FastPass+ Is Lowering Your Wait at Popular Disney World Rides
In April we took at look at How FastPass+ Is Affecting Your Wait In Line At Disney World. We saw a small increase in standby wait times for secondary attractions like Spaceship Earth and Pirates of the Caribbean, and a small drop in standby wait times for super-headliners like Space Mountain and Expedition Everest.
Here’s an update, using 278,000 wait times from Walt Disney World between February 1 and June 19, 2014:
Attractions shown in green have seen their average standby wait time decrease with the introduction of FastPass+. Attractions in orange have seen an increase. For the attractions in white, we don’t have a 95% confidence level on the change in standby waits, so we can’t yet conclude anything about them (the confidence level is 89.25% for Kilimanjaro Safaris and 73.82% for Kali River Rapids).
Our analysis takes into account the approximately 8% increase in Disney World attendance since the beginning of 2012. We also took into account that some holidays, such as Easter, fall on different days each year. Our analysis compares holidays to each other, instead of simply comparing calendar dates across years. See below for more details on both topics.
FastPass+ Is Distributing Guests More Evenly Around the Park
Take a look at the attractions whose waits have dropped with FastPass+, and you’ll see mostly headliner attractions. Most of the attractions whose waits have increased slightly are secondary rides, like Living with the Land, Jungle Cruise, and Maelstrom. We think that more people are aware of these attractions now, because they’re being offered in My Disney Experience, and that’s helping Disney move crowds around the park.
Another interesting observation is that 3 of the 4 attractions with the biggest increase in wait times are high-capacity ride: Spaceship Earth and Haunted Mansion have continuous-loading ride systems that handle more than 2,000 guests per hour, while Pirates of the Caribbean can serve more than 2,500/hour when it’s running at top speed. It makes sense for Disney to dedicate a big chunk of these rides’ capacities to FastPass+, simply because they’ll be able to handle so many people per hour.
Adjusting for Increased Attendance
Walt Disney World attendance increased about 2.2% in 2012, another 3.3% in 2013, and we estimate it’s up about 3% in 2014 – or a bit above 8% since the beginning of 2012. To put that in perspective, if the average wait at Soarin’ was 60 minutes at the beginning of 2012, it’d be about 65 minutes adjusted for 2014’s higher crowds. When we’re looking at the impact of FastPass+ on standby wait times, the first thing we have to do is factor out the higher attendance.
For this analysis we’re comparing standby wait times from February 1 through June 7, 2014, with data from the same months in 2012 and 2013. To factor out the 8% increase in attendance from 2012 to 2014, we’ve increased 2012’s wait times by 8%. To factor out the 3.3% increase in attendance from 2013 to 2014, we’ve increased 2013’s wait times by 3.3%. (Disney doesn’t release official attendance figures, but even if our estimate of attendance increases is off by 1-2%, the results don’t change significantly.)
Adjusting for Movable Holidays
For each attraction, we’re comparing the average wait on each day between 10 am and 5 pm. And to make things as equal as possible, we try to match the wait times from 2014’s holidays with the same holidays in 2012 and 2013. So while Presidents’ Day was on February 20 in 2012, February 18 in 2013, and February 17 this year, we compare all of the wait times from all Presidents’ Days together. We do the same for other holidays, the Disney World Marathon days, other events, and then everything left over.
The t-test: How To Compare Before and After FastPass+
Suppose we flip a coin 10 times. You’d expect it to land on heads about 5 times and tails about 5 times. But we wouldn’t be surprised if it came up heads 6 times and tails 4 because of simple random chance. And if we flipped the same coin 1,000 times, we’d expect it to turn up heads about 500 times and tails 500 times, but we wouldn’t be surprised if it was 501 heads and 499 tails, or 499 heads and 501 tails, or anything around there.
But what if the coin came up heads 600 times and tails 400? If it’s a fair coin, what are the chances that we’d see that kind of difference from the 500 heads/500 tails average we expect?
