We’re more than halfway through the month of February, so let’s take a glance at the Walt Disney World crowds so far this month:
|Park||Predicted Crowd Level||Actual Crowd Level|
|Magic Kingdom||5.2||5.7 (+0.5)|
|Hollywood Studios||4.0||6.8 (+2.8)|
|Animal Kingdom||4.6||8.3 (+3.7)|
Predictions for Magic Kingdom and Epcot did really well. If you visited those parks in February, you experienced a crowd level within one index point of what the Crowd Calendar predicted on most days.
Hollywood Studios and Animal Kingdom are a different story. In particular, Animal Kingdom’s posted wait times indicate crowd levels nearly double what was expected. Does this mean guests flocked to that park in record numbers during February? We don’t think so, based on three things we’re seeing:
- The rides seem to be operating near 100% capacity
- The difference between actual time spent in line and posted wait time is larger than it has been in the past
- The posted wait times seem to change faster than is physically possible – they’re more “volatile” – than at any time in the past
We don’t think the issue is with ride capacity. We spent several days counting the number of guests exiting the rides per hour at Animal Kingdom. For every ride, the number of guests we counted was within a few percent of each ride’s maximum hourly capacity. So we don’t think the issues we’re seeing are due to staff or ride capacity cutbacks.
Actual vs. Posted Wait Time Differences
Recall that the Crowd Calendar represents the average posted time between 11 a.m. and 5 p.m. at a park’s core set of attractions. Usually, this is a good proxy for the number of people in the park. Even when it doesn’t directly correlate to attendance it is still a useful metric because it represents how long (Disney says) a guest has to wait in line. This month at Animal Kingdom we are seeing patterns in posted wait times that we’ve never seen.
An example of this is DINOSAUR on February 10, 2019:
There is a lot of information of this chart, so let’s break it down:
- The blue-green area with the green border represents the set of posted times observed on that day.
- The red line represents what the Crowd Calendar predicted and the dotted line represents an average day.
- The circular markers represent an actual time recorded by a guest in line.
You can see a similar representation of this data on each of the Crowd Calendar Daily pages like this one for February 10, 2019.
Posted wait time curves like this are unusual. DINOSAUR saw six spikes in posted time throughout the day. The largest spike was during the 1 p.m. hour, where the posted time went from 25 minutes to 90 minutes and then immediately back down to 30. Calculating the observed crowd level on a day like this is problematic because these spikes push the average wait time higher than it probably should be. We predicted an average of 30 minutes for DINOSAUR on February 10, which represents a crowd level of ‘4’ for that attraction. But we observed an average posted time of 41 minutes which represents a crowd level of ‘9’. Between 2:30 p.m. and 3:30 p.m. posted times spiked up from 30 minutes to 60 minutes, then back down to 20; however if you lined up at 2:50 p.m. you only actually waited 30 minutes. (We don’t think this is physically possible, so it’s an example of both problems we’re seeing.)
Here is Kilimanjaro Safaris on February 15, 2019.
We collected several actual wait times for Kilimanjaro Safaris throughout the day on February 15, and they all match the predicted curve relatively closely. However, the posted times are higher. This is a good example of what seems to be happening at Animal Kingdom in February. We see observed crowd levels at the upper end of our scale; however, those levels are not always reflected in the actual time guests are waiting.
The wait times at Kilimanjaro Safaris were also unusual on February 12:
These posted times seem to be legit, but the long waits early in the day are odd. This could be an indication of the attraction operating at a reduced capacity in the morning hours. A 64-minute wait for Safaris at 10 a.m. is unusual – especially for a Tuesday in early February. The capacity seemed to increase in the afternoon, but the wait times never fully subsided until the park approached the closing hour.
Here’s the chart for DINOSAUR on February 7.
We see drastic spikes for DINOSAUR’s posted times on this day, as well. Notice how those spikes represent wait times much higher than an average day. This is surprising considering the other three parks showed normal crowds on February 7 (MK 5, EP 3, HS 4).
We see a similar pattern the day before, on February 6. Neither one of these days has a recorded actual time during the busiest part of day (can’t say I blame our community: who wants to wait 50 minutes for DINOSAUR?!), so we don’t fully know what the wait times were like, but it seems likely that those posted times are inflated. The few actuals we do have show wait times in line with predicted levels and pretty much in line with the average. Let’s look at some other Animal Kingdom attractions.
