Elevator Etiquette

Kazzy

One man. Two boys. Twelve kids.

Related Post Roulette

49 Responses

  1. Chris says:

    I know that not everyone can take the stairs, but if you can, and you take the elevator down 4 flights to take it back up 2 floors from where you started, what the hell are you doing?

    I’ve never been in a situation quite like the one you describe, so I’m not sure how I’d feel in context, but from far away, it doesn’t seem so bad to me to take the elevator down to go up (or up to go down) if a.) you need to take the elevator and b.) either time is an issue or you are likely to be stuck on your initial floor for an extended period of time. If, however, the elevator is that crowded and you can take the stairs, you damn well should.Report

    • Kazzy in reply to Chris says:

      Assume the stairs are not an option. I never take the elevator solo and take my kids on the stairs a decent amount (though I have the oldest in the school). Elevator riders tend to be A) seniors, B) large groups of children, or C) caregivers toting very little ones… all of whom I would consider to have legitimate cause to use the elevator.Report

  2. Damon says:

    I don’t consider this a etiquette violation. There’s must too much demand for the single elevator.

    “It’s a dog eat dog world and from where I sit, there just ain’t enough damn dogs.”Report

  3. Not exactly on point, but I used to take a subway on my daily commute, with the work end of the trip two stops from the end. There were occasions when going home, I would intentionally get on a train going the wrong direction.

    Sometimes this would be if the weather was unbearably hot and humid in the station, while the air conditioning in the trains was actually pretty good. A few times it was because the system was in semi-failure mode, and the platform was packed with people. Getting on a train going the wrong way was easy, while getting on the very same train going the right way would be problematic.

    Was this cutting in line? Not really. There was no line. I had as good–and as bad–a claim on a spot in the train as anyone else. And anyone else who thought of it could have done the same thing. And there was some risk that I would miss a train going the right direction and lose about ten minutes as a result. I thought about it, and decided I could get on that wrong-way train and live with myself.Report

    • Murali in reply to Richard Hershberger says:

      I’ve done the same with trains myself (if I’m near the start of the line, go in the opposite direction and wait for everyone to clear out giving me a place to sit) I’ve also done it with elevators when my knees are giving me trouble. Kazzy is the first person I’ve known to see this as a breach in etiquette. It’s common sense. If you were to ask me what to do in a situation where the elevators are always crowded, I would tell you to take the elevator going in the opposite direction (if has space) and grab a spot for yourself in advance.Report

      • Kazzy in reply to Murali says:

        @murali
        Just to be clear, I’m not certain it is a breach of etiquette… just that it feels like it is.

        In reading some responses and reflecting more, I realize that my response may be motivated by the fact that I am highly unlikely to use this tactic — either alone or with my students — but am highly likely to be inconvenienced by it. So perhaps I am attempting to rationalize my personal annoyance and/or bitterness at being burned by it by claiming it to be an etiquette violation?Report

        • Murali in reply to Kazzy says:

          Maybe, or maybe its a cultural difference. Burping is rude in one culture but required by etiquette in others. You’ll have to ask other new yorkers. That said, if everyone is doing it so that the lift is too full to board even at the first floor, then maybe it might be overdone. I sometimes feel frustrated when that happens, but since I do the same thing I don’t really have grounds to complain.Report

          • Kazzy in reply to Murali says:

            @murali Elevators are also a thing that many (most?) people don’t interact with on a regular basis and likely not in the specific scenario I do. In fact, this may be a problem relatively unique to older east coast cities. Someone (I’m forgetting who) said they see this during very busy times in a building with multiple elevators. But I suspect it may be an infrequent enough occurrence that expecting etiquette to evolve is misplaced.Report

            • Murali in reply to Kazzy says:

              Its also a thing in some shopping malls and hospitals* in Singapore

              *I’ve seen nurses do this. Well, not necessarily pushing the opposite direction button, but opportunistically getting on an elevator moving in the other direction to reserve a spot.Report

              • Kazzy in reply to Murali says:

                @murali

                Some malls have elevators but most in my area tend to have escalators. Or, if they do have elevators, they are usually out-of-sight somewhere and available for handicapped people and others in need. Hospitals often have elevators but I rarely find myself in one. Zazzy might know better.

