I was wondering: If something fails on its first try, we’d say it has a 100% failure rate. But, if it fails on the second try, we still say it has a 100% failure rate. But, I argue…shouldn’t this really be a 200% failure rate? How come there is no accounting for this when we work with percentages (other than, say a price hike of 200%)?
Just thinking outside the box, or cage on the funny farm. OK, I’ll go back to my cage now! 
- Jinx
You can’t have 200% of something without increasing the amount of something; it would be absurd to say that 200% of cats have breath that smells of cat food.
Anyway, is this about Cecil’s divorce frequency column?
If something fails on the first try all the time, it has a 100% first-time failure rate; if, in half of all instances, it fails on the second time (the other half still failing on the first), then it has a 50% first-time failure rate and 100% rate of failure across two ‘tries’
Percent means “out of 100.” You can’t fail more than 100 times in 100 tries, no matter how hard you try.
The number of attempts has increased 100% (if you try once more and fail, you will have increased the number of tries by 200% total), but the failure rate is still 2 out of 2, or 100%.
Let me take a stab: 0% means ‘none of it;’ 100% means ‘all of it;’ 200% means ‘two lots of it.’
So a 100% failure rate means ‘all of the attempts have been failures.’ (‘rate’ here means expressed compared to the total.)
If you mean ‘failed twice as often’ you could say ‘there are now 200% failed experiments (compared to before)’
Percentages are FOR giving somthing compared to the total. If you want to do something else you use something else, (or choose a stupid thing to take percentages compared to.)
wasbla meem! bloogleflob, smelb worblyflarm!
(ask silly questions and…)
In the example you give, a 200% failure rate would suggest that the product failed two times for every one attempt (2/1 = 2 = 200%).
Never mind trying to decide what that means - it’s not what happened here. The product failed two times in two attempts.
The failure rate is (failures) / (total tries). 2/2 = 1 = 100%.
Yeah, if the OP wants to get technical about it, they should wait until they’ve had a hundred failures before calculating a percentage. Solves that problem.
[QUOTE]
*Originally posted by Jinx *
I was wondering: If something fails on its first try, we’d say it has a 100% failure rate. /QUOTE]
Shouldn’t be 50%?
No, it should be 100%.
Something fails on it’s first try: It was tried 1 time. It failed 1 time. 100% failure rate.
I meant, a hundred tries. Of course. 
I thought the OP was asking shouldn’t there be a way to show the number of trials - or maybe our confidence in the percentage. Well, percentages are really too simple for that. Statistics is a lot more than just percentages, and there are standard notations for showing sample sizes as well as confidence. You may have seen a simplification of this in a magazine where they tell you the margin of error for a telephone poll for example…there are specific ways of calculating the margin of error and it in part depends upon the sample size.
dood, when someone starts talking about percentages, my ears go up. Why talk in percentages when you can use ‘real’ numbers?
oh, and scupper, that’s not true. you can have fractions of a percent, so if you have a lot of 1000 light bulbs, and 1 doesn’t work, then the lot is considered 99.90% functional.
if it fails on the second try, we still say it has a 100% failure rate. But, I argue…shouldn’t this really be a 200% failure rate?
No, because it didn’t fail twice as often as it was attempted (which would of course be impossible). It has failed 100% (all) of the times it was tried. If there had been success on the second try, the failure rate FOR THE TWO ATTEMPTS would be 50%.
Here’s the key concept I think you’re overlooking: When you add an additional attempt, you change the number of things you’re taking a percentage of.
Try once, fail once, 100% failure rate.
Try twice, fail twice, 100% failure rate.
Try twice, fail once, 50% failure rate.
On the other hand, since this has been explained over and over pretty much the same way, you have about 500% of the answers you need.
[sup]Remember my post in the other thread? It really does apply here.[/sup]
I agree, he’s scraping the bottom of the…keyboard?
Ah, one of my favorite books & still sold, ‘How to lie with statistics’…
you might like it Jinx.
Don’t they teach anything in school any more? This stuff is just so basic. . .