Anyone wanna graph the occurrence of each number?
It's already graphed. There are little bars next to each option, showing graphically how much it has been chosen.
But based upon what's already up there (currently 98 votes recorded, as I start this analysis), the following trends (or lack, thereof) come out...
(Exactly!) 50% of numbers are not yet voted for; 28% have a single vote; 11% two votes; 6% 3 votes; 3% 4 votes; (skip a few) one (thus 1%) with 7 votes (number 7, itself) and another (1%) with 11 votes (42). That's graphable, but I'm not bothering to do so. (Maybe if we get several hundred votes, in total, it'll have some other interesting features.)
60 votes went for odd numbers, the other 38 to even ones, a 3:2 ratio. Although eradicating the upper outliers (lucky number 7 and The Ultimate Answer), that leaves 53 odd and 27 even, which is very close to being a 2:1 ratio.
Tens and units preferences show as follows
| Tens | | | | Units | |
| 0? | 22 | | | ?0 | 5 | <- both include "100" |
| 1? | 14 | | | ?1 | 8 |
| 2? | 5 | | | ?2 | 15 |
| 3? | 4 | | | ?3 | 15 |
| 4? | 17 | | | ?4 | 6 |
| 5? | 7 | | | ?5 | 11 |
| 6? | 7 | | | ?6 | 2 |
| 7? | 10 | | | ?7 | 17 |
| 8? | 9 | | | ?8 | 8 |
| 9? | 3 | | | ?9 | 9 |
I would surmise from that that apart from the 42-effect, and a certain bias towards the 1..20 range, it's largely consistent on the tens side, albeit that the sevens and eights might be slightly inflated by birth-years of the participants (although there's a low result for the 90s range, where I'd presume a lot of participants' birthdates might originate.
On the units side, odd numbers seem to rule, save for the 42-effect, and blips that might be consistent with 2
n values from geeks, except that those numbers are
actually not represented well at all, beyond 8 and 64. And ?4 didn't stand out, despite that...
Apart from the 38 already recognised as being multiples of 2, there are 36 that are multiples of 3, 16 are 5x,
24 are on the 7-times-table (that includes those seven 'sevens'), 4 are 13 multiples, 6 are 17 multiples, 2@19x, 4@23x, 1@29x, 1@31x, 3@37x, 1@41x, 1@43x, 2@47x and among the 50+ primes (where only the prime itself could be a valid multiple), 53, 59, 67, 71, 79, 89 and 97 are unrepresented, there's a single vote for 61 and three votes
each for both 73 (my own 'nearly' vote!) and 83. Did I miss any primes, in my haste? It's a bit of a mish-mash, though.
(Also, this really needs adjusting according to the
opportunities to find multiples for any given factor. i.e. 2-multiples doubled, 3-multiiples tripled, 49-multiples figures adjusted by roughly a factor of 50, or some other method of normalising...)
Counting only those choices landing
exactly on primes (and a good opportunity to see if I've missed any, and prod me to include it in any future analysis!) the following frequencies occur:
| 2 | =0 |
| 3 | =2 |
| 5 | =2 |
| 7 | =7 |
| 11 | =2 |
| 13 | =4 |
| 17 | =4 |
| 19 | =1 |
| 23 | =1 |
| 29 | =0 |
| 31 | =1 |
| 37 | =2 |
| 41 | =0 |
| 43 | =1 |
| 47 | =1 |
| 53 | =0 |
| 59 | =0 |
| 61 | =1 |
| 67 | =0 |
| 71 | =0 |
| 73 | =3 |
| 79 | =0 |
| 83 | =3 |
| 89 | =0 |
| 97 | =0 |
No real pattern there... Perhaps the sum of all prime-landing numbers could be compared to numbers that are "prime+/-1", and so on, but right now I can't see a reason to go there.
Whole number exponentials?
| x2 | 12 | /10 opportunities | =1.2 |
| x3 | 8 | /4 opportunities | =2 |
| x4 | 2 | /3 opportunities | =0.6... |
| x5 | 1 | /2 opportunities | =0.5 |
| x6 | 3 | /2 opportunities | =1.5 |
| x>6 | 1 | /1 opportunity (each) | =1 |
(There I've tried the 'normalising' that I suspect should have applied to the factor results. Should I have even included 0
x in each count, though? Take a single counted item from each one, if you think that'd be for the best. It doesn't adjust the 'adjusted' results significantly, though, at a preliminary glance to that effect.)
Shall I leave the likes of Fibonacci sequence numbers, and other groupings like "fortunate" numbers, etc? I suppose so. Albeit that with Fibonaccis in particular there's nicer aesthetic significance in the ratio of adjacent digits, the further up-stream you get, etc, which might affect things. But affect things more than "Lucky 7" and "Answer=42" app0ears to socially skew the result more than any properly mathematical reason? I doubt it.
(All the above done largely by hand/visual inspection, so open to errors. Plus ninjaing, as I doubt it'll still be 98 votes cast by the time I've tidied this post up. In fact, I can see it isn't, but going through it all to add will mean
more ninjaing while I do so, so maybe later.
)
Poll
Author
Topic: Pick a number, any number (Read 5489 times)