What about the application of Benford's Law? Granted, you are looking at the last digit instead of the first digit, but you are counting/enumerating addresses.
@aikimark1955 Interesting. After the first digit Benford's Law predicts each digit appears about 1/10th of the time pretty quickly, but there may be an effect for the second and third digits of a house number.
@singingbanana I looked at this from the standpoint that houses in the US are generally numbered either from the center of town (100 * number of blocks from center + house number on block) or starting with one, like my little cul-de-sac. This does give a 99 house limit to houses/block and house numbers are 1 origin. For full blocks, that gives at least one more odd number than even number (51:49 bias).
My analysis would have removed the block indicator and only analyzed the 1-99 numbers.
Just as a thought, how would this probability balance out in certain old streets. When those who were developing the buildings were more superstitious and left out house numbers like "13"
(I've seen many examples of streets missing the building number 13, so was just curious)
@JustMeHavingFun1 you use a sampling distribution. In stead of plotting one trial, you use many trials. ie, 100 trials of 10 coin tosses each. and you count how many heads you got in each trial.
So, just being curious, since I'm having a test on statistics tomorrow... Where did you get the σ? Our teachers always told us that the σ is either given, or you have to calculate it, knowing that a certain value is x times σ away from the average(in a question on a test.. I don't know a lot about "real life" statistics). So, how did you calculate the σ here? Thank you :)! (I'm sorry for my bad mathematical English, but doing science in your second language is difficult ;))
@xxSichaye You can calculate sigma if you know what sort of distribution it is. The illustration of a coin toss is a binomial distribution with probability of success p=0.5. Sigma can then be calculated as sqrt(np(1-p)), where n is the number of trials.
Isn't it so that this specific deviation tells you something about about the average length of the streets, counted in number of houses? A lower average would give you more deviation whereas a higher average would give convergence toward 50%. Or put another way, if all streets in the sample had, say, 10 or less houses, the odd houses would have a bigger impact.
While looking at the orangutang, I was thinking that the % of being in an even numbered house was 50% plus a VERY miniscule amount. Does Europe start its street with an odd numbered house? Cause my neighborhood starts its street with an even numbered house o_O for example, my street starts with 100 (my house) and ends at 109.
Hmmm.... do you really get 0.502 as a proportion if you assume a uniform distribution of street sizes up to 1000? If I calculate back from the 0.502, I get an average street size of about 125. That's not what you get when the distribution is uniform up to 1000. What's going wrong? (I'm quite sure the predicted ratio for the average is the same as the average of the predicted ratios.)
As previous posters have said, this probability greatly assumes that every street starts with 1. That is not necessarily the case. Many go by 10s or 100s. What makes this probability really work is that there are many houses that are the only ones on their street, and thus just call themselves #1. I've known rural areas like that.
But the science of it is that odd numbers should be slightly more prevalent, though I don't know about 50.2%.
Many places the numbers start on which block like between 3rd and 4th St would be 300-400 numbered houses. Some places also start with latitude degrees/minutes as the starting point of numbers. There is no science or math to back up your bullshit observation.
You are pretty dumb to think house go 1, 2,3 or 1,2 or 1,2,3,4,5.
That isn't why at all. You could argue it's because we don't have houses numbered zero. But in most places, north/east or left side of street are numbered even and the other side of the street is odd. So there is no reason why roads have more houses on one side or the other.
So your idiotic justification for an observation is completely wrong.
I would seriously though like to know what the odds are that a randomly chosen person lives in a house with a name. Houses in the countryside often do have names rather than numbers.
:3 is it 0.502 because we never count the number 0 for street numbers, and the highest digit for a house usually has a 9 for the last digit just as the highest 4 digit number is 9999 :) taking these into account in a sample of 0-99, there are 100 digits if you count the 0. But because house numbers never start at zero and 9 is an even number this boosts the probrability by 2 thousandths, making it 0.502. :D how was that for awake at 4:44 am math :D
HEY! just curious... this could acctually be used to find out what the average number of houses is on each street. like this: 0.5 / ( 0.50218 - 0.5 ) = 229.346
But if we assume that the numbering of houses on each street generally starts at 1 (not 0, not 2) and that there is a 50/50 chance that the numbering will end on an even or odd number, then it is clear that there will tend to be more odd numbered houses. Then, the fact that the ratio of odd to even numbers is .50211 simply means that the average end number is about 237.
If we suppose that there are as much streets with an odd number of houses as with an even number of houses, could we calculate the theoretical average of houses per street? With 50.2%, it would be 124.5? I consider this as a fail. In France we don't have the same system of numeration. Odds are on one side of the street, evens are on the other side. The poll could be interesting to be done in Great Britain only... ?
I always love your videos. Your enthusiasm about math is contagious! This was an especially interesting one. I love learning things about math that goes against what you would expect, or is counter intuitive, and learning why that is.
I doubt many Americans will be able to figure this ahead of your explanation. Every street here in the US, that I have ever seen, organizes the even and odd numbered houses on opposite sides of the street. If the poll was only taken by Americans, I would think that it would be much closer to 0.500.
@Mr. Banana have you ever done a vid on the birthday paradox? That one is tons of fun. I plan to do that with my math classes this year.
@MrChinleungyau still .500 because there is no guarantee of starting the series of flips with a certain side of the coin. This did not answer your question precisely but rather how I think you meant to ask it. The literal meaning of your question could give an answer of 0.00 to 1.0 depending on the number of times you flipped. For example, you may have a series of flips that never yields a lead in the number of heads. Conversely, if you flipped heads the very first toss in a series the condition
singingbanana, what about hotel rooms? As there there are a lot of hotels without room No.13, shouldn't there be a slight bias towards even numbered rooms?What is your estimate for the same?
Singing banana, please make videos about gambling :) I would esspecialy love one about roulette. I want to know about the casino's profit from the 0 and 00 :)
It seems like the sample he's speaking about is a voluntary response sample and, as a result, may be a biased sample (i.e. non-response bias). Inference techniques such as confidence intervals should be based on random samples. I have a tough time trusting the outcome.
