The Unluckiest Kirby Air Ride Speedrun Ever
Вставка
- Опубліковано 13 чер 2024
- Sackboy Challenges: bit.ly/3tbXjPd
Discount Codes (Thanks to @PlayStation )
EU Code for Sackboy 50% discount - F2GH2LJ4BM
US/LATAM Code for Sackboy 50% discount - BJHFN468DR
(If these don't work go here to access the PS Direct sale direct.playstation.com/en-us/...)
Race your friends and the global PS5 community in this month’s Sackboy: A Big Adventure Knit-Speedrun Challenge. There are loads of great prizes and exclusive rewards up for grabs, from cool collectibles to all-new exclusive costume pieces.
Special thanks to Bluekandy!
Check out his livestreams at: / bluekandy
Follow EZScape on Twitter: / ezscapeyt
Follow EZScape on Twitch: / ezscape
Subscribe to the Second Channel: / ezscapespeedruns
Subscribe to the VOD Channel: / speedrunsaturdays
Join EZScape's Discord: / discord
Subscribe to Bescape on UA-cam: / bescapeyt
Follow Bescape on Twitch: / bescape
Follow Bescape on Twitter: / bescapeofficial
During one of our Speedrun Saturday weekly live-shows, we featured Bluekandy who holds World Records in Mario Golf: Toadstool Tour and Kirby Air Ride. For Kirby Air Ride, he showed off the City Trial Any% run, which involves completing 100/120 checkboxes as fast as possible. The run is extremely luck dependent, and requires a player to react to all sorts of different events that might be present. Bluekandy encountered some of the worst luck he's ever experienced in the hundreds of hours that he's put into this speedrun, and it was hilarious that this all unfolded on our livestream. While the luck was bad, it was highly entertaining and we decided to make an entire video about it, in addition to how the City Trial Any% works and how the World Record was beaten recently thanks to a new discovery. We hope you all enjoy!
*Offer valid in North America, South America, United Kingdom, and European Union only. Must be among the first 2,000 persons to redeem discount codes. Offer ends the earlier of 28/09/2021 at 23:59 GMT & 23:59 PDT or when customer limit is met. Account for PlayStation™Network and internet connection required to redeem.
Videos Used:
[Old Route] Kirby Air Ride City Trial Any% (100 Checkboxes) Speedrun in 47:22
• [Old Route] Kirby Air ...
【RTA】カービィのエアライド シティトライアルAny% 44:57 (WR)
• Video - Ігри
Might have another video out in a few days, hope you guys enjoy this one :) also special thanks to Playstation for sponsoring this video! (discount codes in description)
Correct me if I'm wrong but wouldn't it technically be ~0.003% or ~1/33333, as it is 0.66^12 * 0.0043... because of the misprediction?
@@nichtsoeren4384 While yes, you're correct. The misprediction wouldn't of been a 1/230 chance if I'm not mistaken, it's an 11/115.
1/230 implies the prediction meant nothing but you still got the desired level, but that didn't happen. There's a 1/10 chance the prediction was false, which would lead to a 1/23 chance you get the level you want, since that didn't happen, we hit the 22/23 chance it's not the level we want.
Meaning: 1/10 x 22/23 = 11/115 or 9.565%
Which would then mean: 2/3 ^12 * 11/115 * 100 = 0.0737%
@@Vilonu ah right sorry, I often forget to take the opposite when calculating probability
@@nichtsoeren4384 No need to apologize, you were just taking the chances that were mentioned in the video and fixing one of the flaws, it was the video's fault you thought it was 1/230.
It's good practise to take a closer look at things and try solving them from different angles, that's how you find a more accurate truth. :)
@@nichtsoeren4384 Okay, update, I was wrong about some things, though I don't know entirely what the chances are yet, I wasn't thinking about what the chance of the DR3 prediction showing up was. Instead I was thinking about the chance of what level could've shown up after the prediction failed. So 1/230 is closer to being correct than 11/115, however, I think 1/240 is even closer and 23/5760 is closest for that one event. But I'm still finding out more information.
