Year polls 2.0: a planning thread

[thoxans]

i'm personally [...] done so

while the scoring overhaul section of my previous post solely addresses the option of automatically including #1s, that's not necessarily an endorsement of that system. i'm also open to diff point distros. just wanted to clarify though that i'm not set on auto-including; it just happened to be the scoring alternative that i used as an example. so debate freely on the subject

[Umbugbene]

I trust whatever you come up with for the scoring system, but my only input is that I'd strongly prefer NOT to include old ballots from former members. Aside from the fact that they no longer represent our present community, their ballots will be outdated in any case. Like the rest of us their picks will likely have changed, so their ballots aren't comparable to our updated ones.

Also, unless you make a thorough effort to include ALL past ballots from ALL non-participating members, there's an unfair bias in choosing which ex-SCFZers get counted. For example, by choosing Popular Ex-Champ X you're giving added weight to the tastes of those who mirror his/her taste.

[sally]

no preference for scoring, don't mind if the results never even get tallied up. what i want to know is whether thoxans minds the chaotic image-filled chat-threads that lencho tolerates or if he wants just a tight-lipped list-dump and all the blather elsewhere

[flip]

my only input is that I'd strongly prefer NOT to include old ballots from former members. Aside from the fact that they no longer represent our present community

what about members who are generally active but who happen not to be around for one poll though? their previous ballot at least gives an approximation of what they would vote for.

[flip]

since #1s are automatically being included no matter what with this system, i would lean towards allotting the #1 vote the same amount of points as every other spot on the ballot. if the #1 vote is already guaranteed a spot on the final list, then giving it additional special powers (such as additional points) would just lead to unintentional vote brigading; and that, class, is how we end up with tragically overlooked and underrated citizen kane winning the 1941 poll. at the same time, the alternative perspective would be that automatically including #1s (while also allotting #1s the same amount of points as every other spot on the ballot) would lead to greater variety in the end results, since i imagine a person might be less inclined to give their #1 spot away to the godfather, if they see that another member has already done so

i don't follow this line of reasoning. it seems to me, if we get say 20 ballots in a poll, if every film gets 1 point, we're just evaluating 'which good films has everyone here seen'. if there's a film only 5 people have seen, but they all think it's the greatest film ever made, it can only get 5 points. it has no chance to do well. so in that system, we get all of the canonical films that scfz thinks are good too at the top, and then a lot of interesting stuff at the bottom. if that film with five views can get 2 points on every ballot it appears on, it can at least get 10 points in total and rank nearer to the top -- not at the top, but high enough that its placement reflects its high regard among the people who have seen it.

i'm generally in favour of more hierarchization when awarding points, rather than less. if we don't want to deviate too much from what we've done historically, maybe we can have two 2-point slots instead of one, but where the voter can indicate with an asterisk a single film that automatically ends up in the final results no matter what. then no one needs to choose between 'this is my absolute favourite film' and 'this is the obscure gem that is my second favourite film but i want it to show up in the final list.'

[Lencho of the Apes]

tight-lipped list-dump

Not speaking for Thoxans, but why shoot ourselves in the foot? The blather is potentially far more valuable -- it's the process of educating ourselves and each other about an extraordinary breadth of material that I most value in this game.

[thoxans]

whether [that new dipshit] minds

nope. go for it. totes agree that the more chaotic image-filled chat-threads, the better

[thoxans]

i don't follow

i guess the way i thought about it was this (to use your example):

if every film gets 1 point, we're just evaluating 'which good films has everyone here seen'

imo we more or less do that now, right? i mean, 95% of a hypothetical full ballot of 20 films is already currently attributed 1 point; and voting a film #1 def guarantees nada

if there's a film only 5 people have seen, but they all think it's the greatest film ever made, it can only get 5 points

again just imo (and i could be wrong) but in the very rare instances where 5 people all vote the same film #1, it's in top-heavy canonical years; so i suppose my idea behind the point limitations for #1s in a setup where we auto-included #1s would be to counterbalance running up the score on well-established favorites, while simultaneously promoting a reason for people to extend their #1 votes to stuff that they'd really like to see on the final list, but know that no one else will support it with the votes needed to secure it a spot

