I just noticed my mid-late August trip is now all 2-3 crowd level days per the Touring Plans app. This is different from last week where they were 4-5. What’s funny is I’ve noticed on the LL calendar for cost (on the Disney app) it is showing the LL per day is getting higher as my dates approach. So I’m not sure who to believe anymore!
2 Likes
Great point. Since they lowered, I was thinking of not purchasing the LL for our week in August as well (except for Guardians & Tron) but I don’t know what to think now. But I sure can’t complain about the lower calendar numbers.
2 Likes
I think he’s just using the term “feel” to convey the information to lay people. I think there is math behind it that he is avoiding digging into due to the technical nature.
I think what he’s saying is that in his hypothetical, 45 minutes may have been the average wait on a day that was in the 50th percentile of waits, but now its in the 70th percentile:
2 Likes
This was exactly my dilemma. I’m still teetering on it, as my LL/ILL purchases for my whole trip in August are about $500, which isn’t cheap. So if I can wean that down at all, I will try… Although I’ll do ILL for Guardians & Tron no matter what (I can’t deal with those lines in particular)
2 Likes
Honestly, I have never found TP’s crowd calendar to be terribly useful. What a crowd feels like versus wait times doesn’t always correlate (although ideally do)…but even if they did, the values are rather arbitrary designations. I would like to see much more useful numbers such as what each crowd level indicates overall average wait time across all rides in any given park. Then, I see a “3”, for example, it would tell me that the average wait is X versus Y.
We often hear it said that if two crowd levels are within 2 or 3 values of one another, that you won’t really be able to tell a difference. If that is the case, what value are the values? I get that the predicted number could be wrong. But if one day is a 5 and another is an 8, I should be able to FEEL the difference.
The strength of TP is in providing details ride wait time predictions for a given plan on a given day. But as crowd calendars go, I have typically found other crowd calendars from other sources to be more useful.
Maybe this update will address some of that?
1 Like
I agree with this statement 100%… maybe I need to go back to my roots with TP and stop focusing on made up numbers with crowd predictions… The reason I joined this community 11 years ago was to help me understand the park as a first timer & how long it would take to walk from Ride A to Ride B with a slow moving toddler, as well as the length of the actual ride or show. It was really effective when I used the parts of TP that included facts.
Thank you @ryan1 for pointing this out. Sometimes I do think it is all perspective.
Nobody asked for this 
but for fun I decided to put together an illustration of how the “clustering threshholds” might work.
Say that in 2015, TP developed their crowd calendar by measuring the average waits at Magic Kingdom on all the dates within a selected time period. Each date is represented by a dot on this line chart:
After plotting the waits, TP has to decide where to draw the line between a 1 CL and and 2 CL, etc. They want the CLs to be relatively evenly distributed - ideally you want 10% of days in each CL. Some boxes are wider and some are narrower. In some cases, the clustering is such that it doesn’t make sense to make the boxes equal in number of days (for example, in this chart, CL 4 only has one day but the clustering makes it difficult to split up the days in CL 3 or CL 5).
Now imagine in 2025, TP redoes their analysis. But times have changed and the distribution of wait times looks more like this:
Notice that on most days, the wait times are higher and clustering is more severe on those higher wait time days. So dividing the calendar into 10 different CLs, each representing roughly 10% of days, you need to shift the CLs so that there are more on the right side of the distribution and fewer on the left.
If we used the 2015 model, a day with a 30 min average wait would be a CL 4. But in 2025 that corresponds to a CL 2. Yes, that makes the Crowd Calendar less comparable over long periods of time. But it also makes it more meaningful going forward.
Imagine a more extreme example:
What if someday, average wait times are almost always 40-55 min, thanks to the way Disney hypermanages their parks? How useful would it be if there was only one CL 1 or CL 2 in an entire year?
As guests, we have to recalibrate our expectations of what a CL 5 “feels like” because crowd patterns (and average wait times) have changed. It’s either that or every day is a 10 and we don’t really know whether that’s a high-crowd 10 or a low-crowd 10.
1 Like
Not sure I agree with this part of your analysis. (Nice analysis, by the way.) I don’t think it makes it more meaningful to spread them more evenly. I think the numbers should reflect something more absolute and fixed. There is this arbitrary decision for there to only be values 1 through 10. But there is no reason to say there couldn’t be a CL 11, or a CL 12, because we are now seeing average wait times increase. In fact, we know that a “10” can have VERY broad range of actual wait times. A 10 in, say, June (or whatever) is going to look very different from a 10 on New Years Even. So…why call it a 10 at all?
The fact that we are seeing fewer and fewer 1s and 2s will be indicative of the reality of Disney’s crowds across longer periods of time. But if what was once a 5 is now a 3…we have lost any sense of usefulness, because it places the burden on me, as the consumer, to have some kind of intuitive understanding of some algorithmic gerrymandering.
The fact that, in the example you provided, a 2 used to mean 15-20 minutes, then 20 to 30, and finally 20 to 40 makes the 2 completely useless and meaningless.
If we say each numeric bin is a (for example) 10 minute block, then a 1 means 0 to 10 minutes, a 2 means 10 to 20, up to a 10 meaning 90 to 100 minutes. But if we start to see 105 minute average wait times during the Christmas holiday, we now use an 11. It is all very useful.
3 Likes
All valid criticisms. I think there are trade-offs that have to be made and TP has chosen to sacrifice long-term comparability in exchange for go-forward differentiation.
For sure. Ultimately, TP will do what TP wants to do. But it just means I pretty much ignore their crowd calendar.
1 Like
I think that this is a good perspective. I’m “annoyed” that my April dates “increased” but nothing about the actual attendance changed! My week is set in stone already and all that really matters to me is that my TPs are relatively accurate and that Lines isn’t telling me to “ride now” when I shouldn’t.
4 Likes
Ryan I agree. That’s how my non analytical mind works too. Just add Crowd Levels. This new Calendar makes it really hard to figure out and I have studied it. I do understand what they’re saying 1 & 2s are nonexistent in the new Disney time we’re in. I miss the old heatmap that showed more info about when crowds started to drop off in the afternoon or build for fireworks. We were always able to adjust our touring to hit those sweet spots. Now I can’t see that info anywhere. After 5pm touring becomes the wild west, especially on EEH nights.
1 Like