It seemed to be as good a way as any to start researching an article on predicting lifespan. I asked Google, “How long will I live?” It offered me some calculators. The fun began.

Even before I started calculating, I had concluded I should plan for a long life. My mother is still alive at 89. My father, a lifelong, heavy smoker, made it to 84, though I wouldn’t recommend his last years to anyone.

I’ve got a lot of positive risk factors: female, college education, healthy weight, exercise most days of the week, plant-heavy diet, no chronic health conditions plus good cholesterol and blood pressure numbers. (Don’t hate me. I’m a health writer. I take this stuff seriously.)

I started playing with the calculators. The easiest one was from the Social Security Administration. It’s based purely on averages, and all it wanted was my age and sex. It said an average woman my age — 63 — would live to 86.5. This was not terribly helpful since I know I’m not average.

I also tried calculations for a medical train wreck of an age peer. She was the height and weight of an average American woman — 5 foot 4 and 168.5 pounds. (That’s a BMI of 29. Thirty is considered obese.)

She smoked, but less than two packs a day, and had diabetes and family history of heart disease, barely exercised, ate a pretty typical American diet, was a heavy drinker, didn’t finish college, slept poorly, and was feeling more stressed out than me. I wasn’t completely heartless; she had a dog, used seat belts, was a good driver, and didn’t use recreational drugs.

Luckily, no one calculator wanted to know all these things.

I tried a calculator from what appeared to be a financial services company based in Australia. It said I would live to 91. My unhealthy alter ego would be dead at 63. Wait. We are 63 now. I was suspicious.

Next, I tried one from Northwestern Mutual. It said I would live to 102. Yikes! I’ll be out of money long before that, I thought. Will anyone I know still be alive?

The alter ego would be gone at 66.

I tried to ask Northwestern Mutual how they created their formula. They didn’t call back.

MetLife’s simple calculator said I’d live to 90, with a 25 percent chance of living beyond 97. My alter-ego would live to 76 or 80, depending on how bad I made her answers.

Other calculators wouldn’t let me play with more than one person’s numbers.

The Living to 100 calculator said I’d live to 97 and offered investment opportunities. That one asked fairly detailed questions about diet and stress. I admitted I had some, including stress about national events.

Then I found one designed at the University of Pennsylvania. I talked to Lyle Ungar, a professor of computer and information science, who helped create it. Any good calculator, he said, has to tell you about your odds of reaching a certain number.

That one gave my predicted life expectancy as 96. I had a 75 percent chance of making it to 88 and a 25 percent chance of surviving to 101.

Don’t discount 25 percent odds, Ungar said. Some pollsters said that Donald Trump had a 25 percent chance of being elected president.

Next came ePrognosis, designed to help doctors make medical decisions, such as whether it’s still worthwhile to do screening procedures like colonoscopies. This calculator is also available to regular internet users.

It asks more questions than others about daily activities, such as walking a few blocks or paying bills.

It expresses your odds in terms of how likely you are to die in up to 14 years, which feels a little more threatening than how likely you are to live.

The calculator said I had only a 3 percent to 6 percent chance of dying in five years, but the range rose to 19 percent to 24 percent by 14 years.

Last, I talked to Jay Olshansky on the recommendation of the American Federation for Aging research. A professor in the School of Public Health at the University of Illinois at Chicago, he helped develop an estimator that is now available to life insurance companies and will soon be used by financial services companies to help people plan for retirement.

The problem with most lifespan calculators, he said, is that they use an “additive” model. That means they start with the average expected lifespan and add or subtract years based on behaviors — which I saw as I used the Northwestern Mutual calculator that said I’d live to the unlikely 102. This doesn’t work well because behaviors or demographic factors tend to travel together, he said. “If you’re highly educated and physically active,” he said, “it’s hard to smoke.”

Normally, his calculator also uses photographs of your face, which Olshansky says is a “pretty good biomarker of risk factors” such as smoking, obesity, and sun damage. We didn’t have the photo option, so he walked me through the survey, which asked many of the same questions as the others, but weighted the answers differently.

The results: My expected lifespan was 87.4 years, and I’m likely to be healthy until I’m 82.3. My alter ego, who avoided some damning questions on this calculator, made it to 82, with only about a year of disability at the end.

I was a little disappointed with my numbers after the other calculators, but Olshansky said I should think of these results as the median lifespan for 100 women just like me.

Hopefully, I won’t know whether any of the calculators were right for at least a couple decades.