Just to one of your points, they do measure actual wait times by going through the line. That's part of how they estimate actual wait times vs. posted wait times. Oftentimes Disney seems to overestimate wait times.
I'm sure it'll need tweaking, but the basics are ready to go. And the thing we're anticipating from the initial rollout of FP+ is that it's going to be confusing to a lot of people. That's why we're building tools.
Len
I don't think FP+ is competition for us. Our products are based on practical consumer advice, such as which rides and restaurants are worth your time and money. Even with FP+, people will still want to know which attractions are worth their 3 choices.
I might have been the one to call it out explicitly but you, @menamechris, and others imply it fairly regularly.
so no one in my family has a smart phone, so how will this work for us? I know they have Kiosks, but i feel like this makes the vacation difficult for non-techies like myself. Also so if you book your fastpasses in advance, if i plan a vacation say a month away, and all the fastpasses are "booked" am I out of luck? Is this going to become like Disney Dining were I have to call to reserve being able to go on Splash without waiting a huge time a year in adavanced? Thats not very condusive to my theme park expierence!
Oh, my gosh!
Oh, my gosh!
It's Len Testa!! Have to pick my jaw up off the ground.
I am a wee bit star-struck.
*swoons*
What ever happened to walking into a park to see what lies beyond the turnstyle.... Our lives are so scheduled as it is.
Aside from Marty Sklar, I can't imagine wanting to see anyone else on the board. Len The Planning God Testa!Breathe,luv, breathe!
Current information (subject to change) is that you'll be able to book your FP experiences online before you leave. If those experiences work for you then you should be set, no need for a smart phone for FP+. (There will be smart phone apps that help with other aspects.) After you arrive, you could make changes to your FP+ selections if something "better" opens up. Alternatively, some kiosks will be available for those without smart phones. When it comes to changing FP+ selections after arrival, it seems obvious to me that someone with a smart phone will have an advantage over those that don't.so no one in my family has a smart phone, so how will this work for us? I know they have Kiosks, but i feel like this makes the vacation difficult for non-techies like myself. Also so if you book your fastpasses in advance, if i plan a vacation say a month away, and all the fastpasses are "booked" am I out of luck? Is this going to become like Disney Dining were I have to call to reserve being able to go on Splash without waiting a huge time a year in adavanced? Thats not very condusive to my theme park expierence!
We seem to be arguing the same thing as far as their methods. However, I think that where we differ is our opinion of the value of their results. I think that the sampling that is done, combined with their behavior analysis, crowd level estimates, pathing algorithms, etc. provide for some fairly accurate results. I would consider what they are doing to be a little more advanced than just an "educated guess."I know.. but again the issue is 'resolution'. You need a ton of people to actually do that beyond 'spot checks'. Simply put.. a person can only be checking one thing at a time, and with 30+ attractions happening concurrently every minute of every hour of every day... unless you have dozens of people all day long, it's all samples, build a reusable behavior model, and then extrapolate. That's why the app represents so much potential for them.. it would increase their number of reference points by orders of magnitude and allows them to dynamically update beyond what the behavior model predicted. If you have one or two 'spies' in the park, they may be able to adjust the top level crowd lever.. which would scale all the other attraction models, but they can't scale themselves to be providing real feedback on all 'actual' times across the entire park all the time. The lines users gives them directly reported info, not just the scaled model.
Let's put some perspective of scale into the problem.. How much volume is really there if you have a true consistent sampling of a park like the MK? Take some rough numbers..
25 attractions, three measurements, standby, actual, fp distribution, one sample per 30mins, park open 14hrs a day = 2100 samples a day. If I boast I have 'hundreds of thousands of samples'.. say even 500,000. That is only 238 days worth of data. Not even a single season... and that's just one park.
The data resolution is not true logging of all history.. it's samples, models, and extrapolation.
Yeah... Since you disagree it can't possibly be true. Disney must intentionally be doing things to screw up the guest experience. That makes way more sense.
We seem to be arguing the same thing as far as their methods. However, I think that where we differ is our opinion of the value of their results
I think that the sampling that is done, combined with their behavior analysis, crowd level estimates, pathing algorithms, etc. provide for some fairly accurate results. I would consider what they are doing to be a little more advanced than just an "educated guess."
I guess I am not sure as to how they are arriving at their educated guesses without analyzing historical data? I think that we agree that they do predictive modeling, which is based in part on historical attendance and wait time data. I'm not a statistician, but I think that you may be overestimating the number of raw data points necessary to develop meaningful information, especially given the relatively controlled environments of the Disney theme parks. The nature of the beast makes it something that can be predicted, if you know the trends to look for.Go back to what I originally replied to... "I still would love to see Len do some serious data analysis of crowd behavior histories. Because I really think he's the only one doing it."
How can you do 'serious analysis of crowd behavior histories' - when you don't have the data samples and resolution? As you agreed, their product is based on models and predicting the future variables. I said they are not extrapolating from a historical data set, but rather building a model, and then using inputs to predict the expected crowd behavior in the future. They then use submissions to update the predictions with 'actuals' as the time comes.
Maybe you took offense to my wording.. but the point remains the same. The estimates are based on prediction using educated guesses (guesses.. informed assumptions.. you pick the wording you want) on the inputs.. not actual hard data extrapolation. They replace their lack of hard data with models to estimate the data. Tying it back to the original post I replied to.. they'd be hard to 'analyze histories' beyond the actual data points they've collected. Everything else would again be an estimate based on modeling or interpolation between the real data they have.
Going back to the scale point.. even if I had a million data points.. that's not much data when we are talking about time resolutions measured in an hour or less over the entire property and all attractions.
I guess I am not sure as to how they are arriving at their educated guesses without analyzing historical data?
Register on WDWMAGIC. This sidebar will go away, and you'll see fewer ads.