Story points vs time estimates is not an either-or question. You can use both. My team at my job does. We do time estimating for each two-week sprint (Scrum) so we know what will fit, and have fuzzier point estimates in the backlog for longer-term planning by the product owners. Both levels of abstraction are appropriate in the right context.
To DanielBMarkham in a sibling comment:
> Or how about the ScrumMaster that empirically determines current velocity and then tells the team how many stories they can do for the upcoming sprint?
That's what we have, but is that bad? Isn't that how the points should be used to gauge predicted velocity? Of course it shouldn't be treated as a concrete inviolate set-in-stone prediction, but that methodology comes in pretty accurate for estimating.
To DanielBMarkham in a sibling comment:
> Or how about the ScrumMaster that empirically determines current velocity and then tells the team how many stories they can do for the upcoming sprint?
That's what we have, but is that bad? Isn't that how the points should be used to gauge predicted velocity? Of course it shouldn't be treated as a concrete inviolate set-in-stone prediction, but that methodology comes in pretty accurate for estimating.