So its a 1/24 chance prediction message for DR3 shows up (assuming all levels have their own message and a massage shows up 100% of the time), 1/10 prediction fails, 23/24 it still doesn't choose DR3 (assuming DR3 is included in the possible levels it chooses after prediction failing)
So: 1/24 * 1/10 * 23/24
Howdy friends! I just want to reiterate how much different the run is now with the introduction of the new underflowed item count glitch (you'll learn a bit about it in this video, but this run doesn't use it. It was discovered literally last week lmao). The runs are tons of fun and the category is wide open for world record improvement, so we're quite active in doing these runs as of this upload :)
Goodluck to all of you!!
you rang?
Looking forward to checking out more about air ride speed runs!
I'm also looking forward to more kontent!
Man the amount of hours spent with friends in city trials.. such a fun unique game it was
For real
And in classic Nintendo fashion, they won’t make a sequel people desperately want. OTOH they’re finally getting around to a 3D Kirby, wish it was a collectathon tho.
Absolutely lost it when the prediction was incorrect, also I want to play this, my comfort game, again
i bet there's a kirby's air ride discord where you can netplay with people on the computer
@@Rael0505 yup, they got rollback netcode going with Kirby Air Ride.
bruh kirby air ride was my shit as a second grader
@@Kaytsey rollback for Kirby’s air ride?! That’s amazing
This stream was so much fun. Even tho it took forever, it did bring so much hype for the drag race 3 to occur
Still to this day haven't seen a run as unlucky as this one lol. Truly something special to have these incredibly bad odds happen in a showcase no reset run
This guy is the anti-Dream, created to balance out his absurd luck with equally absurd misfortune
Nightmare
this guy is so anti-dream that he didn't cheat
Now, any future Speedrun saturdays need to:
"The GDQ RNGjects"
Fun niche speedrun categories that are gated by an infinitely revolving door of RNG, that make them not very marathon friendly.
Chance of RNG this bad (13 straight fails on a 1/3 chance event) is 0.541%.
Chance of RNG this bad *and* getting trolled by the prediction message at least once is 0.0402%. (if a prediction message comes up once per attempt)
Damn. I just did the math then came to see if anyone else did it before I posted.
0.514%
@@BananaDynastyX ...yes that is what i said
You're more likely to pick a random person from the earth and have them be from Mongolia (referring to the 0.0402% chance).
Man I absolutely loved this game as a kid.
This 1/3 RNG is making the Triple Weathertanko look like a casual strat out here lmaoo
**Spoilers Below**
Here's the odds of this speedrun going so wrong:
* Odds of not getting a 1/3 after 13 attempts: *0.5138%*
* Odds of getting baited once and failing on 13 attempts: *0.0022%*
I don't think that second probability is correct. Getting baited in any single attempt has a 1/230 chance of happening but getting baited once somewhere in a string of 13 runs is not nearly as unlikely, as 1 - (229/230)^13 is about 5.5%, or only 1 in 18.
So, I think a better estimation of the likelyhood of a run like this is about 0.0283%. Still very unlikely, though
Whats that 1:X ratio
@@Mikeastro and dont forget that any run that is a bait is guaranteed to not be the correct stadium... so there is definitely some complicated math here. I dont really even trust the 1:230
Pedantic point: those are probabilities, not odds
@@kingoftherevolution4855 1/0.000022 is 45454.5454. So for a cleaner number, it's a little worse than 1:45000 odds.
Honestly the funniest part was the guy in the background, had me rolling lmao.
WR: EZScape spoiling his subscribers (Any%)
Google seems to think you and apollo legend are the same person. It said you were dead when you google "what happened to EZScape?" I got scared for a little. Keep up the good work.
the mathematical madness of this video
To provide some context as far as the odds here, this is almost as bad as getting 15 blue balls in A Link to the Past from Aga 1. And oddly enough there is actually a clip of 1 of the Alttp runners getting that in a run.
All those runners that get a 1 in a Million chance in a sub-frame perfect input while New Moon is over Canadá just stole all the luck this runner ever had. My god, i would just turn off the game.
Funnily painful, yet amazing altogether.
Why do I want to watch a speedrun marathon of Shrek games so much? Now that I know it's a thing that could happen, I need it!
I have some of my best gaming memories with friends playing Kirby Air Ride.
Bluekandy is a LEGEND 🔥
Finally City Trial any% gets some big online attention! My favorite speedrun to do : D
EZ you are seriously a cut above the rest. Your videos are of a consistently high quality, very informative yet also very entertaining, on a wide variety of topics and subtopics. I always look forward to your uploads and your videos always get an instant click from me. Keep up the amazing work!
The unluckiest RNG runs ever, now there's a category.