A common way that statisticians answer these questions is with a t-test. A t-test measures two groups and tells you how likely it is that that the groups are really different. In our case, the two groups are these:
- Posted standby wait times before FastPass+
- Posted standby wait times after FastPass+
In the example above, there’s less than a 1% chance that 1,000 flips of a fair coin will result in 600 or more heads. So if that happens, you can be pretty confident that you’re not dealing with a fair coin.
In the chart above, the column labeled “Pr > |t|” is the probability that the result we’re seeing is due to random chance. So for Space Mountain, the odds of you seeing a 13.2-minute drop in wait times due to random chance is less than 0.01%. In other words, we’re more than 99.99% confident the result isn’t due to random chance.
For more information on FastPass+ and how to optimize your use of FastPass+ check out our FastPass+ Tips Page.
75 thoughts on “FastPass+ Is Lowering Your Wait at Popular Disney World Rides”
Excellent post! However, what I’m wondering, and I don’t know if this is quantifiable or not, is if FP+ has reduced the total number of attractions seen per day by those who took full advantage of the old FastPass system, and gathered maybe 5-6 per day. I suspect now that you can add more than 3 FP+ options once they are exhausted, this isn’t the case. But, when I compared two identical offpeak times we visited, before FP+ and after FP+, it seemed as though we did less rides overall than before (mostly because we don’t do standby waits of more than 20-30 minutes *offpeak season*). I’m just wondering if this was my mind playing tricks on me or if there was a way to know if we were actually doing less or not.
Also, I’m wondering if there’s a way to know if the decreased wait times are due to less people visiting some of these attractions, even though park attendance may be up.
It’d be hard to get those data for pre-FP+ attraction visits, but our statisticians would probably agree with you that people are re-riding the headliner attractions less b/c of the difficulty in getting repeat FPs for some headliners.
And this, to me, is the great brilliance of the new system. I love the fact that people can’t re-ride headliners over and over again, exploiting the old system like they used to do.
As always great work Len….
I just showed this article to my lovely wife Keryn who is a quantitative researcher (she doesn’t make me call her ‘Dr’ that often) and I just got a very excited 20 minute explanation of the math behind your results and how impressed she was 🙂 (Considering the rants I have heard regarding reliability math in various papers in the past that is high praise indeed 🙂
(For some reason she wasn’t as excited when I tried to explain how the touring plans as a version of the Traveling Salesman algorithm… maybe I shouldn’t have shown her some example code from an old mapping program we did a few years ago 🙂
Very interesting and super helpful as ever, although I think the spreadsheet might have been better with an Excel man in the middle 🙂
Excellent Post! I never go to Disney without using Touring Plans! I just subscribed to Universal Orlando Touring Plans! We’re going to take our first 4 day trip there later this year. And I recommend Touring Plans all the time to others as well!
I’m so excited that someone loves math enough to do all this work for me! I am a Touring Plans devotee and really appreciate that you keep everything so current. I wonder if it would ever be worth it for Disney to try this (or a version) at Disneyland.
Thanks Jodie! I’d be surprised if Disney didn’t roll this out to Disneyland. They’re making more money and reducing wait times for guests.
Very interesting. Of course “your mileage may vary”, depending on the particular day, month, time of day. Still makes one hopeful that standby times will be down somewhat. Curious what happened on the two with P > .05. Were the data just more widely dispersed?
Len, you fantastically-named devil!
Yeah, not sure what’s happening with the Safaris or Kali. We’ll have to look at it more. They’re both outdoor, weather-susceptible attractions, perhaps more so than Everest? That’s a total guess.
Safaris has issues with fastpass plus because of the huge volume of guests that come through, so they had to adjust their process there. Also, ironically, safaris is probably the least weather dependent attraction out there- they don’t close for any weather at all.
Ah, yes…the new queue process that winds out near the train. Good point.
Do the safari trucks really run in the rain? Lightning? I’ve never tried it, and I’d be moderately surprised that they’d put guests in a metal can on an open plot of land.