                It’d actually be interesting to figure out how often people typically end up in an elevator. My guess is you’d see bulges at the extremes… either you are like me and live in a city and tall apartment building and work in a tall building and ride elevators multiple times a day or you NEVER ride elevators. I bet very few people are in the middle. And I’d bet the “never” bulge would be much larger than the “all the damn time” bulge.Report

              • Murali in reply to Kazzy says:

                Depends on where you are. 80% of people living in Singapore live in public housing all public housing in Singapore consists of flats. Flats are usually 10 stories or more tall.

                So for most singaporeans, they take the lift all the freaking time.Report

              • nevermoor in reply to Murali says:

                Sacramento Superior Court has this issue (despite four elevators) since everyone is using them at the same times (9am hearings, lunch breaks, etc).

                I think getting on an elevator going the wrong direction is different from calling an elevator in the wrong direction, since the former isn’t affirmatively reducing efficiency (just taking a human’s worth of space).Report

    • Kazzy in reply to Richard Hershberger says:

      I considered this scenario and it felt different. I think because you aren’t actually claiming a spot on the train you plan to ride prematurely. But maybe that is just rationalizing?Report

  4. DensityDuck says:

    All I know is that it’s considered very rude to attempt to crush people by pushing the “CLOSE DOORS” button. (Also it doesn’t work.)Report

    • Damon in reply to DensityDuck says:

      Totally!

      Do you know how many of my managers I tried this on only to realize later that the safety mechanism prevents that?

      So many….so many..Report

      • Troublesome Frog in reply to Damon says:

        I have to say that watching a friend of mine with an uncasted fractured forearm reflexively stick his arm in an elevator door to keep it from closing was pretty hilarious.Report

    • And then there’s those embarrassing moments when you see someone running for the elevator and hurry to push the “doors open” button…only to realize you’ve pressed the “doors closed” button instead and now look like you were deliberately being a jerk.Report

      • dragonfrog in reply to KatherineMW says:

        This is why I just stop the doors with a hand or foot – I figure I have about a 75% chance of hitting the ‘close doors’ button (if the button actually does anything – I suspect a lot of those are placebo buttons).Report

      • North in reply to KatherineMW says:

        If it is any consolation I have it on good authority that the “Close Door” button is not connected to anything on most elevators. The doors closes automatically and the button is pure psychology.Report

        • Damon in reply to North says:

          Yep. I think it only works when the fireman inserts his key.Report

          • Kazzy in reply to Damon says:

            Or when moving in.

            Most crosswalk buttons — esp in major cities — are similarly phantom.Report

            • Damon in reply to Kazzy says:

              God forbid we have demand driven cross walks. And stop lights too.

              Shesh.Report

              • dragonfrog in reply to Damon says:

                Mostly I think what’s happened is that the crosswalks were originally demand driven for absolutely no reason – that is, they have no effect on the traffic light timing, but even so you had to push a button to make the crossing light turn on.

                This only served to inconvenience those on foot if they forgot to push the button or if the button was broken. There was no advantage granted to motorists if no one pushed the crosswalk button, the light cycle timing was optimized to traffic flow no matter what. It just came from a mentality of walking being second class transport.

                So they reprogrammed the crosswalk to just always work, but have often not sent people around to uninstall the obsolete hardware.

                In my city, the buttons are not placebos – they’re one of
                – unnecessary and humiliating beg buttons
                – controls for a traffic light that otherwise stays green forever
                – “for audible signal only”Report

        • DavidTC in reply to North says:

          OTOH, you *still* look like kind of a jerk, because the ‘Door Open’ button that you aren’t pushing is connected. So while you didn’t make them close faster, you still let them close at all!Report

  5. Troublesome Frog says:

    That’s an interesting question. I think it’s allowable. In fact, it seems like the most equitable way of handling elevator queuing. People at the bottom floors have a natural advantage getting to the top floors because they get on first. People at the top going to the bottom have the same advantage. Given sufficient demand, people in middle floors will never go anywhere. There’s no way to queue.