i'm generally in favour of more hierarchization when awarding points, rather than less

all of this is to say that i agree with ^ . i gave a rundown on the auto-including #1s option cuz i wanted to flesh out my thoughts, as it's the option i'm most skeptical of. i'd prefer a system of diff point distros, whether it be through tiers, or whatever. in relation to all of the above, maybe we could just do a #1 that's awarded extra points, as well as a 'wild card,' which would be the film that's guaranteed a spot on the final list. idk. this is why i wanted to ask these questions. if i just churn this shit over in my head on my own, we'll end up with a clusterfuck of a poll, with like 5 different tiering options, 25 different point distro options, and 125 different loopholes, including the konami code

[ofrene]

So do we start year polls 2.0 this month? Just wonder...

[Angel]

Could the next year be announced in advance in the current poll? That would be useful for organizing schedules.

[Lencho of the Apes]

announced in advance

I like that idea. Strong second, with both hands raised to vote "aye."

[rischka]

ME THREE

[sally]

minus me pretty sure that if i was planning for the next month i'd entirely stop paying attention to the current year poll if the next month was a year i was more keen on...

[nrh]

i used to figure what the year poll was in advance based on what lencho put on his letterboxd watchlist, which helped so either thoxans make a cryptic letterboxd watchlist or i guess announce it officially

[thoxans]

i mean, i could give y'all the entire order of the next 124 polls, but i'm pretty sure that sally's head would explode

[Lonnie]

Can we please get next month's poll year in advance? 1943 is no banger.

[sally]

oh thoxans, pls don't feel any restraint on my account, my head is already pretty exploded anyway

[Evelyn Library P.I.]

Don't feel strongly either way as to giving folks a heads up of the next month, so if it's just Lonnie's request for the next month reveal, then I don't mind. I'd definitely vote against having the entire order of all 124 revealed now, though, as I really enjoy the suspense of not knowing, and there'd be no suspense at all if it was all available now.

[greennui]

Yeah, I enjoy the suspense of not knowing the next year until the turn of the month. It forces me to dig deeper into the current year and watch stuff that I probably wouldn't have watched if I had another year presented to me. Think the threads would have less activity as a result as well.

[Lonnie]

I can only assume thoxans was dizzied by his new responsibilities to (jokingly?) suggest announcing all years. I too enjoy the surprise of finding out what I'll be watching for the next month, but I sometimes get a little year fatigue, and it would be nice to explore another year while still being productive. Really I was just hoping to restart the request above from the veteran champs. I fully support the group's decision on how to conduct 2.0.

[thoxans]

in lieu of the nonfunctioning spoiler tag, here's next month's poll in advance for anyone who's curious: ((369/3)^2-11179)/2=

[Silga]

Me looking at math problem

https://www.youtube.com/watch?v=ZRe_ACYrJ0A

[Monsieur Arkadin]

Man... my high school math teacher would be pissed that I literally have a calculator in my pocket, which I used only to copy and paste that math equation into google, which just told me the answer without me engaging mathematically with the concepts in any way at all.

[---]

One can do it mentally with only a bit of complexity

The easiest way to do the exponent is probably

100x100 + 2x100x23 + 23x23

100x100 + 2x100x20 + 2x100x3 + 20x20 + 2x20x3 + 3x3

Is easier calculations but a lot more to store

Want to know what flip thinks. Should the natural inclination actually be

aa+bb+cc+2(ab+ac+bc)

??

[flip]

if i needed to calculate 123^2 mentally, i'd probably do (120 + 3)(120 + 3) = 120^2 + (2)(3)(120) + 3^2 = 14,400 + 720 + 9 = 15,129. then all the calculations are fairly familiar and it's just keeping track of digits in the sum that can be annoying. the other way i might do it is 123*123 = (100 + 20 + 3)(123) = 100*123 + 20*123 + 3*123 = 12,300 + 2460 + 369, but that's annoying to add all at once, so i'd be doing something like 12,300 + 2000 + 460 + 300 + 60 + 9 in steps in my head. with a slightly more awkward number, i'd find that too difficult to do probably.

i did the whole thing a bit differently though - not sure this is interesting at all, but since bure asked (i can explain why any of this works if anyone is interested in more detail) :

123^2 ends in 9, because 3^2 does. so 123^2 - 11179 must end in zero. then when we divide by 2, we must get something ending in zero or 5.