@singingbanana Am I missing something? I don't see why being on a block with an odd number of houses would make it more likely for my house to have an odd number. For example, if I live on a block with three houses, but the first house is labeled as #2, then there are actually two even-numbered houses on my block (#2 and #4), compared to one odd-numbered house (#3). Now, clearly we should assume that the first house on every street is labeled #1. But how do we divvy up the blocks?
at 4:55 cant the same go the other way, say the red houses are odd and blue are even, if you start with a red house then yeah theres more oddnumberd houses(considering theres an odd number of houses on the street), but if you start on a blue theres more even houses, so doesn't that just cancel out and the percentage is 50/50?
At first, I thought, "of COURSE it's 50/50!" But after seeing the facebook poll results, I was shocked...the difference is way too big! Fascinating. Interesting explanation about the odd and even rows of houses, I think it's very *probable*...ahha excuse my bad pun :)
Oh my god... I was watching intently, because I was for some reason thinking that you meant, odd number of people living in your house. I would've gotten it right away if I understood correctly. I feel so ridic.
They don't have to go odd-even. That last house, say the 51st, can be on the "even" side... I do see how everything would cancel out... But to me it should cancel out to 50-50. Maybe architects or whoever have some tendency to build on the odd side first to balance it out... I donr know.
good article that gets you thinking. I can think of a couple of other exceptions but again those can probably be cancelled out or are so rare as to be negligible.
p.s. I think you're getting even more enthusiastic with every new video..... keep it up!
I'd just like to point out that here, in Quebec, odd numbers house are on the right and even numbers on the left, so having more odd numbers would mean there are more hourses on the right of the street, which is 50/50 as it is completely random from street to street, some streets starting odd, and others even.
@DarkPuMa1981 No it's not random as the counting would be weird. Having more even numbers than odd numbers makes the numbering quite weird (like 1 2 3 4 5 6 8). Just read the description.
@DarkPuMa1981 This is the way it is in many US cities as well-odd numbers on one side and even on the other. The problem is that zoning can totally hose the random aspect--one side of a street is residential while the other is commercial or industrial so barring a zoning change that street will NEVER have houses that are even or odd. Futhermore, if the number is proportional to the distance from some baseline then houses of different sizes also totally hoses the random factor.
@DarkPuMa1981 Are you sure it's completely random which side is odd? The statistics suggests it may not be. That's the point you see. I'm only guessing, but if the long side is numbered odd then the houses will be numbered consecutively, which is not true otherwise.
@DarkPuMa1981 Ooh, I've just found a wikipedia article that describes house numbering. I've put the link in the description. It says the odd side is put left/right (depending on country) as you look up the street with numbers in ascending order. I'm surprised it's not the idea of numbering houses consecutively. Again, the statistics show there are more odd numbered houses than even, justifying that is secondary. Interesting.
I'm confused. Why does the total number of houses on a street matter? In the US, all the addresses on one side of a street are odd and all the addresses on the other side are even, using the "hundred block" system where the 4400 block is roughly 44 blocks from downtown. I realize Europe uses a different system, where houses are numbered consecutively -- but it appears odds and evens still alternate sides of the street. Only very old streets are numbered consecutively. So, huh?
@gmsherry1953 That doesnt matter. The street has to end somewhere - either with an odd or an even number of total houses. If there is an even number of houses in the street there will be the same amount of even and odd numbered houses but if there is an odd number of houses in the street there should be 1 more odd numbered house. (People should really have to fully watch the video and read the description before being allowed to comment :/)
@GammahooX I watched the whole video. I may not have understood it, but I watched it. When there is an even side and an odd side of the street, and an odd number of houses, the last house is just as likely to be on the even side as the odd. If the lots are larger on the odd side of the street, there could easily be 3 odd numbered houses and 10 even numbered in a block. If there are houses on the odd side and apartments on the even side, there could be 10 odd numbers and 100 even numbers.
@gmsherry1953 I see now what you meant about reading the description. "But if the street had 101 houses I am assuming the majority have 51 on the odd side and 50 on the even side." But there is NO reason I can think of why that assumption would be VALID. He's a genius mathematician but I was an ordinary architect and that's not how street numbers work. At least not in the US, and maybe not in the UK.
@gmsherry1953 My statistical analysis shows the answer *is* 0.502. Justifying that answer is secondary. My argument works brilliantly with streets on one side and blocks of flats. With streets on two side I imagine they number the long side with odd numbers otherwise it will be strange. For example, 7 houses will be numbered 1 2 3 4 5 6 7 rather than 1 2 3 4 5 6 8. If that's not true it doesn't change the fact the answer is 0.502 and we have already come up with enough reasons to explain it.
@gmsherry1953 "I believe exceptions will either be insignificant, or cancel each other out " - Quote from the discription. Apartments are equally likely to be on both sides. If there is one more house on one side that side should be the odd one, because that way the counting makes more sense. Look:
Odd side is longer:
1 3 5 7 9
2 4 6 8
No problem.
Even side is longer:
1 3 5 7
2 4 6 8 10
Where is the number 9?
It makes much more sense to have the odd side to be longer.
@GammahooX@singingbanana This isn't worth the attention we're giving it. It may be entirely cultural, US vs. Europe. In my neighborhood the odds are south side of the street, the evens north, and the numbers are not consecutive. One side 4401, 4407, 4413, depending on how large the lots are, other side 4400, 4402, 4404 if the lots are smaller. If the Facebook poll was started by a European I guess mostly Europeans responded.
Dude, I wish you were my lecturer for maths at my uni, you seem legitimately interested in what you're talking about, and that's alot more than I can say for some of my current lecturers.
I have an address database for an association in New Jersey. These are all business addresses. They skew to even: 307 even to 258 odd. There were a lot of address that ended in '00'. Businesses love corners. But we're talking homes, not business.
In my town, the addresses on the northern or eastern sides of the streets are odd and those on the southern or western side are even. And there are address like 1000 and 4400, so many blocks start out with an even number.
Interesting. One would think it would be the other way around since most streets I've seen start with an even number. (e.g. 100) Why is this not the case??
In the US, even numbered houses are all on one side of the street; and odd numbered houses are all on the opposite side of the street.
Thus, the explanation given here doesn't apply. No matter how many houses are on one side of the street or the other, they are all even or all odd, with no bias.
@joblessjunkie um in my country too there are even houses on the one side and odd houses on the other but if the street has and even number of houses in total they will range 50/50 but if the street has odd number of houses the range will be slightly more for the odd houses if u get what i am saying :D
Part of the final explanation on "drawing it back to 0.50" is what I've used to describe my grading method to fellow math teachers. They didn't get it. As a result, they had a narrower spread in grades. When asked about this difference I explained the obvious (again) and they still didn't understand. Apparently, "my method was less consistent and caused great confusion, causing the wider spread of grades.".