    At the same time, assuming the elevator traverses from the top floor to the bottom each time (a rough approximation), it passes every floor equally and presents everybody with an equal opportunity to queue up and occupy empty spots as they pass by. Just think of the elevator as something that’s moving in a big circle allowing people to jump on if there’s empty space. That way, the wait time for an empty slot should be roughly equal regardless of what floor you’re on and there’s no way for excess traffic to completely ‘starve’ some floors.Report

    • dragonfrog in reply to Troublesome Frog says:

      I’m not sure if that’s right or not – I’m having a bit of trouble sorting out the theory of the two models (only board the elevator moving in the direction you want vs board the elevator at first opportunity)

      I think I have got this much:

      In the board-at-first-opportunity model, if there is heavy enough traffic from ground to 2 and 2 to ground, then nobody from 3 to 7 will be able to board the elevator at all – even if they are trying to travel only among those upper floors. Those seeking to go from 2 to ground will fill the elevator on the way up, it will then travel from 2 to all upper floors where it has been called, nobody will be able to board, then it will descend to ground before emptying. That could go on indefinitely as long as enough people are coming and going from the 2nd floor.

      In the board-only-in-your-direction model, that would not happen. It is possible that the people wanting to descend from 2 to ground would find the descending elevators full at times, but that couldn’t go on indefinitely – at some point, the upper floors would be empty.

      That’s obviously the extreme case, hopefully the real building doesn’t have demand that badly unbalanced.Report

      • DavidTC in reply to dragonfrog says:

        That could go on indefinitely as long as enough people are coming and going from the 2nd floor.

        Not only that, but now the elevator is stopping on every floor, in every direction, often completely full. So all the round trips take *forever*, and have accomplished moving exactly one load worth of people, as opposed to the normal operation of elevators constantly losing and gaining more people.

        This isn’t just a hypothetical failure mode, BTW. I’ve literally seen elevators operate like this for hours.Report

        • dragonfrog in reply to DavidTC says:

          They should put in slides, so everyone wants to take the fun way down and the problem takes care of itself…Report

          • And solve the other direction with jetpacks. Seriously, think about it.

            “How do you like the new job”?

            “Well, the money’s not great, and my boss is kind of a jerk. And they don’t give me as much freedom to build a curriculum as I’m used do.”

            “So, are you looking around for something else.”

            “Are you kidding? We get jetpacks!”Report

  6. j r says:

    I work in a large office building on the highest floor serviced by a bank of four elevators. We have the same problem. I don’t try to leave the building between 1150 and 1220, because that is when everyone is leaving for lunch and almost every elevator that reaches my floor is full of people who caught the up elevator on a lower floor.

    I’m tempted to think about this as a breach of etiquette, but the wrinkle is that I’ve just moved to a new city on the other side of the world. So, I can’t really say what etiquette is here. I’ve just come to accept that this is how it is. If people only got on elevators going the way that they were headed, it would probably be better for me, but not so great for the person on the floor closest to the lobby.

    There are situations where small sacrifices by all individuals leads to a collective realization of greater efficiency. This doesn’t seem like one of those situations. Seems like a zero sum game.Report

    • Murali in reply to j r says:

      Wait, where are you now?Report

      • j r in reply to Murali says:

        I moved to Hong Kong last year. It’s like Singapore, but much less orderly and without those awesome hawker centers. We’ve got cooked food centers and good restaurants generally, but my one weekend in Singapore was some of the best eating that I’ve done.Report

        • Brandon Berg in reply to j r says:

          Are the drug dealers there still really brazen? I went to Hong Kong in late 2011, and one guy whispered “hashish” to me as I was going back to my hotel in Chungking Mansions late at night. I went back in early 2014, and people were yelling “Pot! Cocaine! Meth!” at me ten times a day. I’m still not sure what happened in those two years.Report

          • j r in reply to Brandon Berg says:

            Chungking Mansions is notorious. That side of the harbor is sometimes affectionately known as “the dark side.” It’s funny, because I’ve never felt the slightest bit unsafe anywhere in HK. Mostly whenever I walk by there I got hassled by dudes selling watches and trying to get me to tailors.

