@Drew i don't trust that counter. The totem has never changed when i've ridden over the loops with my urban arrow, and the growth in cargobike ridership since 2015 has been explosive

@bikepedantic OK, but you'd still expect the data to reflect an increase in non-cargo bike usage too.

@bikepedantic I bet remote work would explain it. I'd like to see the mode split data. Maybe traffic overall has gone down but bikes' share has increased.

@Drew @bikepedantic if you want a better sense, look at Wednesday numbers; we moved from a flat weekly pattern to Monday/Friday being ~60% or less, due to a ton of the businesses in the area moving to 3-day schedules for required in office time.

@Drew @bikepedantic I did a little more detailed look at the data before. From 2015-Covid there was a steady seasonally adjusted increase. The numbers cratered with Covid and have been pretty steadily increasing again, though not to 2019-early 2020 peak.

I changed up how I save the data, so that graph isn't something I can regenerate without fixing up some code, unfortunately. Definitely on the list to fix up after the election.

@Drew @bikepedantic The counter is in a kinda weird spot in terms of what it is counting. Useful since it's a long data set, but not the best at understanding the overall biking in the city. BlueBikes data is probably a closer analog.

@BostonBikeData there's a real data quality issue with all loop counters. We know they don't detect scooters and other new micromobility devices. If they're systematically not detecting cargo bikes like mine either, that's a massive chunk of ridership on the road now that wasn't on the road a few years ago, let's say pre 2017ish @Drew

@bikepedantic @BostonBikeData @Drew does your bike trigger Somerville's "stop here for green" spots?

@vathpela Never tried... i don't encounter any on my most-common Somerville trips. WIll keep an eye out and give a try though! @BostonBikeData @Drew

@bikepedantic @BostonBikeData @Drew Aren't there validation counts done? Here (Madison, WI) city staff have done manual counts at the counter location and then communicated that with Eco Counter to get an adjustment factor. RE cargo bikes: Yes, the counters do appear to have an issue with cargo bikes and tandems.

@haraldkliems Perhaps they do them, but i don't know that adjustment factors have ever been published here, and i'm assuming the released data is raw, not adjusted in any way. @BostonBikeData @Drew

@bikepedantic @haraldkliems @Drew Not sure about exactly what they do with this counter, but the city is testing intersection camera counters and those counts have severely under-counted bikes to the point they won't publish the data. I assume that means they are happy enough with the totem accuracy.

@bikepedantic @haraldkliems @BostonBikeData from the data portal Cambridge links through, yes it seems like raw unadjusted data

@Drew @bikepedantic @BostonBikeData I remember a conversation with someone from Eco Counter that made it sound like they provided at least some data cleaning before serving data to their clients. But I don't have any specifics, and this was in the context of one of our counters temporarily malfunctioning and inflating counts.

@haraldkliems They do - they scrub out spikes due to electromag interference, and will apply adjustment factors if the customer provides them. @Drew @BostonBikeData

@itamarst HI!