Maybe since every street has a number one, the odd number houses start ahead. For example, if I went do the street I would begin with one, and then two, and then three. So it seems that it is the even numbers that are catching up with the odd numbers. The odd numbers are either one ahead or the same as the even.
Did you consider the fact that there may be more people living in an odd number than an even numbered house? or indeed vice versa? i.e. On a street with 3 houses (for simplicity), there may be 3 people in house (1), 5 people in house (2) and 2 people in (3). Although there are more odd houses, the number of people in odd and even numbered houses is the same.
@PurpleUkuleleGuy The number of people is not related to whether your house is odd or even numbered. That means it doesn't affect the poll, because the avarage odd house will contain the same amount of people as the avarage even house.
your results are flawed. it isnt .502 chance that any random person will live in an odd numbered house... it is the chanc that the next person who votes in that poll will say odd
people could be lying or voting with the majority, people who live in even numbered houses could be voting less than odd numbered houses, any poll (like on facebook) will be flawed in this way
i see your logic in explaining the results, but there could be people like me who live in even numbered houses that didnt vote
@gurioper There is no good reason why the poll would be biased. And with over 400000 responses, the large amount of data should be giving an accurate result with negligible error. The idea of statistics is then to take the data from this (large) sample and apply the results to the population as a whole. If you do that, the correct conclusion is then to say that the probability of living in an odd numbered house is 0.502.
@singingbanana your explanation is logical and i agree with it however, .002 is a small number and very close to .5
i know that polls can be flawed, for example Gallup polled a fairly large group of people in the 1800s for who they were voting for president, almost everyone answered republican the the democrats won by a landslide. the flaw was they polled by phone, if you recall back then phones were expensive so only the wealthy had them most of which were republican. you say the 400000...
@gurioper 0.002 is a small number, but that was what the bell curve stuff was about. The answer 0.502 falls outside what you would expect with 400000 coin tosses, with 95% of such experiments being within 200000 +/- 300. Therefore the difference 0.002, although small, is statistically significant. That is to say, extremely unlikely to be a simple variation due to chance, but a genuine difference. The conclusion is strong since 400000 is a large sample, regardless of the population of the world.
@singingbanana you make some very good points, and i agree with everything you said. I am biased though because i dislike facebook haha, but we were thinking about the data in a different way.
(based on your explanation) you saw the data and thought about how the world influenced the data, i did it the other way around, i thought about how the data changed the world (from what was actually true)
@gurioper Thank you for your comments, they were good, and forced me to make my case better. I have added it to the description, so it was genuinely useful for me. (And we had a polite conversation which is extra bonus points on the internet!) Thanks again.
@singingbanana Youre welcome, and it was nice to have a decent conversation (which is rare on the internet) and i learned a bit about statistics which is good because the only math i use practically is algebra (for programming)
@singingbanana (ran out of room) ...people were a large sample, but with over 300,000,000 people in the united states and 500,000,000+ on facebook, the sample was fairly small.
@gurioper 1. The poll isnt flawed as there is absolutely no connection between you voting and your house being even/odd numbered. 2. The sample is fairly big. Usually only a few thousand people are asked in these kind of polls and that is enough. 3. The explanation shows that this is pretty much the only way the outcome could've been. 4. Even tho 0.502 is pretty close to 0.5 if it was actually 50% the result would get closer to the 0.5 with bigger samples (law of big numbers) which it didnt.
Thinking about it a little bit more it doesn't make sense to me with the explanation given why there are more odd numbered houses as even numbered houses as the side of the road defines if it's an even numbered house or an odd numbered house. This means there are more houses build on a certain side of the road, which is different from placing the houses in a queue as done in this video 1, 2, 3
Thanks for the videos. As they always make me think about subjects I would normally not think about
@WhatforNameIsThat There's a bit in description about that. If the street had, say, 101 houses, then I imagine you would number the side with 50 houses as even and the side with 51 houses as odd. The numbering would be strange otherwise. In the same way, if one side is a few houses longer I can imagine that will be the odd side for the same reason. The video argument works brilliantly for a close or a block of flats though.
@singingbanana If it works like you say it works you could determine the average length (number of houses) of a street by looking at the amount of difference of the odd numbered houses compared to 0,5
With the number being 0,502 it means that the average street will be very long. I'm not sure if the 1 to 1000 which you refer to in your description means the average number of houses is 500 on a street which seems a lot.
@WhatforNameIsThat That's just a rough calculation to justify the answer 0.502. And that calculation assumes all streets sizes from 1 to 1000 are equally likely - which they obviously are not. But I have no data on how street sizes are distributed. All I know for sure is that the answer appears to be 0.502, and that answer is statistically significant (i.e. not just down to some natural variation from 0.5, but a genuine difference). Like I said, justifying why it's 0.502 is more difficult.
how would this change is the question was ONLY for people living on odd numbered streets?would the probeblility for odd go up without the even streets?dont know if i explained that well,like only streets with a total number being odd,and being polled for if you were even or odd.
Interesting. Now one silly (and somewhat unrelated) question. Suppose we have streets with an infinite number of houses. Would then the ratio between even and odd numbered houses be exactly 50:50? I suppose it would, but can that be proven mathematically?
@Screaminbug Yes, imagine the road length to be x and lets only look at one road. If x is even then it's 50:50, if it's not then there will be (x-1)/2 even houses and (x-1)/2+1 odd houses. Which leads to a (x-1)/2x vs (x-1)/2x+1/x. When x goes to infinity 1/x goes to 0 and therefore both sides will have equal chances (which means it must be a 50:50 chance). When all roads are equal (all being infinitely long) that means the chances will be 50:50 in the whole population.
Oddly, the number of a house around where I live is based on whether it is on the left or right side of the street. This should make the results a little more even.
I would like to see this poll redone with the question "Do you live in an even numbered house?" I would think there is a slight psychological advantage towards answering if it means you are part of the group.
@Crashkiller62 Like I said, justifying the 0.502 figure is more difficult. The average proportion of odd numbered houses in streets of size 1 to 1000 is 1/1000 sum (4k -1)/(4k-2) summing from 1 to 500. Which is 0.502. But that's only a justification now that I know the answer appears to be 0.502.
