The ability to predict the future does not exist. That doesn’t stop us from trying — which brings me to the matter of the Democratic race for the party’s presidential nod. Voters repeatedly — and understandably! — tell pollsters their top goal is ensuring President Trump doesn’t get to serve a second term, and that they would prefer a candidate who can do that even if that means he or she doesn’t agree with them on a host of political matters.

This is a terrific sentiment — come November 2020. But it’s not something that voters should focus on while making selections in the Democratic caucuses and primaries, beginning with Iowa. Voters are not prognosticators, and selecting a candidate in the primaries that you are not enthusiastic about because of a possibly misplaced belief in that candidate’s electability is, to say the least, an unproven proposition.

In other words — don’t try this at home. Or a polling booth. Or a caucus gathering. Just don’t.

What we call “electability” is actually conventional wisdom. Conventional wisdom is right — till it isn’t. The best example? Donald Trump. Even as he led the polls in the Republican Party, pundits and politicos alike insisted Trump simply couldn’t emerge victorious in his party’s primary, never mind in November. Viewed as entertainment, Trump wasn’t seen as a serious political threat by either Democrats or Republicans. According to some reports, not even Trump appeared to believe he could pull it off.

You know what happened next.

All too often the concept of electability can be based on a tautology or wish fulfillment. Voters think Joe Biden is the most electable because … other voters think Joe Biden is the most electable? Sen. Bernie Sanders (I-Vt.) is the most likely to pull it off, because he’s leading a movement, which is something more than a campaign? Maybe they are right and maybe they are wrong, but either way, these sentiments are not exactly foolproof. People citing these things ignore inconvenient facts, hiding them under a carpet of inevitability. Things can change, but in neither case does this make for an ironclad case for electability.

The concept of electability also falls prey to what behavioral psychologists call recency bias, which is a fancy way of saying we believe what happened in the recent past is the most likely to occur again in the future. In investing — where this term is frequently used — this means we assume that when equities are heading down, they will continue to do so, and when they are on an upswing, such a change is semi-permanent. (This explains why so many people make the mistake of selling low and buying high.)

The 2020 political equivalent? Assuming a woman can’t win the presidency because Hillary Clinton did not. This is no small matter. A poll conducted by Ipsos last year found that 74 percent of voters said they would be “comfortable” with a female president, but only a third believed the same was true for their neighbors. Either a lot of people are lying to pollsters about their own beliefs and projecting outward, or they are more skeptical of their neighbors than warranted. Reminder: Clinton received almost 3 million more votes than Trump.

Finally, electability assumes there is a large base of voters who are looking for an excuse — any excuse — to abandon Trump for a reasonably acceptable Democrat. This could be true, but the more frightening converse might also be true: There are voters who don’t personally like Trump but could ultimately find a reason to vote for him. One reason Trump defied the odds and won the 2016 election was that some people decided party loyalty was more important than personal preference. It could happen again. And — here I will make a prediction — if the primaries result in a candidate that people think can beat Trump but few are actually excited about, it will make it much harder to defeat him. Enthusiasm matters.

No one can tell you who can best ensure Democrats will win the White House in November. Rather, let me suggest this: Vote for the person you want to see as our next president. Instead of attempting to predict the future, show faith in it instead. It might just surprise you.

Read more: