Happy New Year, and an invidious comparison
6 views
Skip to first unread message
Rauschpfeife
Jan 1, 2026, 5:22:34 PM (5 days ago) Jan 1
to PB...@googlegroups.com
Annum novum faustum felicem, as the ancient Romans used to say, to one and all.
For comparison purposes, I did some poking around for New York City data comparable to the dataset Maine DOT makes available, which Myles called our attention to. It does exist:
Navigating the site is (as usual with government agencies) frustrating. But if you want to while away an idle hour...
The database, I must say, compares very unfavorably in its organization with the Maine version. There's no field for whether a cyclist or walker was involved in a crash, for example, and the vehicle type seems to have been a free-form entry field, which gave the cops wide scope for whimsical improvisations.
Rauschpfeife
Jan 3, 2026, 2:34:04 PM (3 days ago) Jan 3
to PB...@googlegroups.com
Just to while away an idle hour, and for comparison purposes, I did a mapping version of the same crash database I mentioned earlier, in New York. Same code, basically, as for the earlier Portland map, tweaked a bit for differences in database organization.
As I mentioned earlier, the NYC database is pretty much dogshit, as compared with the Maine version. Many crash records don't even have a lat/long at all, and so aren't mappable. There's no indication of whether a walker or cyclist was *involved* in a crash -- though there's a field for cyclists and walkers injured or killed, that's not the same thing, of course. Still, what there is makes for interesting browsing.
There are so many points in the database (it goes back 10 years and more) that the markers take a while to load, so be patient.
John Brooking
Jan 3, 2026, 2:40:47 PM (3 days ago) Jan 3
to PB...@googlegroups.com
Lack of accurate and precise crash information makes it very difficult to suggest and evaluate counter-measures in the way that happens for, say, airplane crashes. This makes for an environment where everyone is free to suggest what they think will fix the problem, without being able to identify exactly how that suggestion may have prevented any given crash. We are left with theories and beliefs, based at best on studies that rely on statistical techniques to try to surface trends in data sets that often have insufficient data points, or insufficient detail in those data points, to see obvious trends.
John Brooking
Cyclist, Cycling Educator, Technologist
--
You received this message because you are subscribed to the Google Groups "Portland Bicycle and Pedestrian Advisory Committee" group.
To unsubscribe from this group and stop receiving emails from it, send an email to PBPAC+un...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/PBPAC/CAERgV98cAxPYSnsz62O0SgiDsxSj2S3JPpfs%2BOAVOZsXqRepbQ%40mail.gmail.com.
Rauschpfeife
Jan 3, 2026, 2:57:53 PM (3 days ago) Jan 3
to PB...@googlegroups.com
Agreed. Of course the dearth of information is largely due to indifference and sloppy practice by the the original reporting source, namely the cop[s] who went to the scene. Also, the databases often don't seem to have been designed with much thought given to the kind of questions we might like answered. It's interesting to me that the NYC database has fields (six of them!) for vehicle type, and the Maine database has none.
To view this discussion visit https://groups.google.com/d/msgid/PBPAC/CAJg%3D2oxsHD20C1FxfL5botGG06t0TTytdZFgaiF5vh3fPKR65A%40mail.gmail.com.
Reply all
Reply to author
Forward
0 new messages