@dotarcher Good point, but you would expect there to be about half odd and half even numbered distances for the houses so the only effect it is likely to have is pull it closer to 50% slightly.
"It's possible, although very unlikely, to get 250,000 tails in a row".
I hear you bro... I hear you...
undergroundo 1 day ago
I moved from 7378 to 692. I beat the odds, HA!
Helldog4890 2 days ago
It's all because they start counting with 1 instead of 0.
supergsx 2 days ago
The probability that I live in an odd numbered house is 100%.
supergsx 2 days ago
Im moving across the road so i can be a part of the minority
Typho0n86 2 days ago
What about the application of Benford's Law? Granted, you are looking at the last digit instead of the first digit, but you are counting/enumerating addresses.
aikimark1955 3 days ago
@aikimark1955 Interesting. After the first digit Benford's Law predicts each digit appears about 1/10th of the time pretty quickly, but there may be an effect for the second and third digits of a house number.
singingbanana 3 days ago
@singingbanana I looked at this from the standpoint that houses in the US are generally numbered either from the center of town (100 * number of blocks from center + house number on block) or starting with one, like my little cul-de-sac. This does give a 99 house limit to houses/block and house numbers are 1 origin. For full blocks, that gives at least one more odd number than even number (51:49 bias).
My analysis would have removed the block indicator and only analyzed the 1-99 numbers.
aikimark1955 3 days ago
@singingbanana
Just as a thought, how would this probability balance out in certain old streets. When those who were developing the buildings were more superstitious and left out house numbers like "13"
(I've seen many examples of streets missing the building number 13, so was just curious)
JJ4eva2 1 week ago
Because it's discrete, wouldn't it be a normal distribution graph? You're probably just approximating it to a bell though.
8JSimo 2 weeks ago
Hey, sorry im a little confused, how can you plot a bell curve with a heads/tails? Thanks
JustMeHavingFun1 1 month ago
@JustMeHavingFun1 you use a sampling distribution. In stead of plotting one trial, you use many trials. ie, 100 trials of 10 coin tosses each. and you count how many heads you got in each trial.
Child0fAtom 3 days ago
You are that dude on numberphile, what are you doing here σ_ σ
blade3412 2 months ago
@blade3412 He was here first ;)
SREproducciones 1 month ago
Are there really streets with only even or obly odd numbered houses? Why?
jvarga222 3 months ago
But what about if your house as a name?
NATHANinGAME 4 months ago
So, just being curious, since I'm having a test on statistics tomorrow... Where did you get the σ? Our teachers always told us that the σ is either given, or you have to calculate it, knowing that a certain value is x times σ away from the average(in a question on a test.. I don't know a lot about "real life" statistics). So, how did you calculate the σ here? Thank you :)! (I'm sorry for my bad mathematical English, but doing science in your second language is difficult ;))
xxSichaye 4 months ago
@xxSichaye You can calculate sigma if you know what sort of distribution it is. The illustration of a coin toss is a binomial distribution with probability of success p=0.5. Sigma can then be calculated as sqrt(np(1-p)), where n is the number of trials.
singingbanana 4 months ago 3
the only problem i think of, is that most streets with more than 13 houses don't include a number 13, due to belief that it will bring bad luck
baggervance911 5 months ago
Isn't it so that this specific deviation tells you something about about the average length of the streets, counted in number of houses? A lower average would give you more deviation whereas a higher average would give convergence toward 50%. Or put another way, if all streets in the sample had, say, 10 or less houses, the odd houses would have a bigger impact.
Gameboygenius 5 months ago
Sorry xD I didnt finish the video when I asked that question; didn't see the exceptions part
TehMathWiztard 5 months ago
While looking at the orangutang, I was thinking that the % of being in an even numbered house was 50% plus a VERY miniscule amount. Does Europe start its street with an odd numbered house? Cause my neighborhood starts its street with an even numbered house o_O for example, my street starts with 100 (my house) and ends at 109.
TehMathWiztard 5 months ago
Hmmm.... do you really get 0.502 as a proportion if you assume a uniform distribution of street sizes up to 1000? If I calculate back from the 0.502, I get an average street size of about 125. That's not what you get when the distribution is uniform up to 1000. What's going wrong? (I'm quite sure the predicted ratio for the average is the same as the average of the predicted ratios.)
hwieldr 5 months ago
As previous posters have said, this probability greatly assumes that every street starts with 1. That is not necessarily the case. Many go by 10s or 100s. What makes this probability really work is that there are many houses that are the only ones on their street, and thus just call themselves #1. I've known rural areas like that.
But the science of it is that odd numbers should be slightly more prevalent, though I don't know about 50.2%.
cbrhawk1 5 months ago
Many places the numbers start on which block like between 3rd and 4th St would be 300-400 numbered houses. Some places also start with latitude degrees/minutes as the starting point of numbers. There is no science or math to back up your bullshit observation.
sifterjoe 5 months ago
You are pretty dumb to think house go 1, 2,3 or 1,2 or 1,2,3,4,5.
That isn't why at all. You could argue it's because we don't have houses numbered zero. But in most places, north/east or left side of street are numbered even and the other side of the street is odd. So there is no reason why roads have more houses on one side or the other.
So your idiotic justification for an observation is completely wrong.
sifterjoe 5 months ago
This has been flagged as spam show
1 out of 2
TheNoisePolluter 5 months ago
I would seriously though like to know what the odds are that a randomly chosen person lives in a house with a name. Houses in the countryside often do have names rather than numbers.
zyzzyzus 5 months ago
i live in an odd numbered house :3
Pwnafier 5 months ago
This has been flagged as spam show
@Pwnafier "i live in an odd numbered house :3"
What are the chances?!
Zigiwy 5 months ago
What if you live in a house with a name but not a number?
zyzzyzus 6 months ago
Thumbs up if you live in an odd numbered house!
darkghostmagic 6 months ago 14
I'm italian, my number is 21! You're awesome! :D
xXextraslastaXx 6 months ago
4 people live in an even numbered house.
