WEBVTT Kind: captions Language: en 1 00:00:01.120 --> 00:00:03.680 A few weeks ago, I asked my  audience some odd questions, 2 00:00:03.680 --> 00:00:06.000 and more than 13,000 people answered. 3 00:00:06.000 --> 00:00:08.160 Some folks tried to work out what  the video was going to be about, 4 00:00:08.160 --> 00:00:09.840 but very few guessed correctly, 5 00:00:09.840 --> 00:00:12.560 and that's because they didn't  have the other piece of the puzzle: 6 00:00:12.560 --> 00:00:17.360 I also paid a professional polling service  to ask the same questions to the public. 7 00:00:17.360 --> 00:00:21.520 What I was actually trying to  answer is: How Weird Is My Audience? 8 00:00:22.160 --> 00:00:24.160 There’s a deliberate  double-meaning to "weird" there. 9 00:00:24.160 --> 00:00:26.000 If I poll people watching this channel, 10 00:00:26.000 --> 00:00:30.149 the results will have the same bias that plagues  psychology experiments: you're WEIRD. 11 00:00:30.149 --> 00:00:33.401 Western, Educated, and from Industrialised, Rich and Democratic countries. 12 00:00:33.401 --> 00:00:38.000 Not all of you, not as individuals, but overall, that's what viewers here tend to be. 13 00:00:38.000 --> 00:00:42.000 That WEIRD bias is why many psychology  experiments that talk about “humans”   14 00:00:42.000 --> 00:00:45.360 are actually just talking about college  students from America. But because of that, 15 00:00:45.360 --> 00:00:47.760 I didn’t compare this  audience to the entire world. 16 00:00:47.760 --> 00:00:49.440 That would have been very expensive, 17 00:00:49.440 --> 00:00:51.440 involve translating questions into many languages, 18 00:00:51.440 --> 00:00:55.061 and it wouldn’t actually have been that  meaningful to most people watching. 19 00:00:55.061 --> 00:00:59.680 Instead, the polling service asked that I pick one  English-speaking country to compare against. 20 00:00:59.680 --> 00:01:00.800 Most of my viewers are American, 21 00:01:00.800 --> 00:01:03.040 and the US was also the cheapest to poll, so, 22 00:01:03.040 --> 00:01:09.040 we are comparing my audience to a representative  sample of just over 1500 adults from the US. 23 00:01:09.040 --> 00:01:12.040 For some questions, I expected to  find a big difference. 24 00:01:12.040 --> 00:01:16.400 For some, I expected no difference at all.  And for some questions… well, 25 00:01:16.400 --> 00:01:18.144 I just found them funny. 26 00:01:18.144 --> 00:01:18.880 Question one! 27 00:01:21.120 --> 00:01:23.200 When I put a call out for questions to ask, 28 00:01:23.200 --> 00:01:24.885 dozens of people suggested this. 29 00:01:24.885 --> 00:01:27.840 So first, let’s see how the general public responded. 30 00:01:27.840 --> 00:01:31.760 There wasn’t a significant difference between any of the demographics there, 31 00:01:31.760 --> 00:01:34.240 “dog person” is safely in the lead. 32 00:01:34.240 --> 00:01:35.440 As for my audience… 33 00:01:37.440 --> 00:01:42.400 And I can’t explain that. I don’t think I’ve ever  made a video about cats or dogs or animals at all. 34 00:01:42.400 --> 00:01:45.760 Maybe if “infrastructure person” was an  option that would have won by a landslide. 35 00:01:45.760 --> 00:01:50.400 But something about this audience makes  you more likely to be a “cat person”. 36 00:01:50.960 --> 00:01:53.840 The internet likes cats, I guess. 37 00:01:56.240 --> 00:02:00.160 Next question. I put a call out for academics  who were looking for a large audience to poll, 38 00:02:00.160 --> 00:02:05.040 and while only one of those questions actually  made it into the survey, I really do find it fascinating. 39 00:02:05.040 --> 00:02:08.800 This came from researchers at the School of  Advanced Study at the University of London, 40 00:02:08.800 --> 00:02:11.520 who are investigating Olfactory Mental Imagery: 41 00:02:11.520 --> 00:02:13.760 the ability to imagine scents. 42 00:02:13.760 --> 00:02:17.391 I got a load of suggested questions from people who were interested in aphantasia, 43 00:02:17.391 --> 00:02:19.600 aphantasiacs are people who don’t have a mind’s eye, 44 00:02:19.600 --> 00:02:21.920 but that’s already been polled a lot: 45 00:02:21.920 --> 00:02:24.880 this is asking whether you have a mind’s nose. 46 00:02:25.440 --> 00:02:27.680 It took a while to hammer out  the question into a form that’d   47 00:02:27.680 --> 00:02:30.000 work for the polling company. So it’s not perfect -- 48 00:02:30.000 --> 00:02:34.320 it does exclude folks who can imagine smells  but can’t imagine pictures or sounds, 49 00:02:34.320 --> 00:02:37.520 but that should just be a  very small number of people. 