Ummmm, Len, you’re safest in a metal can in a lightning storm. It’s called a Faraday cage. Charge travels on the outside of the metal. It is this property of a metal car (not the rubber tires) that make you safe in a lightning storm in your car.
Lee: A car can sometimes make a faraday cage if hit by lightning. A safari vehicle will not, for the same reason that a convertible won’t: it’s not a closed box. Also the faraday cage doesn’t help if you’re touching metal connected to the cage, which, from my memory of the safari vehicles, at least the people on the edges certainly will be.
No clue for Kali, but while Safari is not weather dependent, it is animal dependent. Don’t forget, the animals have the right of way, and sometimes those dang giraffes just won’t get out of the freakin’ way!!! That’ll jumble things up in an instant. I once sat in a jeep for nearly 7 minutes waiting for one to move. Sure, I got amazing pictures, but none of the other jeeps behind us were moving either, and therefore created a blockage that can’t be accounted for!
As a destination marketing executive with a love for all things statistics and research based, this post makes me giddy. Love all the information you put into it and it really shows the details. Maybe MM+ is worth the investment they made if the reduced wait times continue to occur.
This post has me curious about some other secondary attractions in the Magic Kingdom. How is it for It’s a Small World? I’ve heard the lines are a lot longer for that one.
We can check. Might take a few days.
There are other variables not mentioned that makes me think it’s a wash. Attractions not open yet in new fantasy land which would dilute rides park wide once open. Major refurbishments like Test Track inflating numbers before. What does the tiered system do in epcot that is causing second tier attractions to have longer waits. Also the average number of attractions per day per person wasn’t factored in. Could not having a fastpass for certain rides completely discourage people from even attempting the ride. Overall ridership would be down resort wide causing lines to be shorter.
What I meant was 2012 fantasy land as opposed to 2014 fantasy land.
Yep. Good point. We’re looking at how the new attractions are affecting other waits. We’ll know more once 7DMT gets through its honeymoon phase.
We’re looking at Test Track’s influence now.
Where do you get your data from to make these comparisons? What set of dates and times are you using to make these comparisons?
Hey Patrick- Data come from our users, our staff, and Disney.
Dates are February 1 – June 19, and times are 10 AM – 5 PM.
Thanks for your reply, but why aren’t you disclosing your source data? Are you privy to data that isn’t available to the public?
These kinds of charts which don’t disclose source data is incredibly frustrating.
It was an oversight. We disclosed it in the previous posts.
Thank you again for your response. Can you please provide a link to the previous post(s) disclosing the source data used to create this chart?
We mention the use of our user data, staff data, and Disney’s data in the second section.
If you’re looking for raw data – the numbers themselves – send me an email and we can talk about it.
Was the attendance increase factored in as resort wide or was each individual park factored separately. With no new major investments in the other 3 parks could Magic Kingdom be on a different scale. Also added days of special events like food and wine would increase attendance numbers at Epcot but not increase ridership because those guests largely are not riding during those events cause across the board numbers to skew!
We took each park’s attendance increase and averaged it out across the entire resort. I’ll ask the stats guys to check whether isolating the parks would make a difference. Good point.
We re-ran the analysis using each park’s separate increases. Only minor changes to the results: 1 attraction (Soarin’) changed from a small decrease to “can’t be certain”, and most wait times went up or down about 30 seconds. I’ll post an updated chart, probably tomorrow.
Very cool. Thanks for checking it!
In essence all that has been accomplished is returning to the ticket system without actually doing individual ride tickets. People used to go to the secondary attractions more because they only had so many E tickets in a book…and you wanted to make sure you used all of your A, B, C, and D tickets. After all, you paid for them! Still, I think things played out better overall in the pre-FastPass days. Sure, we Disney fans loved original FastPass. But lets be honest, it was because we knew how to work the system.