123 is also divisible by 3 (its digits sum to a multiple of 3), so 123^2 is divisible by 9. when we add the digits of 11179, we get 19, which has a remainder of 1 when divided by 9. so 11179 has a remainder of 1 when divided by 9 (or you can just see that when you subtract 1 from 11179, to get 11178, the digits sum to 18, so 11178 is divisible by 9 and 11179 is 1 greater, so has a remainder of 1). you can subtract remainders, so when we divide 123^2 - 11179 by 9, the remainder will be the same as 0 - 1 = -1, which is the same as 8 (adding 9 to get into the usual remainder range). if 123^2 - 11179 has a remainder of 8 when divided by 9, then when we divide by 2, we'll have a remainder of 4 when we divide by 9 (this wouldn't work the same way if the 9 and the 2 shared any divisors, besides of course 1).

so the year has a units digit of 5 or 0, i.e. a remainder of 0 when divided by 5, and has a remainder of 4 when divided by 9. that number could be 2020 (ends in zero, digits sum to 4), but that's too recent. if we want to preserve remainders by both 5 and 9, we can just add or subtract 5*9 = 45 (here it's important that 5 and 9 share no divisors - if they did share any divisors we'd use their least common multiple). subtracting 45, you get 1975, 1930, or 1895. i ruled out the latter two, because 1930 was too close to 1943, and 1895 too early, but there are other ways to confirm it's 1975. 123^2 has a remainder of 1 when divided by 4, because all odd squares do. 11179 has a remainder of 3 when divided by 4, because 11176 is divisible by 4 (since its last two digits, 76, form a multiple of 4). so when we subtract 123^2 - 11179, we get a remainder of 2 when divided by 4, so that number is not a multiple of 4, and when we divide it by 2, we get something odd. so 1930 is impossible and that only leaves 1975.

[MrCarmady]

there's a trick my grand-dad taught me which is slightly useful here:
any number ending in 5 squared = [the numbers preceding the 5]*[itself+1] and then 25 at the end
e.g. 25^2 - 2*3 = 625
105^2 -10*11 = 11025
so 125^2 - 12*13 = 15625
which makes 123^2 = 15625 + 4 - 500 {(2*125*2)} = 15129
15129 - 11179 = 3950
3950/2 = 1975

I'm gonna watch the Zulawski for sure, have to think of other stuff

[sally]

there's a trick my grand-dad taught me which is slightly useful here:

that's so cool! i love it when numbers do things like that!

[flip]

interesting, that works because a number x5, where x is some string of digits, is just equal to 10x + 5, and (10x + 5)^2 = 100x^2 + (2)(5)(10x) + 5^2 = 100x^2 + 100x + 25 = 100(x)(x+1) + 25, and multiplying by 100 shifts digits two spots over, so you get 25 in the last two spots and (x)(x+1) to the left of that

[Silga]

Thoxans, are we going to get some elaborate math formula to uncover the upcoming year in our Year poll 2.0?

The advantage of knowing the next year is that I can shuffle my year poll viewing with emphasis on the ongoing year, but mixing-in the follow up year to avoid the fatigue of one specific year.

[thoxans]

mixing-in the follow up year to avoid the fatigue of one specific year

since the spoiler tag doesn't work, i attempted a workaround so peeps who want to know the upcoming year poll could find out (simply copying and pasting the formula into google will give the answer), while shielding the info from those peeps who don't want to know the upcoming year poll. but the answer was posted anyways, so i'll just post the upcoming year poll midmonth from now on. that should give roughly two weeks for folks to focus on the ongoing poll, and an additional two weeks for folks who are ready for the upcoming poll