In 100 man, you can't skip if you aren't on the floor / at the death animation. Bumpers does not count as a floor, so you literally can't skip this stage. I guess the players tried to star it to be able to skip it,but that's not how it works. Even if you're very limited, I think it's possible to do a decent level that is unskippable but still fun to play, then I guess your level will get a reasonable amount of stars. Good video as always Ceave :p
Is the clear rate also really high on this? That would mean it would be put on easy 100 man so newer players would run into it and be more likely to star it not knowing that doing that doesn’t work.
These levels have stars, cause you can’t die and if you play the level but press start over or quit it doesn’t count as a play. It only counts as a play if you die or complete the level and if you can’t die it’ll only count as a play if you complete the level. These ”best levels” are long so you’d exit the level if you don’t like it and if you complete a long level it’s pretty likely that you liked the level and give it a star. I know that was complicated and had poor grammar but I hope you understood.
Oh! So the trick is not to make a level that would be starred by the most people, but to make a level that is only completed by the people that would star it! The bumper level seems to have done a good job hitting that target, since people that don't like it will quit immediately instead of completing it.
There is a confounder here; you can only star a level you finished. For example, a very hard level may be hated by most people that play it. Only a very specific subset of people that finish it, i.e. skilled people or emotionally invested people, finish and rate it. Such a level may get a high ratio, but still not be enjoyed by the average player / majority of players. If you correct for this by dividing not the amount of unique players but the amount of mario lives spent by the amount of stars, you will only get levels in which players can not die. Either way, you still don't know which level got the most amount of enjoyment per play or per player. And by dividing the amount of time spent by the amount of stars, you will only select short levels... this is a much harder problem than it seems. The best way would be to let everyone rate a level, even those that don't finish a level. But then, people that are less inclined to rate a level will be underrepresented. If people that rate levels are more conscientious (for example) than those that don't, your results will be skewed towards that metric, and still not represent the 'average' player (selection bias). So you need to ensure everyone rates every level they play at all times. (This is a perfect example of the effect of confounders, and exactly what happens in bad medical research, where it actually costs lives.)
This should be higher. Kinda bummed it wasn't mentioned in the video, great real world example of why statistics often requires careful thought (and how it can easily be manipulated).
@The_Mighty_ Boole I know right! It is so original, and is very cool. I'll bet Dean is an executive from Nintendo, because he has made the Mona Lisa of Mario Maker Levels.
Which is probably why he's laughing, Bum... he's saying that the Splat 2 girl is his favorite, or at least moreso than the Splat 1 girl. Or maybe he likes Pearl more, or one of the Sisters? Personally, my favorite is Callie, though I should note that Octolings aren't technically squid girls...
Of course, we could have thousands, even millions of alternative universes where the "best" Mario Maker level is good... But in this one, it's just...bumpers.
Hey, data scientist here. I truly enjoyed this video, the fact that you gathered this dataset is awesome. Would you make it openly available? Regarding the grading, I believe you could have used some bayesian posterior mean formula to have a more significant result. For example, the one used originally by IMDB (according to wikipeida) could be slightly modified to get: Weighted rating for a given level = (R*v + C*m)/(v+m) With R = percentage of stars, v = number of players, m = threshold to be considered (what you set at 20k), C = mean of all percentages of stars across all dataset. I wonder if you’d get the same results using this formula.
The scraping of this dataset isn't horribly complicated (the most time consuming part is having to throttle reads, avoid duplicates and deal with UTF8 characters). If Ceave hasn't responded to you yet, let me know and we can discuss getting this dataset.
You cannot skip a stage while in mid-air, and trying to hit Restart when pissed off at being stuck will often make you star something on accident. That level traps Super Expert players until they beat it or lose every single one of their lives. Over such a long time, it is extremely likely to star it on accident.
That fake ending made me laugh so hard. d _And we can see that this is the BEST rated Super Mario Maker Level. I hope you enjoyed this little video, and if y-_
This is basically basic statistics mixed in with mario maker. But I really, really like it. This guy knows what he is doing. This is honestly a rarity in these sort of videos.
For spot of best stage in creativity, I would nominate "Pause for Effect." It uses puzzles that you have to solve by pausing the game at the right time.
Agreed on "Pause for Effect." This is why when it comes to judging, subjectivity/quality has to play a part over objectivity/quantity. That said, what would be the highest liked over views video for Mario Maker videos on TH-cam? That would give a different source of data all together (which would include non-Mario Maker players like myself :P) (that would also increase the demographic pool)
Dude I laughed so hard when you said "Either I just don't get the level, or we just officially found the limits of statistics" LOLOL You're the man dude.
