You're saying that a 1/10 to 3/10 second delay will not have some effect on the statisitcs? That's, by definition, for any situation wherein the first clue that a driver has that something is about to happen that requires his/her action to avoid an accident is the lighting of this light, time stacked onto all other delays inherent in response time of a driver. So you don't think it would have even a 1 in 100,000 statistical effect (i.e., accidents that would not have happened with 0.2 seconds more advance warning)? That's all I was saying.
Possibly, but I still think there has to be some statistical effect - assuming the driver is alert enough to notice a light at all, 0.2 seconds advance warning is still 0.2 seconds off the reaction time - not measureable (re: accident vs. no accident) on a case-by-case basis, but still statistical..
Bill Putney (to reply by e-mail, replace the last letter of the alphabet in my address with "x")