20

It may be the eye of the beholder, but Winter Bash 2020 seems to be much more popular than years before. Some evidence for this can be found in SEDE; I've tried to display the progress of some values x days after Winter Bash started.

For example, the score of the Hat List question:

enter image description here

or the number of answers to the Show Off Your Hats topic:

enter image description here

or even the number of questions about Winter Bash:

enter image description here

I know during last year's Winter Bash many users were really fed up with the Code of Conduct changes and the company removing diamonds, which might explain the low score of the Hat List question, but 2019 wasn't a special year for the other two metrics, and they have grown enormously in 2020. But perhaps I'm looking at the wrong figures?

I'm not sure if the number of earned hats will tell us anything; that could be very much dependent on how easy they are to earn and my gut feeling says most hats are earned by users who don't even actively participate in Winter Bash and perhaps even ignore it altogether (just like most of the network's traffic is from anonymous users visiting the sites from a search engine).

So, what's the story behind the success of Winter Bash 2020? Is the pandemic causing everybody to stay at home, with more time on their hands to participate? Is this year's hat selection offered by the company much better than years before? Or is everybody enjoying Hat Dash or the other Unicorn challenge a lot?

9
  • 28
    I think 2020 in general could use some light entertainment ...
    – rene
    Dec 29, 2020 at 11:42
  • How do these numbers correlate again usage of the SE network as a whole? I think as a whole this year, people are seeking calmness and pleasure wherever they can get it. I certainly feel there’s generally more positivity around.
    – user960635
    Dec 29, 2020 at 11:44
  • People like Winter Bash. I wish there was a slogan for Winter Bash each year :)
    – user895495
    Dec 29, 2020 at 19:24
  • What does the score of 197 for 2017 (in light blue) mean on your Hat List Question? I checked and the Hat List Q on 2017 has a positive score of 221. In your graph, 2019 (yellow) scores just 67 while I see the Hat List Q for that year scored 77, so we must be looking at two different things. Jan 2, 2021 at 12:44
  • @Mari-LouA it had a score of 197 fourteen days after posting. It would make sense to extend the x-axis to the length of Winter Bash; I might do that tomorrow (next Sunday's data won't be covered in SEDE, but that's just a minor thing). Even then, those posts might receive upvotes during the year or even during next year's Winter Bash.
    – Glorfindel Mod
    Jan 2, 2021 at 13:41
  • Also for some reason it seems it's easier to get hats this year. I currently have 16 hats without even trying to get any of them whereas 2 years ago I got fewer and actually tried a bit :P Jan 2, 2021 at 21:11
  • There are many reasons why so, and I don't want to put my point in an answer, but a short form of my answer is "more got closer". Jan 3, 2021 at 18:50
  • There are some interesting statistics posted at Do we have comparative statistics for Winter Bash 2019?. Would be worth to consider updating them after the /stats page is up.
    – Panda
    Jan 5, 2021 at 14:28
  • For me, it's the first year I've really been able to participate. I normally take a few weeks off around New Year, so usually miss most/all of hat season. But this year, stuck with not much else to do, as Scotland is one of the countries affected by the 2020 coronavirus outbreak. :-( Jan 5, 2021 at 21:12

7 Answers 7

26

Like most of the things in 2020, we can just blame it on the corona pandemic.

But first, let's talk about last year, 2019:

  1. Code of conduct controversy

  2. Moderator removal and resignation

  3. Recycled Winter Bash

So participation was in decline due to all the backlash and lack of enthusiasm.

Now if we talk about 2020:

  1. Spare time on hand: Work from home and lockdowns.

  2. Better Winter Bash than last year: new hats, hat dash, etc.

It's been a hectic year for so many reasons, so maybe people are looking for such fun events they can participate in without risking their health, so they opted in.

22
+100

As the product manager for Stack Exchange, I also did some internal analysis to gauge the popularity of Winter Bash in 2020 vs. the last two years.

I approached it a bit differently -- instead of looking at Meta posts/participation, I looked at some server-side events. We only had data going back to 2018, so we were able to analyze 2018, 2019 and 2020.

TL;DR: In 2019, there was a significant drop in Winter Bash participation relative to 2018. In 2020, there was a big bump relative to 2019, but no statistically significant differences between 2018 and 2020. My interpretation is that the community issues in 2019, coupled with the fact that the hats and triggers were recycled, led to decreased participation that year.

Awareness

I defined awareness as people who visited any Stack Exchange site and then opened up the Winter Bash inbox that shows which hats you've earned. (Clicking the Snowflake icon opens the inbox.) Site visits are sampled at 1%.

Year Visited site Opened WB inbox % opened inbox
2018 70,211 3,850 5.5%
2019 81,774 2,315 2.8%
2020 64,285 5,105 7.9%
  • In 2019, awareness dropped by 48% compared to 2018.
  • In 2020, awareness increased by 181% compared to 2019. There was no statistically significant difference between 2018 and 2020.

Participation

I defined participation as people who opened up the Winter Bash inbox and then proceeded to wear a hat.

Year Opened WB inbox Wore hat % wore hat
2018 258,017 59,444 23.0%
2019 151,410 24,139 15.9%
2020 380,401 76,140 20.0%
  • In 2019, participation dropped by 31% compared to 2018.
  • In 2020, participation increased by 26% compared to 2019. There was no statistically significant difference between 2018 and 2020.

Aversion

I defined aversion as people who opened up the Winter Bash inbox and then opted out by clicking "No hats for me, please."

