Jolly Roger wrote:
SafeSpeed wrote:
The page says "Taking a simple case we saw that making existing accidents into impacts at just 30 mph would kill 42 times more car drivers than die at present. This is a way of illustrating that driver response is more than 42 times more important than pre accident speed in the real world." Since the lowest free travelling speeds in general use are in about or beyond 30mph, it is reasonable to compare the outcomes of idealised 30mph crashes with the population of real crashes.
But comparing the outcomes of idealised 30mph crashes with the population of real crashes only demonstrates the relationship between driver response and crash frequency, as the average speeds are the same in each case. It is valid to say that we would kill 42 more car drivers if driver response was reduced to zero, but this doesn't tell us anything about how pre-incident speed affects the outcome.
In the real world the wildest dreams of crash reduction through speed controls are about 50%. (Oliver Carsten). In the context of the gross estimates we're working with here any achievable real world reduction is close enough to zero that I've felt comfortable neglected it.
Suppose a non-wild estimate is 20%. One the one hand we have speed controls delivering a reduction from 1,000 to 800, and on the other we have driver response delivering a reduction from 42,000 to 1,000.
Jolly Roger wrote:
In fact, your own figures show the "driver response is 42 times more important than pre-incident speed" claim to be bogus: at an average pre-incident speed of 30mph we'd kill 42 times as many as we do at present; but at an average pre-incident speed of 12mph, we'd only kill about the same as we do at present. So we've achieved a 42-fold drop in deaths by reducing pre-incident speed by around two-thirds.
I think you're forgetting the purpose here - we're working on estimating the contribution to safety of driver response. We're doing that by looking at real world outcomes compared with estimated rates of death. You can't seriously be suggesting that we have the option of reducing free travelling speeds on our road network into the 12mph range - if we did we'd probably have many thousands dying in delayed ambulances!
You last sentence there is simply restating the Joksch equation - it does nothing to advance the debate.
Jolly Roger wrote:
SafeSpeed wrote:
Jolly Roger wrote:
There's a problem here. Let's say in a particular situation, I am just about able to come to a safe halt with a pre-incident speed of 20mph ( I'm not a very good driver.) A driver who is twice as skilled as I am ( with a driver response twice that of mine ), in exactly the same situation, would be able to come to a safe halt from a pre-incident speed of 840mph ( 20mph x 42 x( 2 - 1 ) ). This surely can't be right ?
That's not a valid sum. You're working on the "wrong side" of the speed / fatality risk equation.
I'm afraid you've completely lost me here.
The driver who was twice as skilled (at avoiding a fatality) would be twice as skilled at avoiding a fatality - we can't infer from that what the speed difference would be unless we predefined a particular speed. So, look at figure 2...
Jolly Roger wrote:
I'm assuming that the risk is directly proportional to the pre-incident speed, and inversely proportional to the driver response. If both these elements had similar effects, then doubling the pre-incident speed would have the same effect on the risk as halving the driver response. However, as you are stating that driver reponse is 42 times more important than pre-incident speed, I take it that halving the driver reponse has the same effect on the risk as multiplying the pre-incident speed by 42.
I think there's an interesting point lurking in there, but I can't quite put my finger on it. Obviously 42 times the speed would have a massive effect on risk, and that's absolutely not anything that I have tried to suggest. 42 times fewer fatalities do not arise from 1/42 of the speed. There's a fourth power in the equation.