50 00:02:37.520 --> 00:02:41.040 Here are the results from the  public. And they’re puzzling, 51 00:02:41.040 --> 00:02:44.240 because they do not match what  the researchers expected at all. 52 00:02:44.240 --> 00:02:46.720 About a quarter of folks said that they can’t imagine   53 00:02:46.720 --> 00:02:49.306 a loved one’s face or play back a  familiar song in their head. 54 00:02:49.306 --> 00:02:52.160 Now, aphantasia like that is a thing, absolutely, 55 00:02:52.160 --> 00:02:55.440 but it’s meant to affect only a small  percentage of people, probably less than 5%, 56 00:02:55.440 --> 00:02:58.880 so that indicates the question  might have gone wrong somewhere. 57 00:02:58.880 --> 00:03:00.960 The results from my audience, meanwhile... 58 00:03:01.840 --> 00:03:05.440 they're more or less what the researchers  expected. That makes a lot more sense. 59 00:03:05.440 --> 00:03:07.680 The most likely explanation, I think, 60 00:03:07.680 --> 00:03:10.880 is that it’s a complicated  question. It requires thought, 61 00:03:10.880 --> 00:03:14.800 and the public who were being asked, well, they're being paid to answer. 62 00:03:14.800 --> 00:03:17.280 A significant number of people won’t think about the question, 63 00:03:17.280 --> 00:03:19.040 and will pick whichever answer will get them   64 00:03:19.040 --> 00:03:22.080 through the survey as fast  as possible. So in this case, 65 00:03:22.080 --> 00:03:25.840 I’d trust my audience to be more  representative of the world at large, 66 00:03:25.840 --> 00:03:27.872 just because there’s no profit  motive there. 67 00:03:27.872 --> 00:03:31.840 So in this case: my guess is that you’re normal. 68 00:03:33.680 --> 00:03:35.120 Now, when I put the call out for questions, 69 00:03:35.120 --> 00:03:37.600 there were a lot of divisive  but meaningless ones suggested. 70 00:03:37.600 --> 00:03:39.520 Whether the toilet paper should go over or under. 71 00:03:39.520 --> 00:03:42.880 Whether the centre or the edge of the  brownie is better. But no-one suggested: 72 00:03:42.880 --> 00:03:47.600 should you put one or two spaces after a period, or a full stop if you’re British. 73 00:03:47.600 --> 00:03:49.360 And that’s the reason I went for the question. 74 00:03:49.360 --> 00:03:51.680 Because it is divisive out there in the world, 75 00:03:51.680 --> 00:03:54.160 but no-one in my audience thought to ask about it. 76 00:03:54.160 --> 00:03:57.520 HTML, the language that the web  is built on, collapses whitespace. 77 00:03:57.520 --> 00:03:58.720 In the code that makes up web sites, 78 00:03:58.720 --> 00:04:00.720 it doesn’t matter if you use  one, or two, or fifty spaces, 79 00:04:00.720 --> 00:04:04.160 they all get crushed down to one  unless you specifically tell it not to. 80 00:04:04.160 --> 00:04:06.786 And a lot of sites don’t  bother doing that. 81 00:04:06.786 --> 00:04:08.880 So often, even if you do type two spaces after a period, 82 00:04:08.880 --> 00:04:10.800 it’ll appear as one to whoever’s reading it. 83 00:04:10.800 --> 00:04:12.320 I’m not going to weigh in on the argument myself, 84 00:04:12.320 --> 00:04:15.600 because it could be a whole video on  its own, but I will say this: wow, 85 00:04:15.600 --> 00:04:17.760 does my audience disagree with the world here. 86 00:04:17.760 --> 00:04:22.160 The public are almost split down  the middle. 55% say “one space”, 87 00:04:22.160 --> 00:04:24.800 45% say two. If you break it down by age, 88 00:04:24.800 --> 00:04:26.640 there is a trend towards one space, 89 00:04:26.640 --> 00:04:30.560 but even among 18-24 year olds,  it’s still only about 70-30. 90 00:04:31.360 --> 00:04:35.520 My audience? 95% one-spacers.  Even for people over 65, 91 00:04:35.520 --> 00:04:37.280 more than three-quarters just use one space. 92 00:04:37.280 --> 00:04:38.720 And my theory on this is simple: 93 00:04:38.720 --> 00:04:40.960 most of the audience for  this poll came from Twitter. 