The Highway Lane Next to Yours Isn’t Really Moving Any Faster | Hacker News Reader

The Highway Lane Next to Yours Isn’t Really Moving Any Faster

citylab.com

13 pointsgkuan20 comments

> And, for sure, sometimes the lane next to you is actually moving faster. But the point is that these cases are happening far less often than you think, and they largely fall outside our powers of detection.

Maybe I am falling into the illusion they speak of, but I am very suspicious of their claims, especially because of the quoted point above.

I realize that anecdotes aren't data, but I makes me think of two things I've noticed...

1) I have a family member with anxiety problems who has trouble just getting into traffic. She hates changing lanes because she perceives it as extra stressful, so she just stays put, content to follow the car ahead of her. While she always eventually reaches her destination, there are times when I've beaten her to a location by 10 minutes or more, all because I'm willing to change lanes.

2) When I am in traffic, I look at the car ahead (of course) as well as the farthest cars I'm able to see without obstruction. When I see the lane 'breaking up' several cars ahead and my own lane staying still, I tend to change lanes to take advantage of the temporary speed advantage I'll have. I can't count the times this has worked very well. The main reason is that people tend to get lazy in bumper-to-bumper traffic, and when cars start flowing there is inevitably one who is looking at his phone or just slow to react. By timing it so that I can overtake these slow-goers, I tend to make quite a bit of progress.

Maybe this study is focusing on people who only change lanes in response to the movement in their immediate vicinity and not up ahead? Either way, it seems like a shortsighted conclusion.

It being "outside our powers of detection" is pretty easily falsifiable too: Just reference a "fixed point" in the other lane (I tend to pick a large truck that's unlikely to lane change, easy to spot at a distance, and driving even more conservatively than I am.)

I tend to drive conservatively. Not out of anxiety, but out of "logical caution" and being in no hurry. Like your family member, though, I frequently stick to a lane. When I think another lane is going faster than mine, that truck tends to disappear into the distance in front of me.

If anything, I underestimate how much faster, frequently wondering if I won't catch up and spot it again as the lanes take turns alternating between fast and slow over the short term.

I think of it like this: traffic lanes are like slinkies side-by-side with each other. Each slinky expands and contracts roughly equally, but not at the same time. If you can hop from lane-to-lane such that you're maximizing your time in the "expansion zone" before it contracts to a standstill then you'll absolutely beat someone who never changes a lane.

More practically: don't change lanes from a "stuck" into a "moving" lane if the moving lane has been moving at a steady rate for more than 30 seconds, because it's probably about to contract and be stuck. If you're in a moving lane, jump ship as soon as it seems like it's beginning to get stuck. This is 1/3 art, 1/3 science, 1/3 likely placebo.

I doubt this has much to do with traffic. It can be explained by another observation: on average you spend more time in slow lanes.

The checkout at a supermarket has the same phenomenon. Let's say with 50% probability you choose a line that takes 10 minutes and with 50% probability you choose a line that takes 20 minutes. On average now 67% of the time the other line will move faster.

I believe your analysis is incorrect. There is a 50% chance the other line will move faster. With no additional new information, once you've chosen a line to stand in, the probability you chose the slower line has not changed.

That's after 1 cycle. Assume you chose once per week for 10 years = 520 trips. So, 50% of your trips (260) you would pick the fast lane and 50% of the trips you would pick the slow lane (260). The sum of your slow trip time is 20 * 260 = 5,200 minutes and your fast trip time takes 10 * 260 = 2,600 minutes.

So, out of a total wait time of 5200+2600= (7,800) minutes you spent 2/3 of it in the slow lane and 1/3 of it in the fast lane.

I think the last statement of the parent is causing the confusion.

>On average now 67% of the time the other line will move faster.

This should be: "On average, you spend 67%~ of your time in the slower lane."

On average the other lane is still only faster 50% of the time assuming the odds are truly 50/50 - but you spend 67% more time in the slower lane.

