Comments

1. Doug

   October 28, 2014

   I must say I found the conclusions of the infographic to be somewhat lacking. What they describe as the most likely accident conditions seem more like a description of the description of the most common cyclist and the conditions most likely to be ridden in. Naturally they account for the most accidents.

   For example, they cite that 25% of accidents occur in wet conditions. Do trips in the rain in the rain account for 25% of total trips? I don’t have any idea, but I doubt it. There might be a higher rate of accidents in those conditions, which, if true, would actually be helpful information.

   1. Paul

      October 29, 2014

      Author of the infographic here, and you make a very fair point, Doug! I put a brief disclaimer about precisely that caveat in the summary section, but this is still good feedback and perhaps that point was not well emphasized. We knew at the onset we were only capturing crashes that folks record on Bikewise, which is a sample population but certainly not necessarily representative.

      I’m continuing to work with Bikewise to see if we can incorporate other data sources beyond these anecdotal records to make the statistics more robust and do something beyond simple percentages. In the case you cite, 25% of recorded incidents did occur in wet conditions, but I agree that these could be just indicative of when Seattlites ride (e.g., rarely in flood conditions) and not necessarily what the commonalities in crashes are. Hopefully with more data we can come closer to that, but this I feel is a good start and is sparking an important dialog – what are the primary causes of crashes, where are they located, and what can we say about the data to help avert them?

      Thanks again for the feedback!

2. Alkibkr

   October 29, 2014

   It appears from the infographic that the death of cyclist Lance David on 1 May 2013 on E. Marginal Way near South Hanford Street was not reported on Bikewise. I think for fatalities at least, you might be able to supplement the Bikewise data from other sources, like the “Deaths” tag on this blog. Your map displays very well that the reported West Seattle incidents are very concentrated near the watershed from the Alaska Junction down to the lower WS bridge and into SODO. That is why improvements to this corridor including the 5 way intersection at Spokane/Delridge/Chelan streets need to be funded if we are ever to get more people comfortable riding the short distance from West Seattle to SODO/Downtown. I agree, too bad that we don’t have a more reliable source for this data since Bikewise reporting is voluntary and not too well promoted. I for one thought it was no longer functional since it was not supported for a while. Fortunately I have not had any serious accidents in a while but have been nearly been hit three times by oncoming cyclists traveling too fast and cutting blind corners in the Harbor Island crosswalks and the Spokane/East Marginal Way sidewalk corner. And I don’t bike this route on a daily basis, more like once a week.

3. Alkibkr

   October 29, 2014

   The bike counters on the Fremont and West Seattle bridges collect base data about cycling volumes by day which SDOT can match up with weather patterns.

   http://www.seattle.gov/transportation/bikecounter_fremont.htm

   1. Paul

      October 29, 2014

      Paul the artist here again – great thoughts Alkibkr! I didn’t think to incorporate the bike counter data, but I agree that that could be a really valuable resource. Leveraging those data for the type of products Bikewise and I are trying to develop might be challenging, but it could definitely help normalize the anecdotal data. Comparing what type of weather conditions crashes occur in versus the meteorology of when the bulk of riders are crossing the Fremont bridge and Spokane Street could be highly informative – thanks for the excellent feedback!

4. Gary

   October 30, 2014

   One other comment on the raw data:
   I would have expected the bicyclists who ride the most to also have the most accidents. And that’s based on the observation that even if your chance of having a single accident decrease with added skill, if you roll the dice enough times you will have an accident. So I’m wondering if you normalized the data for time/distance ridden whether the new cyclists now stand out hugely as a high risk group.

   I thought I read somewhere that motorcyclists after 3months and before 12 months hve a high accident rate. It was attributed to that they felt “comfortable” but were in fact not “skilled” and hence relaxed their guard a bit and thus had more accidents in that first year.