The Art of Waiting: Why We (Gladly) Queue for Some Rides

You round the corner toward your favourite attraction, and there it is, glowing on the sign: “60-minute wait.” On some days you spin on your heel and inform your companion that the ride is “overrated anyway.” On others you stand in that exact same 60-minute line and would swear afterward it was 25 at most.

That isn't self-delusion, because both times you were the same person in the same queue. It's the product of two disciplines that mesh perfectly in a good park: the psychology of time perception and the mathematics of capacity. One decides how the wait feels. The other, how long the line really is.

The topic won't let go of me for a very specific reason. park.fan was invented in a Taron queue, out of pure frustration over what felt like an eternity (the whole story is here). What was actually happening to me back there was something I wanted to know exactly. Two formulas come into it, and the two of them fit together on the back of a single beer mat.

The psychology of time: why minutes stretch

The Harvard economist David Maister dissected the core problem back in 1985 in his classic "The Psychology of Waiting Lines". His first and most important rule: unoccupied time feels longer than occupied time. If all you do is stare at the back of the person in front of you, the clock crawls. If you have something to see, hear or do, it keeps ticking at almost normal speed.

On top of that come three more Maister rules that every park planner knows by heart:

• Uncertain waits feel longer than explained waits. That's why the queue is marked every few metres with time-remaining signs, and a wait-time display hangs at the entrance, which, by the way, loves to lie to you. More on that shortly.
• Unfair waiting is unbearable. Nothing ruins the mood faster than the feeling that others are slipping past, which is precisely why parks route their express lanes as discreetly as physically possible.
• Waiting for something valuable is easier to endure. The greater the anticipation, the more patient the line. We won't stand 20 minutes for a fairground flat ride. For the new mega-coaster, we'll happily talk ourselves into 90.

Your sense of time exaggerates, by a good third

Just how badly we judge unoccupied waiting can be measured. In the field experiments of marketing researcher Jacob Hornik, people waiting overestimated the time that had actually elapsed by around 36 percent on average. Ten measured minutes become almost fourteen in your head. Your inner clockmaker is a virtuoso of rounding up, and it always rounds against you. And back in 1991, the MIT queueing researcher Richard Larson and colleagues showed, under the lovely title "Entertain, Enlighten, and Engage", that even simple distraction markedly improves the perceived quality of the wait. In the experiment, a news screen in a bank branch was enough.

The mechanism behind it: our brain can only count time meticulously when it has nothing else to do. Once attention turns to music, theming details or a show, the internal counter simply runs out of resources, and the overestimation melts away. That's exactly what everything you see in a good queue is designed to exploit: soundtracks, animatronics, interactive elements, hidden details. What looks like decoration is in fact applied cognitive psychology.

Pre-shows: the experience starts before you board

The most elegant weapon against dead time is the pre-show, which simply redefines the wait as part of the attraction. The textbook example sits in Disney's Hollywood Studios. At the Tower of Terror(Disney Hollywood Studios, USA)35 Min you don't shuffle down a corridor; you move through the dusty lobby of the Hollywood Tower Hotel into a library where a video in the style of the “Twilight Zone” tells the building's story. You're checking into a hotel you'll later check out of in freefall, and you never notice that you're currently “playing” queue. Technically you're still waiting. It feels as though the ride started long ago. The European counterpart stands at Disney Adventure World in Paris, with the same dramaturgy: The Twilight Zone Tower of Terror.

The peak-end effect: the last few metres count double

The psychologist and Nobel laureate Daniel Kahneman showed that our memory doesn't rate experiences as an average, but by two points: the emotional peak and the end. That's the so-called peak-end rule. For queues that means the last minutes before boarding shape the memory of the entire hour before them. Which is why the finale of a queue is almost always the most lavishly staged part. The densest theming, the best animatronic, the goosebump moment all arrive just before the station. If the ending is strong, your memory forgives the sluggish middle section remarkably readily. Anyone who's ever recommended a mediocre film because of a great finale knows the principle.

