For example, the average risk of contracting HIV through sharing a needle one time with an HIV-positive drug user is 0.67 percent, which can also be stated as 1 in 149 or, using the ratios the CDC prefers, 67 out of 10,000 exposures.The risk from giving a blowjob to an HIV-positive man not on treatment is at most 1 in 2,500 (or 0.04 percent per act).

Yet not even the Nate Silvers of the world would be wise to gauge HIV risk based on statistics. Numbers and probabilities can be miscalculated and misinterpreted.

Case in point: Having a 1 in 70 chance of transmitting HIV does not mean it takes 70 exposures to the virus in order to seroconvert.

Statisticians, in case you’re curious, do have a formula for cumulative risk: 1 – ( ( 1 – x ) ^ y ) in which x is the risk per exposure (as a decimal) and y is the number of exposures.

But let’s face it, many of us can’t tabulate the tip at a restaurant, so it’s unlikely we’ll whip out the advanced algebra during sexytime.

The risk of contracting HIV during vaginal penetration, for a woman in the United States, is 1 per 1,250 exposures (or 0.08 percent); for the man in that scenario, it’s 1 per 2,500 exposures (0.04 percent, which is the same as performing fellatio).

As for anal sex, the most risky sex act in terms of HIV transmission, if an HIV-negative top—the insertive partner—and an HIV-positive bottom have unprotected sex, the chances of the top contracting the virus from a single encounter are 1 in 909 (or 0.11 percent) if he’s circumcised and 1 in 161 (or 0.62 percent) if he’s uncircumcised.

A report by the Black AIDS Institute states that African-American same-gender-loving men have a 25 percent chance (which is one in four odds) of contracting HIV by the time they’re 25 years old—and a 60 percent chance by the time they’re 40.

Other researchers have predicted that half of all gay men in America who are 22 years old today will be HIV positive by the time they’re 50.

It’s also important to realize that during acute infection, the immune system has not yet created the antibodies that lower viral load, at least for a few years.

HIV tests that rely on antibodies may give a false negative reading during an acute infection, also known as the “window period.” The presence of another sexually transmitted infection (STI)—even one without symptoms, such as gonorrhea in the throat or rectum—can raise HIV risk as much as 8 times, in part because STIs increase inflammation and thus the number of white blood cells that HIV targets.

Phrases like “Pr EP can reduce your risk by 92 percent” tell us relative risks, but most people want to know absolute risks.