Yeah, that's bollocks. Someone's buggered up the explanation. Methinks the journalist wasn't paying attention when he carried out the interview.

Assuming no other information exchanged during the date about whether each man wants to see the girl again (which is of course ridiculous), the chance is 5/22, or about 22% (about 1 in 5, humorously). The chance for the guy is a measily 9%.

I agree with Octavious- the journalist misunderstood, wasn't paying attention, or both.

Yeah, the lady has a roughly 50/50 of finding *a* guy who wants to see her (5 in 22 on the first and 5 in 21 if she has to ask again). But it is always roughly 25% chance the specific guy would want to see her. The guy also has a roughly 50/50 chance of finding a lady (would have been easier with 20 dates) and a roughly 10% chance a specific lady would want to see him again.

Actually I think you don't account for the second date since it is sampling without replacement
For women, the odds the first choice isn't a hit is 17/22, while the second is 16/21 (since her first choice isn't in the pool and her first choice was a dud.). They odds that she strikes out with both is the product of these two odds or roughly 59%. This the odds that one or both is a hit is 41.2% roughly.

For men, same logic applies. The odds the first choice is a dud is 20/22. The second is 19/21, etc. that product for all five choices is also roughly 59%

If I understand right this is because assuming there are equal numbers in the pool men and women have to have the same overall odds on a second date because there's a man and a woman involved. I could be wrong on this last part though

Ogion - that makes sense. It shouldn't matter overall if women or men make more choices - the odds of them getting duds is also governed by what the opposite sex do. To illustrate this if every woman only picked 1 guy but every guy picked every woman the odds of a successful match are 100% for each sex.

Of course it would be an absolute 100% for women and a theoretical 100% for men.

All of the above is incorrect (trust me, I teach math).

If each man randomly choses 5 women from a group of 22 then a particular woman has a chance of (22-5)/22 = 0.773 to not be amongst the choices of any given man. Therefore if she choses two men the chance that neither reciprocates is 0.773^2 = 0.597 and thus the chance that at least one reciprocates is 1 - 0.597 = 0.403 or 40.3%.

Similarly for a man their chance not to be amongst the choices of any given woman is (22-2)/22 = 0.909. If they chose five women the chance that none reciprocates is 0.909^5 = 0.621 and thus the chance that at least one reciprocates is 0.379 or 37.9%.

Can I just point out that the title of this thread doesn't make any sense?

"Ogion - that makes sense. It shouldn't matter overall if women or men make more choices - the odds of them getting duds is also governed by what the opposite sex do. To illustrate this if every woman only picked 1 guy but every guy picked every woman the odds of a successful match are 100% for each sex."

Nope. That would only be the case if each woman picks a *different* guy. If they don't than only the chance for women would be 100% while for men it would only be 1 - (21/22)^22 = 64.1%

The 5/22 vs. 2/22 thing definitely doesn't imply the result maniac mentioned, but it worth mentioning that the result is possible, because the 2/22 is an average, and so some women pick zero people. For example, If two women picked all 22 men and all other women picked none, while 11 men picked each of the two women, it would be the case that women has a 50/50 chance of their offer being reciprocated and the men would still have about a 10% chance.

Sure, if you're using this as an example to understand probability/selections, but we know for a fact that these events are not independent - the men's and women's choices are infact informed by a date; so if you actually perform this experiment you'll find that you can get people picking 'good' dates, and throwing all your figures off.

It's a prob and stats thing only and the articles author clearly didn't understand something. I think we all agree.

I don't think we all agree. Pretty much everything except what orathaic and I said is irrelevant.

Also, I just reread what Maniac posted, and it says that each person has 22 dates, not that there are 22 of each gender. The 5:1 ratio between men and women's success rates can only be explained by there being 5 times as many men who pick more than 0 women as women who pick more than 0 men. And then orathaic is also right that the number of matches likely is not at all random.

I should clarify- that is for a man to have some match and vice versa. It's meaningless to talk about the probability of a specific offer, as all offers are not even close to equal.

It's speed dating. If you know what speed dating is, it is one man and one woman and am equal number of men and women. So there mist be at least 22 of each gender. It is possible there are more so that not every man saw every woman and vice versa, but no one would Swe the same person twice.

So I guess the odds change if there are 110 women and only 22 guys, but speed dating is realistically going to be pretty balanced. The hosts are to that.

See to that...

I'm not sure if that's the case. I had a friend who went to speed dating event that was "literature themed" about a week ago, and she said it was about a 2.5 to 1 ratio women to men. I have no idea if that is typical, but it seems very possible that other "themes" could strongly skew the gender ration one way or the other.

I guess. Sports themed would probably skew male centric.

I wish the OP had included the original article.

http://www.bbc.co.uk/news/magazine-26172314

At your service, Draug

I need to reread all the posts above when I'm not on the phone, but the wife and I are off to a delayed Valentine's dinner (nasty weather in the great lakes states last night) so I'll have to do it later. I did note that this wasn't done as a sampling from just one event but from 80 something events, but when you take into consideration the other things pointed out by the researchers, you'll note people picked others similar to themselves, so it's about observed statistics, not probabilities.

As we can link to the actual research it would be interesting to see if that was wrong.