@@nope4180 that's hella work though. You could make it a lot faster by running multiple instances through different vpns or tor exits, but youd still have to write the script
He said he's going to make more content using the statistics and he linked to the levels without actually linking the database so I doubt he's going to just drop this database, at least not now. Maybe after he's made the videos he wants to make using the data lolol.
You could probably do it yourself in a day or two if you wanted. It's trivial to loop through the different search results with curl, but it would take a moderately long script to parse though the html for the statistics and then output them to a database. I don't even know much about html but I could look through the curl results and pick out how it writes the liked numbers and so forth (typography-). Also it seems Nintendo are blocking a lot of tor exits because they refused my connection when I used tor. So you would probably have to limit the rate of requests to avoid getting blocked. Which means actually scraping the site might take longer. If anyone feels like pitching in to help get a community script/database of SMM levels that is updated every so often let me know I might be willing to help write the script, but I'm too busy to do the whole thing myself.
@@koolaidjerk a quick search finds this comment: www.reddit.com/r/MarioMaker/comments/8y4ylq/is_there_a_database_of_mario_maker_statistics/e2c3p67 If any1 needed the data they could probably pm that person and ask for it. Sucks that his script isn't OSS afaik tho :/
The fact that you guys created a script like that just for this video really shows how passionate you are! Just discovered your channel but already loving it! Keep up with this awesome work man!
It's not always sarcastic, but I find it just as enjoyable either way... except for the time he went "HOOO cares" instead, the subversion was hilarious!
Lol great video Ceave! :D My buddy and famous hacker Jaku has a lot of data on levels, btw. Let me know if you need anything! FYI I plan to post a video this weekend showing people how to get some modded items uploadably onto their Wii U :) (Psycrow)
I'd really like to see a video where you show your code and explain how it works. I am a bit of a Rookie at Javascript so I would really appreciate seeing it being used in such cool scenarios.
Huge huge fan. It's been a while since a channel has me this enthralled on TH-cam and the incredible amount of work and detail you put into your videos really shine
As someone who really loves spreadsheet stuff, leader boards and ranking data and Mario, this video is like a wet dream. Amazing video! Also, it is all sooo worth it for that ending
@Doctor Whouse Holy shit you're right. ___ But in addition to feeling bad about myself, I also felt the need to punish myself for letting myself get into this situation at all. In other words, a masochistic rage ended up taking over my mind. I wanted to hurt myself physically, but that’s against my personal rules. So instead, I settled for psychological harm. So I went onto my Facebook account and unfriended about 80% of my friends, because I didn’t feel that I deserved to be friends with them. Later, I took a bath, but when it was over, I ended up staying in the tub cold and naked for hours, since that’s what I deserved. During this time, I also ended up going on a mental monologue detailing how worthless I was by going over all my past failures and shortcomings, and for each one going over my personal traits that caused those failures. Needless to say, this was a very dark period of my life. This is the type of situation where under normal circumstances, I would channel some mental resources into making myself feel better, such as reassuring myself that I would get through this and that everything would be okay. But did I do this under the rolling rocks mindset? Nope. Remember, the rolling rocks mindset had been created specifically to get the ABC videos done, especially the rolling rocks video that had been eating away at me for months. So instead of channeling any resources into making myself feel better and combat the masochistic rage, the rolling rocks mindset ended up taking a more exploitative approach. The rolling rocks mindset managed to convince my masochistic rage that the perfect punishment for me would be to force me to work on the rolling rocks video. The logic was that I didn’t deserve to relax and enjoy my winter break. Instead, I deserved to slave away all day on the rolling rocks video, even though I hate commentating. While all my friends and family would be having fun over winter break, I deserved to be locked up in my room just working all day. And that was how the rolling rocks mindset (whose goal was getting ABC videos done) and my masochistic rage (whose goal was punishing me) ended up working together. ___
I found your stuff like 3 months ago and now I'm binging the hell out of it, love the effort you put in them and the neat little trick you show in every vid, keep up the good work!
Your number of subscribers has exploded lately. I’ve been watching for years and can’t believe we came this far. Hands down one of the best channels on TH-cam
Great video, as always! I remember encountering the top rated stage during a 100 Mario challenge. I was so annoyed that I had to reset the game since I couldn't just skip it..
