There are two dimensions to a politician. One is her politics and the other is her policies. The ratio of those two determines whether she is ineffectual or beloved, dangerous or harmless.
Policies are the wonky stuff. What sort of educational reforms do the most for inner-city kids? Which educational reforms are the hardest to sell to parents? What are the best ways to lower commute times? What sort of training programs give workers the most flexibility in a dynamic economy? When should we intervene in foreign affairs and when should we stand back and watch?
If you get policy wrong, people notice. Our troops get stuck in quagmires on foreign continents. Wages stagnate. Banks need bailing out. Drug addictions rates rise.
Get policies right and everyone is convinced that they've worked magic. Home prices go up even as homelessness goes down. Businesses are more likely to form and thrive. Suicide rates drop. 401(k) accounts rise. Unemployment is low. People even make more babies.
Politics can be judged in the short-term. Your performance in a debate. The way that audiences applaud when you make a point or laugh when you crack a joke. Your poll numbers in swing states.
If you get politics wrong, you aren't much of a politician. You might get hired to work in someone's cabinet or back office but you won't be elected.
If you get politics right, you win. You head off to the state capitol or city hall or to Congress. These politicians - the ones who get politics right - are the only ones we hear from. The only way you impact policy is to be good at politics.
So with that in mind, here are the four quadrants of politicians: the dangerous and beloved, the harmless and ineffectual. Imagine a scale of one to ten for politics (ten, you win any election you enter, 1, even your own mother hesitates to endorse you) and a scale of one to ten for policy (ten, your initiatives improve prosperity and quality of life even for people they are not intended for and one, your initiatives ruin the lives of people you hadn't even considered).
If your ratio of political skill to policy skill is 0.1 (1 / 10 = 0.1), you are ineffectual. You'll never get into office and no one much knows your name.
The trick for voters is to make sure that the ratio is close to 1. If you are bad at politics and policy (1 / 1 = 1), you are harmless, expounding on your political philosophy to some patient waitress who certainly hopes there is tip waiting at the end of your rant. If you are great at politics AND policy (10 / 10 = 1), you get into office and make magic happen.
Of course the problem is that there is no policy equivalent to polls. We can tell at a glance that Donald Trump is more electable than Bobby Jindal. It's harder to tell that Donald's policies are less credible than John Kasich's.
Still, this is not an impossible problem. It would be fascinating to have the equivalent to a poll number for policy: scores from policy experts for various policy proposals. Almost like a Consumer Report to evaluate policies.
Meanwhile, we'll have to plot the candidates on our own and just hope that this year we don't let the politics to policy ratio drift too far above one.