Comment by a-dub | Hacker News Reader

Comment by a-dub | Hacker News Reader

I don't buy this argument that they were explicitly installed as a placebo. Instead I think they're basically a form of technical debt. In the elevators, they probably once did work, but then the ADA stuff came about or the maintenance people just didn't see the point of hooking them up. For crossings, some of them do work, and perhaps many. What's easiest, roll them out uniformly or install a hodge podge? ...and for the motels, it wouldn't surprise me one bit if it's cheaper to buy a residential thermostat with a knob than a commercial temperature sensor. Maybe that even worked as well...

Engineers and builders don't care about "placebo effects on users" except for in the most carefully built UX scenarios. We're talking about basic infrastructure here, no one really cares about UX that much as long as it meets basic requirements. All these buttons exist as an accident, or they're detritus from an era when they did work, or it was just cheaper to get the panel with a button than a blank.

"Let's put a placebo button in to make people feel better." said no traffic engineer, elevator installer, motel HVAC installer, ever.

Lots of devs are now using fake loading bars to emphasise certain tasks e.g. Spinning up a Droplet on Digital Ocean.

I heard that loading bar is just a placebo to emphisise the creation of a new droplet as being important. Apparently the droplet is finished initialising long before the loading bar's 4 or 5 second length. It's probably why this loading bar apprears to load so perfectly -- it's not timed on anything but a counter.

In fact I had to draw out the experience of topping up in our app. It happened too quickly and people didn't know what was just happened. They clicked their pay button and were brought to the balance screen.

They all panicked for some reason and ran back to history to check the transaction went through. We put in a success message before being brought to balance screen and panic is gone. The success messasge is just elaborting on what already happened -- it's a placebo.

I think you're right about the real world. Real buttons are often just broken but in the world of software theres lots of bright and dark patterns behind the buttons -- no software button broke from mechanical failure.

Visual feedback is not the same thing as a placebo effect though.

A placebo would be adding a button that says "accelerate droplet launch speed" that does nothing. It's something that the user takes action on that is irrelevant.

This is definitely not true on Digital Ocean, or at least it's not just a ui thing. The API actually takes this long to return a value, and even once droplets are shown in the dashboard it can take a minute before the network interfaces are bound.

If anyone at DO is listening why does this take time at all? Why not have instance types pre-provsioned and booted and just run cloud-init when the user makes the droplet? Instant provisioning would be incredibly useful for scaling and fault tolerance.

Earlier NYT article explicitly supports your point: https://www.nytimes.com/2016/10/28/us/placebo-buttons-elevat...

Given that the estimated useful life of an elevator is 25 years, it is likely that most elevators in service today have been modernized or refurbished, rendering the door-close buttons a thing of the past for riders, Ms. Penafiel said.

But as The New York Times reported in 2004, the city deactivated most of the pedestrian buttons long ago with the emergence of computer-controlled traffic signals. More than 2,500 of the 3,250 walk buttons that were in place at the time existed as mechanical placebos. Today there are 120 working signals, the city said.

About 500 were removed during major construction projects. But it was estimated that it would cost $1 million to dismantle the nonfunctioning mechanisms, so city officials decided to keep them in place.

Most of the buttons were scattered throughout the city, mainly outside of Manhattan. They were relics of the 1970s, before computers began choreographing traffic signal patterns on major arteries.

But note the bit about dummy thermostats!

Depending on where you work, you might find the thermostat in a plastic case under lock and key, but if you’re lucky you might have control over one.

Well, you might think you have control.

The Air Conditioning, Heating and Refrigeration News reported in 2003 that it asked readers in an informal online survey whether they had ever installed “dummy thermostats.” Of 70 who responded, 51 said they had.

My building has just replaced the lift systems with brand new "smart lifts" and the buttons are gone completely, only the emergency button remains. This has made me err more towards the placebo button theory though, being in one of the button-less lifts feels like being trapped in a dystopian nightmare with computers running everything, the one's run by the same computer that haven't had the buttons removed yet feel much safer.

Also, these new "smart lifts" are horrible, much slower than the previous ones and seemingly not built with the concept that 1 elevator request could involve multiple people travelling.

For elevators at least, there’s a much better explanation for the “door close” button: it’s intended to be used when the elevator is in a mode which holds the door open by default, like when someone’s moving into an apartment building. In this case, you have to hold that button down until the door closes all the way.

> basically a form of technical debt

Or left there for future flexibility.

A crosswalk near me was just switched to where the button does nothing. (You can tell because now the walk signal shows on every cycle of the light. Previously, it would be skipped if you didn't push the button.)

If they had removed it, it would make things harder if they ever want to reverse this decision.

They'd have to reinstall the buttons, which could involve locating it in storage or even buying a replacement, plus sending out a crew, etc. And there are signs that go along with the buttons, so you've got to make sure those are reinstalled correctly. Not insurmountable problems, but they have a cost, and it's not clear this would be the best use of limited money.

It's actually incredibly common in audio controls to give people a "volume knob" that barely has any range at all. I've seen similar tiny-range controls on commercial thermostats. Even if the system is perfectly calibrated and automatically adjusts levels on its own, people want to feel like they are in control.

As for elevators, I have encountered exactly one building where the door close button works even in the default mode (not independent service).

I agree. Definitely more a "Heinlein's Razor" variation on "Never attribute to malice that which can be adequately explained by stupidity".

*Hanlon's razor

Both names are in use.

From The Jargon File (http://www.catb.org/jargon/html/H/Hanlons-Razor.html):

At http://www.statusq.org/2001/11/26.html it is claimed that Hanlon's Razor was coined by one Robert J. Hanlon of Scranton, PA. However, a curiously similar remark (“You have attributed conditions to villainy that simply result from stupidity.”) appears in Logic of Empire, a classic 1941 SF story by Robert A. Heinlein […].

Right, but since most knowledge bases (including Jargon File, Wikipedia and others) use "Hanlon" as the main title, it seems easier and more understandable to go with that. I should have included that in my comment though.

That’s Hanlon’s Razor, Heinlein’s is, “Never attribute to malice that which is adequately explained by stupidity, but don’t rule out malice.”

In both an element people often overlook is the keyword adequately. Sure anything can be attributed to stupidity, but in many cases it’s too beneficial to the “stupid” party and/or too persistent to be adequately explained by something other than malice.

I call what you're referring to Hanlon's Safety Razor: Never attribute anything to malice.

It seems like it deserves a name since in nearly any thread online involving a supposition of malice someone comes out of the woodwork with Hanlon's razor.

PS: I think I had a better name for this in the past but have forgotten it.

Reply on news.ycombinator.com