ThePraash 6 months ago
:3 is it 0.502 because we never count the number 0 for street numbers, and the highest digit for a house usually has a 9 for the last digit just as the highest 4 digit number is 9999 :) taking these into account in a sample of 0-99, there are 100 digits if you count the 0. But because house numbers never start at zero and 9 is an even number this boosts the probrability by 2 thousandths, making it 0.502. :D how was that for awake at 4:44 am math :D
FlyingJynx 6 months ago
@FlyingJynx
At 4am in the morning, I'll forgive you for thinking that 9 is an even number.
fatsquirrel75 5 months ago
HEY! just curious... this could acctually be used to find out what the average number of houses is on each street. like this: 0.5 / ( 0.50218 - 0.5 ) = 229.346
imkrimbaldur 6 months ago
Maybe it is 50/50. but you are a bit more likely to answer the poll if you live in an odd number house...
TheBasicStamp 6 months ago
@TheBasicStamp No, because of the reason he explained.
anticorncob6 5 months ago
But if we assume that the numbering of houses on each street generally starts at 1 (not 0, not 2) and that there is a 50/50 chance that the numbering will end on an even or odd number, then it is clear that there will tend to be more odd numbered houses. Then, the fact that the ratio of odd to even numbers is .50211 simply means that the average end number is about 237.
rveenstra1 6 months ago
@rveenstra1 and if the ratio is extremely small the average end number would be lots higher that that.
minijimi 6 months ago
Am I the only person who just noticed that he records in this exact same room for most of his videos?
anticorncob6 6 months ago
You should play some of the Professor Layton's if you haven.t already.
PublicEnemyN7 6 months ago
If we suppose that there are as much streets with an odd number of houses as with an even number of houses, could we calculate the theoretical average of houses per street? With 50.2%, it would be 124.5? I consider this as a fail. In France we don't have the same system of numeration. Odds are on one side of the street, evens are on the other side. The poll could be interesting to be done in Great Britain only... ?
boumbh 6 months ago
I always love your videos. Your enthusiasm about math is contagious! This was an especially interesting one. I love learning things about math that goes against what you would expect, or is counter intuitive, and learning why that is.
masterlink101 6 months ago
I doubt many Americans will be able to figure this ahead of your explanation. Every street here in the US, that I have ever seen, organizes the even and odd numbered houses on opposite sides of the street. If the poll was only taken by Americans, I would think that it would be much closer to 0.500.
@Mr. Banana have you ever done a vid on the birthday paradox? That one is tons of fun. I plan to do that with my math classes this year.
jones3420012001 6 months ago
If you flipped a coin and stopped when the number of heads exceed the number of tails, what will the average ratio of heads to tails be?
MrChinleungyau 6 months ago
@MrChinleungyau still .500 because there is no guarantee of starting the series of flips with a certain side of the coin. This did not answer your question precisely but rather how I think you meant to ask it. The literal meaning of your question could give an answer of 0.00 to 1.0 depending on the number of times you flipped. For example, you may have a series of flips that never yields a lead in the number of heads. Conversely, if you flipped heads the very first toss in a series the condition
jones3420012001 6 months ago
I couldn't start thinking because the orangutan showed up...
topnsecret 6 months ago
I live in an odd numbered house.
bigboss643 6 months ago
singingbanana, what about hotel rooms? As there there are a lot of hotels without room No.13, shouldn't there be a slight bias towards even numbered rooms?What is your estimate for the same?
siddharthjoshi91 6 months ago
Singing banana, please make videos about gambling :) I would esspecialy love one about roulette. I want to know about the casino's profit from the 0 and 00 :)
your fan, best wishes from Lithuania
BachasLT 6 months ago
@BachasLT
If your casino has 00, then there are 38 possible outcomes (1-36,0, 00).
If you bet on 1 number the return is 35-1.
Therefore the casino profits 3 out of every 38 bets.
fatsquirrel75 5 months ago
It seems like the sample he's speaking about is a voluntary response sample and, as a result, may be a biased sample (i.e. non-response bias). Inference techniques such as confidence intervals should be based on random samples. I have a tough time trusting the outcome.
dindelicato 6 months ago
Sureley the proberblity is less than that because there are some houses like mine, which just hav a name...
sheepnumber12 6 months ago
Ummm....I crunched in a few numbers....(assuming probability =.502 for odd no. of houses). I got is an average of 125 houses on the odd street....
well practically that isnt the case... so what am i missing here?
geeta172 6 months ago
Why can't streets with an odd # of houses have more Even numbered houses than Odd numbered houses?
johnnyappleseed12345 6 months ago
@johnnyappleseed12345
I'd say they certainly can.
My belief is that if you are the first person to occupy a street, you get number 1 regardless. Therefore there is likely to be more odds than evens.
Happy to be proven wrong.
fatsquirrel75 5 months ago
So, SingingBanana, do tell me.
Do you live in an odd or even numbered house?
GrannyPig21 6 months ago
@GrannyPig21 Is this stalking my gradations? My number is even.
singingbanana 6 months ago 5
@singingbanana Am I missing something? I don't see why being on a block with an odd number of houses would make it more likely for my house to have an odd number. For example, if I live on a block with three houses, but the first house is labeled as #2, then there are actually two even-numbered houses on my block (#2 and #4), compared to one odd-numbered house (#3). Now, clearly we should assume that the first house on every street is labeled #1. But how do we divvy up the blocks?
macnolds 5 months ago
@singingbanana You are the 49.8%
Aviatorsmith 1 month ago in playlist Uploaded videos
at 4:55 cant the same go the other way, say the red houses are odd and blue are even, if you start with a red house then yeah theres more oddnumberd houses(considering theres an odd number of houses on the street), but if you start on a blue theres more even houses, so doesn't that just cancel out and the percentage is 50/50?
milosh1942 6 months ago
DAMN YOU ODD NUMBER OF HOUSES STREETS!!!
acdc10133 6 months ago
At first, I thought, "of COURSE it's 50/50!" But after seeing the facebook poll results, I was shocked...the difference is way too big! Fascinating. Interesting explanation about the odd and even rows of houses, I think it's very *probable*...ahha excuse my bad pun :)
doggylikesoup 6 months ago
I both percentages are less than 50%, because of:
1) houses with numbers like 95 1/2 (plenty of those in Los Angeles!)
2) houses without a house number
3) people who don't live in a house at all
CreatedByBrett 6 months ago
Okay, I really don't like, find interesting, or care about probability questions when one could argue it's all just down to chance.
