I often find it more informative: the proportion of YES positions to all positions. Such a line represents opinion distribution with no regard to the size of people's bets and how it was changing over time (thus not giving weight to gamblers).

If such a plot of positions over time appears in every binary market before the close date, then resolves YES.

Might be an overlap or a separate plot.

Resolves YES immediately if that happens, NO when the close date.

Example:

When the market opens the plot is empty (as both the amount of YES and Total positions are 0), if the first predictor bets YES, then the first point of the plot is 100%. After new people start making predictions, the line would converge to show a number which is like a Poll result.

Sometimes one whale changing his mind can move the market % more than 20%. With the bettors distribution plot the line after some predictors already joined the boat would move significantly only if there is radical shift in opinions.

Such a plot would allow to see whether the market dropped some time ago because something actually happened to change prediction or because somebody withdrew his huge position. Without any need to parse through trades.