[Talk-us-massachusetts] New MassGIS Trails data

[Greg Troxel]

Tom Parent <tomparent at gmail.com> writes:

> I applaud that they created this layer.  However, some things are wrong or
> out of date after a brief look.  I updated/removed (due to closure) some
> trails near me >6 months ago and they have not been updated in MassGIS
> data.  I'm curious about the frequency of updates.

It would not surprise me if it were manual and occasional.

> I'm also curious about what type of vetting they're doing on imported OSM
> data.  As everyone knows, there is a lot of bad data in OSM.  We (OSM
> community) are familiar with this because it's a part of crowdsourcing data
> from the masses with various levels of experience (and intent) and treat it
> accordingly.  However, when brought into an (authoritative) MassGIS
> dataset, presumably there is/should be a higher bar for things being
> vetted/correct(?)

There is a difference between being formally authoritative and being
accurate.  I have looked at previous MassGIS trails layers and they have
been much worse than OSM in my town.  Now, that is true because I have
made the data correct in OSM.  But still, the layer did not have
published accuracy standards, like 99% of trail segments in the map
actually exist, and 80% of segments that exist on the ground are in the
map, and that when there is a matching data segment, that 95% of points
on the map are within 10m of where the trail really is.  Just
articulating an example points out how far away the data is from
accurate.  (The USGS topo maps used to have at least some such
statements.)

Overall, I find OSM to be much more accurate than any other statewide
dataset for trails, by a very wide margin.
Counterexamples most welcome of course.