When the queue is the show

Three modern masterpieces show how far you can push this, all in Orlando, all with queues people voluntarily turn up early for.

Avatar Flight of Passage: a museum as a queue

The queue for Avatar Flight of Passage(Disney Animal Kingdom, USA)105 Min at Disney's Animal Kingdom is essentially a walk-through museum with a flight simulator at the exit. First the path winds through Pandora's landscape and caves painted by the Na'vi, then through an abandoned research lab, complete with a life-sized avatar floating in an amnio tank, breathing so convincingly that people regularly stop and hold up the line. A queue that clogs itself because it looks too good. After that come two pre-show rooms, where you're “scanned” and “linked” to your own avatar.

That's doubly clever psychologically. The story rooms meter the crowd into groups and turn the final quarter of an hour into part of the attraction, which is exactly the phase the peak-end rule says will be remembered most strongly. They're also the capacity backbone: you fly in theatre-style link chambers with 16 seats per level, three levels stacked, four theatres in parallel. Just under 200 guests at a time, around 1,400 an hour. Even after a 120-minute wait, most people afterwards don't talk about the queue but about the breathing avatar in the tank.

Cosmic Rewind: first the museum, then the coaster

EPCOT takes it a step further with Guardians of the Galaxy: Cosmic Rewind(Epcot, USA)75 Min. The attraction sees itself as a complete “pavilion” experience: on your way to the ride you pass through the Galaxarium, a planetarium-like exhibition about Xandar and Earth, followed by a briefing from the Nova Corps. The ride itself, a story coaster with freely rotating vehicles and a backwards launch, is the finale of a production that started twenty minutes earlier.

The capacity side is remarkable too. Cosmic Rewind opened in May 2022 with no conventional standby queue at all. Access came 836 days in a row exclusively via a virtual queue in the app. For two and a half years, EPCOT's most famous line was one nobody ever stood in; a permanent standby queue only arrived in early 2025. The physical queue existed practically as a story route rather than as holding space, and that on a ride shifting around 2,000 guests an hour with trains of 20.

VelociCoaster: waiting in the raptor paddock

Universal plays the same card at Jurassic World VelociCoaster(Universal Islands of Adventure, USA) in Islands of Adventure, only with teeth. The queue leads straight through the raptor paddock from Jurassic World: past torn-open enclosure fences, animatronic raptors within reach and a planted atrium where even the concrete walls look like Isla Nublar. You're not waiting for a coaster, you're on a paddock tour that happens to end in a quadruple launch. This is exactly what Maister means by “occupied time”: occupied time feels shorter, and anyone busy checking whether the raptor behind the fence really did turn its head isn't looking at their phone every ten seconds.

And then there's what may be the most elegant capacity trick of all: the single rider line. If you ride alone and let yourself be slotted into the odd seat a group of four leaves behind, you cut your own wait drastically and help the park fill every car to the brim along the way. Every otherwise empty seat is wasted capacity; the single rider line is the elegant way of plugging those gaps. A rare case where self-interest and throughput pull in exactly the same direction.

Shows: the quiet capacity helpers

And then there's a trick that doesn't happen at the attraction at all: entertainment as a crowd sponge. A parade, a firework or a stunt show ties up thousands of guests in one go, people who aren't standing in a single ride queue during that hour. The amphitheatre for Fantasmic! at Hollywood Studios holds close to 10,000 people per show including standing room. So a single show swallows an entire wave of visitors: ten thousand people guaranteed to be sitting somewhere other than in front of your favourite coaster for the next half hour. Wait times across the rest of the park breathe noticeably as a result. You can watch this effect live in the charts on park.fan: during the big evening show, headliner wait times measurably dip, and anyone who's already seen the show rides at the best value of the day in that hour. Moments like that are exactly why we update wait times by the minute.

Capacity and throughput: the mathematical secret

Throughput simply means how many people an attraction actually moves per hour. That number decides how long the queue really is.

