Apodizing filter

[danadam]

Can you walk me through what the chances are of finishing 2 tests out of 5 consisting of 20 trials with at least a 14/20 by just guessing?

I don't know a formula for that but it's easy enough to write a simulation: python @ godbolt, C++ @ godbolt.

I run it on my desktop with 10'000'000 scenario runs and I got:

• 4.429% for scenario 1 (1 test of 100 trials with 59 or more correct answers, just to confirm that the simulation works ok. P(X>=59) = 4.431%)
• 2.956% for scenario 2 (5 tests of 20 trials and at least 2 of them with 14 or more correct answers)
• 0.412% for modified scenario 2 (15 or more correct answers because P(X>=14) = 5.7%, so still slightly above 5%)

 

[Ghoostknight]

I don't think so but it's hard to tell because there is no reasoning behind why you came up with these numbers.

im unsure myself with the numbers, feel free to correct me

0,95*0,95*0,95*0.95*0.95 = 0.77378% , so to hit once 5% in 5 trials, its 1.00 - 0.77 = 23% , fairly certain here

after that my calculation is probably wrong now thinking about it, im unsure how i would calculate "hitting 5% two times in 5 trials"

I only accept tests conducted to find out the truth, not ones conducted specifically to prove an argument. I think you have an extraordinary level of bias and it’s coloring everything you do and say.

well i try to find out the truth,
the truth so far
1. subjectively audible
2. DBT "suggests" (so far) its audible, sure, more DBT give a more bulletproof result
3. logic/objective facts tell there is a difference, question is can it be heared

you guys are only biased by studys.... ignoring all other "facts"
imo i made enough points that atleast subjectivisits maybe try to listen for themself

 

[Ghoostknight]

I don't know a formula for that but it's easy enough to write a simulation: python @ godbolt, C++ @ godbolt.

I run it on my desktop with 10'000'000 scenario runs and I got:

• 4.429% for scenario 1 (1 test of 100 trials with 59 or more correct answers, just to confirm that the simulation works ok. P(X>=59) = 4.431%)
• 2.956% for scenario 2 (5 tests of 20 trials and at least 2 of them with 14 or more correct answers)
• 0.412% for modified scenario 2 (15 or more correct answers because P(X>=14) = 5.7%, so still slightly above 5%)

ah, thanks!

and now please 3 out of 5 trials with 15 or more correct answers, it should be pretty marginally

 

[Ghoostknight]

but thinking about it, since the formular to come up with a 59/100 result for 95% confidence, actually takes the number of overall trials into account, i should maybe just do a larger scale DBT and see what result i get...

 

[gregorio]

well i try to find out the truth,

So you think that not verifying what you think you’re hearing, making-up BS to explain it and then defending it with fallacies is trying to “find out the truth”? Again, that’s very backwards compared to “us guys” who would consider that pretty much the opposite of the truth!

the truth so far
1. subjectively audible
2. DBT "suggests" (so far) its audible, sure, more DBT give a more bulletproof result
3. logic/objective facts tell there is a difference, question is can it be heared

1. And here we go again. You just made that up, there’s no such thing as “subjectively audible”, there’s either audible or inaudible.
2. No it doesn’t. The only thing an invalid DBT “suggests” is that you don’t know what you’re doing!
3. Logic/Objective facts tell us there is a difference and it cannot be heard.

So there we have it, out of your own mouth. Your “truth so far” is in fact not a single ounce of truth, it’s ALL just BS you’ve made-up!

you guys are only biased by studys....

Yes, “us guys” are “only biased” by science (that includes studies), which given the name of this forum is surprising to you why exactly?

ignoring all other "facts"

What other facts? You making-up BS is neither “the truth so far” nor “other facts”, it’s just made-up BS! How on earth can you believe that your made-up BS represents the truth/facts when you’re the one who made it up? In that case, it’s a fact that the moon is made of peanut butter because I just made that up … or is it only the truth/fact if you make it up?

imo i made enough points that atleast subjectivisits maybe try to listen for themself

Sure, you made up enough BS that some deluded or ignorant subjectivists would do a sighted listening and some would be biased by that enough to think they’re hearing a difference. How does that demonstrate any truth/facts beyond the fact that you’re making up BS and they’re gullible and ignorant enough to let that BS bias their listening test?

Round we go in circles, you demonstrating backwards thinking compared to science and the rest of us and that you making up BS is somehow the truth/facts!

G

 

[VNandor]

I don't know a formula for that but it's easy enough to write a simulation: python @ godbolt, C++ @ godbolt.

I run it on my desktop with 10'000'000 scenario runs and I got:

• 4.429% for scenario 1 (1 test of 100 trials with 59 or more correct answers, just to confirm that the simulation works ok. P(X>=59) = 4.431%)
• 2.956% for scenario 2 (5 tests of 20 trials and at least 2 of them with 14 or more correct answers)
• 0.412% for modified scenario 2 (15 or more correct answers because P(X>=14) = 5.7%, so still slightly above 5%)

I'm pretty sure I found the correct equation for it which is P(x=k) = (n choose k)*p^k*(1-p)^(n-k). I don't think I could figure this one out myself.
That yields 2.96% when p=5.77% n=5 and I sum k=1 and k=2. When p=2.07% it goes down to 0.411%.

So it looks like it's slightly more unlikely to get at least 2 14/20 tests out of 5 than getting a 59/100. That said, I'm nor sure if I should trust the second test more because the regular test is like flipping a coin 100 times while the test proposed by Ghoost is like rolling a 17 sided (or 48 sided for the second case) dice only 5 times. I know this can be accounted for by looking up how bayesian statistics work but why open that can of worms when it can be so easily avoided.

Scoring a 15/20 at least 3 times out of 5 tests just by chance is extremely unlikely though. It's 0.00859%, so less than 1 out of 10'000 times