ok so i may be the opposite here but i have been reading and wanted to share a couple thoughts.
The following is about ON THE DAY wait times and is an ideal that may not be realistic.
I dont want it to guess for me behind closed doors.
I want to see when i look at Barnstormer:-
Disneys Posted current wait
The within the last 30 mins waits posted by users of TP
So mocked up example
Unnatural! Witchery!
Should we put something below the TP Estimated Wait along the lines of “75% of actual waits should be this long or less, based on past actual waits at this time of day”? Essentially tell people what we computed the odds as and how?
I am not sure that the user recent waits should necessarily be taking up valuable real estate. I think they could also set false expectations. They aren’t really relevant anymore. It’s fantastic if someone at 8am waited 10 minutes, but if it’s now 8:30am I don’t think I will only wait 10 minutes! Maybe just make it easier to get to the recent waits. I would think yesterday’s wait at 8:30 would be more relevant than a stale wait time.
ok so i may be the opposite here but i have been reading and wanted to share a couple thoughts.
The following is about ON THE DAY wait times and is an ideal that may not be realistic.
I dont want touring plans to guess/estimate! ok that sounds stupid..
I dont want it to guess for me behind closed doors.
I want to see when i look at Barnstormer:-
Disneys Posted current wait
The within the last 30 mins waits posted by users of TP
So
I think there should be a key at the top of the page yes explaining that where you have “wait times last updated”
That might be good to have in general, but maybe not with every single ride, unless you’re going to have a different percentage on a per-ride basis.
Every ride card already takes a decent chunk of scroll real estate
Maybe one time just below the header or near the Optimize/Evaluate/Modify buttons
This is a fair point and i considered would I rather know and make that decision myself and for me chose yes.
But I do agree its going to be heavily dependant on the accuracy and frequency of data. If i see 3 mins ago someone had a 20 minute wait then great.
The ideal (for me) would be seeing user data more upfront when i click on the ride/show
Guess is probably the wrong word, but isn’t the whole point of TP to provide you with a strong estimate?
I agree with this but I think we are among the savviest of TP users. Most users aren’t digging for data, they want to be provided with the estimate. I also don’t think there are even that many user posted waits. I think if we could just get a one-click to the user submitted waits it might be simpler.
How about the currently hidden by a click user times are just exposed in full when you click on the ride and listed at the bottom?
currently the text doesnt work great on iphone and its an extra click, see below
instead just have
I tried to make a distinction about pre planning and day off. but actually lets say future and current!
I want TP to estimate tomorrows times for riding HM at 1pm and give me a estimate. and i want it to estimate what JC might be if i ride it at 7pm today.
But if im looking at SM now id like to see Disney time and recent actual times not an estimate..
So i can make my own choice. of course actually all three are possible, I think if its clearer whats happening and guest recorded times are one click away then im certainly happy 
I agree that if the algorithm is the same across the board, put that info in one place, not on each.
I do think that is important to have somewhere, even if the people complaining / confused don’t see it until it’s pointed out.
Or add the user submitted wait times to that estimated time graph, like other TP graphs we’ve seen?
An update on work so far this week.
We’ve updated the intraday forecasting process to use the last set of forecasts, not the original set of forecasts, as the baseline for the next set of forecasts (Issue #2 and #3 above).
The next set of steps is for Issue #1:
I’m now loading the DAK data to my laptop and to our staging server. I’ll use my laptop to test how these new forecasts work with the updated intraday process.
Some interesting findings
Many (most?) attractions have actual waits that are usually less than the posted wait. So the sign says 50 minutes and you wait 40 or whatever.
Expedition Everest is an example:
In the graph above:
Other attractions have actual waits that are often 
more than the posted wait. Guardians of the Galaxy: Cosmic Rewind seems to be an example:
(These are a newer version of the graph - the light blue line is the point where the acutal wait equals the posted wait - that is, the sign is correct.)
