Why don't more teams try the double steal?

Here's a link to the expected runs table for MLB based on every possible combination of base runners and number of outs.

http://www.nssl.noaa.gov/users/brooks/public_html/...

According to this chart, if you get runners at 1st and 2nd base you will score an average of 1.50 runs that inning. If you pull off a double steal which puts those runners at 2nd and 3rd base with nobody out, your average run expectancy jumps to 1.98. However if your runner is thrown out at 3rd base leaving you with only a runner at 2nd and one out your run expectancy drops to 0.69

So on average you gain 0.48 runs when you succeed, but lose 0.82 runs when you fail. So at first glance it sure doesn't seem like double steals are a great strategy - at least not when you have nobody out.

Of course the manager also has to consider things such as the speed of his runners, the skill of the opposing catcher, the score, the inning, etc to determine if the risk is worth the potential reward. However since the risk is 2x the reward its going to be a rare & unique situation where a smart manager will take the risk.

Most teams are playing for one run, and that usually scores with a well hit ball; so why risk a pretty sure run