# 2636: What If? 2 Countdown

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
 What If? 2 Countdown Title text: If you don't end the 99 Bottles of Beer recursion at N=0 it just becomes The Other Song That Never Ends.

## Explanation

 This explanation may be incomplete or incorrect: Created by a BOT - Please change this comment when editing this page. Do NOT delete this tag too soon.If you can address this issue, please edit the page! Thanks.

# Discussion

I've started the table to explain all the calendar entries. Barmar (talk) 00:19, 23 June 2022 (UTC)

Is the dog minutes calculation backwards? 777,777 dog minutes should be 777,777 x 7 human minutes, which is over 10 years. Randall seems to be dividing instead of multiplying. Barmar (talk) 00:36, 23 June 2022 (UTC)

No - 1 human year = 7 dog years; 1 dog year = 1/7 human year; 1 dog minute = 1/7 human minute; 777,777 dog minutes = 111,111 human minutes = 77 days, 3 hours, 51 minutes. 172.70.90.173 11:32, 23 June 2022 (UTC)

First entry is probably mistake by Randall, e^pi would give value of 84.5 162.158.203.38 11:57, 23 June 2022 (UTC)

That would be too high, though. 82.xxx days (from midnight at the start of launch day) would fall within the 83rd day before it (Jun 22). 84.5 would fall within the 85th (Jun 20). 172.70.91.58 12:15, 23 June 2022 (UTC)

Not sure if this is even worth mentioning, but he forgot the box around the date number in the top corner for August 29th. 172.70.126.151 12:49, 23 June 2022 (UTC)

Fyi, used wolfram alpha for most of the calculations. Seems to be able to handle anything I throw at it (nanocenturies, megaseconds, fortnights etc) Aditya95sriram (talk) 13:02, 23 June 2022 (UTC)aditya95sriram

Some of the calculations done forward (assuming what Randall means as a Generation, for example) might be best done as "to get this many days, what does Randall think ilhe is starting from. And see if 365, 365.25 or even 365.24 days per year works best, where relevent. Although I think in many cases you'll find the fractional differences negligable, when done right. (I'm also a bit surprised by the off-by-one errors in days-to-go and derived value, but I suspect that this is because of assymetric rounding effects that would be revealed by running the assumption backwards and seeing how different (or otherwise) the decimals actually are.) 172.70.85.211 13:32, 23 June 2022 (UTC)

I would suggest using 365.2425 days per year, as that's consistent with current leap year conventions. Dansiman (talk) 21:49, 23 June 2022 (UTC)
Did not see your comment, but already done trivial replacement. No recalculation that goes more complicated than magnitude, though.
(For the mathematically curious, in the Gregorian calendar it's normally 365 days, but a leap day every four years (+0.25 => 365.25), except no leap day every century (-0.01 => 365.24), except there is every fourth century (+0.0025 => 365.2425). Which is very very close to the more astronomically-precise figure of 365.2422, at least at this point in our planet's history and definitely over the timescale of the Gregorian calendar itself. edit-to-add-convoluted-musings: A successor system might need to de-reinstate three of the Four-Millenial leap-days in every 10,000 year period, or perhaps by re-removing four of its various leap-days then re-reinstating one of those back again, but by the time it's relevent I doubt that 365.2422 is going to be as valid for whatever reason... Hey, by then, maybe we could just deliberately adjust the Earth in or out a bit to make it a better fraction/not a fraction at all! )
On the other hand, the old adage is "no use being precise over imprecise details". One can perhaps apply it to nominal decades (the true average decade; though a given decade might be 10*365 days plus either two or three leap-days, for 3652.5±0.5 days in that instance... not equally likely each way, though) but the Generations calculation already assumes 27 years per generation (not even 27.5, exactly half way between 22 and 33, which already seems a dubious backformation to suit other purposes) and gets a good-enough approximate number. Using a factor precise to around 1 in 146000 alongside one that's unlikely to be even as accurate as 1 in 54 is a bit rich and overly anal (rather than analytic) in the long-run.
But this is explainxkcd, so I'm not saying it's misplaced, just that those who would be pedantic about everything (myself included) might find themselves even more out-pedanted in very reasonable circumstances... ;) 172.70.162.77 22:47, 23 June 2022 (UTC)

Not sure about most numbers but at least the order of magnitude seemed plausible. I can't quite find a proper way to read August 28th. π^π^π is roughly 80662.666 - if you read πcoseconds as "picoseconds", that's way less than a second. I have no idea what π * coseconds are supposed to be. π * c * o * seconds doesn't look much better - there are values associated with "c" (speed of light, for example) but I have no idea what "o" could be and certainly nothing that would make this a unit of time. Sixteen days would be 1,353,600,000,000,000,000 ps (picoseconds). π^π^π^π is three orders of magnitude too small, π^π^π^π^π is many orders of magnitude too big a number. Am I missing something (really obvious, maybe?) here? 627235 (talk) 14:52, 23 June 2022 (UTC)

Exponent towers are by convention evaluated top-down, so pi^pi^pi should be read as pi^(pi^pi), which is ~1.34e18, which in picoseconds is ~15.51 days. 172.70.114.71 15:21, 23 June 2022 (UTC)

10,000 minutes in Heaven is making out for a week. I was able to find a record for the longest kiss (58 hours, 35 minutes), but not the longest make-out session. I think Randall may be indulging in some nerdy wishfull thinking. Barmar (talk) 15:27, 23 June 2022 (UTC)