            The only place I know of where drugs are on offer out in the open is in LKF, which is Hong Kong’s version of Bourbon street. There is a corner where a bunch of African immigrants hang out and apparently they sell cocaine. I’ve never been offered anything myself. The cops recently busted a bunch of Gambian immigrants, so perhaps they now wait to be solicited. I think that there is a sizable population of African and South Asian asylum seekers whose immigration status prevents them from working, so some of them end up populating the lower ladders of the elicit trades.Report

  7. DavidTC says:

    Man, it’s weird how completely random things I’ve spend *large amounts of time thinking about* end up here. Not so much the ‘etiquette’ direction, but ‘Does this come out even’. Here is my claim: Getting on the wrong way might personal benefit you slightly, or not. And it causes net harm to the system.

    But first, a story as to why I’ve spent time thinking about this: The last two years I have been going to DragonCon in Atlanta. Now, never having been to other large nerd cons, I gather that DragonCon is pretty unique, in that it takes place within five hotels. Not, like, people get rooms at a hotel and go to the con…it’s literally taking place in the hotels.

    Basically, here is the average ‘layout’ of a hotel: Three floors connected by escalators, where all the stuff happens, and then the room up above. And elevators connecting all the floors. So floors 1-3 is the con, where everyone is, with almost all elevators trips having one endpoint on floors 1-3, and one endpoint on floors 5+. (Floors 4 and 5 generally take the stairs.)

    And the first time, I went to an orientation on the day before, which explained that you should get on the elevators whenever you can, even when going the wrong way. Being a commuter, I didn’t think about that…I didn’t take the elevators at all. Moving between the lower floors, you take the escalators or stairs.

    And then I went home, and. months later, for some reason, was thinking about this, and I realized how *stupid* that advice was. So I actually spent an hour of so at the next DragonCon studying the elevators (THE NERDIEST THING POSSIBLE.) and realized I was correct.

    Every time you are occupying space on an elevator, and you are travelling between floors you don’t need to travel between, and other people cannot get on because of you, you have just reduced the efficiency of the entire system by a certain amount. One individual doing that doesn’t reduce it by much, but when *everyone* does it…well, at DragonCon, it seems to have degraded the entire thing to less than 50%. Seriously.

    As @j-r said above (although he didn’t state it explicitly), one of the ideas of getting on in the wrong direction seems to be the idea that people on the middle floors are screwed in normal operations. The problem is, if those people on the middle floors get on the wrong direction, now the people on the *higher* floors are screwed if the general direction is down. And lower floors are screwed if the general direction is up…and both upper and lower are screwed if the traffic is *both* ways. So the idea is that the middle floors are ‘cheating’ at the expense of the other floors.

    But, what’s worse, the middle floors haven’t actually made things *better* for themselves, because now the elevators are still going all the way up the building, stopping at floor after floor where no one can get on…and then doing the same thing going *down*. I’m not sure people who do that are even get where they’re going faster, as opposed to just getting on the elevator faster! And even if they did get there faster, they just made everyone get there slower *on average*.

    So, it’s not a zero-sum system…it’s a prisoner’s dilemma. And, for some reason, occasionally everyone in a building starts defecting.

    There are almost certainly way to program elevators to actually deal with this sort of situation, so I’m baffled as to why they don’t seem to be. An *obvious* solution would be to run specific elevators to specific floors. But that requires notifying people and they mess it up, and also elevator banks often don’t have individual call buttons.

    Another less obvious solution might be to have elevators disregard call buttons for a specific amount of period of time (Only when they have passengers, obviously), which would result in elevators *skipping floors*, instead of these dumbass current ‘let’s open on every floor, up and down, even when we got full ten floors ago’ method. Instead, they’d only open on floors where they had passengers wanting to get out. Assuming some sort of equal-ish random distribution of destinations, and when the elevators do get around to answering unanswered call buttons they get priority to the ends, it would mostly be fair.

    Another option would be to let *passengers* somehow indicate to the elevator they weren’t taking more people on. (Which would reset every time someone got off, obviously.)

    But I am not sure how to solve it. I am sure, however, that getting on elevators in the wrong direction isn’t like ‘cutting in line’, which keeps the average amount of time in line the same. Getting on the wrong direction elevator objectively *wastes* ‘elevator spacetime’, just throwing that resource away as you occupy it when you don’t need to and, hence, keep other people from occupying it.Report

    • j r in reply to DavidTC says:

      So, it’s not a zero-sum system…it’s a prisoner’s dilemma. And, for some reason, occasionally everyone in a building starts defecting.
      Don’t see how this is a prisoner’s dilemma. Pressing the up buttom when you want to go down is only defecting if you have established button norms, which is what Kazzy was asking.