A world with no FastPass makes the parks an equal playing field for all guests. As a person who does not often do the big week-long, planned out for months, Disney vacation (I do more last minute trips for a day or two), why am I valued LESS? Shouldn’t I be able to have an equal chance of getting a FastPass for the latest attraction (SDMT, or dare I say, Anna and Elsa)? The same can be said for the situation with dining reservations. Why is my money not as valuable? Why don’t I have the chance to eat dinner at Be Our Guest?
Anyway, its the world that we now live in and we are stuck with it…but I can’t help but reminisce about those days before the FastPass infection arrived.
Your money may be slightly less valuable to Disney. The long-range planners will have paid for at least a portion of their trips before you, so there is the time value of money. Also, it may be that week-long guests spend more per-person per-day than short-stay guests, which makes them more valuable to Disney.
I would venture a guess that Disney is happy to take anyone’s money. And regardless of the various systems in place, my responsibility to get the most for my vacation dollar depends a great deal on figuring out all my options and making smart choices.
Any crowded venue will always favor the early-riser, the light traveler, the smart planner. My ticket doesn’t guarantee me much of anything aside from getting in the gate. (Read it some time. Standard boilerplate.)
Yes, as much as I appreciate the analysis here, and appreciate a big-picture explanation of what the heck MyMagic+ is supposed to accomplish, I agree with you that I feel like my money (as a day-tripper) is less valuable to Disney, and they’re more interested in maximizing the maximizers who stay longer and spend more money in more ways at the resort. Knowing more and more about the behind-the-scenes management of ride capacity and guest experience makes me feel, well, managed, rather than welcomed. While I’m sure the management of the guest experience was always there, this new system makes it feel more utilitarian, business-like, and less magical. Disney’s always been a business; I realize that. But the business of making people happy is not easy or simple. As luck would have it, my best Disney experiences came after the ticket books and before MyMagic+. I’m not critical of those who are enjoying the new system. But I think there are a whole lot of us out here who may think it’s time to find our happy experiences elsewhere.
A yup! We are doing a 4 day 3 night at Universal Orlando late fall. Subscribed to Touring Plans Universal!
Len, you’ve somehow managed to take math and make it relatable and interesting! 🙂
I find myself fascinated by the level of depth you and the researchers put into your work: mixing Disney geekdom and math geekdom–and I say that with the utmost respect.
I’m curious. With all the hand-wringing that’s been going on since FP+ testing first started (and I still see on some forums), it seems like time has become the system’s friend. Kinks and hiccups still exist, but that is to be expected. As both Disney and its guests become more familiar with the system (how to set it up correctly, how to troubleshoot more effectively) I see a continued trend toward a smooth experience.
I believe the people who have the biggest issues with the FP+ are the ones who knew how to manipulate the legacy system the best. Confession: I played the system a few times myself, so no judgement on right or wrong here whatsoever. The idea of spreading people out is a good one, except for those of us who may not be used to seeing larger crowds at “B” level attractions. That may be adding to veteran guests’ frustration level. But, what do I know?
I appreciate seeing the numbers that illustrate what is actually happening in the parks to go along with park visitors’ description and perception.
I don’t know if these numbers are due to FP+ so much as they are due to the new DAS card. People not having unlimited FP use (per the old GAC system) would shorten lines more than anything.
I’ve heard that GAC abuse was a driver of wait times at some attractions. Let me see if the stats folks can factor that in. Good point.
I remember the old GAC system with no fondness. There was one time I was standing outside Toy Story, waiting for my group to arrive. In less than ten minutes I counted fourteen groups come through with GACs. In not one group was a person obviously disabled. Now the normal caveats about people having non-visible disabilities must apply, of course. But there’s no question in my mind GAC abuse was a huge driver of wait times at headliner attractions.
GAC abuse is very much a big problem with wait times. I’ve hung out in the parks with plenty of different people who would fake a disability in order to get the card but the problem is that legally those in guest relations are not allowed to ask for proof of a disability and can only rely on the guest’s word.
If this is all due to redistributing guests, do you think this will incentivize Disney to build more high capacity rides and be able to better fill FP+ demands? I wonder if they would build Toy Story Mania now if they had known its requirements on FP+.