So for Guardians from about 3 hours after the park is open (minute 180) until late in the day, your wait is likely to be the posted or slightly higher.
That’s unusual. Until mid-2025 and Epic’s Battle at the Ministry, I don’t think we’d ever seen this on a regular basis before. So we’ve learned some stuff about how Disney’s operating rides in 2025.
I’m thinking we could have these new models and processes in place by Dec 22. It’s going to take a lot of testing before then, so it’ll be the thing I’m working on every day.
Thanks again to all y’all for your input here, especially in timing actual waits.
This by itself is pretty fascinating.
Thank you for the work on this! I look forward to using and timing waits in April!
One of those offices in the Battle at the Ministry queue should be able to handle this. Time to start knocking on doors.
I think, on my last trip, there were a couple of instances of me stuck in a queue much longer than TP predicted. One was directly do to someone assaulting a CM and they shut down the ride so that person could be arrested and removed. I don’t think TP has a way to plan for that.
The other I genuinely think was just sheer bad luck. I don’t really know what happened, but I refreshed the Disney app and saw the wait time jump shortly after we got in line - it wasn’t even a predicted time climb in the day. I think we waited 25 minutes longer than TP predicted?
I think my most frequent problem is with ToT, across all of my TP wait times. I think I’ve said it before, I wish there was a way to report that one side was down, so it could adjust the wait times. If I look up and only see one side running (something you can see from outside the ride) - I know TP has probably under-predicted?
As for how I think the app works - I just constantly refresh whatever app I’m in.
And Bravo @len to the under 2% refund request rate! That’s impressive!
Glad you are looking into this! I did a plan for Magic Kingdom and the times are showing as incredibly quick… and it’s during March Break so I know that shouldn’t be the case. Hoping it’s fixed soon! Thanks!!
That is one of my biggest complaints with MDE and Disney websites. So. Much. Scrolling.
I’ve checked AK rope-drop touring plans using the new wait times, and they produce reasonable-looking plans (FOP is first, NRJ is next, etc). That’s an important sanity check. So, good news there.
Tomorrow we’ll see how the intraday adjustments look starting at park open. 
Trying to get WDW done before around December 20, then DLR, then UOR. I’m confident for your trip in March!
I’m running the new version of the intraday adjustments on my laptop. It looks decent. Here’s today for Flight of Passage:
The column “Submitted Time” is Disney’s posted wait.
The column to the right, “Running Est of Posted Wait” is our estimate of what the posted wait should be, based on the last set of intraday adjustments we made.
Ideally those two numbers should be the same. It would mean our predictions match what’s happening.
The “Running Delta” column, highlighted in orange, is the difference between those two numbers. The goal is for that number to be as close to 0 as possible.
One of the things I like about the modified intraday adjustments is that it’s faster to adapt to those differences. For example, the posted wait jumped from 50 to 70 minutes in 9 minutes of real time, from 11:14 to 11:23 this morning. And it took the intraday process 6 minutes to verify that wasn’t a one-off fluke and adjust the forecasts accordingly.
Reminder that the estimates of actual waits are using the 75% value - so 3 out of 4 times you get in line, your actual wait will be that number or less.
Here’s the ratio of actual-to-posted at that 75% value for FOP:
i’m actually surprised that ratio is so high. I suspect Disney’s shooting for a 50% number, which is an entire separate discussion.
just starting to prep for our trip next week.
Thought it might be interesting to talk through the process I May have followed if I hadnt been reminded here! A simple change IMO would be to add “Dont forget to optimise or evaluate your plan before pressing print”.
Also I just realised even if I had viewed it on the site i also may not have pressed evaluate or optimise as I could have assumed it updates/refreshes when opened
So i set this plan up a couple weeks ago and just went to it, and i can press print with no reminder to update it! if i DONT update it i get:-
pressing Evaluate today gives me :-So 1 hour Worse predicted across those rides than what was predicted a couple weeks ago.
