Welcome to Nonrival, the newsletter where readers make predictions about business, tech, and politics.
This email is the first set of scores for Season 2. It's a question from May about the GOP primary. I'll be scoring Bud Light forecasts from May soon, too. If you didn't make predictions on those two questions, no problem! Next Sunday will be a new question, so you can make a prediction then. Going forward, Sundays will typically alternate between weeks with new questions and weeks with scores from questions that have resolved. (More about how the scoring works here.)
Thanks for forecasting. Send feedback to newsletter@nonrival.pub.
In this issue
Scores: Will DeSantis have the highest polling among Republican presidential candidates, excluding Trump, by July 31?
Outcome: Yes (so higher forecasts score better)
Average forecast: 69%
Your forecast: Sign up to see your forecast here.
Your score: Sign up to see your forecast here.
Your total Season 2 points: Sign up to see your forecast here.
Is DeSantis still #2?
In May, Nonrival asked readers to predict the likelihood that DeSantis would have the highest polling average among GOP primary candidates, excluding Trump, at the end of July. Most of you thought that he would—and you were right. DeSantis has faded in the polls, but even now, post debate, he's number two.
But back in May, readers were divided over whether DeSantis would have staying power. Consider the rationales behind two forecasts, both of which predicted he'd still be in second place in August:
90%: DeSantis seems like the only politician willing to mount a serious challenge to Trump. I doubt he'll succeed, but unless something dramatic changes, I expect he'll be the only one to try.
70%: Not enough time for anyone else to surge
Which is it? Is DeSantis Trump's biggest competition? Or, is he the racer who goes out fast and inevitably gets caught by the rest of the pack?
Primary polling the year before the election is usually very predictive of who actually becomes the nominee. In 2019, Fivethirtyeight looked at 44 years of data and found that candidates with a polling average of 35% or more in the first half of the prior year went on to win the nomination 75% of the time. Sure enough, the "superforecasters" at Good Judgment Inc. currently put Trump's chances of winning the nomination at 72% (paywall).
But there was one big caveat to Fivethirtyeight's analysis: Name recognition matters. If everyone knows who you are and you poll badly, your chances of winning are remote. But if you're unknown and polling badly, you may still have a chance.
Fivethirtyeight found that adjusting a polling average for name recognition was even more predictive. The gist of their analysis was to divide polling average by a measure of name recognition. I've done something similar below, using a June poll from Marist that included “Never heard of them” as an option. I'm taking candidates' most recent polling average and dividing it by the share of Republicans who had an opinion on them in June.
For example, Nikki Haley is polling at 6% and 36% of Republicans and Republican leaners said they'd never heard of her. So her adjusted score is 6% / 64%. The less people had heard of a candidate, the more they're boosted by this measure.
This exercise comes with many, many caveats—including that I’m comparing Sept. 1 polling to mid-June name recognition, which is not ideal. Nonetheless, this back-of-the-envelope measure puts DeSantis neck and neck with Vivek Ramaswamy. 82% of Republicans and leaners had a view of DeSantis in June; most still did not know who Ramaswamy was.
For comparison, in Fivethirtyeight's historical analysis, a candidate with DeSantis's current numbers and high name recognition went on to win the nomination about 10% of the time. Candidates with low name recognition and Ramaswamy's numbers won about 14% of the time.
Judged by polling average, DeSantis remains in second place. But history suggests Ramaswamy might have as good a chance of winning, if not better.