In the video you talked about how a player needs to complete 100 out of the 120 possible objectives to get the victory credits and time for the run. The runner himself said that Drag Race 3 contains 4 objectives, and unlocks Drag Race 4 which has 4 more objectives. As we saw in the video he was at 99 objectives, with 100-107 being locked behind getting the Drag Race 3 RNG to work for him - but what was *NOT* talked about at all was what about the other objectives 108-120 that he could be doing instead while waiting for those 3 minutes to count down every time? Could he have done ANY OTHER OBJECTIVE in 39 minutes? I don't know the game at all so I don't know, but it's precisely the thing a viewer wants to know!
Bluekandy's the man! Love his runs
The mode: obviously party based
The community: i have a 1 in 3 chance to speedrun this.
For the record, the odds of him not getting DR3 in all those attempts are ~0.3%
Its 0.514%.
I loved this game so much, so fun to watch
I remember watching this stream! Absolutely insane!! The hype was so real though when he finally got Drag Race 3
Interesting video analysis! Thanks for uploading!
Loved that the dude remained professional and took the crap RNG like a champ.
Eyyyyyy, love to see the upload!
This was pretty funny bro haha
That stats overflow glitch is really interesting. I wonder if there are any vehicle g,ouches you can combine with that glitch to do anything else that’s interesting?
BlueCandy went from down bad to down astronomical
Also, when will we see Any% No Drag Race 3 runs?
2 videos in 1 week? colin is speedrunning youtube now
I miss this game so much
1:55 honest to god, for a second I thought the "glitch" was going to just be to go to the menu and picking the stadium you want
That was hilarious wish I was there 😂😂😂
Hey babe, new EZScape video! I gotta crash my airplane into a volcano to watch this shit.
Wow! That's Caleb Hart levels of bad luck!
Take a shot every time he says 'stadium'
Being stuck for over an hour not being able to get a 1/3 chance to happen is some crazy bad luck!
Ayo EZ wya?
smash bros brawl 100% uses the kirby air ride engine
I don't think so but melee does
Where you been man you missing the once in a lifetime oppurtunity to get the bag
Might want to submit this to GDQ as a 1:50:00 now
If my math is correct (big if), you have a higher chance of getting a shiny pokemon than not getting Drag Race 3 13 times in a row using the method shown.
Wow, that rng... Actually insane
When you play through every Shrek Video Game, will you do Super Slam's Mega Challenges?
I will never be able to understand Kirby air ride. It just baffles me
The probability that the prediction is wrong is not 1/230. One in 230 is the probability that the prediction is wrongly given as Drag Race 3, *given* that the next event is NOT Drag Race 3. However, that is not the scenario you are in. The probability you want to calculate is the probability that it's a wrong prediction *given* that you have received a prediction for Drag Race 3. Without crunching the numbers, that probability is *much* closer to 1/10 than 1/230.
Funnily enough, the situation you said I was not in is exactly the situation I was in lmao. Might need to rewatch that part of the video my friend :)
@@Bluekandy Please do me a favor and hear me out. I'm not trying to be an asshole, I'm not trying to "dunk" on anyone, I'm just trying to do my part to stop misconceptions about probability from spreading. The topic is complex and unintuitive and lots of people get it wrong all the time. One famous example is the Monty Hall problem, where many people established in the field of probability came up with the wrong answer and firmly asserted that wrong answer was correct for a long time. Being wrong about probability puts you in good company; there's no shame in it. There IS shame in refusing to listen to someone trying to set you straight when you're wrong about something, though. :)
If you want to understand how to correctly assess probability, you should check out Bayes' Theorem. The concepts behind it are related to why your 1/230 figure is inaccurate. Veritasium has a good video on it if you look. I won't try to fully re-explain it here, but, basically, if you test positive for a disease, and the test is 99% accurate, that does NOT mean your chance of having the disease is 99%. It is deeply unintuitive, but it is true. If you're curious why that is, go check out that video.
This situation is similar; just because the probability that another Stadium incorrectly shows up as Drag Race 3 is 1/230 does not mean that the probability that the prediction is incorrect, given the information you had at the time, was 1/230.
If you want a more detailed explanation, read on.
The way to express the probability of one event X occurring given that some other event Y occurred is Pr(X | Y), which reads "the probability of X given Y." There are several different events here; let's give them names for brevity.
