The 'What are you playing right now?' Thread

[samurai black]

I've been playing some Metroid: other M and it's wicked.

how is that? i was keen as for it when i first saw it, but the sideways wii-mote thing doesnt really do it for me

[stereomonkey]

Should be playing Halo but currently occupied with Dead Rising 2.  So awesomely good

[redtie]

Playing STATS Practice test.

The accuracy of a cutting machine is such that the mean length is the length it is set to and the standard deviation is 2.2mm. The lengths follow a normal distribution. The engineer wants to make sure that no more than 20% of the lengths cut are longer than 350mm. (It is less of a problem if the lengths are a little short.) What should she set the mean length to be?

If someone explains how to figure this out I will give them $5

[massi4h]

Well it seems you're trying to get the highest mean possible if I read it correctly. You want to be as close to 350mm I guess. So I guess if you worked out the value of 20% left to the mean using mean 350mm and SD = 2.2mm then setting that to your mean would be your answer, as the bell curve is symmetrical. I don't remember how you calculate that stuff in high school(?) but looking on Wikipedia you can use this massive as formula: f(x) = 1/sqrt(2πσ²) * exp(-(x-μ)²/2σ²

Sound like it makes sense but I haven't done any stats since high school 2 years ago. So use that mean and SD and find the 20% mark (I'm sure you learn some methods that do that or I think the GC can do it) then subtract the mean off that new value and use that as your answer, since the 2 bell curves will be the same, just shifted that 20% value should go up to 350mm.

Hmm so I tried putting it in my GC (can't remember exactly how to do it though), but for some reason that method returning ~30%. Either I forgot how to use the area finding one, or the one that finds the point or something.

[[NIUE] BIRRY WONG]

348.1484

Thats the answer.

This is the normal distribution bell curve. Roughly. The middle value on the x axis is the mean, and each other value on the x axis is how ever many standard deviations away from the mean a particular percentage is.



To work it out, you need to work out where on the x axis the 80th percentile sits. Fucked if i can remember HOW to do this formula wise, i just, by coincidence, happen to remember the value from a problem i did in school. Which probably doesnt help you, particurlarly if you need to show how you worked the value out in your paper.

Anyway, this is the point where you want 350mm to be, since 20% of the results will be above it, and the exact value is 0.8416 standard deviations. Bout time my ridiculous memory for numbers payed off.

To get your average, you simply multiply your standard deviation of 2.2mm by 0.8416, which gives you 1.85157. If you subtract 1.85157 from 350, you get the value that youre after. 348.1484

Edit: I THINK Masons formula for working out where that percentile lies is correct, but im not sure, and cbf testing it. I mustnt have been paying attention that day in high school. Although admittedly, there are a few values in that equation that you dont actually have at the stage you would be using it, so i dunno.

Edit2:Im confident with my answer. I just checked it on a standard deviation calculator online. its 348.14843 if you want to be super accurate.

[geneterror]

NERDS!

[samurai black]

yeah, fuck you guys.


i'll keep playing 'dumb cunt'

[massi4h]

Yeah that was the number I got, then I wasn't sure how to use that number and check if it was also 20% which it definitely is.

[redtie]

@BW and massi4h, geniuses!

Damn everything is good and yeap the answers are correct but I got to show some sort of working. Oh yeah this got cut off on the the cut and paste.

Hint: Find the value of z for which Pr(Z>z)=0.20 then use the z-score formula to calculate the mean.

[[NIUE] BIRRY WONG]

Just did a massive MASSIVE write up, which the work computer just deleted when i tried to post it.

Basically, look at this page. http://en.wikipedia.org/wiki/Normal_distribution

Look at the Standard Deviation and Confidence intervals section.

Look at the second table.

For the purposes of this equation, = 0.6

(Needs to be 0.6 because we need to find 20% on either side of the area that this calculates. 0.2 x 2 = 0.4 and 1 - 0.4 = 0.6)

To work out N, the equation we need to use is:

sqrt2(0.6/erf) = N

BUT to work out erf (error function) we need to use

erf = 0.6/(N/sqrt2)

If we substitute in the value for N that we know (because we cheated and didnt use the formula) then we're sweet to work out erf. (works out to 0.5951), but we still have the problem of working out what N is in terms of a formula. And i never did error functions (erf) at school, and i cbf figuring that shit out now, especially given that the error function equation looks like this



Thats what i got to in a nutshell. A big fucking circle. And now im tired, ive been up since 5am, and im going home. Good luck dude.

[redtie]

My feeble mind was enlightened by Birry Wong. Ended up getting A- on the test AND there was a similar equation on it too.

Where am I sending this $5 too? Or maybe I should get 'Birry Wong > You' printed on a shirt and rock it around uni.

[electric]

'Birry Wong > You'

This x 10000000

[[NIUE] BIRRY WONG]

BIRRY WONG is happy to have helped.

But my vote goes to T-Shirt + pics. XD

I might start a guerilla campaign. Throw some "Who is BIRRY WONG?" and "BIRRY WONG > You" posters around town. XD

[samurai black]

playing some alpha anthologies. mostly dicking around in alpha and pocket fighters.

[redtie]

BIRRY WONG is happy to have helped.

But my vote goes to T-Shirt + pics. XD

I might start a guerilla campaign. Throw some "Who is BIRRY WONG?" and "BIRRY WONG > You" posters around town. XD

IT STARTS HERE! Going mall tomorrow sus out some priceys. White font on a black T, unless someone pulls out the meanest design that can be printed

Halo Reach, Jet Pack is too fun