The basic formula, known in the industry as THRC (theoretical hourly ride capacity), is simple:

Throughput = seats per dispatch × dispatches per hour

The word “theoretical” is meant seriously. The calculation assumes every seat is filled and nothing jams; it knows nothing of the family deciding at the train who sits next to whom, or the one loose shoe holding up a dispatch. Real, operational throughput is therefore practically always lower. The formula gets interesting because parks combine the two factors completely differently, either on the buffet principle or the tapas principle. Two Orlando rides that delivered almost identical numbers on paper show it perfectly:

Example A, the buffet. The Incredible Hulk Coaster(Universal Islands of Adventure, USA) at Universal Islands of Adventure goes for large portions: eight cars, four people abreast, making 32 seats per dispatch. Dispatching roughly once a minute lands the ride at a theoretical capacity of 1,920 people per hour (32 × 60). One train swallows more people in one go than fit inside some dark rides at once.

Example B, the tapas bar. The counter-model ran right next door until August 2025. Hollywood Rip Ride Rockit(Universal Studios Florida, USA)Closed at Universal Studios Florida sent small cars with only 12 seats onto the track, but what felt like every few seconds. To reach a comparable ~1,850 people per hour, seven cars were circulating at once and the station had to dispatch nearly three times as often as the Hulk. Same result, completely different operating concept. (The ride is history now; its plot is becoming its successor, Fast & Furious: Hollywood Drift.)

Block sections: why not just “more trains”?

Seven cars at once raises the question of why they don't eventually pile into each other. The answer is the block system: the track is divided into sections, and only one train may be in each section at a time. The controller only releases the next train once the block ahead is clear. Every block therefore needs a point where a train can come to a complete stop if necessary.

So more trains don't automatically mean more throughput. Without enough block sections the trains back up before the station, and enthusiasts know the dreaded “stacking”: the train sits in the final brake, the riders wave at the station, the station waves back, and nobody moves. The bottleneck is almost always the dispatch in the end. How fast are all the restraints checked and the train back out?

How seriously modern rides take this is shown by Voltron Nevera at Europa-Park. According to the Mack Rides manufacturer factsheet, seven trains run there at once, with a target dispatch of one every 36 seconds. That makes 1,600 guests an hour, even though a single train seats only 16. Frequency beats size.

Standing still is the enemy: rolling launches and rolling stations

The latest escalation of this thinking: trains that don't stop at all. Voltron uses rolling launches, where trains aren't halted before the launch and then fired off, but accelerated as they pass through, “flying” like a relay handover. To let the LSM motors sustain that on a 36-second cycle, Mack simply gave them four rows of stators instead of two. Every stop avoided means the block clears sooner, the cycle holds, and the queue keeps rolling.

And Walibi Holland applied the same principle to the station in 2025. On YOY, Europe's first single-rail duelling coaster, where you choose between the sister tracks YOY Thrill and YOY Chill, the trains don't stop in the station at all any more. They roll through at walking pace while guests board: a rolling station, essentially the omnimover principle applied to roller coasters. With only eight seats per train (in single file, as though on a very determined bicycle), every second the train isn't standing still counts.

Little's Law: the formula behind every wait-time display

The formula that ties both sides together is one of the most elegant results in queueing theory. The MIT professor John D. C. Little proved the relationship in 1961 that is known today as Little's Law:

L = λ × W: the number of people waiting (L) equals the arrival rate (λ) times the waiting time (W).

For a park visit you simply rearrange it:

Wait time = people in the queue ÷ throughput

If 640 people are queueing for the Hulk and the ride handles 1,920 an hour, you wait 20 minutes (640 ÷ 1,920 = ⅓ of an hour). Those same 640 people in front of a ride with 800 capacity? 48 minutes. The elegance of Little's formula is that it holds for any stable queue, no matter how irregularly guests turn up. And that exact calculation sits, in refined form, behind every wait-time display. Parks estimate the number of people waiting and divide by current throughput, or they measure the time directly, for instance with timing cards a guest receives at the queue entrance and hands back at the station.

