WTF?! The Zipper-Merge?
I don't normally write on driving. But I am a commuter and do it often enough that when I read stuff like this I tend to form a strong opinion.
"Zipper-merging" is merging right where/when the road ends as opposed to merging prior to it, supposedly at some point between the warning sign that the lane ends and when it actually does.
Being a commuter, as I read through this I was waiting for the point that explains what I'm missing. Because people on my commute drive this way all of the time, and to no avail. And sure enough, the explanation comes, quickly and almost dismissively not long before the end of the article. For this tactic to work, the people in the lane being merged into need to play their part and let them in. In other words, the technique in and of itself actually doesn't work at all.
That sounds like a harsh criticism of the technique, but it is true. On my commute, people don't give a damn if a lane is even closed for unexpected reasons like construction or an accident. They still work their hardest to make sure not another living soul gets in front of them. As a result, either someone in the merging line cuts off someone in the live lane causing a near total stop of traffic, or someone finally decides to be nice, but the person being cut off has been being cut off for so long that they are already at a complete stop, and once again traffic needs to halt entirely, in both lanes, just to let them in.
They say that is isn't maximally efficient without compliance... it doesn't work AT ALL without compliance. I live in a world where people try this without widespread adoption and I can tell you... it doesn't work.
My argument has been a simple one... traffic would flow better if people didn't follow too closely. On a highway, going 100km/h or more, you should be leaving several car lengths between you and the car ahead... you should leave AT LEAST 2 seconds worth of space. Let's assume people on the highway only want to go the speed limit. To figure out what the theoretical minimum safe distance people should leave apart is we need to know what fraction of an hour 2s is. An hour is 60minutes*60seconds, or 3600s. So, 2s is 1/1800h. If we're going 100km/h the minimum distance would be 100km/h*1/1800 which is 0.05556km or about 56m. FIFTY SIX METERS! Yes, at 100km/h you're travelling roughly 56m every 2s. Many humans don't even react fast enough to begin breaking in 2 seconds (though the odds of the car in front of you stopping on a dime at 100km/h are also pretty darn slim).
Anyway, the point is this... if we travelled a safe distance apart, there would be ample room for anyone to merge just about anywhere without any concern. The problem with that is that, just like the zipper merging approach people need to actually be on board with it for it to work. In fact, I would wager there are hundreds of strategies drivers could take which would combat congestion effectively if only people would adhere to them.
As far as strategies go, the zipper-merge is actually one of the worst. A handful of negligent drivers (jerks, out of towners, etc...) can easily negate the entire system. This is largely because the system intentionally advises against foresight, or fore-action. You aren't supposed to merge early, which means you aren't looking early for people who might plugging gaps early. So once it does happen there is no counter-measure. This system is actually only maximally efficient if both lanes are going around the same speed and people are alternating letting people out of the ending lane. And the only way that be possible 100% of the time anyway is if people are already travelling a safe distance apart. If both lanes were going the same speed, and there was not enough room in front of me for someone to merge, I would need to slow down to make the room and then both lanes are no longer going the same speed.
Basically, if either of those parameters is out of whack it drags the other down with it. If either lane comes to a complete halt the system reaches 0% efficiency. And you don't even need to be a jerk... you just need to not feel confident enough with the amount of room someone has left you to merge in to create this scenario. We are not all equally competent or confident drivers. The suggestion to simply leave a safer amount of room actually work relatively well as long as a substantial number of drivers obey. In my experience, it doesn't even need to be the majority. The tactic alleviates congestion even when all lanes are moving at different speeds and it takes many drivers to break the benefits.
The ONLY thing that the zipper-merge has going for it is that it plays to our strengths. We like to be ass holes. That is why virtually everyone follows too closely to begin with. By making zipper merging a publicly acknowledged and recommended approach, all you're really doing is trying to brainwash the people who either thought in advance or were by happenstance already in the "correct" lane to not see the move as being what it was otherwise and promoting them to to not stop, but also encourage it. In that respect, while it may actually be a fairly weak recommendation, it may end up being a stronger solution, simply on the basis that it is more likely to actually catch on.