Expedition Everest saw some strange wait time patterns in the last few weeks. Here is Everest on February 10, 2019.
Wait times throughout the first half of the day were in line with predicted levels, but then at around 3 p.m. the posted time climbed steadily, peaking at 140 minutes! These are wait times we normally see on the busiest days of the year, not on a Sunday in early February. Expedition Everest can have strange wait time patterns because it has a high frequency of temporary breakdowns, but we didn’t see any of those on February 10. It is difficult to find a reason why the wait time would spike so high at the end of the day like this. Did it switch from operating five trains to only two? Clearly, this spike cannot be due to an influx of guests after 4 p.m. – the curve is too steep. Something else is happening here.
Wait Time Volatility
The third unusual thing we’re seeing is large swings in posted wait times, especially large drops within a short time. The chart below shows DINOSAUR’s posted wait times from February 14 as an example:
At 1:30 p.m., DINOSAUR’s posted wait time was 90. Fifteen minutes later, it was 20: Person “A” got in line at 1:30 p.m. and got on the ride at 3 p.m. Person “B” got in line at 1:45 p.m. and got on the ride at 2:05 p.m.
It’s not possible for both of those wait times to be true. It means Person B got in line later than Person A, and yet somehow got to ride first – i.e., Person B passed Person A in line. (Remember, we’re pretty sure the the ride is operating at full capacity, because we’ve counted the number of people leaving the ride.) So one of the posted wait times is wrong.
We see a similar drop starting at 4:24 p.m.: the posted wait goes from 70 to 25 in the span of 6 minutes.
We also see large increases in posted wait times, far beyond what can be explained by normal traffic patterns. For example, the posted wait went from 20 minutes at 2:10 p.m. to 65 seven minutes later. For that to happen, around 1,300 people would have had to walk to DINOSAUR in 7 minutes.
It’s highly unlikely that 1,300 people walked to DINOSAUR in 7 minutes at 2:10 p.m. To put that in perspective, the number of people who line up for Seven Dwarfs Mine Train at park opening was around half that the last time we counted. DINOSAUR isn’t as popular as Mine Train, especially not at 2 p.m. Also, a line of 1,300 people walking to DINOSAUR would be somewhere between 0.3 and 0.6 miles long.
We’ve collected wait times continuously for the last 78 months (since August 2012). If you count the number of these unusual wait time changes across all Walt Disney World attractions, the months of January 2019 and December 2019 rank as the #7 and #2 worst out of those 78. January 2019 was the worst January we’ve ever seen, and December 2018 was the worst December we’ve ever seen. (November 2018 was the worst November we’ve ever seen.)
Effects on the Crowd Calendar
When we observe wait time curves like these, it really messes with our calculation of “How did the Crowd Calendar do?”. Despite seeing crowd levels of ‘3’, ‘4’ and ‘5’ at other parks, Animal Kingdom shows ‘9’s and ’10’s. But are they really ‘9’s and ’10’s? The curve for DINOSAUR on February 10 represented a crowd level ‘9’ instead of the ‘4’ we predicted, but was it really a ‘9’? It is fair to say that Animal Kingdom wait times have been higher than predicted so far in February, but it is also fair to say that those higher wait times don’t seem to represent the entire truth either. These wait time curves are so different from what we normally see, it is difficult to compare with days in the past.
We took at look at the volatility of these posted wait time curves (the up and down-ness) over the years and found that Animal Kingdom attractions are more volatile in 2019 than they have ever been (our database goes as far back as 2006). Why that is, is anybody’s guess. But, it is something we will have to deal with if it continues.
Crowd Calendar Updates in March
When we update the Crowd Calendar at the beginning of March we may see some adjustments for Animal Kingdom based on these patterns. Magic Kingdom and Epcot seem stable so we don’t expect any major changes there, just the usual tweaking. At Hollywood Studios we will see some significant changes. Tower of Terror is going to operate at reduced capacity throughout Spring and Summer. This has an impact on the wait times and therefore also impacts the Crowd Calendar. In addition, we are adding the new Toy Story Land attractions to the list of core rides at Hollywood Studios that we use to determine the crowd levels. This will help stabilize the numbers there.
Got any questions for the number gurus? Theories about the odd spikes at Animal Kingdom? Let us know in the comments.