How Marketers Twist Statistics | Don McMillan Comedy
HTML-код
- Опубликовано: 26 сен 2024
- If marketers can twist statistics to support a claim within their marketing campaign, then what is it that's stopping me from doing it too?
🌐Social Media🌐
📶Facebook: / donmcmillan
📶Twitter: / donmcmillan
🤓Don’s Bio🤓
America's #1 Nerdy Comedian
What do you get when you cross an Engineer with a stand-up comedian? You get Don McMillan. This former chip designer has been doing his one-of-a-kind, PowerPoint-Driven comedy show for audiences for over 20 years. In his show packed with graphs & charts, Don will show you the funny side of your world that has been sitting right in front of you - you are just too busy working to notice. Don graduated from Stanford University with a Master’s Degree in Electrical Engineering. He then went to work at AT&T Bell Labs where he was part of the team that designed the world’s first 32-bit microprocessor. He then moved to Silicon Valley where he helped launch the start-up company, VLSI Technology. Then after 15 years in the tech world, Don quit his job to become a stand-up comedian. That year he won $100,000 as the Comedy Grand Champion on “Star Search”. Don’s been seen on “The Tonight Show”, “HBO”, and the “Comedy Central”. These days, Don spends most of his time writing and performing customized corporate comedy shows for companies like Google, Apple, Amazon, Microsoft, Ford Motors, and Exxon/Mobil. Don has performed more than 800 corporate shows in the last 20 years and he was named the #1 Corporate Comedian by the CBS Business Network.
My favourite statistic from Australia:
20% of fatal car accidents involved alcohol.
That means 80% of fatal accidents didn't, therefore it's much safer to drink and drive!
Sounds logical😂
I'm screwed, i don't drink alcohol.
Juuust in case anyone is thinking that this means that drinking and driving isn't a problem, it's important to consider what the unmentioned factors were in the other 80%. Other drugs, exhaustion, texting, etc
@terranovarubacha5473 it's simpler than that. How many people drink and drive? 1 in 10? Probably less, most people in rush hour are sober, but let's say 1 in 10. That would mean that 10% of drivers account for 20% of crashes. That means that drunk drivers are more than 2 times more likely to crash. If you consider that 1 in 20 drink then drunkards are 4.75 times more likely to crash (rate = percentage of crashes / percentage in the population)
OK, vodka is BACK on the menu guys... 🥳🥳🥳
As an engineer working with marketing on claims literature right now, this is closer to reality than most people would believe.
It's also how pretty much all news coverage of science works.
Who wouldn't believe it? It's only funny because it's true
@@terranovarubacha5473 good point...
Who is disbelieving this? It's common knowledge.
I don't think it's necessarily that most people wouldn't believe it, it's that most either don't know how math works or aren't smart enough to understand. Marketing is designed to catch someone's attention at a specific point, and it works all too often. I heard a statistic that something along the lines of "your teenage driver is 50% more likely to get into a car accident if there are two or more passengers in the car." Parents don't hear that and ask what the likelihood is of a car accident in the first place. They don't ask how many are fatal vs a simple fender bender. They hear 50% and their first thought is "my god, if my kid drives his friends around he will get into an accident." Not all obviously, but you get the idea. The fact that people lack the brain power to think about these things scares me more than the fact that marketing companies do it in the first place.
The old joke is a statistician looked and found that the chances of a bomb being on the plane you’re on is 1 in a million. But the chances of you being on a plane with TWO bombs is 1 in a billion. So he simply carried a bomb with him on every flight because the odds there would be a 2nd bomb is so much higher.
unclecreepy. It's like hitchhiking..... I always feel safe because what's the chance of 2 psychopaths being in the same car.....?😅
*lower
Reminds me of Baldrick in Blackadder Goes Forth: ruclips.net/video/y8wdynZ0iWg/видео.html
This is a joke for the many people who don't get probability. A statistician would know that the chance of a second bomb being on the plane when you are already sure there is one is exactly the same as one being on the plane in the first place.