      Also, short of significant hardware and/or software upgrades, pressing the “right” button doesn’t bring any gains to total efficiency. The elevator still has to go up and down from one floor to the next and it still can only accommodate a fraction of the folks who want it at any given time. The only question is whether you make the cut or don’t.Report

      • Murali in reply to j r says:

        Pressing the up buttom when you want to go down is only defecting if you have established button norms, which is what Kazzy was asking.

        Don’t we define cooperation and defection by the payoffs? Everyone would do better if everyone took option A than if everyone took B, but any one individual can do better by unilaterally picking option B regardless of what anyone else does. So, what counts as defection is not defined by what norms there are. But, I don’t think that there is a genuine prisoner’s dilemma. While the elevator is a commons, there should to be some mixed equilibrium of behaviour.Report

      • Kazzy in reply to j r says:

        @j-r and @davidtc

        So leaving aside the question of etiquette, the system I described is less efficient overall?Report

        • DavidTC in reply to Kazzy says:

          So leaving aside the question of etiquette, the system I described is less efficient overall?

          The system you describe can degrade to *total crap*. I have seen it with my own eyes. I have ridden elevator trips (as an experiment) starting floor eight, where we spent a good thirty minutes riding *up* a forty story building, and then back *down* the same building, stopping at *every* floor in *both* directions, letting in a few people on the way up (as people got off…presumably those people having first ridden *down* and then back up), and absolutely no one on the way down…to get to the floor three.

          All because people on middle floors tend to get screwed in times of heavy traffic, because the elevator stops at their floor full, and the system assumes everyone got on, resetting the call button and, thus, the call button timer.

          All it *actually* would take to stop this is for the system to assume, when a call button was *immediately* pressed after an elevator visited the floor, that it was still the same people waiting, and keep that call button on a priority. (Which means they should start *there* at some point, instead of going all the way to one of the ends to start servicing the other direction.)Report

      • DavidTC in reply to j r says:

        Don’t see how this is a prisoner’s dilemma. Pressing the up buttom when you want to go down is only defecting if you have established button norms, which is what Kazzy was asking.

        That’s not what a prisoners dilemma is. It doesn’t matter if there are ‘norms’, it only matters if there’s a system where if both people do X, everyone’s outcome is okay, and if both people do Y, everyone’s outcome is bad, and if one does X and one Y, the person who did Y has a good outcome and the person who did X has a worse one.

        The term ‘defect’ doesn’t require any norms at all, it’s just what the ‘win at the expense of the other player’ choice is called. You can explain the game to people without stating what they should do.

        Now, this isn’t *exactly* prisoner’s dilemma, which sorta requires everyone to *understand* the payoff and what different options do. But as has been proven by people here, no one seems to notice that this isn’t zero-sum…you’re not just taking a trip from another person and causing no other harm. It’s not ‘cutting in line’. It’s actually degrading the entire system, both reducing capacity *and* adding delays. But people *think* it’s just cutting in line, because they don’t understand the payoffs correctly.

        Also, short of significant hardware and/or software upgrades, pressing the “right” button doesn’t bring any gains to total efficiency. The elevator still has to go up and down from one floor to the next and it still can only accommodate a fraction of the folks who want it at any given time. The only question is whether you make the cut or don’t.

        No.

        If it was just people getting on in the wrong direction when the elevator happened to open at their floor, that might make sense, but that’s not how it works. If everyone is getting on in the wrong direction, they’re *pushing the button* in the wrong direction also, and hence all the elevators stop at all floors in *both* directions. (Hell, even if everyone gets on at a floor, and no one new person shows up to wait, the stupid elevator will *still* stop at that floor in the other direction, because that button was pushed and as far it it knows, there hasn’t been an elevator in that direction yet so people are waiting. Or, hell, some *other* elevator will.)

        Whereas without that dumbass behavior, if mostly everyone was going down, elevators would only stop at a few floors going up. (Whatever floors the ‘backwards’ people wanted.)