A: The event that you see an incorrect prediction for Drag Race 3.
B: The event that you see prediction for Drag Race 3.
C: The event that a Stadium other than Drag Race 3 is the next Stadium.
You calculated Pr(A | C), which is 1/230. However, this is the wrong calculation based on the information you had at the time. The correct calculation is Pr(A | B), which is much closer to 1/10 than 1/230. This is because, at the time, event B had occurred, and you knew about it, and you did not know if event C had occurred.
Your rebuttal, paraphrased, is "Ah, but event C did, in fact, occur! So I was right!" Event C did occur, but this rebuttal misses the point. You did not *know* C was true at the time. Once the knowledge that event C occured enters into the picture, then... well, the probability that A is true is not 1/230, it is *one hundred percent*, because event A is the intersection of events B and C. Thus, your 1/230 figure is even further from the true probability in this case. You're doing this weird post-hoc analysis where you retroactively grant yourself the knowledge that event C occurred, which you didn't and couldn't have known at the time, but conveniently omit the knowledge that event B occurred, which you did know at the time. It doesn't make any sense.
@@WildMatsu I'm a meteorologist, including a specialization in climate science, as well as a mathematician. I took an advanced statistics course last year in the frame of meteorology and aced it. I fully agree with your desire to stop spreading misconceptions lol
There's some critical info about the game you may not know in your stance here-the game selects the stadium that will appear the moment you enter a city. Any stadium prediction that pops up is either correct to the already granted stadium (90%) or incorrect (10%). In the event it's incorrect, it will pick a message that reflects any stadium but the correct one (1/23 for any given one). In this situation, I don't care about whether the stadium is or isn't Drag Race 3. I only care about the odds of a stadium prediction being incorrect, and then which incorrect prediction it showed. This is the 1/230 rate for only the stadium prediction event. If you want the overall probability of that occurring in the city, stadium predictions are a 32.41% chance to be rolled each time an event is flagged to happen (randomly between 55-125 seconds). Events can't be duplicates in a single city.
"The way to express the probability of one event X occurring given that some other event Y occurred is Pr(X | Y)"; thus, event X was the fact that I saw a prediction for Drag Race 3, and event Y is not getting Drag Race 3. Thus, the probability of seeing a prediction for Drag Race 3 and not getting Drag Race 3 is 1/230, due to the fact that the game has already chosen which stadium will show up before the prediction could occur.
Essentially, it's arbitrary to draw the line at saying Pr(A | B) is the furthest extent I could calculate the probability of what happened given that I know a stadium prediction is wrong in 1/10 cases and the message it chooses would then be 1/23 for any particular one. If this weren't true, why would this occurring in the game be so unlikely and warrant such a response?
@@Bluekandy I've read your comment but haven't thought about it in detail yet. I still think I'm right, but I'll make an earnest attempt to understand what you're trying to say and if it turns out I'm wrong, I'll cop to it. Hope you have a good weekend, man!
@@Bluekandy As you know, the probability that a received prediction is for Drag Race 3 _given that_ the game had already decided the next stadium will _not_ be Drag Race 3 is 1/230. Within the specific city instance where you received your Drag Race 3 prediction, the probability that the received prediction was for Drag Race 3 was 1/230, because the game had already decided the next stadium would be Drag Race 1. That's indeed unlucky. I can see why you care about this probability: it's what would be included in a calculation of how unlikely this run really was. It's why you said "that was a 1/230" after what happened.
Now, in _general_ (i.e. not within the specific city instance where you received your Drag Race 3 prediction, but instead within an _arbitrary_ city instance that has a 1/3 chance of being followed by Drag Race 3), the probability that the game _had already decided_ the next stadium will _not_ be Drag Race 3 _given that_ you receive a Drag Race 3 prediction-a probability you might not care about at all-is 2/209 (which isn't closer to 1/10 as Wild Matsu suggested). I used Bayes' Theorem to get this result.
Its trippy that this game just reuses all the menus from Smash Bros Melee
RNGesus? No, this is RNGG Allin 👹
This is hilariously cursed
Don't we all love RNGsus ;P
What was that outro song? Imma need to hold that
Best game mode ever. Only problem was 7 mins was never enough time
Dang oxygen and our ability to ride it
would be cool if you credited some of the music used in the description :)
Inb4 people beg for kirby air ride 2 on switch or a port.
