Expected Crowd Impact Of The June 15 Opening Of Cars Land, Disney California Adventure

15 thoughts on “Expected Crowd Impact Of The June 15 Opening Of Cars Land, Disney California Adventure”

• Yes, I follow and it is an excellent point, thought-provoking in fact. I agree with all of what you say but just to make things more complicated I would add that the wait times do not exactly increase year over year as you describe because the parks grow as well.

  The one thing we are thinking about is using the wait times themselves as a metric. For example, the average peak wait time, the maximum peak wait time or something similar. This solves all the issues you’ve raised, but of course, it causes some new ones. What does it really mean if a day has an average peak wait time of 57.5 minutes?

  Reply

  • Fred, thank you for your reply, for a minute there I thought I was “losing my marbles”, lol. And I agree that attendance/”wait times” do not exactly increase year over year, that is just how they would “trend” if capacity wasn’t added to the parks to account for the increasing numbers of bodies. Yes, it is more complicated (and you would know more than anyone else!).

    I just had a thought of how you might approach it. The good news is that I think the changes on your end might be minimal.

    The major goal is to give people an indicator of “how busy will the park be?”. And the logical next question is, “compared to what”?

    You are actually in a good position to answer back with, “how about compared to your last trip?”, I think. If the guest has been to the park before, they have an idea of how busy “it was then”. They either enjoyed themselves, or not. It was either “too busy then” or it “was ok”. If you used your same algorithms, but encouraged them to go back and find out your “current” rating of how busy it was then, it would give them a benchmark. “Apparently the last time we visited is now considered by TP to be a 5”.

    Basically, a guest really wants to know if their next visit will be better or worse than last time, right? They need to also get the current number for when they want to visit. If that number is the same or lower, coolio! Their next trip should be the same or better than their last one, in terms of wait times. But if the number is higher, they know that it will likely have longer waits, so they might want to consider changing the dates.

    Just a thought. I think the changes technically would be minimal, but you’d have to change the approach that most folks are currently using it. But the fact that you could give them a sense of better or worse than their own previous trips is rather powerful, IMHO, and they’d likely be excited to have that kind of customized information available t them (again, IMHO).

    I do think your idea to get user feedback before implementing any changes is a good one. There are professionals who conduct usability studies, which is the best approach IMHO. That way you can get actual feedback to see if users are successful or not using the product. Sometimes users think they know what they want, but turns out they are wrong and don’t know it! Even gasbags like me 🙂 JK, I’m always right so you should always listen to me. Smile.

    Reply

• Frank:

  Thank you for your responses! I think that Keith and Marc have done a great job explaining why I thought it was based on a 365 day year. In my opinion comparing summer 2012 crowds against 2006 or 2007 crowds is meaningless. I am only concerned with planning a trip in the future. I can’t go back and take a trip last fall on a “1” day. If a 2012 “1” is busier than a 2006 DCA “4”, so be it. I’ll live with it.

  I don’t know how you’re “deleting data” if you only look forward. Doesn’t the “old” data give you insight into your prediction? I just don’t want the old data to have most of the “1s”.

  If you don’t want to adjust your scoring model so that you have equal distribution of digits over the next 365 days, at a minimum you should change the references cited by my prestigious fellow commenters. 🙂

  BTW I am a very frequent WDW visitor but have been to DLR only three times. I wouldn’t expect a 2012 “3” to feel like a slow day when I last visited DCA and it had fewer popular attractions (2003 and 2005).

  Reply

  • Thanks Jim,

    By far, the most common negative feedback we get is that the estimates on the calendar change. If we were to base the calendar on a 365 day cycle, that would make it worse.

    The second most common complaint is that the levels we use don’t match people’s expectations of crowds. Again, this would be worsened by going to a 365 day cycle. Imagine a family who had a great trip in 2011 on a ‘5’ day who select a return trip on another ‘5’ day in 2013 only to arrive to crowds that feel more like a ‘7’.

    The point is that there is no easy way to do this. We have a pretty good system and the vast majority of our guests use our products with great success. That being said, we are always trying to improve and love hearing suggestions from our readers.

    So tell us, how would you do it?

    Reply

    • Thanks for the replies, Fred, was worried this was too old to get your attention. I am aware that the numbers are based on wait times, and not total people in the park. It is clear others are not, and after writing what I posted earlier, I wondered if a simple name change might go a long way. The one that came to mind is “Wait Time Index”, more words than Crowd Calendar, but the same number of syllables. At the very least, a name that reflects that it is based on wait times may make it clearer. Especially in DL, it doesn’t take a whole lot of people for it to feel crowded due to the geographic realities of the park. Admittedly, many of the people you’re dealing with probably don’t consciously grasp that a long line and a long wait are not the same thing.

    • Funny you mention that, internally we have always referred to the numbers in the calendar as the “wait time index” but we felt it sounded too technical. You are right, that is essentially what it is. We are discussing some new ideas for the calendar and one thing I would like to do this time is consult more with our readers for feedback before making changes. That will help us design something that minimizes confusion and maximizes usability.

    • Just for the record, I was also aware that your algorithms were not using attendance numbers, per se, though I wasn’t really aware of the exact internals of what you were doing. But my language was colored by the same language that you use to try to get the points across with your users. In other words, if the thingy is a “crowd calendar”, it is a natural to use a word like “attendance” (even if that is not exactly what is going on under the covers).

      So let me rephrase.

      Are you retroactively revising previous “wait time” estimates for yesterday and all previous days? And because attendance at the parks grows each year, wouldn’t your algorithms tend to skew higher for days in the future versus days in the past (i.e. as time goes on, I would think that the “smallest wait time numbers” will generally be in years gone by, when “wait times” were naturally lower because of the size of the earth’s population).

      You wrote:

      The bottom line is this: The calendar is a rank. A rank of the wait times in the parks. For days in the past the numbers represent the rank that the wait times “were” and for days in the future the numbers represent the rank that we “expect” the wait times to be. I don’t know how to break it down any simpler than that.

      I completely get that. I also understand that you get complaints when the numbers change, and so you are trying to minimize that. The thing is, I think you are still changing numbers, but you are doing it in such a way the numbers in the past are the ones that are changing, not the ones in the future. And because the numbers in the past have much less visibility, nobody really “sees” that you are doing that, or even understands it.

      You wrote: “If someone experienced a “7″ in 2009 then a “7″ in 2013 should be about the same experience.”

      The thing is, if I experienced a “7” in 2009, I learned that it was a “7” back in 2009 when I used the calender to plan my trip. But in 2013, based on all the new “marbles” that have been added to “the box”, the ranking of that trip will likely no longer be the same as it was back then. I believe that your algorithms will skew items in the past lower, so the previous trip will more likely be something like a “6” as of 2013. And that is because of population growth, even though you don’t use “attendance” per se. When more bodies are in the park, it IS going to affect “wait times”. And there are more bodies on the planet and therefore in the park today than there were in years gone by. So, to meet your objective to allow a user who visited back in 2009 to understand how that experience will compare to the new one in 2013, you *should* be advising that visitor to not use the old, stale value of “7” that s/he might remember from that old trip – no, they *should* compare the “current” ranking (as of today) for that date in the past with the “current” ranking expected for the date in the future.

      I hope you can follow what I am trying to say.

• Eek! At first when I saw your phrase “a renaissance unlike any other”, I thought perhaps you might be poking fun at peoples’ concerns about crowds.. Unfortunately, as I read further, I could see you were serious 🙁 I was there last year in Disneyland for the Little Mermaid and Star Tours official openings, and although there was the craziness of the posted 240 minute wait time for ST on opening day (with the accompanying queue snaking down through Main Street!), we still had a very enjoyable week at DLR and were able to ride both attractions numerous times through the wonder of FPs. I plan on being at DLR again this summer for the mid-June opening, so hopefully it will still work to avoid the new area on opening day, and then on later days grab early morning FPs.

  Fred, it’s interesting to see your explanation about the 1-10 scale being based on data going back to 2006. As long as I’ve been a fan of the Crowd Calendar, I hadn’t ever realized that the 1-10 was based on data across multiple years. It’s not stated explicitly one way or the other, but for example the blog at http://blog.touringplans.com/2011/10/18/using-the-crowd-calendar/ seems to hint that it’s the numbers *within* the year, rather than *across* years:
  – “The Crowd Calendar is a list that ranks from 1 to 10, the wait times that we predict will occur on every day of the year. ”
  – “For those that are interested, the distribution of wait times throughout the year is bunched up between the rankings of ’3′ and ’8′.”
  Or as an older example at http://blog.touringplans.com/2010/07/23/individual-park-crowd-levels-now-available:
  – “These individual park crowd levels are based on wait times and are measured on a relative scale from 1 to 10. If the Magic Kingdom Crowd Level is a ’1′ — it’s one of the lightest days of the year for the Magic Kingdom (and when it’s a 10, it’s one of busiest for the Magic Kingdom).”

  I had just always assumed that the references to “of the year” meant that it was based on data for the current year. Thanks for clarifying!

  P.S. Maybe the main page at http://touringplans.com/walt-disney-world/crowd-levels could be updated to include this clarification. For example, it starts off with “crowd level estimate for the next 365 days in the future” and then sorts the marbles into 10 groups (so I thought it was those 365 days being sorted into 10 groups).

  P.P.S. As DCA expands with more attractions than in the past, and also as it gains credibility as a good park, doesn’t this likely mean that DCA will always be more popular now than it was in the past when so many scoffed at it? Basically, when comparing 2006 – 2012 data, it seems that we’ll be less likely to ever see 1s again for DCA.

  Reply

  • Thanks Keith, this is a great discussion (maybe fodder for another post of itself).

    The decision to use data from multiple years came out of a couple different strategical thoughts.

    1) We are often asked the question “Are the crowds busier this year compared to last year?” so designing the calendar this way helps to answer that question.
    2) Visitors to Disney parks often compare the projected crowds to what they experienced on their last trip. Structuring a calendar that relied only on 1 year’s data would exclude changes year-over-year. Besides, not all of us are lucky enough to visit the parks every year so having some context to compare crowds going back to 2006 helps. If someone experienced a “7” in 2009 then a “7” in 2013 should be about the same experience.
    3) If the calendar were based on a 365 day cycle there would be more volatility day to day. Every day we would be eliminating one days data from the data bank that goes into predicting crowds. Eliminating data is generally not a good idea if it is still valid, especially when it causes undue change. Updates and changes to the calendar are one of our biggest ongoing concerns, using data back to 2006 minimizes this.

    For Disney California Adventure, don’t worry, I expect that we will see plenty of 1s on the calendar once the popularity of the new attractions die down, but it may not be until 2013.

    Reply

• I have been wondering about this for some time. How is that statistically accurate/honest? If your 1-10 system is based on percentiles over a 365 day year, then you CANNOT have more than 36 days of 10’s. You also must have 36 days of 1’s. If you are projecting 60+ days (6/15-8/15?) of 10’s, then the percentiles are statistically meaningless.

  Reply

  • The crowd calendar is based on percentiles from all days of wait times since 2006, not in a 365 day period. So, if the set of days in the summer end up being the busiest since 2006, they may all be a 10 without violating the methodology.

    Just out of curiosity Jim, we get that statement a lot from people but I have not been able to find anywhere where we state that the calendar is based on a 365 day period. Do you recall where you read or heard that?

    Reply

    • I’m going purely by memory, so I’m a bit uncomfortable speaking up to begin with. And I’ll admit freely that what I (and others) interpenetrated may not be exactly what was said or intended. But my impression of the calendar is that it’s based on a 365 day concept. And I came to that impression because of what I’ve heard from either Len, Henry or yourself on the wdwtoday podcast.

      As a consumer, here’s the observation I’ve made. A LOT of people, including the ones that pay for the information, either don’t trust it or don’t understand it. It astounds me how many people on the Lines chat will ask if they can trust these numbers. Possibly they don’t trust it BECAUSE they don’t understand it. There also seems to be a lack of consistency when explaining the calendar. Every new explanation I’ve heard tends to make things less clear. This is certainly the first time I recall ever seeing anything about going back to 2006. I can’t find this mentioned anywhere on this site (other than here of course) and don’t see anything in my latest copy (2010) of the UoG either.

      I think that explaining this calendar in a way that people truly understand is a challenge that, for whatever reason, still hasn’t been met. And I play a pretty good game of theme park. My teen has spent 40-50 days in Disney parks and waited more than 30 minutes for an attraction exactly once. I won one of the Theme park Insider best ride pools a few years back. So I think I have a pretty good grasp of this stuff, and yet I am still not completely sure how to interpret the crowd calendar.

    • Hello Fred, I tend to agree with Marc. Especially this comment, “Possibly they don’t trust it BECAUSE they don’t understand it.”

      I myself have been thrilled with the UoG and your site. I always consult the Crowd Calendar when scheduling a trip, and always use the touring plans/strategies too. If you could figure out how to better explain the numbers, though, I think more people would have higher regard for your service.

      FYI, re the 365 day number, I was under that impression too and this is where I personally think it came from: http://blog.touringplans.com/2010/05/17/crowd-calendar-2-0/
      “Imagine you have a box of 100 marbles of different sizes. If you sort the marbles from smallest to largest you can divide them into ten groups of ten, according to size. The largest ten marbles get put into Group Ten. The next ten largest get put into Group Nine, etc.

      In our case, we are sorting all the possible crowd sizes at Walt Disney World into ten groups. The highest ten percent get a rank of ten; the next highest ten percent get a rank of nine, etc.” Looking closely, I realize that you don’t actually say “per 365 days” there, but that is where my mind went automatically. In fact, I’m rather confused by the whole notion of going back to 2006 (and I have an EECS degree from a prestigious university). Are you retroactively revising previous crowd estimates for yesterday and all previous days? And because attendance at the parks grows each year, wouldn’t your algorthims tend to skew higher for days in the future versus days in the past (i.e. as time goes on, I would think that the lightest days will generally always be in years gone by, when attendance was naturally lower because of the size of the earth’s population). It confuses me, and I’m a bit of a Math geek! 🙂

    • Thanks for the comments Marc et al.

      I’ve been at this for several years now and each method for presenting the calendar has its drawbacks. This current method is the best way that I can think of to answer the question that it intends to answer, mainly “On any given day, how will the wait times compare to other days”.

      The bottom line is this: The calendar is a rank. A rank of the wait times in the parks. For days in the past the numbers represent the rank that the wait times “were” and for days in the future the numbers represent the rank that we “expect” the wait times to be. I don’t know how to break it down any simpler than that.

      One of the main obstacles to the calendar is that it is based on wait times and not attendance. Partially because that is the information we have access to but also because our main business is Touring Plans, we care more about predicting the wait in line than the number of people in the park and trust me, there is a difference in those two concepts.

      We are exploring other ways to present the calendar (lists of wait times themselves, ratings of crowd density, customized calendars, etc.) and we are always grateful for the feedback we get. Thanks and keep it coming!