        It’s possible there’s some sort of ‘ideal’ overloaded system where all elevators are so large and there are so many people that the elevators would be stopped at every floor, in every direction, already, so there would be no additional stops from ‘wrong direction’ people to slow things down, but that’s not generally how it works in practice.

        And on top of that, there’s the *other* way the system is being harmed:

        Someone getting on an elevator early is occupying space that others could be using *at that point*. (This is only a problem if the elevator system is full, but if it’s not, people already have no reason to get on in the wrong direction!) If I am on the sixth floor, going to the second floor, and I get on upward and ride up to the twenty-fifth before the elevator turns around and goes down, I have just occupied space from six to twenty-five, which means I just denied a trip to anyone going from the seventh *upward*. And, for that matter, I’ve erased a trip *down* down to any floor above seven, which the system could have dealt with without impacting me.

        The elevator system has finite operational capacity. It can only move X people Y floors per every hour. People going in the wrong direction occupy space in it they don’t need to, (in addition to the space they *do* need to occupy for their trip), thus denying *other* people that space. Likewise, people causing elevators to stop at floors when the elevator is going the wrong way add delays to a trip, aka, reducing the total amount of people it can move.

        Both of those objectively slow the system down, and appear to provide no benefit to the system at all that could possibly cancel that harm out. At best they supposedly get people somewhere faster, although in reality so much capacity has been removed from the system that I seriously doubt *anyone* is getting anywhere faster than if everyone just used the thing correctly.Report

      • Brandon Berg in reply to j r says:

        In the canonical prisoner’s dilemma, there are no established norms. The prisoners are informed of the payoffs, but they’re not allowed to communicate. Lack of communication isn’t really an essential part of the prisoner’s dilemma, since they can still defect even after agreeing on a strategy, but it’s still defecting, with or without norms.Report

      • j r in reply to j r says:

        You guys are all thinking in very static terms. Think iteratively. Why do criminals develop norms against snitching? Because police play them off of each other. Why do police separate people that they are questioning? Because criminals have developed norms against snitching. The optimal prisoner strategy is to keep the norm against snitching. The optimal police strategy is to get one or both of the prisoners to break the norm.

        Sure, you can take the prisoner’s dilemma purely as thought experiment, stripped of all context. But since Kazzy’s post asks specifically about etiquette (ie norms), I’m not sure why you would do that.

        If people could coordinate their behavior across floors, then maybe they’s be able to reach a more efficient solution. But they can’t. And that’s why I said that, short of more elevator capacity, the only way to make these situations more efficient is with a software improvement. And that’s why this isn’t a prisoner’s dilemma situation. There is no adoptable norm that improves the outcome for everyone. Someone is always getting shut out of a full elevator.Report

        • DavidTC in reply to j r says:

          And that’s why this isn’t a prisoner’s dilemma situation. There is no adoptable norm that improves the outcome for everyone.

          Your ‘everyone’ is doing a lot of work there. The entire premise of a prisoner’s dilemma is that each individual is better off if they *do* defect, regardless of what the other person does. So a ‘norm’, or, rather, a rule, that improve the outcome for everyone *individually* exist…they should defect. That’s the *premise* of a prisoner’s dilemma…everyone should individually defect!

          I think what you’re trying to say is ‘improves the outcome on average’, as in, even if it’s not best option individually, it is better in total. But even that isn’t right.

          If, in a prisoner’s dilemma, people are offered a million dollars to them and $0 to the other person if they’re the only defector, but get $100 if neither defect, and $10 if both sides defect…that’s a perfectly valid prisoner’s dilemma. The best choice is, regardless of what the other does, is to defect. (And this is the sort of prisoner’s dilemma where the two sides would tend to ‘cheat’ the rewards by having only one side deliberately defect and they agree to share the money outside the game.)

          Meanwhile, not only *is* there actually an adopted elevator norm that improves the system for everyone, and a method of communication, and it *so clearly* improves the system that *we built extra hardware to enable it*.

          Or do directional call buttons exist because they look nice?

          Someone is always getting shut out of a full elevator.