When the beer song reached F(0) how would you 'take one down' from -1 bottles of beer? Would they be imaginary bottles of beer? (Joking) At F(n-1) would there be a matter/antimatter annihilation, where Randal could do a riff of What-If #1 and describe the play by play of the bartender turning into exotic forms of matter? 172.69.68.88 15:58, 23 June 2022 (UTC)

(Not ✓-1, it's just straight repeated subtraction, not a power function...) After so much beer, you probably think it a good idea (even necessary) to fill cans up and start to put them back up on the wall... Not sure you could sustain it, to the point of F(-99), but I think someone'd be more than ready to start the process when F(-1) is invoked, for any group of just a few likely individuals.. 172.70.91.58 16:23, 23 June 2022 (UTC)
This begs the question of what beer bottles are doing on a wall, rather than a shelf. Barmar (talk) 16:26, 23 June 2022 (UTC)
Randall already considered what happens at F(0), refer to the title text. Paddles (talk) 08:16, 24 June 2022 (UTC)
If you wanted to take down an imaginary bottle of beer, you'd have to take it from another wall that runs orthogonal to the original wall. 172.70.85.211 08:50, 24 June 2022 (UTC)

We've finally filled in all the units columns in the table. Hopefully someone can automate turning that into a transcript. Barmar (talk) 16:51, 23 June 2022 (UTC)

Funfact: This comic mentions Cyndi Lauper by name, and it was published on her birthday… 162.158.38.27 20:51, 23 June 2022 (UTC)

Sweet! I'm a big fan of playing Time after Time on repeat to get into a flow state, so I loved that one. 162.158.166.183 20:43, 24 June 2022 (UTC)

Looks like someone's math is wrong on the explanation for July 18. I calculated using 4681 and 4763 years and they came out to 51.29 days and 52.19 days, respectively. So then I worked backwards and determined that Randall would actually have to be using a number closer to 5200 years to arrive at the correct result of 57 days. Dansiman (talk) 21:49, 23 June 2022 (UTC)

Randall seems to be wrong about "It's a Small World". The song is about 2 minutes long, so at 1/10,000 speed it's 20,000 minutes = 14 days. He seems to be using a length a little over 3 minutes. I found a YouTube video of the ride that's 3:45, but the song ends at 2:15 and the rest is silent. Barmar (talk) 22:16, 23 June 2022 (UTC)

This video of it on YouTube lasts 3:02. It was uploaded by Universal Music Group (allied with Disney), making it some kind of 'official' version, and its length fits Randall's calculation. (Also, thanks for making the table!) DKMell (talk) 22:38, 23 June 2022 (UTC)
That could be it. It has a long instrumental coda after the singing is done. Barmar (talk) 14:07, 24 June 2022 (UTC)

Could "eπ Ionian months" also be a very subtle reference to the Euler identity given the first two characters of Ionian? Or am I reading/visualising a bit too much into it? Paddles (talk) 08:16, 24 June 2022 (UTC)

Aug 26 needs editing, but I just reset my password and can't fix it. At 4 breaths per minute, 100,000 breaths is 17.36 days. To get 17 days exactly, Randall would need to assume about 4.085 breaths per minute. Wjw 08:24, 24 June 2022 (UTC)

Most of the calculations are very approximate. Barmar (talk) 14:07, 24 June 2022 (UTC)
Many of them were given to subsecond digits of precision, too, so I rounded everything off to two significant digits of days unless there was some compelling reason to have 0, 1, or 3. Don't @ me, because I filled up and homogenized all that column, finally (except for 100,000 breaths, which are slow enough to be what I'm guessing is probably Randall's error.) If someone wants to get a better value for the total duration of Star Trek than the January, 2021 reference I found by counting all the released episodes since up to the date of the cartoon, please do. 162.158.166.183 20:26, 24 June 2022 (UTC)
For Star Trek total run time, it might be best to count all episodes scheduled for release up until August 14, the date of that specification. 172.70.210.233 21:02, 24 June 2022 (UTC)

There's a closing /div HTML tag on the front page after the transcript (but not on this page). Nitpicking (talk) 17:21, 24 June 2022 (UTC)

Thoughts on including a "% of error" column in the table?172.70.130.121 15:54, 27 June 2022 (UTC)

An additional column would make it look worse on mobile portrait, and a residual error wouldn't really explain anything that readers couldn't get a gist of by eyeballing. 172.70.210.233 01:31, 3 July 2022 (UTC)
Most of these don't really have a margin of error, they're all based on specific numbers, beyond that most of them can go to a ridiculous number of decimal places (mostly due to the infinite decimal places of pi, e and phi). Rounding to 2 decimal places is sufficient and doesn't count as a "margin of error". NiceGuy1 (talk) 23:54, 3 July 2022 (UTC)

Several entries were rather unspecific, particularly the ones just saying "See day whatever" (and one "Refer"), so I plugged in some numbers. Also, it seems like the Star Trek entry should specify what's included - strange that unlike the others, it doesn't end up under a day less than the target. Also, that entry referred to/used an article which summarized CBR, I replaced it with the detailed article from CBR itself, then listed all the shows and movies, and added the things the article missed. NiceGuy1 (talk) 23:54, 3 July 2022 (UTC)

Great work. I wonder if Randall is counting some fan-made Trek. 172.70.211.36 01:10, 5 July 2022 (UTC)