Quote:
Originally Posted by buffoon OKno thank you
boyohboy, what is this, desperation?
If one wants to apply premises that devoid of logic one could even take 10 pct infection via homosexual contact vs. 90 pct infection via hetero contact and still come out with gay men having the higher propensity of infection (of the overall population). 
That is correct. For the homosexual men to have the same probability of being infected, they should represent 16% of the infected population just as they represent 16% of the total population.
Quote:
I don't know whether you don't realize it or are being intentionally obtuse but the relevant figure is the number of infected and their path (by percentage) of infection.

No, the relevant figure is the ratio of infected to not infected people in each group. It's simple, yellow and blue beads in a bag type conditional probability.
Let's have 8 yellow beads and 17 blue ones, so overall the chance of getting a yellow is 32%. P(Y) = .32
If you have four yellow beads in bag A and 1 blue, the chance of getting a yellow bead is 80%. P(YA)=.8
If you have 4 yellow beads in bag B and 16 blue, the chance of getting a yellow bead is 20%. P(YB)=.2
Here too we've got 50% of the yellow beads in each bag, but the chance of picking yellow is wildly different in the two cases.
The actual numbers will make the case more strongly later, but I'm just trying to remind you of the perfectly legitimate field of probability statistics.
Quote:
By the same logic I can mathematically make a case for drunk driving being the safer form to driving sober in that 10 pct (just making that up) of traffic accident related deaths are caused by lushes driving. From which follows that the other 90 pct are NOT caused by lushes and thus by sober drivers. Thus showing sober driving to be for more unsafe and giving cause to propagate that everyone shouldn't get behind a steering wheel until completely legless.

Well you could,
if that were the same logic. You're not considering the total number of drunk and sober people driving (or, probably more usefully, miles driven by drunk and sober people).
And of course, you could prove anything if you can make up the numbers.
Quote:
But, following the chart for UK, if we assume the number of infected men to be around 60,000 (overall 85,000 minus 26,000 women and without making allowances for children and others in the overall figure) and 40 pct of them infected thru heterosexual contact, we have 24,000 men and 26,000 women having infected each other as in one group the other. Which at least makes a case for men sleeping around more than women by about 2,000. 
OK, let's use those numbers then.
The UK working population is around 62 million. Assume that half are male, we've got 31 million. Some of those will be too young, but it doesn't actually affect the proportional comparison.
Assume that 10% of men are exclusively homosexual, and 90% are exclusively heterosexual.
That gives 3 million homosexuals and 28 million heterosexuals.
According to your data, we have about 24,000 infected men through heterosexual contact and about 23,000 infected men through homosexual contact.
That gives 1 in every 1167 heterosexual men is HIVpositive.
While 1 in every 130 homosexual men is HIVpositive.
Again that's being as generous as I can with the number of homosexuals. If it were indeed only 2%, the figures would come out at about 1 in 1266 and 1 in 27 respectively.
If your point is to demonstrate that people can and do get infected through heterosexual contact, it's granted of course. 24,000 men is a lot of lives. Further, if you knew that the particular homosexual in question were in a stable, faithful long term relationship with one man while the heterosexual has the sexual integrity of Al Capone, taking blood from the homosexual man would probably be safer. But if you're concerned about the risk of contracting HIV from a random unit of blood, you'd be a lot safer, ceteris paribus, taking it from a heterosexual man than a homosexual one.