          Yes, and either some floors will be shut out of a full elevator in a *working* system that will clear up in short amount of time, or some different floors will be shut out of a full elevator in a barely functioning degraded system that will take twice as long to clear up because everyone is defecting.Report

    • DensityDuck in reply to DavidTC says:

      “But, what’s worse, the middle floors haven’t actually made things *better* for themselves, because now the elevators are still going all the way up the building, stopping at floor after floor where no one can get on…and then doing the same thing going *down*. I’m not sure people who do that are even get where they’re going faster, as opposed to just getting on the elevator faster!”

      The story in the OP is that the middle-floor people never found space on the elevator. It wasn’t a matter of just waiting a little bit longer.

      The question is what you weigh more strongly; a higher average, or a lower average with higher variation. It’s been pretty well established that the former results in higher throughput for traffic movement; and, as we see from the story, variation in the latter can be high enough that some users don’t complete their trip within the timeframe of the scenario.

      *******

      Here’s a proposal: At peak travel times, doors only open going up. To get to the ground floor, you get on the upward elevator and ride it to the top, then ride it all the way down (which it will do in one movement, since it doesn’t stop on the downward leg). Presumably, on the upward leg of the route you will have users reaching their destinations and exiting the elevator (which would be rare for the downward leg) and this will help manage the overall load.

      This turns the whole “go up to go down” into a feature of the system, rather than an erroneous usage; and it handles the issue of “no space on the down elevator” by removing the ability to get on an elevator going down.

      If you want to take the elevator to a lower floor, then you ride it all the way up, then all the way down, and then up to the floor you want to be on. This is inefficient compared to just taking the elevator down, but if you could just take the elevator down then we wouldn’t be having this conversation.Report

      • DavidTC in reply to DensityDuck says:

        Here’s a proposal: At peak travel times, doors only open going up. To get to the ground floor, you get on the upward elevator and ride it to the top, then ride it all the way down (which it will do in one movement, since it doesn’t stop on the downward leg). Presumably, on the upward leg of the route you will have users reaching their destinations and exiting the elevator (which would be rare for the downward leg) and this will help manage the overall load.

        Yes. There are an awful lot ‘doors opening to full elevators’ that serves no purpose at all.

        Your idea does help, but only by reducing the number of double door openings. Which, don’t get me wrong, is useful, but it doesn’t help at all with the fact that people at the top still can’t ever get an elevator, because they are already full of people trying to go down. (Likewise, you could run the elevator all the way to the top and let people off/on going *downwards*, and now it’s the people closer to the bottom who can’t get on.)

        That’s why I was saying, ignore all the call buttons for a bit, which will cause the elevator to go to *random* floors where people could get on.

        So…let’s combine the ideas.

        How about the elevator only lets people off going *up*, and then only stops for call buttons going *down*? I.e, it skips call buttons upward, and elevator buttons downward.

        Slightly less efficient than your idea, but it *does* allow people at the top to get an elevator. People often take elevators in groups, after all, so there’s a chance that everyone in an elevator could be going to just two or three floors. They get out, a few people get in, and then it goes to the top and starts letting people in from there.Report

  8. j r says:

    @davidtc
    Thanks for telling me what a prisoner’s dilemma is. This still ain’t one.

    Let’s simplify this. Consider a building with three floors and a lobby. It has one elevator with a max capacity of two people. There is a person waiting on every floor to get to the lobby and there are no accessible stairs. And every time someone gets on the elevator, another person shows up to wait for the next one. That’s pretty much the situation that we are talking about.

    If everyone pushes the down button, the elevator goes straight to the third floor and then makes its way down. If the person on the first floor keps pushing the down button, then every time the elevator gets to him, it will have already been to the second and third floors and full. If that person pushes the up button, then the elevator gets to him first and he can get in. Then the elevator goes to the third floor and the person on the second floor gets stuck every time. The person on the second floor figures this out and starts pushing up. Now the person on the third floor will never be able to get on.

    The only way to maximize the total efficiency is to increase the elevator capacity or to give the elevator some software that allows it to go the floor where the down button is pushed first and then skip floors once it’s filled. Then the elevator could alternate which two people it takes on the basis in when they showed up. Otherwise, one floor will keep getting left no matter what. That’s why I wait until 1220 to go to lunch every day.Report

    • DavidTC in reply to j r says:

      Thanks for telling me what a prisoner’s dilemma is. This still ain’t one.

