[dsr_]

I apologize for conflating the rhythm method and the sympto-thermal method.

It's still a bad idea for most people.

All from the same Planned Parenthood site:

Twenty-four out of every 100 couples who use fertility awareness-based methods each year will have a pregnancy if they don't always use the method correctly or consistently.

whereas:

Vasectomy is the most effective birth control for men. It is nearly 100 percent effective.

and

Less than 1 out of 100 women will get pregnant each year if they use an IUD.

---

I want to control my systems. I have many choices: Puppet, Chef, cfengine, writing my own system... you're offering me a system that requires maintenance every single day, I can't pay someone else to do it for me, and if I screw up, it's 76% likely that I won't have unintended consequences. But there are a bunch of other systems on the market where I configure them once and they work for years without attention, and even systems where I just have to be picky when I'm conducting operations, not every single day.

[wayward]

The 24% failure rate cited by Planned Parenthood includes ALL couples who claimed to be using fertility awareness, from those who had taken classes and were using a modern method, to those who were just guessing. IIRC, over 80% of the couples were not properly trained in a modern method of fertility awareness. Not surprisingly, they had very high pregnancy rates. Someone did the math (can't find the article) and figured untrained users had about a 28% pregnancy rate while trained users had a 7% pregnancy rate. (80 * .28 + 20 * .07) = 23.8 The 7% pregnancy rate is comparable to the real-world use of the Pill.

As for your system analogy, if your automated system was a significant resource hog or ran the risk of corruption of data or crashing the system, would you use it? You focus on efficacy without factoring in side effects.