A classic logical falacy. If he always has a bomb with him, then the odds of one bomb being on the plane are now 100%. So he was obviously a stupid statistician. LOL etc.
I think this guy is probably my favorite comedian, like he doesn’t talk about anything controversial. No sex, no politics, just nerdy cool stuff. And I am all for that
Two out of ten people agree.
He doesn’t need to swear all the time either.
Try Brian Regan - he's clean too
So no funny business, just facts.
Manipulating data is politics.
Statistics say the reason for divorce :
5% out of boredom
30% monetary problems
25% infidelity
100% marriage
And statistically divorced people are more likely to be lonely, so to avoid loneliness, you should not get married.
Statistics is like a bikini. What it reveals is suggestive. What it conceals is vital.
Nice...I'll quote that one!
"There's 3 kinds of lies. Lies, damn lies and statistic." Mark Twain
@@monkeybusiness673 Just don't use it for any monkey business 😉😆.
@@flamenmartialis6839 While I was thinking of the same quote, thank you for writing it, I do like the bikini idea for the problem.
@@flamenmartialis6839I’ve seen that quote attributed to two different Prime Ministers of England so, statistically, it probably wasn’t Twain.😊
Grandma Grace's 2nd law: Figures don't lie, but liars can figure.
Yes. Numbers don't lie, but you can lie with numbers.
I’ve used the Alzheimer’s and smoking one as an example to my friends or students for how stats can be twisted. Heard it first by a prof back in grad school who said something along the lines of “Smoking has a lot of negative health effects, but it’s one of the few things known to drastically reduce your chances of getting Alzheimer’s.” I don’t think most people got it.
You should look up how much Russian Roulette lowers your chances of Alzheimers.
@@markvanpopering4598 statistics say, Russian roulette winners have a far higher chance of getting any desease afterwards than the ones who loose
One of my favorites is this: Pretty much every bar graph you see doesn't actually start at 0. They'll go from like 3,000 to 3,100. Suddenly a 50 point difference looks a lot bigger now doesn't it?
The worst offenses of this is when comparing 2 graphs, one has the scale completely changed. Very common.
Also omitting axis labels, also using nonlinear scales for axis values
What this reminds me of is the quote from Fight Club "on a long enough timeline, the survival rate for everyone drops to zero." Graphs are done that way on purpose, yes, but also to show performance metrics that make sense. If you start every single one at zero and have specific intervals the graph would look flat all the time. Hence the quote. I have no doubt that people manipulate them all the time, but many also don't. If you don't do that you get a false picture one way or the other. Good numbers don't look good at all, and vice versa.
@@BoraHorzaGobuchul Nonlinear scales for axis values can be reasonable, most notably logarithmic scales but there are certainly other ones useful for very niche applications.
Discontinuous scales though? Pure evil.
Favorite marketing meme; “up to” really means LESS THAN. “Removes up to 100% of grease.”
for those in the region of nyc, phil rizutto for the money store where you could "borrow up to $25,000 or more."
Technically it means "less than or equal to"
Also, "up to X or more" conveys no information whatsoever, which is why it shows up so often in advertisements.
@@JohnDoe-nq4du Got to love it when people use so many words to say nothing at all, while also pretending they're saying something. And when I say love, I probably really mean loathe. Yes, that's the word...
In your example, it's not so much the use of "up to" , but rather the use of "up to" in combination with "100%" . If it had been "removes up to 50% of grease", the statement would have had some informative value.
Survivorship bias on the potato chip statistic. The ones who died of chipping can't answer the survey.
Plus, “people” answered, not people who ate pretzels.
Always be wary of the survivorchip bias.
Cheese doodles could be the real death trap; the statistics don't even mention them, so we may infer that they could very well have anywhere up to a 100% mortality rate associated with them.