The objectively best level will be a "don't let the Goombas drop" level where Mario has to prevent Goombas from hitting the ground by placing a spring where it can bounce up the Goomba. That way Mario has to escort Goombas from the left tower to the right tower. Occasionally bob-ombs will be spawned instead, which will - instead of disappearing like Goombas will - be thrown back to the left tower and blow up part of it. 3 bombs and you win. After every bomb the spawn rate increases. Contraptions required include: - Don't let Goomba/bomb hit the ground (that then also prevents picking up). Seems hard, as bumpers seem to be the only SMM thing that enemies can pass and Mario cannot - but that's then be a weird hybrid with New Junk City :) and it would also require throwing the spring up and catching it to push the Goombas forward, which seems rather unforgiving. - Distinguish Goomba from Bob-omb (a lazy solution could involve small Goombas and big bombs, or vice versa, making them easier to take apart by a contraption). Goombas fall offscreen, bombs get lit and thrown back by a spring. - Spawn Goombas with an occasional Bob-omb in between. - A note block loop playing "Italian Medley". - Ideally spinjump-killing the Goombas should be forbidden. I bet +JoCr2 aka Psycrow will _love_ this level. Too bad I absolutely can't think of how to make a "don't allow Goomba to drop to the ground" contraption. In SMW a first idea could be using switchable blocks for Mario to walk on, as enemies fall through these (and from there can trigger note blocks) - but then Mario would even have to throw the spring up to bounce the Goomba up which would be annoying. Not sure how to get the enemy spawning and the bomb/Goomba distinguisher in SMW though. Basically at the end this would be a port of the Earthworm Jim level "Puppy Love". If that has been ported to SMM, SMM is complete.
He didn't make this video, inside a mind did. Check his channel for the compliments. Inside a mind does a lot of research for channels like game theory and checkoutthis.
When examining the threshold for how many plays a level should have, you should view it the same way you determine sample size. For example: I want a 95% confidence level., I want a 0.5% confidence interval (margin of error), As of March 2018, 4 million copies of Super Mario maker have been sold, Therefore I can say (based on those numbers) that a level with at least 38,051 (or just 38 thousand) plays is representative of what the entire community thinks of the game with 95% confidence and 0.5% margin of error. 20k is almost half of 38k, so I'd argue that's awfully low even though it sounds high. The point of finding a proper sample size is that you can say with confidence that it covers the opinions of everybody, or that it's representative. It's possible that levels with fewer plays really would be more objectively liked, but without the amount of plays to back that up it's kind of a stretch.
What you're missing is that not every play is random. People can manually find levels, and certain levels are going to draw different crowds of players. For example, that 38,051 players might not be a sample of the entire Mario Maker community, just a sample of the fanbase of an internet personality.
I was so afraid near the beginning that you wouldn't write a script/get someone to write a script to aggregate all the data you needed. Thank you for not killing my soul.
Big point that needs to be adressed. When players search for a level to play they search can search by catagory. What you are really measuring is "what is the level that best pleases the people who used search terms that would return it as a result?" For example if the vast majority of players HATED levels of type x, and would never even begin to play them, they wouldn't count towards the stars per pay metric as they would never open the level to begin with. While players that do like level x may often star that level because it is exactly what they were looking for. In this way, a level that most players hate could have the best star to play ratio.
20,000 plays isn't high enough, that's why it's messed up. Mario Maker has sold 4 million copies not counting 3DS sales. A more sensible threshold would be 40,000 - 1% of the WiiU playerbase. If a level is good than word is going to spread and it should have a lot of plays. Mario Maker has sold almost 7 million copies, accounting for people owning the game on both consoles and those who didn't really play the game, 40,000 should be representative of 1% of the people who actually played the game. This would exclude all 3 levels. The first level is clearly not supposed to be there and ,as others have pointed out, the 2nd place level is a rip off of a more popular level (~100,000 plays) and isn't really a good mario maker level. It's a concept that is used in many games just created in Mario Maker. It's a good idea, not a good level.
Except that's not how it work either, you cant find the objectively best levels, and increasing the required amount of plays won't help, there are hidden gems stuck at a few plays, especially now that the community size decreased. What you'll find will just be a list of the most popular levels
I almost immediately went into the comments to post about this! I hadn't known which level was created first but I guess it's nice to know people appreciated the concept?
I thought it would not happen but after they said in a Pokémon let's go they said "If you're on the *handheld* mode of the switch you can change the hairstyle." which means it wouldn't be a problem to use the touch screen!
It's not possible to have an objectively "best" video game level. But there are objective measures and concepts we can try to implement into our levels that will generally improve them: Variety, scaling to the skill level of the player, etc.
I already didn't like the idea if trying to find the objectively best level ever, fir I believed it still had tobe made or will never be made because there will alwayscome better ones. But this, this truly convinced me that there will never be a better level.
I laughed so hard on when you said some trashy one letter stage was the best and the outro music began playing. What makes it so much funnier is that it's just called "d"
the explonation for the nr.1 best level is that the player can't restart a level when he is in the air. this also counts when he is stuck in between bumpers. most people don't know that you can't restart it and instinctivly try to do so, skipping the "start over" button and giving the level a star.
Haha nice video! So for the number one, I believe an actual translation of the name means "in many different sense, this is hell- please play this as a time attack challenge" hahah I guessed many people wanted to beat the best time maybe?