Hey Len – I think I heard on a podcast that you were recently in the parks yourself. Are you still seeing problems with the FP+ lines backing up?
I saw back ups at Thunder Mountain & Space Mountain the weekend of June 14th. Space Mountain backups are still pretty long whenever I see them.
It wasn’t as bad as before, and I think it’s getting better. They still have some issues to work out there, especially with guests who just don’t understand how it all works.
One of the things I’ve seen them do when FP+ return lines are long, is just accept 1 person’s FP+ band from an entire group. That unclogs the line pretty quickly.
I was just visiting June 9 – 12 and it was my first time using fast pass+ and it was a complete bust. We only got to ride one ride at magic kingdom. We had to wait 30 min in line at the fast past kiosk upon entering the park. We arrived right when the park opened and by the time we picked our rides the ones we wanted were full and the others were in the late afternoon. The stand by lines we’re up to 45-90 min waits. Not to mention space mountain and pirates were shut down for the day due to mechanical reasons. It was just highly disappointing
As an FYI, everyone is able to book FPP in advance now, so you shouldn’t have to wait at the kiosk at opening anymore.
I never thought the SAS course I did last week would help me read TP articles! 😉
Woo! Our stats guys are fans of SAS.
As they should be 😀
I’m just building up my skill set so I can come and work for you Len!
Payment via Dole Whips please.
Lots of posted jobs in celebration fl and around WDW for jobs in the field of “big data” for people with SAS experience. Disney is definitely using the RFID data gathered from the magic bands. Whether or not it is put to actual good use for guests is yet to be seen. But they are looking for SAS experience.
The mathematician in me struggles a little bit with the use of a t-test since the underlying data is rounded to the nearest 5 minutes so distribution is questionable :-p But I still love that you guys took the time to do this and good to see it does what it says on the tin!
Wilcox-signed rank in that case, right? continuous non-normally distributed data.
If that’s the test, we might be able to run it. Let me ask the stats guys what the issues might be.
SAS can certainly run a Wilcox-signed rank test; there’s all kinds of statistical tests in proc univariate. But one has to know the appropriate test to use. And only then does it give statistical significance. Given the sample size, once a large enough sample is used statistical significance tests are always passed.
The real issue is that this is only a correlation and not causation. (As with all “economics” studies.) Controls were only used for increased attendance and matching “like” holidays. There could be confounding factors. Obviously there is now way to do some kind of double blind study or truly isolate causation. So, correlation is the next best thing.
Here’s another possible confounding factor: With the introduction of the DAS system, anyone with a DAS card must “return” if the standby wait time is more than 10 minutes. It seems like rides are now a “minimum” of 15 minutes even if it’s less than 10 minutes.
Were the touring plan time estimates used or the inflated posted Disney times used?
And this was only a comparison of 10 am and 5 pm. A new statistic like total aggregate wait time throughout the day or the total aggregate wait time in an ideal touring plan comparing before FP+ and after FP+ might be more ideal.
But just because an analysis can be complicated, it doesn’t mean it’s better.
I love that we’re able to see somewhat definitely that post FP+ standby times are down compared to the expected attendance adjusted times.
I view this article as definitive proof that FP+ reduces wait times.
The real problem with FP+ is that the aggressive park visitor is not as able to use FP, and the most zealous Disney fans, insiders, and touring plans readers are the ones being disadvantaged. I wonder if that will translate into more abuse of DAS.
Thanks Ron! I like the idea of using total aggregate wait time in a touring plan.
We used Disney’s posted waits, inputted from our users, staff, and My Disney Experience.
Agree on your “complicated analysis doesn’t mean it’s better.” 🙂 At the end of the day, we’re looking at wait times that we know are sometimes manipulated higher on purpose. And let’s face it, a lot of these times are set by overstressed staff without a lot of help.