08firthf 6 months ago
@08firthf I'm going to start a casino just for people who think the same.
singingbanana 6 months ago 14
Can I have your Orang-Utan, if you don't need it anymore?
donheller1 6 months ago
Oh my god... I was watching intently, because I was for some reason thinking that you meant, odd number of people living in your house. I would've gotten it right away if I understood correctly. I feel so ridic.
Mewigi 6 months ago
They don't have to go odd-even. That last house, say the 51st, can be on the "even" side... I do see how everything would cancel out... But to me it should cancel out to 50-50. Maybe architects or whoever have some tendency to build on the odd side first to balance it out... I donr know.
AnCoSt1 6 months ago
@AnCoSt1
I believe that back in the day if you are first in the street, you are number 1.
Therefore there will always be more odds than evens.
fatsquirrel75 5 months ago
good article that gets you thinking. I can think of a couple of other exceptions but again those can probably be cancelled out or are so rare as to be negligible.
p.s. I think you're getting even more enthusiastic with every new video..... keep it up!
sasfcps 6 months ago
@sasfcps Thank you! This one does have people thinking!
singingbanana 6 months ago
I suspected it would be slightly more than 50%, considering from 0001 - 9999 there are 4,999 even numbers and 5,000 odd numbers.
Assuming that half of all streets have even numbers, a street should have an average of 125 houses.
anticorncob6 6 months ago
I love you
nelsyeung 6 months ago
I thought 0.51.
Berntisso 6 months ago
left the video at picture time -.-
DontserXakaDX 6 months ago
In some places you may get 13a and 13b. Also are you counting
GSA14101996 6 months ago
I'd just like to point out that here, in Quebec, odd numbers house are on the right and even numbers on the left, so having more odd numbers would mean there are more hourses on the right of the street, which is 50/50 as it is completely random from street to street, some streets starting odd, and others even.
DarkPuMa1981 6 months ago 4
@DarkPuMa1981 No it's not random as the counting would be weird. Having more even numbers than odd numbers makes the numbering quite weird (like 1 2 3 4 5 6 8). Just read the description.
GammahooX 6 months ago
@DarkPuMa1981 This is the way it is in many US cities as well-odd numbers on one side and even on the other. The problem is that zoning can totally hose the random aspect--one side of a street is residential while the other is commercial or industrial so barring a zoning change that street will NEVER have houses that are even or odd. Futhermore, if the number is proportional to the distance from some baseline then houses of different sizes also totally hoses the random factor.
Maximara 6 months ago
@DarkPuMa1981 Are you sure it's completely random which side is odd? The statistics suggests it may not be. That's the point you see. I'm only guessing, but if the long side is numbered odd then the houses will be numbered consecutively, which is not true otherwise.
singingbanana 6 months ago
@DarkPuMa1981 Ooh, I've just found a wikipedia article that describes house numbering. I've put the link in the description. It says the odd side is put left/right (depending on country) as you look up the street with numbers in ascending order. I'm surprised it's not the idea of numbering houses consecutively. Again, the statistics show there are more odd numbered houses than even, justifying that is secondary. Interesting.
singingbanana 6 months ago
My house doesn't have a number at all!
Tealover0010 6 months ago
I'm lucky because i lived in an even numbered house :D
mudkippy998 6 months ago
I'm confused. Why does the total number of houses on a street matter? In the US, all the addresses on one side of a street are odd and all the addresses on the other side are even, using the "hundred block" system where the 4400 block is roughly 44 blocks from downtown. I realize Europe uses a different system, where houses are numbered consecutively -- but it appears odds and evens still alternate sides of the street. Only very old streets are numbered consecutively. So, huh?
gmsherry1953 6 months ago
@gmsherry1953 That doesnt matter. The street has to end somewhere - either with an odd or an even number of total houses. If there is an even number of houses in the street there will be the same amount of even and odd numbered houses but if there is an odd number of houses in the street there should be 1 more odd numbered house. (People should really have to fully watch the video and read the description before being allowed to comment :/)
GammahooX 6 months ago
@GammahooX I watched the whole video. I may not have understood it, but I watched it. When there is an even side and an odd side of the street, and an odd number of houses, the last house is just as likely to be on the even side as the odd. If the lots are larger on the odd side of the street, there could easily be 3 odd numbered houses and 10 even numbered in a block. If there are houses on the odd side and apartments on the even side, there could be 10 odd numbers and 100 even numbers.
gmsherry1953 6 months ago
@gmsherry1953 I see now what you meant about reading the description. "But if the street had 101 houses I am assuming the majority have 51 on the odd side and 50 on the even side." But there is NO reason I can think of why that assumption would be VALID. He's a genius mathematician but I was an ordinary architect and that's not how street numbers work. At least not in the US, and maybe not in the UK.
gmsherry1953 6 months ago
@gmsherry1953 My statistical analysis shows the answer *is* 0.502. Justifying that answer is secondary. My argument works brilliantly with streets on one side and blocks of flats. With streets on two side I imagine they number the long side with odd numbers otherwise it will be strange. For example, 7 houses will be numbered 1 2 3 4 5 6 7 rather than 1 2 3 4 5 6 8. If that's not true it doesn't change the fact the answer is 0.502 and we have already come up with enough reasons to explain it.
singingbanana 6 months ago
@gmsherry1953 "I believe exceptions will either be insignificant, or cancel each other out " - Quote from the discription. Apartments are equally likely to be on both sides. If there is one more house on one side that side should be the odd one, because that way the counting makes more sense. Look:
Odd side is longer:
1 3 5 7 9
2 4 6 8
No problem.
Even side is longer:
1 3 5 7
2 4 6 8 10
Where is the number 9?
It makes much more sense to have the odd side to be longer.
GammahooX 6 months ago
@GammahooX @singingbanana This isn't worth the attention we're giving it. It may be entirely cultural, US vs. Europe. In my neighborhood the odds are south side of the street, the evens north, and the numbers are not consecutive. One side 4401, 4407, 4413, depending on how large the lots are, other side 4400, 4402, 4404 if the lots are smaller. If the Facebook poll was started by a European I guess mostly Europeans responded.
gmsherry1953 6 months ago
0:01
Srxjer 6 months ago
My god, you are nerdy. I somehow really like it though.
