Superdelegates are mathematically relevant when a candidate has 41.2 percent to 58.8 percent of elected delegates. Below that range, a candidate couldn’t win a first-ballot majority even with the votes of every superdelegate; above that range, the superdelegates’ help wouldn’t be necessary to clinch the nomination.

That’s still a fairly wide range, however. In theory, for example, a candidate could lose elected delegates 58 percent to 42 percent — equivalent2 to losing the average state by 16 percentage points — and still win the nomination through superdelegates.

My guess, especially given what we saw in 2008, is that superdelegates wouldn’t feel comfortable weighing in anywhere near that much on Clinton’s behalf. In the case where she’s won only 42 percent of elected delegates, she’ll have lost to Sanders all over the map, and any conceivable “electability” gains from nominating Clinton would be outweighed by alienating at least half of the Democratic base.

If it’s closer, however, superdelegates could make a difference. Suppose that Clinton wins 47.5 percent of elected delegates to Sanders’s 52.5 percent — equivalent to her losing the average state by 5 percentage points. According to our formula, Clinton would then need only about 64 percent of superdelegates to win the nomination, a figure that seems realistic.