A major determinant in standby times is the ratio of fast past guests/standby guests admitted per hour. I suspect the ratio has been lowered either by design or effect of spreading guests throughout the park. I recently saw a long line for Peter Pan which, by experience, looked to be at least a 60 minute wait. The posted wait was 30 minutes. It is reasonable to conclude that since the ride capacity is constant, cast members were admitting more standby guests per hour.
Great post. I was wondering if you have any statistics on ride lines at park opening. For example in the old days for DHS I would make a b-line to TSM for FP’s and have the family head to RR. I am wondering when I go in Feb if I should FP+ TSM or RR and run to the other. I am thinking the line for TSM might be shorter at park opening since you can FP in advance now days. I plan on checking the wait times during less crowded times this fall to get an idea, but wanted to see if anyone had any info.
Thanks, Len! I’m a Word Girl, but I love a good set of statistics. If I can quantify something (or see it quantified) it’s as good as giving it a name and writing it into a story!
Great article! Thanks for the insight.
1) This is dealing with REPORTED wait times only and not actual wait times. So this is assuming a company that spent billions on a system wouldn’t “round down” their wait times since that is the main metric of success. If you were part of the FP+ team, you might be tempted to round down too. It’s the human factor.
2) Can you please help me understand your coin flip analogy? I don’t understand how the probability is less than 1% that the coin would show one side 600 times. Each time the coin flips, there is an equal chance of either side coming up. So it’s the same probability that there would be 600 heads as there would be 500 heads. This is the gambler’s roulette fallacy. If he sees that black has come up the past 8 spins, he will get on red because “it has to average out”. But in reality, there is still a 50/50 chance of red or black on the next spin. In other words, heads/tails/heads/tails/heads/tails has the same chances of heads/heads/heads/heads/heads/heads. Not logical to the human brain, but true.
Uh, nevermind. This answered my question. http://en.m.wikipedia.org/wiki/Gambler's_fallacy
I’ve been fortunate to enjoy the WDW parks on many occasions. I can honestly say I never once waited in any line that was longer than the posted wait time. More often than not, it was significantly shorter. So, on the contrary, I think Disney would have every incentive to deal with the data honestly. Fudging it would only be eventually outed.
I <3 you guys. Disney + Statistical Analysis = Heaven.
Agreed. If this crew taught math we’d be a nation of genyus, geeniou…smarter people!
Mention of the t-test and ANOVA stuff brings me back to my college psych stats class. I feel young again. Thanks for the memory and the data!
Why does Touring Plans not use my Fast+ selections that I have reservations for? This plan does not use the Fastpass+ reservation for Voyage of the Little Mermaid, for instance. I thought personalized plans would recognize my Fast+ reservations.
The Optimizer won’t use a FP+ reservation if it thinks it won’t save you time. In this case, it thinks that you don’t need a FP+ for VOTLM. You can try getting FP+ for another attraction, and seeing if that helps.
While I get that the tourning plans are trying to tell you that you don’t need a Fast Pass + for an attraction, sometimes the user (me) has a very good reason for wanting to use a Fast Pass + for a certain attraction and a specific time. It would be nice if you could chose to “over ride” the tourning plan and make it incorporate your Fast Pass + selections if you want it to. I apprecaite the suggestions from the tourning plan, but sometimes I like what I have and I want to use it. Thanks!
Ahhhh…I get it. Yeah, try using the ‘Evaluate’ button. It won’t re-order your attractions, so you can put VOTLM exactly where it needs to be in order to use the FP+.
“Optimize” will re-order your steps and make decisions for you. “Evaluate” lets you build the plan, and only tells you how long it’s all going to take.
Let me know how that works. Thanks for using the site!
Ok, I will try that. Seems like you need a third option or something. One where you could imput your three Fast Passes, and lock them in and optimize around the locked in items. You could even lock in things like lunch/dinner ADR’s and optimize around those.
Thanks Len. It is hard to resist change but I know you guys are right.
Len if we got crowd tracker updates with new forecasts should we redo our personalized touring plans?