I loved playing it... but I don't play enough of any game anymore to warrant begging for releases
Sakurai: So, you want Drag Race 3?
Would be a shame if something were to happen to it......
I can't find ANY guides on that item underflow glitch or others like it rip
When you think about it.. the crazy thing is you COULD be there for the rest of your life and never see Drag Race 3...
Mathematically it is a 0.001% to not see drag race 3 until the 14th attempt (with the false prediction of that specific stage)
You should cover psychonauts
I mean you could show it as a worst case scenario to gdq, after all, it's only 15 min over what he said he set the estimate for
That’s the same menu style as smash 👀
So decided to crunch some numbers for fun...
Since one of the failed attempts to get Dragrace 3 was a 1/230 chance. You get the following numbers:
((2/3)^12)X(1/230) which comes out to 3.35102e-5. Which in standard notation is 0.0000335. This is a 0.0035% chance. or 335/10,000,000 (1/29,850)
So there was a 1 in 29,850 chance (0.0035%) of the RNG being this bad.
thank you for bringing attention to this game, one of my all time favorites.
It only took me leaving the chat for the rng to be in favor
Kirby Air Ride one of the best Kirby games, glad to see that you are a man of culture.
How haven’t they made an air ride 2??
Kirby Air Ride: “drag 3s n**s in your face”
this RNG is so insanely bad it's actually unbelievable lmao
make the shrek pls i neeeed
Okay, so I ran the numbers.
If Drag Race 1 or 2 is chosen (2/3 of the time), the odds of seeing the prediction for Drag Race 3 are 1/10 × 1/23, or 1/230. If Drag Race 3 is chosen (1/3 of the time), I assume the odds of seeing the prediction are 9/10 exactly (that is, the 1/10 "false prediction" event will always give a _false_ prediction, not just a random prediction which could happen to be correct). In total, the probability of the Drag Race 3 prediction appearing (assuming Drag Race is selected) is 209/690.
If A is "Drag Race 3 was selected by the game" and B is "we see the prediction for Drag Race 3," then we want to calculate the probability of A given B. We know the probability of B given A is 9/10 as discussed previously, the probability of A is 1/3, and the probability of B is that 209/690 value we calculated previously.
We can then use Bayes' Theorem to calculate the probability of A given B, which is (9/10 * 1/3) / (209/690), which is 207/209. In other words, the probability of seeing the prediction for Drag Race 3, and then not getting it, is 2 out of 209.
Sudden ad ends at 5:29
Hope you're well. (somehow never watched this video.
i’m far too lazy to do the math, but i’m hoping the community is / did?
like the statistical analysis of every recorded checkbox run to see the actual stats on when drag race 1/2/3 come up and in what order and how many number of choices?
looking at the code and seeing that its supposed to pick a 1/3 is one level of analysis
but when you run into variability this heavily in your actual runs you need to account for the existence of bugs in the software or bugs in the RNG functions specifically
i’m assuming they already checked the game’s source for possible extraneous factors like “player recently won a drag race, give them a harder one etc”
But how the fuck do you do the challenges where I swear you need two players for, like having everyone end the trial whilst on grind rails etc.
Eggs
Yo shout out to the runner 1davidj. He's an OG and super cool.
Was literally about to submit my game for speedrun saturdays and then i hear you say that you are no longer doing them. Bruh
...Why does this UI look EXACTLY like Melee?
both are hal labs
Art critique no one asked for but people definitely need: I'm just gonna say it's super weird to put blue tears like that on Kirby because the way the original art is shaded and colored. Tears won't reflect blue like that, looking weirdly viscous. It's a really unsettling, low effort looking thumbnail for such a great channel. A quick search for "Kirby tears" brings up some decent clear tears references. Keep it up
The chances of this unlucky of a speedrun is 2.39305887e-9 which is about 4 times as likely as winning the lottery. Better not go to Vegas with those odds
why couldn't they just do a different checkbox?
Your subtitles are broken at 7:30, sir.
i wish i liked air ride more. the concept is fun but it's so half-baked. a sequel can elevate it a lot because the foundation is good
As someone who speedruns Hades I absolutely sympathize with the guy lmao
Sup EZ
Engagement and stuff
Thr fastest Unluckiest speed run ever tho that's I want to know
Yo EZ check out Liam's new mario 64 120 star WR if you havent yet its pretty awesome.
