Wednesday, November 07, 2012

the uncomfortable situation room

During my final tour in Nevada I was present for a number of Air Wing debriefs led by a senior Commander who, about halfway through my tour, headed back to sea to lead an Air Wing. What was most impressive about his ‘precise’ and hilarious debriefs was his ability to distill to the younger pilots exactly how it was the world of aviation and strike execution actually worked. His most important lesson, and something I’d long ago learned but never quite got into words, was the idea of situational awareness (SA). SA is a military term that applies to the world around your little bubble or weapon system. As he so simply stated, often, “There are three types of SA: good SA, bad SA, and no SA.” He’d then go on to quiz most of the new pilots, and the new(er) strike leads, throughout the debrief – Where are you? What are you thinking? What do you know? Where, exactly, is your SA needle at this point? The kid would look up at the dry-erase board, peer down at his kneeboard, look up, and mumble something like, “Well, I thought…”, and then proceed to ramble just long enough to fully realize that his cockpit was stuffed with bad SA. Here’s a quick rendition of the gas gauge the Commander would draw on the board (he refused to use the required debrief PowerPoint):
His training point was this: If you are going to base your decisions on what someone has told you, or data that someone has provided, you better know where you are on the SA meter. Measured against the ground-truth you are gold if your SA is good. Bad SA will kill you – you think you know what’s going on, you trust what’s been said, and you then make the appropriately bad decision. You are far better off to have No SA then to have bad SA – if there’s no SA then at least you are aware of having nothing, and everything you do is then driven by keeping focus on the develop the situation; at least you aren’t sitting around fat, dumb, and happy with a big pile of bad data.
I bring this up as the background to my only post about the election. The disservice done by the GOP/Conservative bloggers and political commentators would, if I were one of the followers, make me homicidal. The final three weeks of the campaign was full of endless diatribes from the right against pollsters in general, and Nate Silver in particular.  Even if you hate the NYTimes, don’t like Silver’s methodology (or his political leanings), attacking the messenger was such a reach that I think it fully dysfunctional. Silver has explained his methods and they make sense to anyone who listens. As an example, if there are 12 polls out of Ohio that show the President leading by between .5 and 1.5 points, while one outlier (see how I use words?) show Romney leading by 6 points, then he assesses the polls and comes out with something that balances out the differences. But, even his methods aren’t important and you don’t need a higher math degree to see that 12 v. 1 means the 1 is most likely out of whack. Numbers don’t lie, but if you willfully choose to ignore basic math to the extent that you are misleading your own ‘party’ then you shouldn’t be allowed to ‘commentate’.
Back to SA. As the talking heads spent weeks trying to get the voters out they decided to attack math, disregard some stable poll reporting, and simply chose to tell the posse that the (wildly vague and incorrect) Rasmussen poll was right - Mitt Romney was 5% up in Ohio! They told people that he was leading the national polls (which he never was), and that Romney would most likely take 300+ of the Electoral College votes. To the followers, that’s bad SA – believing you know what’s going on when, in fact, you have shitty data. You know what’s better? Nothing, no SA. At least the minions can then decide for themselves. You know what’s even better? Good SA. Here’s how I would use the math:
“All right folks. Here are the numbers: we are losing by about 2 point in Ohio and Colorado. We are down less than a point in Florida, a point in Virginia, and 3 points in New Hampshire. These are the facts. In order to win we need turnout – thousands of voters to get out and make up the difference. Believe the numbers.”
As opposed to Peggy Noonan saying something like, “No worries folks. I did the point-and-click mappy thing at and Mitt should easily win this election. Rasmussen shows our guy winning. Also, I think Nate Silver is effeminate.” You know what this might lead to? People not showing up to vote because it’s in the bag. Well done, Peggy. I thank you and your ilk.
Hey, Fox News, where’s your needle pointing?

No comments: