'd need to prove the significance before you start, and let it run even if it's "obviously winning" (or losing).
I'm not the author of the article, but I sometimes use the same approach because in my experience saying 5 people out of 100 does not have the same effect as 1 in 20 even though mathematically they are same. It provides a mental check if you are still ok with your original perception.
Though obviously equivalent, I think as a percentage 95/5 and as a reduced fraction 19/1, they both have value in different ways to the non-mathematician. Giving 2 intuitive handles on a thing instead of one is helpful.
He is not changing scale. He is just illustrating what the probability means in terms of the size of the set you need to expect the event to occur once, which has a more practical value than a probability.
At 95% certainty, you have 95 people saying "yes" and 5 people saying "no"?
So it's easier to make an apples to apples comparison? What point is the author trying to make with changing the scale?