I hope you don't take this the wrong way. My goal is to convince drivers that cyclists are not as bad as they think they are and that they should drive safer.
> You only see cases where they lose that bet, but it's just a numbers game.
I don't follow. Seeing both is pretty easy; they drive away. What I question is that the 2 minutes that one saves at most (likely closer to 15 seconds or so on average) is worth the extra risk to other road users. The asymmetry is the problem. You say cyclists regularly risk death, and sure, they do, though the risk is probably lower than you suspect, and drivers risk death too. A large fraction of cyclist deaths occur late at night with a drunk cyclist; remove those factors and your risk is a lot lower. For someone like me who goes out of my way to ride safely, my main (perhaps only) risk on the road is bad drivers, who are by large tolerated or encouraged by western society.
> As to speeding, we have actual data on this stuff. So it's easy to demonstrate the risks are very low directly from accident rates vs # of speeders.
The NTSB says speeding contributes to about 10,000 fatalities per year in the US:
https://www.ntsb.gov/news/events/Pages/2017-DCA15SS002.aspx
Doesn't sound low to me. Maybe it seems low when you compare against how frequently people speed, but that's a red herring as far as I'm concerned. It's unlikely you'll die while smoking any particular cigarette, but clearly the net effects of smoking at significant. I focus on the net effects. I don't think the time the speeders save justifies the life lost, even neglecting those disabled, etc. As I said, I think speeding provides mostly psychological benefit.
> As to personal data I only have ~45,000 cyclists which may be high or low. But it feels like a very large number.
You have seen about 45,000 individual cyclists on the road? I find that hard to believe. That's comparable to the total number of transportation cyclists in Austin. For comparison, I suspect I've seen at least several thousand individual cyclists on the road. If I see only two new cyclists on the road per day I'll see about 7000 after 10 years. Seems reasonable to me.