Re: Contiguous US 14ers
Posted: Mon Aug 17, 2020 10:46 am
Great explanation, Scott. I see the reasoning in only wanting to include peaks you're "sure" about with clean prominence (peakbagger). However, like you said, a pursuer of this list would very, very likely be missing out on some peaks that are actually ranked. Using the interpolated prominence is statistically the more accurate way to calculate the prominence of peaks. And again, as you mentioned, if someone would like to be fairly certain they've climbed all the ranked peaks, they should use the "optimistic prominence" (or over-estimating) list. That way, they'll likely have climbed too many peaks, but will VERY likely have climbed all the ranked ones (assuming no map contour errors, etc). Just because the majority of sites/people use one method doesn't make another method incorrect, however, it does urge scrutiny when evaluating the less common method. And the less common method in this case does indeed seem to be less statistically accurate for estimating prominence values of peaks (leaving out too many that would actually be ranked).Scott P wrote: ↑Sun Aug 16, 2020 10:35 pmI'll use Peakbagger and Mt. Muir as an example.
Saddle elevations usually aren't marked on topo maps (but in a few cases they are). Most saddle elevations also don't fall exactly on a contour line either.
What everyone (I know of) world-wide, except for Peak Bagger does is to use the average of the two contours; the one above and the one below the saddle. This is known as prominence or interpolated prominence. Statistically, this is the most accurate way to do things.
Peakbagger doesn't do this. They use only the contour above the saddle. This is known as clean prominence since there is no averaging involved. Peak bagger still does list the avearge (interpolated) prominece, but they use only the highest saddle contour (and/or highest peak contour for peaks that don't have a listed elevation) for their lists.
See here for the Peakbagger page on Mt Muir and the box I outlined:
Only Peakbagger uses clean prominence on their lists, everyone else (I know of) world wide uses the average (which is usually known as just prominence or interpolated prominence), which is also in the box and which is the last number in that box. This is what everyone else uses. See listsofjohn for example:
Average is the most accurate way to do it, even if it isn't perfect.
What the topo map of Mt Muir actually says is that Mt Muir rises somewhere between 298 and 364 feet above the saddle (it doesn't give an exact elevation of the saddle; only contour lines).
That means there is a 2.9% chance that Mt Muir has less than 300 feet prominence and a 97.1% chance that Mt Muir has more than 300 feet prominence and thus is a ranked peak.
Even though the there is only a 2.9% chance that the peak has less than 300 feet prominece, that is what Peakbagger uses because clean prominence would say that a peak has to have a 0% chance of not being ranked to be included. This is why I consider using clean promince to be flawed. A lot of peaks that really do have 300' prominece are eliminated.
Using prominence (or interpolated prominence) is much better to use, but it isn't perfect.
Let's say a peak has 270-310 feet of prominence using the countour lines. Using prominence/interpolated prominence, the peak would have 290 feet of prominence. The prominence would be listed as 290 feet, but since there is still a 25% chance (in this example) that the peak has over 300 feet of prominence, the peak would be given a soft rank. If you see a soft ranked peak in a guidebook (such as Roach's) or on an online list, this is what it means. Any soft ranked peak has less than a 50% chance of having more than 300 feet of prominence, but still has some chance that it would be ranked had an exact elevation been available.
Hopefully that made sense.
No, it's just that this isn't a hill I need to die on. I gave you the reason, you want to argue it. You don't like it, take it up with Greg (again). That's the beauty of free will. I can work any list I want, and I don't have to answer to anyone.