Gotta be careful with the word 'time'. 50% of your trips, the other lane is faster. 67% of the time, its faster.

But this isn't what the parent wrote. I think it is what he/she meant but it isn't what was written. The probability that the other lane is faster is still 50%.

There's a 50% chance of a given lane being the fast one, but you'll spend twice as long in the slow lane if you're stuck there.

Thinking of total time for grocery store lines may feel unnatural, but it's certainly the perspective that makes sense on a road. I wouldn't start a 200 mile trip by saying "I'm going to spend this trip in the right lane," I'd do some of my driving in each.

And if I drive 100 miles in the fast lane and 100 miles in the slow lane, I'll have spent >50% of my time in the slow lane, even if my odds of picking the right one are 50/50 at any given moment.

"The utility of switching lanes while stuck in traffic"

https://news.ycombinator.com/item?id=7622626

  Counterintuitively, in stop-and-go traffic everybody spends more time
  watching others pass them than passing others.

  Let's say you have to drive 100 km. Half of the distance your lane
  moves at 100 km/h, and the other half your lane creeps along at
  10 km/h. So  you'll spend 50 / 100 + 50 / 10 = 5.5 hours in the
  car, at an average speed of 18.18 km/h.

  For every one minute you spend passing other cars, you'll spend 
  10 minutes watching other cars whiz by you!

  Even though the fast and slow distances are evenly distributed,
  all drivers perceive that they're in the wrong lane.

In California the lane to my left should be moving faster than the lane to my right. Big rigs tend to stay in the right-side "slow" lanes and typically aren't driving the 70-75mph everyone else is (although the speed limit is supposed to be 65mph)

People in the far left lane or carpool lanes are sometimes going 80mph~

Small side note:

When I think highway, I think a 2 lane stretch of road without any stop signs or traffic lights. The simulation makes it sound like they're talking about a freeway.

What's your definition of a freeway? I think this may be a regional difference, perhaps - in Ohio, freeway and highway are generally synonyms, and "freeway" may even be used for toll roads at times.

This sums it up nicely.

http://www.diffen.com/difference/Freeway_vs_Highway

It strikes me that this is a good explanation for "imposter syndrome". Let's say you are a Dev Lead looking to understand your team relative to others and you mentally measure "solid new features". Each team that releases a solid new feature seems like a slip - and there will be a lot of those, but your release (skip) will barely pass notice in your mind

Anecdata coming, so keep in mind that this is 100% scientifically verifiable fact:

I find that this behavior is more typically the case when there is congestion due to the number of lanes expectedly decreasing.

In the case where the number of lanes decreases unexpectedly (e.g. due to an accident), the unobstructed lane seems to be slightly faster on average.

When the congestion is due to heavy merge/exit traffic, then I tend to find the traffic moves more quickly as you move away from the merge/exit lane.

I always spot against other cars in my vicinity to see which lane is moving more quickly.

I certainly do see the behavior illustrated in the article, but I don't know that it's the dominant traffic behavior.

Although I didn't look really carefully, it looks like the data only took a single lane change into account.

What about weaving into the lane that seems faster, passing several cars, then switching back? It seems like the snaking effect would work to your advantage if you always get in the lane where there is more distance between cars. Then again, maybe it's perception.

Imagine a theoretical highway where all of the cars except you are totally stopped, and they are all very widely spaced. You could easily weave in and out and make good time.

Clearly the ability to freely and easily change lanes is advantageous to a single driver.

The problem is, in gridlock, this ability is pretty unlikely.

Unless they took a very large sample this might not be entirely true. Or put differently the lane next to yours over a long time frame isn't moving faster, but it does move faster at various points. By switching to the lane that is moving faster at that moment, you can move faster than your original does over the average.

They give a reason why people might think they go faster, but don't prove that they don't go faster.

See http://www.anthropic-principle.com/preprints/lane/cars.html

Reply on news.ycombinator.com