madrid match re-watch

[yeats]

first of all, thank you to everyone who filled out the original poll about this idea -- i'm so glad to see that people are excited to participate! i apologize for how long it's taken me to make a follow up post...these numbers were not always easy to crunch! here are the answers to the poll:

do we want to watch old madrid matches over the summer, as we wait for the season to start again?
overwhelmingly, yes!

do we want to watch games from earlier seasons, and not just 2010-2011
raúúúúúúúúúúúúúúúúúúúúúl...aka also yes!

when should we have the watching parties?
...this is trickier.

CAVEAT LECTOR: i am not a statistician, by any stretch of the imagination. i took statistics, once, in college, and skipped most of the classes due to the lecture hall being on a hill. i am not even a particularly mathematical person, so all that follows may very well be completely wrong or illegible to others.

the problem was, basically, that everyone who answered is from different parts of the world, meaning that their votes for "what time of day" translate to different times. in order to solve this, i put together a massive spreadsheet (200 rows, 20 columns, five sub-sheets) and did some number-crunching.


1. of the 163 people who filled out the poll, 155 provided sufficient data to allow me to compute their preferences. i first sorted everyone into one of six timezone tranches from the original poll:


tranche 1 (europe, africa and the middle east): 43 respondents
tranche 2 (russia, most of asia, and western australia): 8 respondents
tranche 3 (the pacific rim aka japan, eastern australia, new zealand, oceania and hawaii): 5 respondents
tranche 4 (the western united states and western canada): 22 respondents
tranche 5 (eastern united states, eastern canada, the caribbean and latin america): 70 respondents
tranche 6 (greenland and some minor atlantic islands): 7 respondents.*

(*i really can't imagine there are seven of you who live in the azores, cape verde and greenland. i ended up snooping on the profiles of everyone who chose this option, and fixing the data set for people who demonstrably made a mistake. but if you don't have a location listed on your profile, i left your answer as it was entered...welcome to greenland!)

2. for each tranche, i first counted whether more people preferred to watch on weekends, or weekdays (people who said either didn't qualify). in each case, a substantially larger percentage preferred weekends. this matched the overall count: 69 for weekends, vs 20 for weekdays.

3. from there, i counted the most popular options for watching in local time:
tranche 1 (europe, africa and the middle east): weekend nights (9:00 pm to midnight)
tranche 2 (russia, most of asia, and western australia): weekend nights (9:00 pm to midnight)
tranche 3 (the pacific rim aka japan, eastern australia, new zealand, oceania and hawaii): weekend nights (9:00 pm to midnight)
tranche 4 (the western united states and western canada): weekend afternoons (noon to 5:00 pm)
tranche 5 (eastern united states, eastern canada, the caribbean and latin america): weekend afternoons (noon to 5:00 pm)
tranche 6 (greenland and some minor atlantic islands): weekend evenings (6:00 pm to 9:00 pm)

of course, these results are based on the local time of the respondents, which makes things complicated -- for example, a majority of respondents in both east asia and europea/africa/the middle east would prefer to watch the matches at night, but "night in seoul" and "night in paris" occur at very different absolute times.

4. therefore, i had to figure out some way of converting the respondents' preferences. my methodology was as follows:

4a. select a single location from each tranche to serve as the "marker" time zone
tranche 1: madrid (GMT +1)
tranche 2: beijing (GMT +8)
tranche 3: tokyo (GMT +9)
tranche 4: los angeles (GMT -8)
tranche 5: new york (GMT -5)
tranche 6: azores (GMT -1)
[PLEASE NOTE: none of the calculations i made take daylight savings into account, which varies country-by-country and even region-by-region (seriously, fuck you, arizona).]

*obviously, this presents certain problems -- most notably, it distorts the votes of people in a certain tranche whose local time differs significantly from the marker city. thus, a person in lahore's vote for "9:00 pm to 12:00 am" their time will be counted as a vote for "9:00 pm to 12:00 am" in beijing...or 6:00 pm to 9:00 pm in lahore.

in an ideal world, i would have allowed people to pick the precise time zone they lived in, to erase these errors, but the resulting data would have been even more unwieldy. in selecting the marker cities, i tried to pick cities in the most populous time zone of each tranche.

4b. convert each of the possible times of day (9:00 am - 12:00 pm, 12:00 pm - 5:00 pm, 5:00 pm - 9:00 pm, 9:00 pm - 12:00 am, 12:00 am - 9:00 am) into their corresponding local time in madrid.

therefore, a vote for 5:00 pm - 9:00 pm in beijing (GMT +8) is actually a vote for 10:00 am - 2:00 pm in madrid (GMT +1), a vote for 9:00 pm to - 12:00 am in beijing is actually a vote for 2:00 pm to 5:00 pm in madrid, and so on.

4c. tally the votes for each time of day, across time zones, as expressed in madrid time.

*this step also presented certain problems -- namely, that the times of day people voted for were not of uniform length. (see the above beijing -- to madrid conversion) i toyed with several methods of resolving these discrepancies, but in the end i just tallied according to my best guess. (this unfortunately means that some votes were counted twice, but hopefully the doubling was uniform across the tranches)

*other potential imprecisions in my method: i wasn't able to weigh the votes of people who could ONLY watch on weekends (or weekdays, for that matter) more than others, or people who could ONLY watch at a single time. i'm sure there was a way to do that if i had structured my questions differently, or if i had better manipulated the data, but to be honest this all felt pretty exhausting as it was.

in the end, the results were...
weekends, 9:00 AM - 12: 00 PM, GMT+1: 21 VOTES
weekends, 12:00 PM - 5:00 PM, GMT+1: 34 VOTES
weekends, 5:00 PM - 9:00 PM, GMT+1: 61 VOTES
weekends, 9:00 PM - 12:00 AM, GMT+1: 67 VOTES, or 43% of the total respondents
weekends, 12:00 AM - 9:00 AM, GMT+1: 54 VOTES

interestingly enough, this was actually the overall preference of a plurality of people in tranches 1, 4, 5, and 6. for anyone who didn't get their preference, please accept my sincerest apologies. i tried to use the most objective and rational method i could devise, and i truly wish i could have found a time that worked for everyone.

okay, but what is the exact time that we will be watching? and what day?
i don't know!

Poll rewatch opinion poll número dos!

if you need to look up your local time relative to madrid, the world clock converter is awesome.

what match are we going to watch?
up to you guys! if you have an idea, please nominate it in a comment, and then other people can "+1" your suggestion. i am pretty confident of our abilities to find downloads for all matches from last season, but we'll cross that bridge when we come to it.

this is insane...is there something wrong with you?
so many things. mods, apology for the avalanche of words....i'm gonna go take a nap.


EDIT: winkiss has just told me that the dates for the marcelo support box have been updated...click the link for more info!