      I have no idea if the rest of this post is suppose to back that statement up, but it doesn’t.

      All you are demonstrating is that it is an *uneven* prisoner’s dilemma. I.e.,if neither defect, the middle floors (A) starts off worse than the upper floors (B), but if they both defect, A ends up better than B. (Although they’re both still worse off than if no one had defected.)

      Are you suggesting it’s not a prisoner’s dilemma because one party doesn’t have a chance to defect? That’s true…but only on the top-/bottom-most floors , who obviously only have one choice of when to get on an elevator. But in most buildings where the end floors are a large percentage of the floors (A 3-5 story building)…none of this is really relevant, because elevators don’t really get overloaded in those buildings…and a large percentage of the population takes the stairs.

      I mean, since you work on the topmost floor, I guess you can argue it’s not a prisoner’s dilemma for you…except, actually, it is. How? Well, you could take the stairs down until you got low enough that you caught an up elevator, thus defecting against the people on your floor and whatever floors you passed. (It’s less payout than getting an elevator at the top, of course.)

      Or you could just choose not to play, which is an entirely rational response when your behavior has no impact on the other side, and the other side has decided to repeatedly harm you.

      The only way to maximize the total efficiency is to increase the elevator capacity or to give the elevator some software that allows it to go the floor where the down button is pushed first and then skip floors once it’s filled. Then the elevator could alternate which two people it takes on the basis in when they showed up. Otherwise, one floor will keep getting left no matter what. That’s why I wait until 1220 to go to lunch every day.

      I have to idea what you are trying to say here. There are plenty of ways to *increase* efficiency, most of which indeed involve strategic skipping of floors. (Although your plan doesn’t appear to deal with the people are trying to go up, you let them off at the second floor and people will get on the elevator….unless the idea is that people would suddenly restart obeying the ‘don’t get on up elevators to go down’ rule.)

      Anyway, you want to discussion how to run an elevator, go ahead. You want to figure that out, you probably want to start with a more complicated building, though.

      My point was that people getting on early *decreases* efficiency. Always. It’s not an even trade off, one person taking another’s place…it degrades the entire system. Even in *your* example, the elevator is *still* making one more stop per trip.

      Normal operation: 1) Elevator starts ground. 2) Elevator goes to top floor, lets people on. 3) Elevator goes to second floor, people cannot get on.

      People going up to go down: 1) Elevator starts at ground. 2) Elevator goes to second floor, people going down get on to go up. (Added stop) 3) Elevator goes to top floor, people cannot get on. 4) Elevator goes to second floor, people cannot get on.

      A single elevator cycle of moving 15 people (Or whatever the max elevator load is) just got 33% longer.

      And note while that example only has one extra stop, that problem scales up as ‘one more stop per floor that has people waiting’. I.e, in a 12 story building, the elevator is making *10 extra stops* per trip.(1)

      Additionally, now notice that people are having a hard time getting from the second to the third floor, whereas before they could do it fine! And a hard time getting from 3-1.

      And *also* notice that 1-3 and 2-1 are now *sharing* space. Which in this specific example would seem fine, because people would get on at the ground floor, and off at the top, which would sometimes would allow people on the third floor to go down. But that’s solely because this has been simplified that there’s only two floors trying to go up. In a ten story building, people on the 6th floor would have a lot of trouble going up, for example.

      People standing inside elevators for any amount of time, when they don’t need to stand inside elevators, in an elevator system that is at max load, is a really dumb idea. People pushing call buttons in both directions when they only need to go *one* direction is also a dumb idea. I understand *why* they do it, but that doesn’t mean it isn’t harmful.

      1) When I said degrades to less than 50% above, I wasn’t just making up a number. In a 40 story building, the round trips are probably 40% longer *by themselves*, and that’s on top of the fact they aren’t moving as many people per trip, because people getting on early are physically occupying space on the elevator that other people are denied, even if *both* their trips could have fit if the first had got on at the correct time. And in *my* example, both up *and* down traffic are overloaded already. ‘Less than 50%’ is probably overly optimistic!Report