Probability is in part a tool for capturing levels of certainty and uncertainty regarding future events.
I’ve seen many exaggerations of the certainty of Hillary becoming the Democratic nominee for president. In several comments on dailykos, there have been estimates placing her odds at 1000-to-1 or greater. Jeralyn Merritt of talkleft.com recently estimated Hillary’s chances at 699,999/700,000. Even diaries friendly to Bernie Sanders on this site have estimated Bernie’s chances of winning as equivalent to rolling a pair of threes with two dice on a single roll — 1/36.
I’m writing this diary because I believe these estimates understate Bernie’s chances to win the Democratic nomination. I believe the chances are closer to rolling doubles (some pair or other) with two die on a single roll — 1/6. If I’m right, Kos made a grave error in calling the election over. Given these odds, the site should still be helping its community of Democratic voters choose the best option among candidates with unequal but still reasonably good chances of winning.
Similarly, given these odds, many Hillary supporters on this site are also mistaken. Everyone agrees Bernie is not currently the front-runner, but many Clinton supporters have been misdescribing the relevant probabilities, potentially with the intent of discouraging Sanders supporters.
Currently, multiple betting markets have Bernie’s chances at about 1/10 (which is likely to shift up after his Wisconsin win). I would argue Sanders’s chances are bit better on the following basis: The election results since 3/15 are indistinguishable from Bernie slowly moving into the lead and building to large margins (circa 65-35) on the west coast — OR (5/17) and CA (6/7) — that are big enough to put him into the pledged delegate lead on June 7. The election results over the past three weeks look like what you’d expect if one presidential candidate who was down was gradually moving into a front-runner position.
To flesh out this scenario more — which I consider Bernie’s most plausible path to the nomination— Bernie might win most states other than CA and OR with decent-to-modest margins so that he nets circa +60 delegates in non-western states. Then he might net circa +160 delegates in OR and CA’s races. Given Bernie’s absolutely dominating record on the west coast in several recent caucuses (HI, WA, UT, ID), he has a reasonable chance to build momentum in the coming weeks and months in order to translate his general west coast strength into slightly less dominating wins in OR and CA’s primaries that nevertheless put him into the pledged delegate lead. Call this scenario 1.
The fact that recent election results are entirely compatible with scenario 1 doesn’t mean Bernie is the favorite or that his chances are even 1/2. The post-3/15 election results are equally compatible with Bernie failing to improve his standings in the polls and getting essentially knocked out in the northeast later this month (a second scenario) or remaining on track in the northeast to become the nominee but failing to build a large enough lead on the west coast primaries to overtake Hillary (a third scenario).
On this way of splitting up the probability space, Bernie’s chances would be 1/3 given the information we’ve received from states voting after 3/15. But I think Bernie’s chances are smaller, because the first scenario would likely be accompanied by more movement in national polls, which still have Hillary up by double digits, and more movement in several state polls with upcoming elections.
But let’s not exaggerate the situation! The national polls are imperfect indicators of upcoming election results and there’s been a lot of state poll movement in the weeks prior to a state’s election, most of it towards Bernie. And Bernie’s outperformed the polling averages on many occasions, including in WI. A reasonable probability assignment is 1/2 that the current polling doesn’t indicate a higher probability for scenarios 2 & 3 over scenario 1. For these reasons, 1/6 (i.e., 1/2 x 1/3) is an entirely reasonable if rough estimate of Bernie’s chances of winning the Democratic nomination for president.
Given these odds, it does not make sense to call the race “over,” unless one also dismisses the probability of rolling doubles on a single roll with two die. Given these odds, Hillary supporters who exaggerate the relevant probabilities aren’t hard-headed realists who are superior at math, they’re advocates cloaking their advocacy in fake, objective-sounding mathematical-ish analysis.
Given these odds, it’s entirely rational for Bernie supporters not to focus on Donald Trump or Ted Cruz (even ignoring that the GOP candidates are already receiving too much attention). And it’s entirely rational for Bernie supporters to keep advocating for their preferred candidate, including emphasizing Bernie Sanders’ strengths and the alternative candidate’s weaknesses. A community site dedicated to electing more and better Democrats can only expect its members to behave rationally in light of reasonably assigned probabilities to future events.
