Horay for math!

[Banshee 105]

The last post of one of the AR's topics was this:

Sufficient contextual data acquired. 871803909 ± 384 hours since event.

Compiling local archives/resources. [Hibernating] until next incident.

Now, I did some math, to find out how long ago that "event" was.

871803909 - 384=159181.341 which is about 436 years

Now, when you do:

871803909 + 384=36325178.875 which is about 99521 years!

Almost, 100000, the same time the Halos were first activated.

[Echelon Three]

Horay!

EDIT: By the way, people found this out a while ago. Nifty, but days old.

[3of9]

And Halo takes place in 2552.

2552 - 2007 = 545 years from now.

Add that to your value, and we get 100,066 years (in Halo time).

This implies that the AR is in 2007 (somehow).

[Jman]

I at first thought that the +/- 384 was significant, but when I did the math:

384 hours / 24 hours / 365 days = 0.0438 years

So is that the correct way to look at it?

Becuase I have a hunch that AdjacentReflex needed that +/- 384, because it wasen't sure exactly how long...and if it were about 500 years off...well that would be interesting, wouldn't it? I was thinking that it might think it's in the future when it isin't...or something. Am I pulling at straws here?

[tak]

I at first thought that the +/- 384 was significant, but when I did the math:

384 hours / 24 hours / 365 days = 0.0438 years

So is that the correct way to look at it?

Becuase I have a hunch that AdjacentReflex needed that +/- 384, because it wasen't sure exactly how long...and if it were about 500 years off...well that would be interesting, wouldn't it? I was thinking that it might think it's in the future when it isin't...or something. Am I pulling at straws here?

Yes, in science you general refer to the uncertainty in a measurement with +/-. In general, you want the magnitude of your uncertainty range to be one standard deviation, so if you were to make repeated measurements you would find 68% of the time your measurements are within your uncertainty range. This simiply stems from most errors fitting to Gaussian function. Does this probably matter to what AR states? No.

Also, in general it's bad practice to state your error range with such precision, you normally want to round to one significant figure (assuming the leading digit is not one). In addition, you don't state a value as 871803909 if your uncertainty range is in the hundredths, the 09 obliviously have no significances.

I'm still not sure how the original poster came to his conclusions (his math does not make sense to me). If you are going to convert what AR stated to years you would have to do the following. First, lets clean up sloppy notation.

871803909 ± 384 (bad) -> 871803900 ± 400 (correct)

Now, let us do the math,

871803909hours*(1day/24hours)*(1year/365days)=99520.9942years

Note, I use the unrounded value above since you carry as much information as you can through your calculations. Also, I don't worry about prorogation of error since the conversions are consider to be exact measurements. However, what is the uncertainty range now?

384hours*(1day/24hours)*(1year/365days)=0.0438356164Years

Thus, we have,

99520.99+/-0.04 Years

Now, what is interesting is that 100,000 years falls well outside of the uncertainty range of this result. What does this mean? I'm not sure. I could calculate the probability of a measurement of 100,000 years (assuming error fits to Gaussian function), but I can tell you just from looking that it will be a very small probability. Does it mean its still possible? Yes, but the chance are *very* small. Basically, think of it like this, if AR was to make another measurement he would get a different result. Now, 68% of the time his measurements will fall within the uncertainty range; however, there is still a 32% chance his measurement will fall out of this uncertainty range. Of course, I'm ignoring systematic errors, which there is now way we can determine those.

Disclaimer: I'm sleepy, so if something I stated is incorrect, forgive me.

[Banshee 105]

Sorry, my math WAS wrong.

A better estimate would be

99521 years

or

99520 years.

Give or take a few tenths.

[ACEfanatic02]

<snip>
99520.99+/-0.04 Years

Now, what is interesting is that 100,000 years falls well outside of the uncertainty range of this result. What does this mean? I'm not sure. I could calculate the probability of a measurement of 100,000 years (assuming error fits to Gaussian function), but I can tell you just from looking that it will be a very small probability. Does it mean its still possible? Yes, but the chance are *very* small. Basically, think of it like this, if AR was to make another measurement he would get a different result. Now, 68% of the time his measurements will fall within the uncertainty range; however, there is still a 32% chance his measurement will fall out of this uncertainty range. Of course, I'm ignoring systematic errors, which there is now way we can determine those.

Disclaimer: I'm sleepy, so if something I stated is incorrect, forgive me.

Halo takes place around the 26th century. 2500's. Add 500 years to 99520 and you have a little over 100,000 years.

-ACE

[tak]

Halo takes place around the 26th century. 2500's. Add 500 years to 99520 and you have a little over 100,000 years.

-ACE

100020.99+/-0.04 Years

As you already know, the above would be the result since 500 years is exact and you do not need to propagate uncertainties. In all honesty, I'm a little confused about the timeline and what AR thinks the current date is.

[ACEfanatic02]

Halo takes place around the 26th century. 2500's. Add 500 years to 99520 and you have a little over 100,000 years.

-ACE

100020.99+/-0.04 Years

As you already know, the above would be the result since 500 years is exact and you do not need to propagate uncertainties. In all honesty, I'm a little confused about the timeline and what AR thinks the current date is.

Likely as not, I think AR knows what the date is. (If for no other reason than Bungie's expressed their distaste for time-travel.)

-ACE

[Sir Topham Hat]

You know what's fun? I figured out the circumference of Earth's rotation around the sun in my head once. And calculated how fast the earth had to be moving relative to the sun.

It's very fast.