Comment by zhte415 | Hacker News Reader

Comment by zhte415 | Hacker News Reader

Parent Full thread

I'm reminded of this train/bus is taking ages to come.

If one comes, say, every 15 minutes, that's an average waiting time of 7.5 minutes (and there's a +/- of around a minute for whether the 'time' is arrival, doors opening, closing, departure, depending on one's perspective of which 'time' is important to them).

If one's already there, or comes within 1-2 minutes, that's perceptively close to time = 0, which pushes out perceived average waiting time to 8-9 minutes with times approaching or exceeding average waiting time having a disproportionately negative perception (your total journey time/connection impact, expectation of unannounced cancellation, etc, increasing).

Nifty modern stuff like indicator boards work to re-align these perceptions.

Passage of time is also quite non-linear - back to the ATM queue. When using the ATM I'm engaged in a series of short-term tasks, just as writing a comment on HN when sitting on the subway causes stations to pass 'faster'.

And back to the ATM - I'm sure there's someone somewhere in a retail banking team that measures average interaction with ATM timings. Do they read HN or would like to comment?

My favorite chestnut regarding this is the one that occurred at Houston Intercontinental: https://archive.ph/Z3gjx

> SOME years ago, executives at a Houston airport faced a troubling customer-relations issue. Passengers were lodging an inordinate number of complaints about the long waits at baggage claim. In response, the executives increased the number of baggage handlers working that shift. The plan worked: the average wait fell to eight minutes, well within industry benchmarks. But the complaints persisted.

> Puzzled, the airport executives undertook a more careful, on-site analysis. They found that it took passengers a minute to walk from their arrival gates to baggage claim and seven more minutes to get their bags. Roughly 88 percent of their time, in other words, was spent standing around waiting for their bags.

> So the airport decided on a new approach: instead of reducing wait times, it moved the arrival gates away from the main terminal and routed bags to the outermost carousel. Passengers now had to walk six times longer to get their bags. Complaints dropped to near zero.

Love this solution. Make it tougher overall but the perceived impact is less. Ooof.

And it has a negative impact for the people without checked bags, but they’ll never realize that something changed — their life just gets imperceptibly worse. Big oof.

It also makes it worse for people with checked bags. In the past, you got to decide what you wanted to do with your time - taking a long hike around the terminal was always an option. Nowadays everyone is forced to partake in the jetlagged death march

Thanks to this, after your flight you no longer walk 3 minutes and then rest for 15.

You get to partake in a half-marathon, dragging along your jetlagged kids and your numerous bags, jackets, etc. It’s a excellent example of optimizing for metrics instead of optimizing for people.

Imagine your hotel moving the valet service from downstairs to a parking lot two miles away. You tell the reception desk you need your car, and sure enough, your car is ready by the time you’ve made the walk.

Imagine being so detached from reality that you think making your passengers take an extra 20-minute walk around the terminal is an elegant solution

I mean, I really doubt this story actually happened the way it did, if only because at most major airports gate capacity is so constrained that there is no such thing as a separate gate for arrivals that you can just move planes to. (Planes arrive and leave from the same gate; any more moving around and they'd be wasting precious taxiing capacity and man hours, so if an airplane is getting moved it's usually for maintenance or to standby somewhere else.)

> that there is no such thing as a separate gate for arrivals

The passengers from international flights must be routed through border control first.

In every international airport I’ve been too, this is achieved with a gate that has entrances to two or three levels. International arrivals on floor A and everyone else on floor B, sometimes also splitting domestic arrivals and departures.

While I agree with the point you're making

> there is no such thing as a separate gate for arrivals

You might be surprised. In many of the newish airports I've been to (outside the US) the gates are designed such that arriving and departing passengers for the same aircraft never bump into each other.

They do this by clever design; immediately after exiting the aerobridge and onto the terminal building, there's a split (staircase or a sloping walkway) of some kind. So, the departing and arriving passengers are on different floors.

> It’s a excellent example of optimizing for metrics instead of optimizing for people.

It seems like the opposite. The change directly led to less complaints from people. It's more like optimizing for human comfort rather than giving people what's supposedly best for them via an objective metric. Though I'd say even more specifically (assuming the story is even true) it's optimizing for keeping the small percentage of squeaky wheels comfortable.

I would replace "comfort" with "perception" but otherwise I agree.

> The change directly led to less complaints from people.

Less complaints about waiting time, that is.

Of course, what they should have done is improved the baggage handling time.

The frustration with waiting at the carousel is not the time it takes but the fact that you have no idea how much time it takes. So you're just sitting there in alert mode looking for your things.

If they had given customers information about _when_ their luggage would be available for pickup, the complaints would probably also have dropped to zero.

There is also the effect that buses arrive imprecisely and consequently may be best modeled as a Poisson distribution of arrival times. If this is the case and buses arrive on average every 15 minutes, then passengers arriving at random will wait not 7.5 minutes but 15 on average.

Intuitively, think of throwing a dart at the time line. We are more likely to hit the line in the bigger gaps between buses’s arrivals than between buses that arrive close to each other.

See https://en.m.wikipedia.org/wiki/Poisson_distribution

With buses you will on average end up waiting longer than the average rate when there's heavy traffic because sometimes they get stacked up right behin each other, and two (or once I had three) buses arriving at once is the same as one big one operating at a fraction of the rate. I remember this being relatively common when I used to take the bus.

Reply on news.ycombinator.com