5 ms·
That is interesting. At the same time I wonder about the data source. I’d imagine these are pulled from insurance/police data. Anecdotally, I’ve been hit by se
by leesalminen 8y ago
That is interesting. At the same time I wonder about the data source. I’d imagine these are pulled from insurance/police data.
Anecdotally, I’ve been hit by senior citizens 3 times. All were very minor fender benders in parking lots or at traffic lights. In each case, I didn’t have the nerve to put them through insurance. We settled it personally and never reported to anyone.
Is it possible there are a lot of people like me that take pity on these older drivers and don’t report accidents?
Another common theme I see around me is when you’re in a corridor where most drivers are speeding by 5-10mph and there’s the one person in the left lane going 10mph under the limit. They’re usually a senior citizen. That creates bottlenecks and potentially dangerous situations behind the slow driver. They could be a contributing factor to accidents and wouldn’t be included in these statistics.
I suppose you could argue that those speeding are causing the accident because they’ve broken the law. It’s a fair argument.
- agumonkey 8y agoI was wondering if (instead of level 5 self driving) there could be a simple inter car protocol to communicate presence/distance/range so that car could warn or even slow down (smoothly) when signal is caught.
- Piskvorrr 8y agoThat sounds workable, and could be implemented in parallel - with two caveats: what of unconnected vehicles (not using this transponder), and what of malicious data (e.g. faking road jams for fun and profit)