I mean, technically, every single consumable has a 100% mortality rate; after all, almost everyone in the world who's ever eaten or drunk anything has died at some point afterward!
Actually, doesn't the statistic in the video suggest that _potato chips_ are three times safer* than pretzels (instead of the other way around)? After all, three potato chip eaters/chokers survived to participate in the survey, vs. only one pretzel eater/choker...
;-)
*Note: I'm aware that the wording should be "three times as safe", not "three times safer" (which would suggest four times as safe), but I'm keeping my wording in synch with that used in the video.
This video was 40% funnier than the last one I saw which was 20% longer, so there's an inverse correlation all up in here
Brevity is the soul of wit.
I was all set to add my worthless two cents on marketing deception but he presented this so well, there's no need! The only thing I would add is they do the exact same thing in politics.
The more I watch this fella, the more I like his schtick. Genius is rarely appreciated in its day.
"What's your name, maths guy?"
"Mathews"
Gesundheit!
@@magicmulder lol :)
Years ago there was a PR drag race between a Russian and an American car. Of course we won, we created the sport. Apparently the Russian newspaper headline read; Russian car finishes in second place, American car finishes next to last. Statistically accurate. Dunno if it's true but it's a great example.
80 year old guy goes to the doctor and says his right leg hurts. The doctor says ‘well, you’re not getting any younger’. The old man replied, ‘my left leg is just as old and it doesn’t hurt at all!’
lol, I must remember that one next time I go to the doctor
'There are three kinds of lies: lies, damned lies, and statistics.' Mark Twain
That was actually british prime minister benjamin disreali.
@@sugarnads It wasn't. That's a common misconception. Twain, as a joke, once said Disreali said it.
He explained exactly what media does every day
This is actually important for people to understand.
Actually, it doesn't matter if you don't see the bear. What matters is whether the bear sees *you!*
And make sure you run faster than the guy next to you ...
The numbers never lie. It’s the people interpreting them that do
Often the people collecting the "data" are also flawed.
@@crissd8283 Like those "studies" run by and for various commercial companies under a veil of being "scientific".
1 in 10 people like our product? Toss that study, run it again. 5 in 10? Better but not what we were ordered to 'prove'. 9 out of 10? Okay, those ten people get to have their results published, cool beans. Nobody needs to know we ran 30 sample groups for this, it's not part of the 'truth' our client company wanted to show.
101 in binary is 5 in decimal, so that is the math/computer science "high five!"
Just wanted to know if 01 and 10 in binary mean something in decimal. How can I find out?
@@patriciat.8121 If only you had access to a vast reserve of information on tap .... 😉BUt yes they do. If you have Windows your Calculator may have a Programmer mode (via the View menu) which will, at the selection of an option, swap between various common bases (10 - decimal, 2 - binary, 8 - octal, and 16 - hexadecimal) and the number shown in the display will change to reflect the base selected.
I remember reading a funny one in Reader's Digest when I was a kid: 'The homicide rate in Switzerland doubled last year. (It went from 1 homicide to 2 homicides.)'
Same as some fear monger click bait I saw. The amount of airplane turbulence over the Atlantic increased by 56% over the past few years. 940 hrs up to 1,680 hrs in total compared to millions of hours of flight time each year crossing the Atlantic Ocean.
Most automobile accidents occur within 50 miles of the address on the vehicles registration. Therefore, register your vehicle as far away as possible
Very, very funny, intelligent and witty. Love it.
I thought I had seen all is routines, but this is brand new for me. It's a treat to see something new.
I always admire his jokes which are mostly PG. You don't have to be using sex, politics or even race to be funny. And to add it all up he can crack you up with slides 😂😂😂
Two out of ten people agree.
Personally especially the sex jokes bore me. Most of us have been there, it's really not a controversial topic. Just something that is better kept private.
@@Volkbrecht a few here or there are fine. When it's your whole set, that's sad.
Somehow this reminds me of that study about colon cancer.
