Full list of memes and stats collected.
CONTEXT: (I know most people here know about it right now, but this is for future readers and others out of the know.) From December 5 to 15, Know Your Meme held a poll where users could select 3 memes of their choice to compete for top spots among "Memes of the Decade" (that being from 2010 to 2019). 105 memes were listed as well as an "Other" write-in option. After the poll closed, each day from December 16 to 20, the top 50 memes as voted were revealed in increments of 10.
Original post links for convenience:
Announcement and poll
#50-41
#40-31
#30-21
#20-11
#10-1
But how many people voted? The only figure explicitly provided in the final post is "nearly 55,000 votes", but also in that post, each of the top 10 memes had a vote count and a percentage (these were regrettably not provided for #50-11). By averaging each of the vote counts divided by the percents (assuming the percents are that of voters), about 22,457 users voted. By these figures, the average voter cast ~2.45 votes out of a possible 3. In other words, ~67,371 votes were possible, 12,371 of which went uncast and/or were write-ins (if write-ins were discarded in the same manner as uncast votes; they weren't quantified or mentioned in any way after the poll).
.
For each of the top 50 memes, a "peak year" was listed. In addition to finding the "peak years" of the other 55 memes (usually from Google Trends), I also found all their "origin years", the years I judged to be their origin as a meme (not always what KYM lists). For the top 50 memes (both year types), there seems to be a slight preference for older memes as the ranks get higher, but the R² is so small for both that there's almost no correlation (Figure 1). Interestingly, no meme originated or peaked in 2019. My assumption is they didn't want any overlap with the 2019 Meme of the Year poll.
.
I then looked at the differences between the peaks and origins. Naturally, most memes originate and peak in the same year (leaving a difference of 0), but some of the more special ones can last a while and peak the year after or even later.
For the top 50 memes, 20 (40%) have a difference of 0 years, the average difference is 1.96 years, and the median difference is 1 year.
For all 105 memes: 49 (~47%) have 0, average 1.6, median 1.
For the losing 55 memes: 29 (~53%) have 0, average 1.27, median 0.
There seems to be something of a tendency for memes with longer shelf lives to have been more popular while flash in the pan memes generally got fewer votes. However, again the R² is tiny and points to no correlation (Figure 2).
.
Looking at the breakdowns of the peaks and origins by year for the top 50, bottom 55, and all 105 candidate memes contradicts the picture given by the correlation between earlier years and higher ranked memes in the top 50 (as low as the correlation was). 61 of the 105 candidates (~58%) peaked in the later half of the decade (2015-2019). Of these, 34 are in the top 50 (68%) and only 27 are in the losing 55 (~49%). Later peaking memes were more likely to be voted into the top 50, but among these, it appears a special few earlier memes were voted en masse to the top of the top 50 and skew their ranks to lean earlier (again, by a small correlation). The origin years have a similar pattern to the peaks. However, 11 of the 105 candidates (~10%) originated before 2010 and of these, 7 are in the top 50 (14%) and only 4 are in the losing 55 (~7%) (this pre-decade pattern contradicts the contradiction and reinforces the earlier pattern). See Figure 3 for some pie charts of these year sections.
For a more detailed look at the same information, see Figure 4 for some histograms of the individual years.
I just faded them together for this last one. Never mind the first word.
.
The last factor I looked at was the page views of all the candidate memes (collected Dec. 28, 2019).
The top 50 memes averaged 1.97 million views with a median of 1.00m.
All 105 memes: average 1.68m, median 0.98m.
Losing 55 memes: average 1.42m, median 0.90m.
Clearly (and as one would expect), the more popular memes in the poll tend to have more views. See Figure 5 for the relationship views has with the top 50 memes' ranks (another tiny correlation) and the top 10 memes' vote totals (a much larger but still less than significant correlation).
.
All in all, the KYM userbase exhibited somewhat predictable voting patterns, but ultimately not so much, reflecting the variety of individual user tastes and their divergence from the wider normie meme populace. I'll end on this observation: the 2nd most viewed entry Doge won the #1 spot while the most viewed entry Slender Man didn't make the top 50. Make of that what you will.