Fartssmellgood 6 months ago
That picture is so cute!!!!
elizze6 6 months ago
my street has more even then odd numbers, there is a park where 3 odd numbers would be
ijustdontcarenemore 6 months ago
at 1:02 i left lol just kidding
fredrocks222 6 months ago
Dude, I wish you were my lecturer for maths at my uni, you seem legitimately interested in what you're talking about, and that's alot more than I can say for some of my current lecturers.
Jezza408 6 months ago
McFly? Probably not a Back to the Future reference, is it?
HaslamCorp 6 months ago
@HaslamCorp It is. The band take their name from BTTF... Not usually the sort of question I get to answer in these comment.
singingbanana 6 months ago
I have an address database for an association in New Jersey. These are all business addresses. They skew to even: 307 even to 258 odd. There were a lot of address that ended in '00'. Businesses love corners. But we're talking homes, not business.
HaslamCorp 6 months ago
You do know that "self selection" surveys are inherently flawed, right?
magicbuskey 6 months ago
i wonder how you talk about about this for 6 mis
nicholishere 6 months ago
In my town, the addresses on the northern or eastern sides of the streets are odd and those on the southern or western side are even. And there are address like 1000 and 4400, so many blocks start out with an even number.
HaslamCorp 6 months ago
Interesting. One would think it would be the other way around since most streets I've seen start with an even number. (e.g. 100) Why is this not the case??
Kevbox2008 6 months ago
4:23 i was waiting for u to point tht out since the start of the video :D
meat283 6 months ago
In the US, even numbered houses are all on one side of the street; and odd numbered houses are all on the opposite side of the street.
Thus, the explanation given here doesn't apply. No matter how many houses are on one side of the street or the other, they are all even or all odd, with no bias.
joblessjunkie 6 months ago
@joblessjunkie um in my country too there are even houses on the one side and odd houses on the other but if the street has and even number of houses in total they will range 50/50 but if the street has odd number of houses the range will be slightly more for the odd houses if u get what i am saying :D
meat283 6 months ago
Part of the final explanation on "drawing it back to 0.50" is what I've used to describe my grading method to fellow math teachers. They didn't get it. As a result, they had a narrower spread in grades. When asked about this difference I explained the obvious (again) and they still didn't understand. Apparently, "my method was less consistent and caused great confusion, causing the wider spread of grades.".
*sigh* Oh well, just a story you reminded me of.
Fun video! Interesting topic :-)
FHomeBrew 6 months ago
Even's got to be more likely
Xerxes4242 6 months ago
Thank you singingbanana for making facebook interesting!
ninjayuppi 6 months ago
haha, I love the Orangutan and kitten xD
Charettecody1 6 months ago
wow thank you so much singingbanana
Prolaxed 6 months ago
Was not expecting that orangutan
Prolaxed 6 months ago
Maybe since every street has a number one, the odd number houses start ahead. For example, if I went do the street I would begin with one, and then two, and then three. So it seems that it is the even numbers that are catching up with the odd numbers. The odd numbers are either one ahead or the same as the even.
EmperorNaman 6 months ago
@EmperorNaman That's my logic too. You must be an awfully clever chap :-)
nytrex2001 6 months ago
but what if one side of the street has odd numbered houses and the other side has even numbered houses?
Rockthemaf1s 6 months ago
@Rockthemaf1s Read the description.
GammahooX 6 months ago
@GammahooX thanks!
Rockthemaf1s 6 months ago
Law of large numbers :)
muffemod 6 months ago
Did you consider the fact that there may be more people living in an odd number than an even numbered house? or indeed vice versa? i.e. On a street with 3 houses (for simplicity), there may be 3 people in house (1), 5 people in house (2) and 2 people in (3). Although there are more odd houses, the number of people in odd and even numbered houses is the same.
PurpleUkuleleGuy 6 months ago
@PurpleUkuleleGuy The number of people is not related to whether your house is odd or even numbered. That means it doesn't affect the poll, because the avarage odd house will contain the same amount of people as the avarage even house.
GammahooX 6 months ago
Kris Draper! from the Detroit Red Wings?! :P
jampk24 6 months ago
I don't know how it works outside the United States, but in the US, odd numbers are on one side of the street and even numbers on the other side here
darththeo 6 months ago
@darththeo Read the description.
GammahooX 6 months ago
Id say they normally build houses in even numbers, but always start with number 1??
00beasis 6 months ago
I live in an even numbered house - I FEEL SPECIAL! :D
GammahooX 6 months ago
Seems obvious to me that there would be more odd number houses because counts start on an odd number.
I wonder though if there are streets without the number 13, for superstitious reasons and have the number 12A instead.
dantheist 6 months ago
your results are flawed. it isnt .502 chance that any random person will live in an odd numbered house... it is the chanc that the next person who votes in that poll will say odd
people could be lying or voting with the majority, people who live in even numbered houses could be voting less than odd numbered houses, any poll (like on facebook) will be flawed in this way
i see your logic in explaining the results, but there could be people like me who live in even numbered houses that didnt vote
gurioper 6 months ago
@gurioper There is no good reason why the poll would be biased. And with over 400000 responses, the large amount of data should be giving an accurate result with negligible error. The idea of statistics is then to take the data from this (large) sample and apply the results to the population as a whole. If you do that, the correct conclusion is then to say that the probability of living in an odd numbered house is 0.502.
singingbanana 6 months ago
@singingbanana your explanation is logical and i agree with it however, .002 is a small number and very close to .5
i know that polls can be flawed, for example Gallup polled a fairly large group of people in the 1800s for who they were voting for president, almost everyone answered republican the the democrats won by a landslide. the flaw was they polled by phone, if you recall back then phones were expensive so only the wealthy had them most of which were republican. you say the 400000...
gurioper 6 months ago
@gurioper 0.002 is a small number, but that was what the bell curve stuff was about. The answer 0.502 falls outside what you would expect with 400000 coin tosses, with 95% of such experiments being within 200000 +/- 300. Therefore the difference 0.002, although small, is statistically significant. That is to say, extremely unlikely to be a simple variation due to chance, but a genuine difference. The conclusion is strong since 400000 is a large sample, regardless of the population of the world.
singingbanana 6 months ago
@singingbanana you make some very good points, and i agree with everything you said. I am biased though because i dislike facebook haha, but we were thinking about the data in a different way.