They compared eating habits from USA vs south africa and found out, that south africans are less likely to get colon cancer due to their diet.
Only problem: Average life expectancy at this time was 55 years in south africa and chances for colon cancer increase at 65 years.
The study may have corrected for that factor, simply by comparing the same age cohorts. "Life expectancy was 55 years" does not mean that there were no 65-year olds (or older) to be found in South Africa.
Oh how this brings back memories of the past 3 years
Yeah, like how the jab was going to reduce your chance of dying from virtually nothing to virtually nothing.
SOOOOO GLAD this guy popped up in my recommendations!!! Definitely my kind of comedy! Extra points for the novel/original jokes/concept! Unique AND rofl hilarious!!!! New sub for sure.
The periodic table tie is ❤🔥
Why is there RDJ from the future in the crowd having a laugh? 2:20
This is absolutely rampant in politics as well lol, and really any other area that people who don't understand statistics try to use them to fit their agenda.
I really enjoy Don's comedy! Great stuff!!!
In the popular statin trial for Lipitor, about 10,000 high risk participants were studied.
at the end of the trial, 1.9% taking Lipitor had a heart attack. 3% of the placebo group had a heart attack. But Pfizer advertised that Lipitor reduced the risk of heart attack by 36% (1.9 is 36% less than 3)
Some years ago my employer ran an internal charitable giving drive, and part way through the campaign the organizer circulated an email showing participation for each department. .... The Marketing department had 104% participation! 😮
I would think that people who have seen a bear are 100% more likely to be mauled by a bear than those who haven't seen one.
Who do I give the props too, Don McMillan or his mom 😂
Hats off to you Don!
As an unreasonable person who also happens to see a lot of bears I now live in fear
Cam Dobson. That sounds unbearable!!
"Don't Look at Bears" is a good alternate title for this show.
over half of Americans think that 2/4 are bigger than 1/2
My dad used to say when you quoted statistics that he had read that 95% of people who die of cancer of the rectum have used toilet paper and he’s not going to quit using toilet paper.
High five ☝✊☝
😆😆
You should be part of "An evening of unnecessary detail!"
He didn’t really talk about it, but that’s the problem with random polls where they ask 1,000 people. If you went to a school with 2,000 kids, or lived in a town of 5,000 people and 1,000 of those people were polled, it might have some relevance.
If you poll 1,000 people and it’s suppose to represent 332 Million people, then anything you gleaned from the poll is meaningless. We polled 1,000 people and 12 of them have been to the moon, therefore, it’s estimated that almost 4 million people have been to the moon.
There's lies, damn lies, and statistics.
It's amazing how data can be manipulated, especially if you disregard the whole causation/correlation thing.
That's the quote I was gunna put up, but you beat me to it, what are the odds?😉
One of my favorite stats from a decades old Norwegian comedy skit: 9 out of every 10 people who slipped on a banana peel and hurt themselves last year drank coffee.
That's an example of one of the most common mistakes made using statistics, even in fairly well organized scientific studies, that just because two things are correlated doesn't mean there is a causal relationship.
THANKS DON,🤗 FOR THE LAUGHS 😂💚💚💚
I remember all the X times more likely when they were marketing the jab. Very few believed me when I said X times more likely means nothing if you dont know what the first number is. 10 times more likely than 1 in a million is only 10 in a million.
Remember, 9/10 dentists recommend our toothpaste!
That’s because they give them the tiny tubes and toothbrushes.
still wanna know why that last dentist dont reccomend anything
They weren't paid...
I can contribute with a few as well:
Some years ago WHO put obesity as #1 on their list of things that endangers human health - AIDS was #2.
Conclusion: The fatter you are, the more your health can improve by getting AIDS!
Statistics say that spending a long time on the road increases the risk of getting into an accident.
Conclusion: The shorter the time spent on the road, the greater the safety. So therefore you must always drive as fast as you possibly can.