94 00:04:40.960 --> 00:04:43.280 I did try to post it on my  YouTube community tab as well, 95 00:04:43.280 --> 00:04:45.120 and y’all crashed Google Forms by rushing to it. 96 00:04:45.120 --> 00:04:49.200 So that’s a new responsibility I have to  worry about. But even allowing for that, 97 00:04:49.200 --> 00:04:52.160 one space is the standard on  basically every social media site. 98 00:04:52.160 --> 00:04:54.560 If you do post with two spaces after a period, 99 00:04:54.560 --> 00:04:58.240 it’s now a pretty big sign that  you’re not used to the internet. 100 00:04:58.240 --> 00:05:00.720 And my audience tend to be used to the internet. 101 00:05:00.720 --> 00:05:02.480 So congratulations to the one-spacers. 102 00:05:02.480 --> 00:05:04.400 It’s pretty clear that you have won. 103 00:05:04.400 --> 00:05:08.330 You just need to convince the folks  who still hit the spacebar twice. 104 00:05:10.800 --> 00:05:12.880 I’ll be honest, this isn’t an academic question. 105 00:05:12.880 --> 00:05:14.800 No-one suggested it. But as I wrote the questions, 106 00:05:14.800 --> 00:05:18.320 and realised that a genuine polling  company would send this out to people, 107 00:05:18.320 --> 00:05:21.520 I just could not stop giggling at the  idea of something this ridiculous. 108 00:05:22.080 --> 00:05:24.880 This is a scruffy owl. It doesn’t look wise. 109 00:05:24.880 --> 00:05:26.640 It doesn’t look like the owl out of Harry Potter. 110 00:05:26.640 --> 00:05:29.040 It looks kinda like it’s  annoyed at being woken up. 111 00:05:29.040 --> 00:05:32.080 I did not expect there to be any difference  between the public and the audience, 112 00:05:32.080 --> 00:05:33.920 I just wanted to see some funny owl names. 113 00:05:34.960 --> 00:05:39.520 But there was a difference. Ignoring  the people who just typed “OWL”, 114 00:05:39.520 --> 00:05:44.160 the most popular names from the public were some  variation on “Hoot”, like “Hooter” or “Hootie”. 115 00:05:44.160 --> 00:05:45.600 Second place, “Olly” or “Oliver”. 116 00:05:45.600 --> 00:05:48.240 Which makes sense, “Olly the Owl” scans nicely. 117 00:05:48.240 --> 00:05:52.560 But from the people who watch my  videos: the top answer is “Henry”. 118 00:05:52.560 --> 00:05:55.200 Hootie is the eighth most popular. Oliver is 15th. 119 00:05:55.200 --> 00:05:59.440 There is a difference, for whatever reason, in the names that this audience chose. 120 00:05:59.440 --> 00:06:02.000 And: almost everyone read the owl as male. 121 00:06:02.000 --> 00:06:06.160 There’s only one definitely feminine  name in the Top 100, and that’s Athena, 122 00:06:06.160 --> 00:06:08.800 because the Greek goddess Athena  is often depicted with an owl. 123 00:06:08.800 --> 00:06:11.360 Even if you filter to just women replying, 124 00:06:11.360 --> 00:06:14.160 the top answers are Henry, Harold, and Gerald, 125 00:06:14.160 --> 00:06:17.520 and there are only four definitely  feminine names in their top 100. 126 00:06:18.240 --> 00:06:22.165 But yes, I would be remiss if I didn’t point out  a few of the individual outliers: 127 00:06:22.165 --> 00:06:25.760 Clovenhorn, Destroyer of Mars; Baron von Murderpillow, 128 00:06:25.760 --> 00:06:29.200 Flappy Ben Soulgazer; Jimmy Tallon; James Van Der Beak;   129 00:06:29.200 --> 00:06:32.194 Miss Scarlet Blumberton of East London; Persephone the Uncanny; 130 00:06:32.194 --> 00:06:36.867 XxX_The_Mouse_Killer_69_XxX, and Former UN Secretar-- 131 00:06:38.640 --> 00:06:41.524 -- and Former UN Secretary Gen... I can't say it! 132 00:06:42.400 --> 00:06:44.800 And Former UN Secretary General Ban Ki Moon. 133 00:06:46.000 --> 00:06:47.840 That'll have to do, I got through it. 134 00:06:50.320 --> 00:06:52.880 This question turned out to be kinda dull, 135 00:06:52.880 --> 00:06:54.720 because it’s exactly the answer you’d expect. 136 00:06:54.720 --> 00:06:58.560 This audience is more familiar with  internet phenomena than the general public, 137 00:06:58.560 --> 00:07:02.080 even for references that go back more  than twenty years, even for Chuck Testa, 138 00:07:02.080 --> 00:07:05.840 an obscure Rhett and Link gag from a  decade ago. Of course we’re in a bubble. 139 00:07:05.840 --> 00:07:08.480 It’s often difficult to remember  quite how much we’re in a bubble, 140 00:07:08.480 --> 00:07:12.480 but… yeah, turns out getting the result  you expect is a little bit boring. 141 00:07:12.480 --> 00:07:16.160 The younger you are, the more likely  you are to recognise any of these. 