> A year or two after a new, more intrusive format became ubiquitous, its performance would fall back to the old format's
When Sweden changed driving side, that is exactly what happened.
"The relatively smooth changeover saw a reduction in the number of accidents. On the day of the change, only 157 minor accidents were reported, of which only 32 involved personal injuries, with only a handful serious. On the Monday following Dagen H, there were 125 reported traffic accidents, compared to a range of 130 to 198 for previous Mondays, none of them fatal. Experts suggested that changing to driving on the right reduced accidents while overtaking, as people already drove left-hand drive vehicles, thereby having a better view of the road ahead; additionally, the change made a marked surge in perceived risk that exceeded the target level and thus was followed by very cautious behavior that caused a major decrease in road fatalities. Indeed, fatal car-to-car and car-to-pedestrian accidents dropped sharply as a result, and the number of motor insurance claims went down by 40%.
These initial improvements did not last, however. The number of motor insurance claims returned to 'normal' over the next six weeks and, by 1969, the accident rates were back to the levels seen before the change." https://en.wikipedia.org/wiki/Dagen_H
One only has to look at huge crowded Asian cities, where signage is present but effectively ignored, to see that removing signage doesn't really help. The accident rates are just as high if not more.
Certainly there's plenty of places I've been a passenger in Asia and India where I would never try to drive -- but I think you're conflating 'remove signage' with 'keep signage (but assume all of it is ignored by everyone, all the time)'.
Keep in mind that's traffic death, not accidents. Accidents are dramatic for everyone who's involved, and that is quite a lot of people: first and foremost the victim (with which exact damage? Anything permanent? Not taken into account by these statistics). Also, the suspect, the insurance company, innocent bystanders who are held up and need to be rerouted, and ultimately the taxpayer.
This list is interesting bu difficult to interpret as cultural context is completely different.
Most of the countries with huge crowded cities where road rules are basically ignored also have a huge amount of people scrapping by and who could just disappear the next day for any other reason (hygiene issues, hunger, criminality, social status)
Looking at the list of countries with 15+ deaths by 100 000 people, IMO the price of a random single life is perceived as way lower than in Switzerland for instance. People in that context won’t de driving the same way as they would in a different country, or their government wouldn’t let people ignore circulations rules in the first place.
It can't just be poor, it must be deliberately misleading signage, "smart", dynamically adaptive, evolving. And more than just signs: a large system of automated aerial observation drones, satellites, the whole sky. All networked together. The branding is simple though, just call it NetSky, or SkyNet.
Okay, fine. Then we ramp it up by employing hundreds of clowns to hide in the bushes on the side of the road and jump out when the driver least expects it.
This actually holds for roads. Large wide roads are driven at higher speeds and more fatal for pedestrians. The technique now is to make the road narrow enough to cause traffic to slow down.
That seems similar to the economics concept of Sharpe ratio (Return divide by Risk). Behaviours will change till the time all opportunities in life offer equal risk-adjusted return.
And when people get used to that, start each week with a new drive side bit polled from a cryptographically sound RNG. If that still isn't enough: quantum uncertainty, everybody does not know wether or not they are on the right side until they meet oncoming traffic.
When Sweden changed driving side, that is exactly what happened.
"The relatively smooth changeover saw a reduction in the number of accidents. On the day of the change, only 157 minor accidents were reported, of which only 32 involved personal injuries, with only a handful serious. On the Monday following Dagen H, there were 125 reported traffic accidents, compared to a range of 130 to 198 for previous Mondays, none of them fatal. Experts suggested that changing to driving on the right reduced accidents while overtaking, as people already drove left-hand drive vehicles, thereby having a better view of the road ahead; additionally, the change made a marked surge in perceived risk that exceeded the target level and thus was followed by very cautious behavior that caused a major decrease in road fatalities. Indeed, fatal car-to-car and car-to-pedestrian accidents dropped sharply as a result, and the number of motor insurance claims went down by 40%.
These initial improvements did not last, however. The number of motor insurance claims returned to 'normal' over the next six weeks and, by 1969, the accident rates were back to the levels seen before the change." https://en.wikipedia.org/wiki/Dagen_H