(based on your explanation) you saw the data and thought about how the world influenced the data, i did it the other way around, i thought about how the data changed the world (from what was actually true)
gurioper 6 months ago
@gurioper Thank you for your comments, they were good, and forced me to make my case better. I have added it to the description, so it was genuinely useful for me. (And we had a polite conversation which is extra bonus points on the internet!) Thanks again.
singingbanana 6 months ago
@singingbanana Youre welcome, and it was nice to have a decent conversation (which is rare on the internet) and i learned a bit about statistics which is good because the only math i use practically is algebra (for programming)
gurioper 6 months ago
@singingbanana (ran out of room) ...people were a large sample, but with over 300,000,000 people in the united states and 500,000,000+ on facebook, the sample was fairly small.
gurioper 6 months ago
@gurioper 1. The poll isnt flawed as there is absolutely no connection between you voting and your house being even/odd numbered. 2. The sample is fairly big. Usually only a few thousand people are asked in these kind of polls and that is enough. 3. The explanation shows that this is pretty much the only way the outcome could've been. 4. Even tho 0.502 is pretty close to 0.5 if it was actually 50% the result would get closer to the 0.5 with bigger samples (law of big numbers) which it didnt.
GammahooX 6 months ago
Thinking about it a little bit more it doesn't make sense to me with the explanation given why there are more odd numbered houses as even numbered houses as the side of the road defines if it's an even numbered house or an odd numbered house. This means there are more houses build on a certain side of the road, which is different from placing the houses in a queue as done in this video 1, 2, 3
Thanks for the videos. As they always make me think about subjects I would normally not think about
WhatforNameIsThat 6 months ago
@WhatforNameIsThat There's a bit in description about that. If the street had, say, 101 houses, then I imagine you would number the side with 50 houses as even and the side with 51 houses as odd. The numbering would be strange otherwise. In the same way, if one side is a few houses longer I can imagine that will be the odd side for the same reason. The video argument works brilliantly for a close or a block of flats though.
singingbanana 6 months ago
@singingbanana If it works like you say it works you could determine the average length (number of houses) of a street by looking at the amount of difference of the odd numbered houses compared to 0,5
With the number being 0,502 it means that the average street will be very long. I'm not sure if the 1 to 1000 which you refer to in your description means the average number of houses is 500 on a street which seems a lot.
WhatforNameIsThat 6 months ago
@WhatforNameIsThat That's just a rough calculation to justify the answer 0.502. And that calculation assumes all streets sizes from 1 to 1000 are equally likely - which they obviously are not. But I have no data on how street sizes are distributed. All I know for sure is that the answer appears to be 0.502, and that answer is statistically significant (i.e. not just down to some natural variation from 0.5, but a genuine difference). Like I said, justifying why it's 0.502 is more difficult.
singingbanana 6 months ago
Thumbs up for the rare amount of people who live in an even numbered house?
weeman7970 6 months ago
I live in odd and even numbered houses
What now?
TKuja1 6 months ago
Busted are better.
ArcanoiseLegend 6 months ago
how would this change is the question was ONLY for people living on odd numbered streets?would the probeblility for odd go up without the even streets?dont know if i explained that well,like only streets with a total number being odd,and being polled for if you were even or odd.
gC222SA 6 months ago
@gC222SA My guess would be the average proportion in odd streets will be 0.504.
singingbanana 6 months ago
May live on the same house
GSA14101996 6 months ago
It could bet things like. 13b and 13a. Also more than one people m
GSA14101996 6 months ago
Interesting. Now one silly (and somewhat unrelated) question. Suppose we have streets with an infinite number of houses. Would then the ratio between even and odd numbered houses be exactly 50:50? I suppose it would, but can that be proven mathematically?
Screaminbug 6 months ago
@Screaminbug Yes, imagine the road length to be x and lets only look at one road. If x is even then it's 50:50, if it's not then there will be (x-1)/2 even houses and (x-1)/2+1 odd houses. Which leads to a (x-1)/2x vs (x-1)/2x+1/x. When x goes to infinity 1/x goes to 0 and therefore both sides will have equal chances (which means it must be a 50:50 chance). When all roads are equal (all being infinitely long) that means the chances will be 50:50 in the whole population.
GammahooX 6 months ago
@GammahooX Good answer.
singingbanana 6 months ago
@singingbanana Thanks, YouTube is already starting to make me enter CAPTCHA's because I posted so much. :D
GammahooX 6 months ago
I live in an even-numbered house; I'm special! Who's with me?
D3m190d 6 months ago 16
I actually took the time to think about it during the gorrilla.
EighteenCharacters 6 months ago
i dont get it what is odd and what is even
MrFjga 6 months ago
@MrFjga odd and even are VERY basic math. odd numbers are 1,3,5,7,9,11 and so on and even numbers are 2,4,6,8,10,12 and so on
emilkris33 6 months ago
Streets with odd number of houses? That's odd.
Th1s1sR3z0r 6 months ago
Oddly, the number of a house around where I live is based on whether it is on the left or right side of the street. This should make the results a little more even.
I would like to see this poll redone with the question "Do you live in an even numbered house?" I would think there is a slight psychological advantage towards answering if it means you are part of the group.
MarcusMacgregor2 6 months ago
@MarcusMacgregor2 The original question was "Is your house an odd number or an even number?"
singingbanana 6 months ago
most main streets in citys dont have a number 13 beacuse of bad luck
Ro6278 6 months ago
i also thought it will be slightly more but why is it exact 0.502? why isint it 0.503 or 0.53 or any other number?
Crashkiller62 6 months ago
@Crashkiller62 Like I said, justifying the 0.502 figure is more difficult. The average proportion of odd numbered houses in streets of size 1 to 1000 is 1/1000 sum (4k -1)/(4k-2) summing from 1 to 500. Which is 0.502. But that's only a justification now that I know the answer appears to be 0.502.
singingbanana 6 months ago
@singingbanana okay :D
Crashkiller62 6 months ago
before I watch this video, I'm gonna say that it's less than 50% because some roads will have 1 more odd than even number
velocityeleven 6 months ago
@velocityeleven sorry, I meant MORE than 50%
velocityeleven 6 months ago
Comment removed
BassmrDK 6 months ago
in some places, the number of the houses is the distance in metres from the starting point of the street.
dotarcher 6 months ago
@dotarcher Good point, but you would expect there to be about half odd and half even numbered distances for the houses so the only effect it is likely to have is pull it closer to 50% slightly.
Alfalotter 6 months ago