A business man saw a beautiful yacht at the harbour with a sign saying "For sale! 300,000 euros", but he didn't want to buy it.
Conclusion: He saved 300,000 by not buying it and money saved is money earned, therefore he will have to pay 50% in tax of those 300,000 he earned.
Similar to when Wash DC politicians boast they cut the deficit. Translation: they decided to only over spend $500 billion when they were originally thinking of over spending $1 trillion. Therefore, they cut the deficit by$500 billion.
There was a real example of that, I can't remember the details, other than it was a transport-related statistic, but the statistic being tracked was fatalities per unit of time, not unit of distance traveled, which caused the fatalities statistic to rise as vehicles traveled faster even as the vehicles got safer and there were fewer deaths.
Simple example, if speed doubled, and journey times halved, if deaths remained constant the number if deaths per hour of travel time, would double.
I love this guy!!!
My dad would have loved this guy, my dad was an engineer, a physicist and a chemist,
Someone please explain the last joke, how do you get 10x? In order to get 10x I'm assuming you'd have to have ppl who get mauled by a bear but never see one, right? How is a person that never see a bear but gets mauled by one calculated?
The math drove me crazy too
im convinced its not a real joke and hes playing the audience ive been thinking about it for 20 minutes
That puzzled me too, and I came specifically to this comment section to find an explanation.
But I think I've figured something out.
Notice that in the given data, 1,677,345 is exactly 5 times 335,459 ; so I think the numbers were specifically invented in order to achieve some desired numerical outcome, and the conclusion "10x more likely to be mauled" doesn't come out of nowhere.
However, I think he made a "mistake"; the third line should have been "8 out of 9 people have never seen a bear" (wouldn't it make more sense that the majority of people have never seen a bear?), in order to arrive at the conclusion that people who have seen a bear are 10x as much at risk.
(Maybe he changed the line in order to play this particular audience, possibly in connection with a pun on the word "bear". Maybe he referred to "bears" or "2 out of 10 people" earlier in this performance.)
If the third line was "8 out of 9 people have never seen a bear", the derivation could go like this:
1 out of 335,469 bears will maul a human; "therefore" if you _see_ a bear, there is a chance of B = 1/335,469 that it will maul you. This rate applies to 1/9 of the population.
For the other 8/9 of the population, who have never seen a bear, the rate is A, which is yet unknown.
However, for the total population, the average rate is 1 in 1,677,345 chance of getting mauled by a bear; which is exactly 5 times as small as B. This gives us the following relation:
(8/9)*A + (1/9)*B = B/5
Solving for A yields
8A + B = 9B/5
40A + 5B = 9B
40A = 4B
10A = B
A = B/10 ==> the chance of getting mauled by a bear is, among people who have never seen a bear, one-tenth of that chance among people who have seen a bear.
I hope that helps.
I guess Weight Watchers is saying that you can lose three pounds a week, and then remain in Weight Watchers to help keep it off after that.
So you spend the rest of your life three pounds lighter. Good diet plan...
Reminds me of a Bill Hicks stand-up bit about how people in Marketing should unalive themselves...
Torture the data until it tells you what you want to hear
You can call 25 people “dozens of people”
24, even!
Yeah why wouldn't you?
Brilliant stuff ❤
4:21 Doom Cyberdemon in the crowd.
4:55 Doom Hell Knight dies of cringe.
I'm always wary of people who use percentages instead of actual numbers of things. For example electric car sales have increased by X% per month hiding the fact that 100 EV's cars sold last month and 107 EV's sold this month.
Right. A serious disease goes up 100% percent overnight among a particular population. "OMG, run for the hills!", "Get the jab!", only to find out that yesterday 1 person was sick, and today 2 people are sick.
Thus halfway through the days of AIDS, we were breathlessly told that women were the group where AIDS was growing the fastest. Problem was that women were less likely to engage in risky behaviors that led to HIV transmissions. So 1 woman with AIDS today and 2 women with AIDS tomorrow made for scary statistics, but a nothing burger kind of pandemic.