Year Opened WB inbox Opted out % opted out
2018 258,017 8,386 3.3%
2019 151,410 7,620 5.0%
2020 380,401 10,290 2.7%
  • In 2019, aversion increased by 55% compared to 2018.
  • In 2020, aversion decreased by 46% compared to 2019. There was no statistically significant difference between 2018 and 2020.
3
  • 1
    Why does the "Opened WB inbox" column in the "Awareness" section have so much smaller values than the same column in the other sections? Does "sampled at 1%" mean that this is 1/100 of the actual values (even in this case, the values do not match with the other sections, so there is something I do not understand...)?
    – wimi
    Jan 5, 2021 at 17:58
  • 2
    @wimi There are two factors: 1) the sampling itself is the main factor and 2) it's only counting events that are preceded by the "visited site" event. In Participation and Aversion, we are looking at the raw "Opened WB inbox" without sampling and without a preceding event Jan 5, 2021 at 18:32
  • 1
    Aren't all events preceded by the "visited site" event? How can one open the Winter Bash inbox without visiting the site?
    – wimi
    Jan 5, 2021 at 19:35
17

Personally speaking:

  1. I missed out last year.
  2. It’s not been a great year for many people, so taking part in some enjoyment makes me feel good.
  3. It’s nice to bond with people in trying to find the secret hats and the secrets behind them.

This year, for me, hasn’t been much different in terms of Winter Bash, but I imagine people simply want to have a little fun.

The questions I’ve seen about Winter Bash have been from users who haven’t really noticed it before - I wonder whether there’s been any different ways of promoting it this year that have increased the engagement.

1
  • Also, I would like to add that, more users joining communities each year.
    – siniradam
    Dec 31, 2020 at 17:50
10

The question is worded cleverly, the OP asks why participation is so high on Winter Bash 2020.

In fact, high participation does not necessarily mean 2020's hat chase was more popular than ever. If we take a look at the following Hat List posts, which I'll use to measure consumer satisfaction, we'll see the score for this year's Winter Bash Hat List is currently in 5th place, narrowly pipped by 2017 but well short of 2014, 2015, and 2016 scores. The latter with its impressive positive score of 613, 75 bookmarks, and 16k views is unrivalled.

Hatlist Views Bookmarked Total score Upvotes Downvotes
2020 Winter Bash 2020 hat list 👒 🎩 8k 46 214 216 2
2019 ❄️ Winter Bash 2019 Hat list ❄️ 5k 19 77 103 26
2018 Winter Bash 2018 Hat list 9k 50 203 209 6
2017 Winter Bash 2017 Hats 10k 44 221 225 4
2016 Winter Bash 2016 - All Hats 16k 75 613 627 14
2015 Winter Bash 2015 Secret Hats 12k 57 365 377 12
2014 Winter Bash 2014 Secret Hats 10k 54 343 348 5
2013 I have a Secret Hat 4k 4 92 100 8
2012 What do you think of Winter Bash? 4k 2 164 181 17
2011 Holiday 2011 Hat Dash: The Hattening N/A N/A N/A N/A N/A

Judging by the scores and the number bookmarks, participation in this year's event has seen an upward trend compared to that of 2019 and 2018, which must mean a considerable number of users enjoyed the challenge of collecting as many hats as possible. And it's not over yet, we may well see a last minute sprint before 2020 Winter Bash ends on January 4.

1
  • Wow. I thought I went back far enough with 2017, but apparently I was wrong :)
    – Glorfindel Mod
    Jan 2, 2021 at 13:42
7

Concerning the number of questions about winter bash:

Year Questions Questions tagged Questions not tagged
2020 96 48 48
2019 47 19 28
2018 53 14 39
2017 55 14 41

So you see the number of questions concerning winter bash have increased, but the majority of the increase is because there are many more bugs this time ;).

The number of answers to Show your Hat Off may have been greater this time because there is was a bounty by @Tinkeringbell on it. And you know, we have the This is Fine hat so...


Not saying winter bash isn't more popular this year, its just that some of your metrics may have been influenced by other factors as well :)

6

Here's another set of stats

Questions Tagged Questions Closed Answers
105 9 266
47 5 236
53 (51*) 10 568 (276*)
55 12 211
102 28 144
85 27 150
110 24 257
154 30 312
113 36 263

* Excluding the two posts: Announcing Our Amazing 2018 Stuff-A-Way! and New swag for Stack Overflow's 10th birthday: wear-ables or desk-ables? with their 252 and 40 answers respectively.

A special mention goes to 2018 when a staggering 292 answers were submitted in two heart-warming competitions but completely unrelated to holiday knitting, leaping unicorns and hats and is therefore not indicative of Winter Bash participation.

If participation is measured in number of answers posted, then 2020 wins the bronze medal with its 266 answers behind the 2018 and 2013 editions when 276 and 312 answers were submitted. However, the first prize for the highest number of questions posted during Winter Bash goes to 2013 for its 154 questions albeit 30 questions have been closed for being off-topic or duplicates.

Hat tip to Paulo.Almeida who created the query number of answers per tag

4

Here are my thoughts,:

Being at home and having more spare time on my hands definitely contributes, when you're sitting at your computer working from home, and you decide to just click over to SE to earn hats and chat about them with other users. I also really enjoyed Yaakov Ellis's invitation for us to pentest Sparkles the unicorn, and the quiz.

I also think the chance to have a good bit of fun at the end of a trying year is very appealing. It's also completely transformed my view on hats. I may have one custom-made.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .