Has Touring plans lost its collective mind?

[TravelGrl]

Thanks both of you! I never realized this feature existed. Just added my upcoming trip to the tracker. Now hoping for no major updates ..

you can sign up for the Crowd Tracker updates for certain dates and it emails you when it changes.

http://www.wdwinfo.com/images/smilies/love30.gif

Using the Crowd Calendar navigate to the first day you want to track. There is a link "Track This Day" that you click to track that day. You will get email updates when anything changes.

Near the bottom of the page there is another link "Crowds for mm/dd/yyyy" to go to the next day and you click the tracking link again. Repeat until you are tracking all your days.

You can also view all the days on one page by navigating to any day using the Crowd Calendar and there will be link near the upper right: "View your Crowd Tracker"

http://www.wdwinfo.com/images/smilies/love30.gif

 

[julirina]

I just started tracking for our upcoming Sept/Oct trip. I found it strange that TP has such a low crowd projection (1.9/10, neutral park) for October 1st at Epcot, which is the 30th anniversary (according to EasyWDW it's expected to be a big event). After reading this, it sounds like I should base my plans on another crowd calendar (and be thankful we have the flexibility that hopping and no dining plan affords, just in case). Thank you all for the heads-up!

 

[Tarheel girl 1975]

I just started tracking for our upcoming Sept/Oct trip. I found it strange that TP has such a low crowd projection (1.9/10, neutral park) for October 1st at Epcot, which is the 30th anniversary (according to EasyWDW it's expected to be a big event). After reading this, it sounds like I should base my plans on another crowd calendar (and be thankful we have the flexibility that hopping and no dining plan affords, just in case). Thank you all for the heads-up!

I haven't heard that there is much planned for Epcot's 30th anniversary.

 

[vincev33]

I have used touring plans for all of my trips over the last several years and they are usally right. This is the first year that I remember them making updates, but it makes sense why they are changing the rankings. It hasn't changed much. I found a free service that was easywdw that does the same thing. I like the lines app though so I will probably still use it.

 

[TXMama09]

I didn't read through the entire thread, so someone might have already said this, but...

Crowd calculators are obviously based A LOT on the days of EMHs. For a very long time, all the parks have held the same AM and PM EMHs based on the days of the week. This has recently changed (for the first time in awhile from what I hear) and I am sure all the "touring" pages are having to recalcuate. I paid for Touring Plans and thought it would be a great resource, however, I can't figure out how to change the duration of each Country Pavillion in Epcot to anything more than 10 minutes and that is just not realistic. Sooooo...I wasted my money

 

[michelle06]

It is called a forecast. As more information is added, it refigures the information. No forecast is going to be 100% as you are essentially attempting to predict the future by using information from the past. There are too many variables that could change things.

I actually am glad to see all the updates because it most likely would mean they are adding information and running the possible scenarios. That said, I would not worry to much as things can change the week of.

If you understand what Touring Plans, EasyWDW, and Tour Guide Mike are... then you know that they will never be the be all, end all of what could happen.

This isn't at all what's going on. These are not normal updates. THey are drastically dropping the crowd levels, like from 9 to 2. And then later, they are reversing back to the old numbers and predictions for parks, etc.

I am finding this annoying too. They changed multiple things for all but one of the days of my trip. Then either 1 or 2 days later, they reversed it all back.

I'm wondering if what's happening is they may be tweaking things for the new fastpass rules? Whatever it is, it's not a normal update.

 

[michelle06]

I didn't read through the entire thread, so someone might have already said this, but...

Crowd calculators are obviously based A LOT on the days of EMHs. For a very long time, all the parks have held the same AM and PM EMHs based on the days of the week. This has recently changed (for the first time in awhile from what I hear) and I am sure all the "touring" pages are having to recalcuate. I paid for Touring Plans and thought it would be a great resource, however, I can't figure out how to change the duration of each Country Pavillion in Epcot to anything more than 10 minutes and that is just not realistic. Sooooo...I wasted my money

If you are willing to accept touring in multiples of 10, you can just put that country in more than once. For example, put that country in twice for 20min, 3x for 30min. Maybe that would help?

 

[Day-Day]

If you are willing to accept touring in multiples of 10, you can just put that country in more than once. For example, put that country in twice for 20min, 3x for 30min. Maybe that would help?

I can't figure out how the duration time comes into play when I look at TouringPlans recommended plans nor custom plans. Does it keep up with the estmated cumulative time throughout the day?

 

[michelle06]

I can't figure out how the duration time comes into play when I look at TouringPlans recommended plans nor custom plans. Does it keep up with the estmated cumulative time throughout the day?

The general plans that they've always had do not show times. The customized plans show exactly what time they expect you to arrive at the ride, how long you will wait, how long to ride, and how long to walk to the next ride. If you make a customized plan, you should see this. You do have to click "OPTIMIZE" after choosing your date, park, and attractions.

 

[lentesta]

This isn't at all what's going on. These are not normal updates. THey are drastically dropping the crowd levels, like from 9 to 2. And then later, they are reversing back to the old numbers and predictions for parks, etc.

I am finding this annoying too. They changed multiple things for all but one of the days of my trip. Then either 1 or 2 days later, they reversed it all back.

I'm wondering if what's happening is they may be tweaking things for the new fastpass rules? Whatever it is, it's not a normal update.

Hi folks,

I'll be asking our stats team to get answers to your questions. If you have a second, posting the following information would be really helpful in doing that:

1) The specific dates of your trip that have either (a) substantially changed crowd levels; or (b) changed best/works park recommendations.

2) With respect to the emails for the crowd trackers, whether the issue is that (a) we're sending too many email for minor updates; (b) we're sending too many emails related to the changes described in (1) above.

I'll post responses here as we get them, assuming the mods are okay with that. Thanks very much for the feedback.

Len

 

[lentesta]

Thanks GatorLady. Let me know the date, please.

I don't know what is going on over there!

They are posting contradictory information within the page for an individual day. Their written explanation paragraph at the top is different than the park by park breakdown on the same page! Here is a taste of the madness! (I didn't post date so as not to break any rules)

"We're recommending the park with the lowest per park level, Disney's Hollywood Studios."


Disney's Hollywood Studios
Crowd Level: 8.3 out of 10
This is a Park to Avoid
Park Hours: 8:00am - 10:00pm
Fantasmic! at 8:30pm
Fantasmic! at 10:00pm
Pixar Pals Countdown To Fun! at 3:00am
...view park wait times / full show schedule"

They also contradict themselves on what park has the lowest crowd level, and they say to avoid MK at the top of page, but in breakdown of each park, have it as a best park???

Just be sure that you read the entire page for your days to see if it makes sense.

 

[DISHoppy]

Len, thanks for jumping into this thread. My intent wasn't to flame TP, just noting that recent planning changes seems more frequent than in the past. I'll watch my own plans more closely now that I'm not on vacation and have my normal Internet access b/w now and June, and post discrepancies as they occur.

-Kevin

 

[lentesta]

I can appreciate that the booths get longer lines than those for the rides, but a 1.1 on the first Saturday of F&W...and its a neutral park. Nah, I don't think that's right. I don't need any numbers to prove it either, that's just common sense. I would be willing bet that this will change to a park to avoid or at the very least, that number will go up.

I love TP and love the lines app, but that is really off.

Interesting question. I went back and looked at the wait times we recorded for Saturday, October 1, 2011 to see how busy Epcot was.

The highest posted wait time we recorded for Soarin' was 80 minutes between 3 and 4 pm. The peak posted wait we have for Sum of All Thrills was around 15 minutes. The highest posted wait for SSE was 20 minutes, and the longest actual wait we have recorded is 4 minutes.

The highest posted wait for Test Track was 100 at 4:30 pm. Test Track, however, went offline from around 1:45 pm to around 2:30 pm. Based on that, the posted standby wait may have been inflated by FASTPASS guests who were denied rides during the outage, who later returned to do so. The posted wait dropped to 30 minutes by 5:45 pm.

We don't have any actual wait times for Soarin' or Test Track on October 1, 2011. A good rule of thumb, though, is that actual waits are 80% of posted. (There's a lot of variation in this, but it's a decent starting point.) That would make the peak waits in line for Soarin' about 64 minutes, Sum of All Thrills around 12 minutes, and Test Track 80.

Translating those to our 1-to-10 scale, Soarin', SSE and Sum of All Thrills would be a 1; Test Track would be a 10, but again, the outage skews that number a lot.

So, broadly speaking, the wait times for Epcot last Saturday, October 1, were either a 1 on our scale, or pretty close. I don't see anything obvious in the wait times to indicate that we were off substantially.

That being said, the lines for the food in World Showcase were long, and the drunk college kids made that experience unpleasant. But the crowd calendar doesn't measure those things (yet). If you're concerned about wait times at attractions, though, October 1 looked pretty good.

Let me know if this isn't clear. Thanks for the question.

Len

 

[rwdavis2]

Interesting question. I went back and looked at the wait times we recorded for Saturday, October 1, 2011 to see how busy Epcot was.

The highest posted wait time we recorded for Soarin' was 80 minutes between 3 and 4 pm. The peak posted wait we have for Sum of All Thrills was around 15 minutes. The highest posted wait for SSE was 20 minutes, and the longest actual wait we have recorded is 4 minutes.

The highest posted wait for Test Track was 100 at 4:30 pm. Test Track, however, went offline from around 1:45 pm to around 2:30 pm. Based on that, the posted standby wait may have been inflated by FASTPASS guests who were denied rides during the outage, who later returned to do so. The posted wait dropped to 30 minutes by 5:45 pm.

We don't have any actual wait times for Soarin' or Test Track on October 1, 2011. A good rule of thumb, though, is that actual waits are 80% of posted. (There's a lot of variation in this, but it's a decent starting point.) That would make the peak waits in line for Soarin' about 64 minutes, Sum of All Thrills around 12 minutes, and Test Track 80.

Translating those to our 1-to-10 scale, Soarin', SSE and Sum of All Thrills would be a 1; Test Track would be a 10, but again, the outage skews that number a lot.

So, broadly speaking, the wait times for Epcot last Saturday, October 1, were either a 1 on our scale, or pretty close. I don't see anything obvious in the wait times to indicate that we were off substantially.

That being said, the lines for the food in World Showcase were long, and the drunk college kids made that experience unpleasant. But the crowd calendar doesn't measure those things (yet). If you're concerned about wait times at attractions, though, October 1 looked pretty good.

Let me know if this isn't clear. Thanks for the question.

Len

Len,
So to boil that down, the number you assign to a park is based on ALL wait times put together by an algorithm? (copyrighted or patented or just a trade secret I assume? )

It seems to me that even when people see a 1 for a park they should still assume a significant wait for headline attractions during at least some part of the day? (love your podcasts too!)
Thanks!
Bob

 

[lentesta]

Bob, I'm sure our statisticians will say I'm butchering this explanation, but the calculation of the crowd level goes something like this:

1) For each attraction, take the submitted wait times and draw a best-fit curve which minimizes the error between the curve and the submitted times.

The peak of that curve is the attraction's crowd level for that day.

2) We assign each attraction a "weight" which indicates how well its wait times map to crowds. Multiply each attraction's crowd levels times its weight and sum the values.

Some attractions are better indicators of crowds than others. For example, Splash Mountain is not a good indicator of Magic Kingdom crowds. New Year's Eve is generally one of the busiest days of the year, but if the temperatures are below 50, not many people will want to get wet on Splash, regardless of how many people are in the park. Our weights are designed to account for things like that.

3) Map that sum to a table of historical crowd levels and find the value that is closest. For example, if the sum of the weighted attractions' crowds is 75.3 and 75.3 maps to a '4' in our table, the overall crowd level for that day is a 4.

I'm sure I've missed some subtleties, but that's the general idea.

Len

ETA: There are also a variety of ways of dealing with attraction outages when calculating crowd levels. For example, when Test Track went down last October 1, there are a number of ways we could have handled it:

a) We could have just said that whatever the peak was, was the crowd level. The problem with this is that it can make an otherwise slow day look very busy for one attraction. There are days where Space Mountain has gone offline two or three times, while the FASTPASS machines continue to operate, for example, and that causes increases in standby times later when those FPs were redeemed.

b) We could ignore the wait times for some period after the closure. For example, if we know that there are wild fluctuations in standby times in the 90 minutes after Test Track re-opens, we can have the model ignore those wait times. We'd start looking at the wait times after things quiesce.

c) We could throw out the entire day. You hate to do this, but sometimes it's the only option. If Test Track doesn't open on time, for example, and then has a couple of outages throughout the day, then it becomes really difficult to figure out what the actual demand was for that attraction on that day.

Generally speaking, we have separate models which look at crowd levels differently when there are outages. We have a technique in place now, but we're constantly discussing how to handle things like this.

 

[Bellamouse]

This is all very interesting. I suppose it probably said so somewhere on your website, but I must have missed it, but I didn't realize that the "crowds" are measured purely by ride wait times, and not by how many people are actually packed into the parks.

As adults who aren't too interested in the rides (we do ride some of them, but it's not our main focus), we are looking for numbers that tell us how bunched up we will feel in the parks, not so much how long the lines will be for rides.

My big beef last year was TP had MK listed as a .7 on Friday Oct 21st. When we got there in the morning it was a MADHOUSE. By my own calculation, I would have put the crowd level at like an 8 or something. There was even a post (by someone else) on this board titled "why is MK so crowded today Fri Oct 21st" (or something like that). So it wasn't just me who thought it was packed. We couldn't stand it. We couldn't even move through the park. We left pretty much right around noon as it was totally unpleasant.

Now our vacation wasn't ruined by that of course, but I was very, very disappointed and frankly very annoyed having gone in with the expectation of a .7 level crowd and finding level 8 (or even 7 or 6 - much more than a .7).

I sure wish there was a calendar that gave me statistics on how many people will be in the parks, not how long the wait for rides will be. The overall crowd levels are of much more importance to me as we like to stroll leisurely trhough the parks and not feel boxed in.

But it does help to know why things seem so skewed to me.

 

[rwdavis2]

Bob, I'm sure our statisticians will say I'm butchering this explanation, but the calculation of the crowd level goes something like this:....

Len,
That makes sense. I figured there was some weighting (no pun intended) going on among the variouse rides. Had not thought about best fits though.
Bob

 

[MikeRx]

Thanks Len!
I’ve experienced strange fluctuations for Animal Kingdom on April 22, 2012. It was reported as:
11/23/11 April 22, 2012: Animal Kingdom Crowd Level updated from 5.6 to 8.7
11/28/11 April 22, 2012: Animal Kingdom Crowd Level updated from 8.7 to 5.6.
3/22/12 April 22, 2012: Animal Kingdom Crowd Level updated from 5.4 to 3.5.
I also know that I checked on-line around March 21st and the tracker for this was over 8 again! It just seems like a bit of wildness to me. I have 11 of the 13 emails sent so far so the 3/21 change must have been deleted… I hope the 3.5 holds!

Mike DW DS (4) DMIL

First visit 1976, last visit WDW 2/2010 BC CL and DL GC 12/2010, future WDW 4/2012 POLY TPV, 7/2012 (off site meeting), 4/2013 POLY with Disney Dream cruise!

 

[GatorLady]

Thanks GatorLady. Let me know the date, please.

Hi Len, I just went to the website late this morning, and the information has already been corrected. (It was April 1st.)

 

[lentesta]

Thanks Bellamouse.

We've considered using a "how crowded the park will feel" number. In fact, for the past year we've asked Lines users to rate, on a scale of 1 to 10, how crowded the parks felt, every single day.

The problem is that how a park "feels" is entirely subjective, and people will answer differently based on everything from how many times they've been to the park, to what kind of mood their spouse woke up in.

For example, yesterday (March 27, 2012) was fairly busy at the Magic Kingdom. In terms of actual waits, it was somewhere between a 9 and a 10 on our scale.

Around 50 people answered our survey question asking them to rate how crowded the Magic Kingdom felt. Of those, 12 (around 22%) indicated a crowd level of 6 or below; a few rated the day as a 2, 3 or 4.

When we've looked across months of these surveys, the problem we see is that for any number you want to pick on any day, the vast majority of people will not agree with the number you chose. If you said the park was a 9 yesterday, for example, the vast majority said it was something else.

Even if you give a range (e.g., "9 to 10"), almost 70% of the people we asked yesterday would not have agreed with that.

You'd have to make the range fairly broad (e.g., "7 to 10") before you got a majority of people to agree with you. And at that point, I think the 10-point scale has narrowed to 3- or 4-point scale.

The problem with narrow scales is that they don't covey enough information to a lot of people who *really* want specific estimates. (In fact, we added decimal points to our scale because so many people asked us to differentiate between the days numbered '5' in our old version.)

Those are some of the problems we see with rating how a crowd "feels." There doesn't seem to be a way to get a consensus on what that would be. I'm open to suggestions though!

Thanks again. Great comment.

Len