All brought to us by Dr Fauci, no less!
3:32 Dr. Gundry be like
Actually a true statement at the end, when you encounter bears you aren't supposed to look them in the eyes since it's a form of challenge🤔
Remindes me of a book or article from the 60's that was titled "How to lie with statistics". Really, you could look it up which was a joke line from a Jewish comedian back in the day.
Number one threat to American: Bears
Who remembers this?
The math on the final slide has triggered my brain. He could've used the correct mathematics and the joke still would've worked. 10x more likely?
My guess is that the third line, "2 out of 10 people have never seen a bear", was changed from something like "8 out of 9 people have never seen a bear". Because wouldn't it be more realistic that a majority of people have never seen a bear?
Maybe the line was specifically changed for this audience, maybe he made a pun on "bear" or "2 out of 10 people" earlier in this performance.
I was thinking of starting an internet weight reduction course with the bonus of you can eat what you like and as much as you like. $20 a month and if you do not lose weight during the month you get $10 back.
1 in 1.7 million people getting mauled by a bear is still a lot of bear maulings.
Please do a segment on toilet paper math!
Mu Alpha Theta ... that was the math team of our school I was on.
Tell your mom you don't need to see the bear to get mauled by it.
On second thought, maybe don't do that, she'd probably have a heart attack. Maybe. I don't know the statistics, you're the math guy, you tell me.
The most important statistic is that 9 out of 10 doctors hate that Doctor that keeps messing up their perfect score.
"75% of heart problems are of smokers 50+ years old" -> If you don't smoke, you won't make it to 50...
Your mom really said out of sight out of mind
No one plays with the numbers and statistics more than politicians.
Marketers lies more
These are great
Dude math teachers that are funny are the best.
Well played!
There's an actual video of a physician claiming smoking lowers your chances of getting Alzheimer's 😂😂😂
'3 times safer' (3:10) - nope. 2 times safer. 2 more than 1 is 3.
Or "3 times as safe".
@@yurenchu Yup. Although safety does not really have a scale- but risk does. Safety is a lack of risk.
@@Tensquaremetreworkshop Touché! :-)
Or maybe we could express it as "999/997 = 1.002 times as safe" (so actually 0.2% _safer_ )
@@yurenchu There is a wealth of nonsense out there these days- 'three times smaller'- so minus two times the size. Negative size is difficult to comprehend...
@@Tensquaremetreworkshop LOL! Might come in handy when you need to bring bad news: "The _profit_ of your investments is three times smaller than last quarter..."
==> Investor thinks "okay, 66.7% less, that could happen", but you know it's actually 300% less...
Hey… you got a laugh outta your picture. It’s worth it.
Binary high five...oh...sniff, beautiful man...stealing that. 101!
He looks like a young and skinny Clancy Brown lol
In using his bear mauling statics times the amount of bears I've seen in the wild, and including the ones at the Zoo, how do I say this accurately and succinctly? Oh yes, I'm in really big trouble.
the smoker voice should be done imitating the electronic voice
I remember having his album back in 2000
Thanks 😂
Statistics are bs. We learned how to alter results in psychology and sociology.
Mum is right
There’s an old saying, “Figures lie and liars figure”. Marketers just prove it.
Since they introduced the national lottery to the UK in the mid-90s, I've been telling people how to double their chances of winning. 😎👍
Are you allowed to tell those jokes after the marketing of a remedy for the most popular cold with 97% efficacy has begun?
My Grandma keeps trying to get my Dad to smoke and drink so he can avoid alzheimer's.
great numbers.🙂🙂
where does the "10x more likely to get mauled" figure come from? Or is it just comedy?
The young man photo is definitively Raphaelian, or rather Joycean if you swap 'artist' with 'nerdist'.
Cigarettes are the main cause of statistics...