The formula also explains why the same queue length can mean completely different waits on two different days. If a ride runs two trains instead of three today, λ falls, and W rises immediately without a single extra guest being in the park.

And the display at the entrance, the one that lies to you? On top of the precisely calculated number, a generous safety margin tends to get added. That isn't sloppiness, it's psychology. The display has to do two contradictory things at once. It should give you a reliable basis for a decision, and it shouldn't disappoint you. A display that underestimates produces a stream of people at the exit who feel cheated; one that overestimates sends a few guests away but makes everybody else happy. Parks systematically opt for the second error. If you budget 60 minutes and board after 45, you leave the station a winner. Peak-end rule, remember: the experience ends better than expected, and that's exactly how it gets filed away. So the display really does lie, but it lies in your favour.

Let me talk shop for a moment, because this is exactly where park.fan lives. Our live wait times show you by the minute what L and λ are actually doing, and when our AI model forecasts wait times up to 365 days ahead, at its core it's modelling nothing other than those two quantities: demand (how many people want to ride this thing today?) and throughput (how many can it move?). Little would probably be astonished at what his formula gets used for these days. How often our forecasts match reality is on the Fancast page.

Why Peter Pan's Flight always “escalates”

Why does Peter Pan's Flight, of all things, a leisurely dark ride of 1955 vintage, no coaster, no thrill, sit at a near-permanent 45+ minutes on the board in practically every Disney park in the world, from Disneyland Paris to Orlando?

There are three reasons for that, and all three of them are arithmetic:

1. Pirate ships are not mass transit. Peter Pan's Flight in Paris sends 16 galleons out over London, and each one holds, generously counted, five people. That's the carrying capacity of a mid-sized lift, spread across an entire night sky. Unofficial counts arrive at roughly 1,200 guests per hour, and the original version at the Magic Kingdom manages only around 800. A single Hulk train carries twice as many people per dispatch as Peter Pan has ships. And for comparison in its own house: Pirates of the Caribbean right next door swallows nearly four times as many with its large boats.
2. Saturation from breakfast onwards. Once demand reaches maximum capacity (saturation = 1.0), the queue grows linearly with every additional guest. It can only shrink again once fewer people arrive than the ride can move. With such a low capacity ceiling, that point isn't reached at lunchtime but roughly when the second coach pulls up.
3. The queue as a seal of quality. Visitors read a long line as proof that the ride must be worth it. The same logic that has us picking the holiday restaurant with the fullest terrace. So everybody joins all the more eagerly, and the wait becomes a self-fulfilling prophecy.

The phenomenon is measurable worldwide, incidentally. Here are the live wait times for the Orlando version in the world's busiest park, straight from our data:

Europe waits differently: Phantasialand and Europa-Park vs. Orlando

How much demand and capacity determine the wait-time level of an entire park is clearest in a comparison of heavyweights, Europe's big parks against the Orlando giants.

Orlando plays in a demand league of its own. The Magic Kingdom is the busiest theme park in the world, and both Disney and Universal pull guests from every continent. On top of that comes a factor Europe barely knows in this severity: over there you can buy your way past the queue. Lightning Lane and Express Pass sell part of the capacity to paying guests, and every express ride is one the standby queue doesn't get. Headliners like Avatar Flight of Passage therefore regularly sit at 60 to 120 minutes despite massive hourly capacity.

Europa-Park is the counter-model. Germany's largest park, and after Disneyland Paris the busiest in Europe, spreads its roughly six million guests a year across thirteen roller coasters plus dozens of themed rides. That sheer volume of parallel capacity works like a pressure relief valve: demand disperses, and hardly any queue runs into permanent saturation. Which is why even busy days in Rust rarely feel like Orlando, and only new arrivals like Voltron regularly break the 60-minute mark.

Phantasialand, on the other hand, is the extreme case in the other direction: one of the most compact large parks in Europe, with few attractions but extraordinarily elaborate theming. The load concentrates on a handful of headliners, and when everyone wants Taron on a holiday Saturday (greetings, I'm usually one of them), saturation arrives just as fast as it does for Peter Pan in Paris. A small audience doesn't protect you from long queues when only a few attractions can absorb the demand at once.

