on the interpretation of losses, mt special election edition

bridge_players_do_it_with_finesse_bridge_humor_invitation-r27c2f5f2d3b945458051f406869682ce_zk9gs_530

Last night, Republican Greg Gianforte won a special election for Montana’s sole Congressional seat. How do we interpret this event? Here’s how the NYT approaches the question:

Voters here shrugged off the episode and handed Republicans a convincing victory. Mr. Gianforte took slightly more than 50 percent of the vote to about 44 percent for Mr. Quist. (President Trump won Montana by about 20 percentage points.) Mr. Gianforte’s success underscored the limitations of the Democrats’ strategy of highlighting the House’s health insurance overhaul and relying on liberal anger toward the president, at least in red-leaning states.

I believe this interpretation is incredibly misleading and reflects a larger problem with how we make sense of binary outcomes in the presence of more information.

As the NYT notes, in 2016, Trump carried Montana 56-36. The House race in 2016 was a similar 56-40. Gianforte here won 50-44. That’s a 10 point shift. In a special election. In Montana. And with something like 70% of votes being cast before the assault that brought national attention. Turnout was about as high in the special election as in the 2014 general. That’s wild. Yes, Gianforte’s awful, and yes that he will be a congressman is depressing. But framing this outcome as having “underscored the limitation of the Democrats’ strategy” or as a big loss for Democrats strikes me as absurd. If you are a GOP rep who won by say 10 in 2016 (55-45), this result should terrify you. And if you’re a Democrat looking at an even marginally competitive district, this should embolden you.

That’s most of what I wanted to say; the rest of this post is an aside about learning, probabilities, continuous information, and contract bridge.

Let me step back a second and talk more generally about inferring the correctness of strategy from outcomes by way of a discussion of contract bridge. Bridge is a bit of a dying game, but I grew up with it and it deeply shaped my understanding of how to learn. Bridge is a hard game to learn in two senses. First, the initial rules are quite complicated. Bridge has many more moving parts and interlocking complications than, say, hearts or chess, or even more modern and seemingly complex games like Settlers of Catan. But you can still learn the rules of bridge in a couple hours. The second, harder bit is actually getting good at bridge. The issue is that bridge is a game of probabilities. In each deal of bridge, you are faced with numerous small decisions. Often, the difference between the right decision and the wrong decision is small – 55% to 45% or even narrower. Good players make the higher probability decision every single time. But in so doing, they appear to get it “wrong” (a bad outcome) almost as often as they get it “right” (a good outcome). Becoming a good bridge player requires learning how to not to learn from a bad decision that happens to work out, or a good decision that doesn’t. You have to actually analyze the whole hand, learn how to calculate probabilities, and learn how to infer the correct strategy in spite of very noisy signals. Novices (or me at age 10) would often point to the outcome of a hand as a defense of a particular choice: Yes, I took the finesse the wrong way, but look, it worked! Experts never do.

The New York Times ought to be an expert at reporting on political strategy, and about understanding how to infer from outcomes, but here they sound like a novice. “Look, the Democrats lost, therefore their strategy must not be working.” But is that what the data show? It’s always hard to say with any single case, and we should be cautious of learning the wrong lessons from a loss. But what I see is a 10 point swing towards the Democrats in an institutional setting that’s usually very unfavorable (special elections with typically low turnout). Focusing on the binary outcome (Gianforte won) ignores the continuous information (the shift in the vote share) – much as most analysis of the 2016 Presidential election radically overinterpreted what we could learn from a very small shift in votes. Should we have expected more from the Democrats right now? Political science suggests that big shifts are rare; partisanship is a helluva drug. It would be interesting to know if the MT result involved many voters shifting their partisan affiliation (unlikely), or if it involved differential turnout (many voters stay home in special elections, but this time most Dems turned out). So, this result may showcase an upper-bound on how well Democrats can do in Montana right now, but all that tells us that is that Montana is a very Republican state, and Republicans vote for Republican candidates almost no matter what. It tells us very little about the specific strategy Democrats employed. In bridge parlance, just because the queen was offside doesn’t mean you were wrong to take the finesse, it means the contract was always unmakeable.

Author: Dan Hirschman

I am a sociologist interested in the use of numbers in organizations, markets, and policy. For more info, see here.

2 thoughts on “on the interpretation of losses, mt special election edition”

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s