Imagine this hate crime happened to you Id be miserable
What kind of Scott Steiner math bs is this that he got the message and didnt get #3? Thats incredible
You know, they say all speedruns are created equal...
People use stadium predictions for speed runs? I used to get completely wrong predictions. It'd say you're gonna fight king dedede and then make you race
Oh okay 10% chance it's lying
There's a couple things I'm questioning about the calculations done here.
How would it be a 1/230 chance to not get Drag Race 3 when that message shows up?
If it's a 10% chance to not guarantee DR3 after showing the message, that would mean there would be 23 levels for it to choose from randomly, 22 of them NOT being DR3.
Making the chance to not get DR3 after seeing the message in City predicting DR3 this: (1/10) x (22/23) x 100, which is 9.57% or a 11/115 chance.
To get a 1/230 chance you need a 10% chance for the message to not mean anything and a 1/23 chance of getting the level you'd want. Basically saying, the message means nothing, but you still get the level you want, which isn't what happened.
Also while it would be true that the chances of not getting Drag Race 3 13 times in a row is a 0.513% chance. His luck was actually far worse than that, you thought of the 13th attempt as a 2/3 rather than a 11/115 (or 1/230 which would've been wrong, but more right than calculating for a 13th 2/3)
This would make the chance: ((2/3)^12) x (11/115) x 100, which equals 0.0737%
This was slightly unclear-23 other stadiums than the one you got. There are 24 stadiums total.
It is a 1/230 chance, because both events are independent of each other. The only possible way you can get this exact combination of Drag Race 3 prediction but not get Drag Race 3 is to first get the 1/10 probability of an incorrect prediction, and then the 1/23 chance that the incorrect prediction is the exact one for Drag Race 3
@@Bluekandy First off, thank you for responding!
This makes a little less sense to me now. If there are 24 stadiums total, would that mean there are 24 different prediction messages that could've shown up in the video? And is there a 100% chance a prediction message of any kind will show up in the City? Cause if so, the chance for the prediction message for DR3 to appear would be a 1/24 chance.
I can see that if you were to get that 1/24 (or 1/23) and it fails that would end up being a 1/240 (1/230) chance for that event.
I have another question though, in the case a prediction fails, how does it choose the level that appears? Does it choose one of the 24 levels? Or just from the 23 that aren't the level it was predicting? Just from looking at the game you wouldn't be able to tell if a prediction got the 10% chance to fail if it also then chose the level it was originally predicting randomly out of the 24 different levels. Unless you were able to see the inner workings of the game.
I'm thinking of it like this now: 1/24 you get the prediction message for DR3, 1/10 prediction fails, 1/24 you still get DR3 (which means 23/24 you don't get DR3 after prediction fails)
So: 1/24 * 1/10 * 23/24 to get the chance of getting DR3 prediction, prediction failing, still not getting DR3
This is only assuming DR3 is in the pool of levels it could choose from after prediction fails, if it's the only level removed from the potential level pool it would be a 1/240 chance to get DR3 message but not get DR3
Hope that made it a bit more clearer how I was thinking about it, cause I don't know how I could make it any more so
@@Bluekandy I see now that I originally wasn't thinking about the chance for DR3's prediction to show up. I was calculating for unluckiness, and since getting the prediction message for DR3 is considered lucky I didn't think about it
@@Vilonu You're approaching the odds of the stadium prediction being incorrect in the wrong manner. It's a flat 90% chance for the stadium prediction to be normal and correspond to the stadium you get after the city. The stadium game is determined the moment you enter the city, so it's already known to the game, but not to you (in the case of Drag Race). The 10% chance of it being incorrect doesn't affect which stadium shows up. It will always be Drag Race 1 in that city, regardless of stadium prediction. So when it decides to be an incorrect prediction as an isolated event (10%), it then chooses one of the other 23 available messages that would be incorrect relative to the known stadium; thus, to land on DR3's message is just a 1/23 on top of the 1/10.
@@Bluekandy Okay, I think I found a better way of putting my thoughts into words.
I don't know if it's correct, I know next to nothing about this game. But I can see the devs making the predictions have a 10% failure rate. If it does fail, the game chooses a random level like as if the prediction was never made.
So rather than thinking of the prediction having a 10% chance to be WRONG, I'm thinking of it having a 10% chance to not have any AFFECT. Meaning that DR3 could potentially still show up with a 1/24 chance (if the game is choosing between all 24 stadiums)