The numbers are almost funny. Taron handles around 1,200 guests per hour, which is almost exactly what Peter Pan's Flight in Paris manages. The difference isn't the ride, it's what the demand around it can spread across. In Paris, Peter Pan stands among dozens of alternatives. In Brühl, on a holiday Saturday, a substantial share of the park has the same single ride in mind at the same time. And Voltron at Europa-Park, at 1,600 an hour, is only a third higher, yet keeps its queue far shorter because twelve other coasters next door siphon demand away. A capacity figure on its own says very little. What matters is its ratio to everything else that's open.

Those profiles are exactly why every park page on park.fan shows long-term statistics alongside the live wait times. Because “busy” is relative: 45 minutes is a bad day in Brühl and a gift at the Magic Kingdom. Look at the same data set for both parks, typical wait times by month and weekday, from the last two seasons.

First Phantasialand in Brühl, the compact extreme case: when a number goes up here, it goes up steeply, because everything concentrates on a few headliners.

And now the Magic Kingdom in Orlando, the busiest park in the world: a higher base load, but spread across dozens of attractions. Note how different the “typical” minutes and the seasonal curve across the year already look.

Patterns like these are exactly what our model learns from when working out whether a visit is worth it.

Closing the gap achieves nothing, and actually slows the queue down

A small experiment for next time: you're in the queue and a two-metre gap opens up in front of you. What does your body do? It closes it. Instantly, reflexively, as though somebody else would otherwise steal the gap. And that is, with respect, completely pointless.

Because Little's Law from a moment ago says it plainly: your wait depends on the throughput of the station right at the front, not on the distance to the person ahead of you. Whether you press up against them or leave two metres of air changes your position in the line by exactly zero places. You move two metres; you arrive at the front not one second sooner.

Worse still: collective shuffling forward makes the queue measurably slower. It's the same physics as the motorway jam that seems to appear out of nowhere. The physicist Yuki Sugiyama sent 22 cars around a circular track with the single instruction to drive at a constant speed and equal spacing. Within a few minutes a stop-and-go jam formed with no bottleneck whatsoever, running backwards through the column. MIT mathematicians call these self-sustaining waves “jamitons”, because they behave like detonation waves. In dense crowds exactly the same thing happens, and there such backward-running waves are even a feared early warning sign.

The culprit is start-up lost time, which everyone knows from traffic lights. When they go green, the whole column doesn't move off together: each driver reacts around a second after the one in front, and the rearmost car starts noticeably later. Every time your queue jerks forward, the same staggered reaction time evaporates. Twenty jerk cycles times sixty people waiting adds up to a surprising amount of nothing.

If everyone simply walked on evenly and slowly instead of standing, closing up and standing again, the queue would move more fluidly and on average faster. Less jostling, more flow. Traffic research knows this as the “faster-is-slower” effect: pushing harder at a bottleneck lowers throughput because everyone wedges into everyone else, a result Dirk Helbing demonstrated in Nature in 2000. The one condition: you have to stay below the critical density. And that's exactly why keeping your distance helps more than closing up.

On the vehicle side, parks solved this exact problem long ago. The Haunted Mansion's omnimover and YOY's rolling station never stop, no halt, no start-up lost time, maximum flow. Only the guests in the queue in front are the last stop-and-go system nobody has designed away yet. Until then, the lever that helps most is the one you already hold yourself: come on a day when the queue never runs into a jam in the first place. Which day that is, the calendar of best days will tell you.

Modern solutions: virtual queues and double stations

Because physical capacity can't be raised indefinitely, parks increasingly move the waiting to where it doesn't hurt: onto your phone.