"Zipper-merging" is merging right where/when the road ends as opposed to merging prior to it, supposedly at some point between the warning sign that the lane ends and when it actually does.
Being a commuter, as I read through this I was waiting for the point that explains what I'm missing. Because people on my commute drive this way all of the time, and to no avail. And sure enough, the explanation comes, quickly and almost dismissively not long before the end of the article. For this tactic to work, the people in the lane being merged into need to play their part and let them in. In other words, the technique in and of itself actually doesn't work at all.
That sounds like a harsh criticism of the technique, but it is true. On my commute, people don't give a damn if a lane is even closed for unexpected reasons like construction or an accident. They still work their hardest to make sure not another living soul gets in front of them. As a result, either someone in the merging line cuts off someone in the live lane causing a near total stop of traffic, or someone finally decides to be nice, but the person being cut off has been being cut off for so long that they are already at a complete stop, and once again traffic needs to halt entirely, in both lanes, just to let them in.
They say that is isn't maximally efficient without compliance... it doesn't work AT ALL without compliance. I live in a world where people try this without widespread adoption and I can tell you... it doesn't work.
My argument has been a simple one... traffic would flow better if people didn't follow too closely. On a highway, going 100km/h or more, you should be leaving several car lengths between you and the car ahead... you should leave AT LEAST 2 seconds worth of space. Let's assume people on the highway only want to go the speed limit. To figure out what the theoretical minimum safe distance people should leave apart is we need to know what fraction of an hour 2s is. An hour is 60minutes*60seconds, or 3600s. So, 2s is 1/1800h. If we're going 100km/h the minimum distance would be 100km/h*1/1800 which is 0.05556km or about 56m. FIFTY SIX METERS! Yes, at 100km/h you're travelling roughly 56m every 2s. Many humans don't even react fast enough to begin breaking in 2 seconds (though the odds of the car in front of you stopping on a dime at 100km/h are also pretty darn slim).
Anyway, the point is this... if we travelled a safe distance apart, there would be ample room for anyone to merge just about anywhere without any concern. The problem with that is that, just like the zipper merging approach people need to actually be on board with it for it to work. In fact, I would wager there are hundreds of strategies drivers could take which would combat congestion effectively if only people would adhere to them.
As far as strategies go, the zipper-merge is actually one of the worst. A handful of negligent drivers (jerks, out of towners, etc...) can easily negate the entire system. This is largely because the system intentionally advises against foresight, or fore-action. You aren't supposed to merge early, which means you aren't looking early for people who might plugging gaps early. So once it does happen there is no counter-measure. This system is actually only maximally efficient if both lanes are going around the same speed and people are alternating letting people out of the ending lane. And the only way that be possible 100% of the time anyway is if people are already travelling a safe distance apart. If both lanes were going the same speed, and there was not enough room in front of me for someone to merge, I would need to slow down to make the room and then both lanes are no longer going the same speed.
Basically, if either of those parameters is out of whack it drags the other down with it. If either lane comes to a complete halt the system reaches 0% efficiency. And you don't even need to be a jerk... you just need to not feel confident enough with the amount of room someone has left you to merge in to create this scenario. We are not all equally competent or confident drivers. The suggestion to simply leave a safer amount of room actually work relatively well as long as a substantial number of drivers obey. In my experience, it doesn't even need to be the majority. The tactic alleviates congestion even when all lanes are moving at different speeds and it takes many drivers to break the benefits.
The ONLY thing that the zipper-merge has going for it is that it plays to our strengths. We like to be ass holes. That is why virtually everyone follows too closely to begin with. By making zipper merging a publicly acknowledged and recommended approach, all you're really doing is trying to brainwash the people who either thought in advance or were by happenstance already in the "correct" lane to not see the move as being what it was otherwise and promoting them to to not stop, but also encourage it. In that respect, while it may actually be a fairly weak recommendation, it may end up being a stronger solution, simply on the basis that it is more likely to actually catch on.
Comments
Post a Comment