142 00:07:16.160 --> 00:07:19.120 And men are more likely to recognise them -- 143 00:07:19.120 --> 00:07:21.760 but that makes sense since I set the questions, 144 00:07:21.760 --> 00:07:25.600 there will be other references that I don’t  know that would cause the opposite result. 145 00:07:26.320 --> 00:07:29.040 However, that wasn’t the  main point of the question. 146 00:07:29.040 --> 00:07:32.000 The real reason was to see how  many people claimed to know   147 00:07:32.000 --> 00:07:35.680 what the Mist Challenge was. There is no such thing. 148 00:07:35.680 --> 00:07:37.520 There’s never been an internet craze called that. 149 00:07:37.520 --> 00:07:41.440 But just under 4% of the public  and 2.5% of this audience   150 00:07:41.440 --> 00:07:43.840 ticked the box anyway. Were they misremembering? 151 00:07:43.840 --> 00:07:47.600 Were they trying to appear more in-touch than  they actually were? I wouldn’t like to guess. 152 00:07:47.600 --> 00:07:51.440 But I’m going to claim that my audience is more  honest and trustworthy than the public. 153 00:07:54.800 --> 00:07:56.880 Which brings us to the final question. 154 00:07:57.440 --> 00:08:00.000 How weird do you think you are? 155 00:08:00.800 --> 00:08:03.840 No guidance. No examples.  Just a scale to pick from: 156 00:08:03.840 --> 00:08:05.760 “very normal” to “very weird”. 157 00:08:06.320 --> 00:08:09.920 And this… I think this was the  most revealing difference of all. 158 00:08:09.920 --> 00:08:11.280 Because for the general public, 159 00:08:11.280 --> 00:08:12.560 there’s quite a spread of opinion. 160 00:08:12.560 --> 00:08:15.280 7% are happy thinking of  themselves as “very normal”. 161 00:08:15.280 --> 00:08:17.520 14% think of themselves as “very weird”. 162 00:08:17.520 --> 00:08:18.880 Around half sit in the middle. 163 00:08:18.880 --> 00:08:21.040 Women tend to rate themselves slightly more weird;   164 00:08:21.040 --> 00:08:24.400 older folks tend to rate themselves  slightly less weird. That’s the public. 165 00:08:24.960 --> 00:08:26.320 But for this audience… 166 00:08:27.200 --> 00:08:32.800 Almost no-one wants to think of themselves as  being “very normal”. Less than half a percent. 167 00:08:32.800 --> 00:08:35.200 And under 5% think of themselves as “very weird”. 168 00:08:35.200 --> 00:08:39.440 More than 40% described themselves as slightly  more weird than average. Don’t get me wrong, 169 00:08:39.440 --> 00:08:41.760 that’s still the majority opinion of the public as well, 170 00:08:41.760 --> 00:08:45.600 it turns out most people like to think they’re  just a little bit out there compared to the world, 171 00:08:45.600 --> 00:08:50.000 but the people in this audience are much, much less likely to think of themselves as being   172 00:08:50.000 --> 00:08:52.640 on the extreme ends of the scale, in either direction. 173 00:08:53.520 --> 00:08:56.480 Is that because you’re more likely to be  spending lots of time on the internet, 174 00:08:56.480 --> 00:08:58.800 to have been exposed to more people and opinions, 175 00:08:58.800 --> 00:09:02.640 and therefore have a different boundary  for what “normal” and “weird” mean?   176 00:09:02.640 --> 00:09:06.038 Is it because, ironically, you are more likely to conform, 177 00:09:06.038 --> 00:09:09.615 and to want to sit in the middle of the chart and not draw attention to yourself? 178 00:09:09.615 --> 00:09:12.800 Or is it just that the public,  when paid to answer a survey, 179 00:09:12.800 --> 00:09:14.403 don’t think it through? 180 00:09:14.403 --> 00:09:16.697 I don’t know the reason why. But big picture: 181 00:09:16.697 --> 00:09:19.475 the answer to “how weird is my audience”? 182 00:09:19.475 --> 00:09:22.640 Probably, more weird than you think you are. 183 00:09:25.501 --> 00:09:28.175 This year's been really difficult for production, for obvious reasons, 184 00:09:28.175 --> 00:09:32.589 there's been a lot more filming against a green screen in a tiny flat than I thought there would be, 185 00:09:32.589 --> 00:09:36.865 so I do want to say thank you to William Marler, who's been my animator through this year, 186 00:09:36.865 --> 00:09:38.808 and who has put together these incredible graphics, 187 00:09:38.808 --> 00:09:41.399 there is a link to him and his podcast in the description.