Europa-Park introduced its VirtualLine at the start of the coronavirus pandemic in 2020 and kept it, because it simply works: through the park app you reserve a free time slot for attractions like blue fire, Wodan or Voltron. Disney ran mega-openings like Rise of the Resistance or indeed Cosmic Rewind entirely on boarding groups for years. The principle is always the same: you queue digitally while you eat, shop or watch a show. The queue still exists, it just happens without your legs. The wait doesn't disappear, but it no longer eats your day, and the park spreads demand across the hours in a more controlled way.

On the hardware side, parks work in parallel on dispatch speed, with three classic patterns:

• Separate unload and load areas, so the train doesn't have to wait for people to get off before the next group boards.
• Double stations, where the track forks into two parallel platforms before the station: one train is loaded while the other is dispatched. A trick Vekoma's flying coasters needed because of their long loading times, and one Efteling also uses on the water coaster De Vliegende Hollander.
• Two complete tracks: Disney's Space Mountain at the Magic Kingdom is the most radical example. Two interlaced, mirror-image tracks (“Alpha” and “Omega”) run in the same building with their own stations, which practically doubles the system's capacity. Effectively two roller coasters passing themselves off as one, and most guests never notice. The model was the 1959 Matterhorn Bobsleds at Disneyland, the world's first coaster with tubular steel track, also with two courses.

And the premier class of throughput manages without a station entirely. On the omnimover, Disney's continuous vehicle chain of the kind the Haunted Mansion uses, the belt never stops. Boarding and alighting happen on the moving vehicle, and the attraction swallows over 3,000 guests an hour that way, more than many a mega-coaster, with a vehicle chain that has been doing its rounds uncomplainingly since the sixties and never asks for a break.

Conclusion: the display tells only half the story

The next 60-minute queue won't get any shorter for it. But it reads differently. Three questions remain, and park.fan has an answer for each:

• How long is the queue really, right now? That's what our live wait times are for: over 200 parks, 7,000 attractions, by the minute. Little's Law in real time, without you having to count the 640 people in front of you yourself.
• Is that a lot or normal? The long-term statistics for each attraction tell you, because 45 minutes is either an annoyance or a lottery win depending on the park.
• And do I have to queue at all? That's usually decided when you pick the day. Which is what the calendar of best days to visit is for, with forecasts up to 365 days ahead, for Europa-Park for example:

The only thing we can't do anything about is the psychology. Whether the hour feels like an hour or like a first act is decided by pre-shows, theming and the peak-end effect, and by the parks that have mastered that craft. So the next time you're standing in a lovingly themed queue and the time passes strangely quickly: that's no accident, that's design. Enjoy it, you're part of a rather magnificent illusion. And whether it's worth it, the data will tell you beforehand.

— Patrick

P.S.: Yes, I still queue for Taron anyway. But now I know to the minute how unwise that is, and I have it in writing that my sense of time is exaggerating by 36 percent while I do.

---

Sources & further reading

• David Maister: The Psychology of Waiting Lines (1985)
• Katz, Larson & Larson: Prescription for the Waiting-in-Line Blues: Entertain, Enlighten, and Engage (Sloan Management Review, 1991)
• On the ~36% overestimation (Hornik 1984): Consumer Perception and Evaluation of Waiting Time: A Field Experiment (Journal of Consumer Research)
• Alex Stone: Why Waiting Is Torture (New York Times, 2012)
• Daniel Kahneman et al.: Peak-end rule
• John D. C. Little: A Proof for the Queuing Formula L = λW (Operations Research, 1961) · Little's Law explained
• On phantom traffic jams: Sugiyama et al., Traffic jams without bottlenecks (New Journal of Physics, 2008) · MIT “jamitons”
• On the “faster-is-slower” effect: Helbing, Farkas & Vicsek, Simulating dynamical features of escape panic (Nature, 2000)
• Technical data: The Incredible Hulk Coaster, Hollywood Rip Ride Rockit and YOY on RCDB · Mack Rides factsheet for Voltron Nevera · The tech behind Voltron (Coaster101) · YOY with rolling station (Freizeitpark-Welt)
• Capacity deep dives: THRC vs. OHRC at Disney World · Flight of Passage facts at TouringPlans
• Europa-Park: VirtualLine official