![[TH] VCT Pacific - Playoffs - Knockout Round // TS vs T1 | PRX vs TLN](http://i.ytimg.com/vi/23EjntCWm_g/mqdefault.jpg)
No 3 best stage - "it's amazing"
No 2 best stage - "it's amazing"
No 1 best stage - "What the %$#!"
Perfekt Portrait our society
Hey lie it is not the best
yep
I'm amazed that mechabowzilla is not on the list.
"Statistics never lies": a statistician drowned crossing a river with an average depth of 1 meter.
Flamewalk dude that is so weird. There is no way that a river could be 18 meters deep in one area and 1/4 meter deep in other areas
hallucinacaust near the banks?
hallucinacaust that is a joke, dude. It is supposed to illustrate that statistics alone sometimes can give unexpected results.
That's why you walrus take the median, it's a lot more reliable :p
But this doesn't disprove the claim though... The statistician never claimed you couldn't drown because the average is 1 meter, you projected that.
The objectively best level's name translates into 'Utter Hell (please speedrun)'.
Or more literally, 'hell in many meanings of the word'.
Why do you have the icelandic flag on your PFP
Very Niceu why not
@@melon4345 LOL. It's a combo of the Empire of Japan and the Union Jack. Because the countries they represent are both a big part of my life.
Please make a video of you saying hooray for 10 hours
I want a compilation of "ouching spikes," "yum yum mushrooms," and "it's actually surprisingly simple."
who rae
Hoo Rae
Hehe
20 hours
In 100 man, you can't skip if you aren't on the floor / at the death animation. Bumpers does not count as a floor, so you literally can't skip this stage. I guess the players tried to star it to be able to skip it,but that's not how it works. Even if you're very limited, I think it's possible to do a decent level that is unskippable but still fun to play, then I guess your level will get a reasonable amount of stars. Good video as always Ceave :p
Is the clear rate also really high on this? That would mean it would be put on easy 100 man so newer players would run into it and be more likely to star it not knowing that doing that doesn’t work.
holy crud that's evil.
So really the #2 is the actual #1, then
If the timer were to run out could you pause and then skip?
Is this a tie for the worst level ever?
You CAN skip in the death animation actually.
These levels have stars, cause you can’t die and if you play the level but press start over or quit it doesn’t count as a play. It only counts as a play if you die or complete the level and if you can’t die it’ll only count as a play if you complete the level. These ”best levels” are long so you’d exit the level if you don’t like it and if you complete a long level it’s pretty likely that you liked the level and give it a star. I know that was complicated and had poor grammar but I hope you understood.
Oh ! Yeah i get it ! That is... disappointing :c
Thank you, I wanted to point this out too.
Well that sucks. Maybe he'll make a part two?
Oh! So the trick is not to make a level that would be starred by the most people, but to make a level that is only completed by the people that would star it! The bumper level seems to have done a good job hitting that target, since people that don't like it will quit immediately instead of completing it.
So you should count the plays and deaths and get from that the star to like ratio
There is a confounder here; you can only star a level you finished. For example, a very hard level may be hated by most people that play it. Only a very specific subset of people that finish it, i.e. skilled people or emotionally invested people, finish and rate it. Such a level may get a high ratio, but still not be enjoyed by the average player / majority of players.
If you correct for this by dividing not the amount of unique players but the amount of mario lives spent by the amount of stars, you will only get levels in which players can not die.
Either way, you still don't know which level got the most amount of enjoyment per play or per player.
And by dividing the amount of time spent by the amount of stars, you will only select short levels... this is a much harder problem than it seems. The best way would be to let everyone rate a level, even those that don't finish a level. But then, people that are less inclined to rate a level will be underrepresented. If people that rate levels are more conscientious (for example) than those that don't, your results will be skewed towards that metric, and still not represent the 'average' player (selection bias). So you need to ensure everyone rates every level they play at all times.
(This is a perfect example of the effect of confounders, and exactly what happens in bad medical research, where it actually costs lives.)
This should be higher. Kinda bummed it wasn't mentioned in the video, great real world example of why statistics often requires careful thought (and how it can easily be manipulated).
madmaniaq Couldn’t this logic also be used for any kinds of media?
This isnt true though, you can pause a level a star it at any time
THIS MAN IS THE CHOSEN ONE
@@theAstarrr both games allow you to like or dislike (in the case of smm2) a level at any time
I just watched the no coin run video and now i'm getting anxiety every time you get a coin
Me too!
I have the same thing but with getting hit in mega man
I watched Gamechamp9000s jumpless super mario odyssey video and I've never quite been the same since
I feel like someone’s doing magic when they do something like that after watching a challenge video.
Lolol me too!
I mean all of those stages were pretty good but in my opinion the best level up to this date still is d by Dean
lela
@The_Mighty_ Boole
I know right! It is so original, and is very cool.
I'll bet Dean is an executive from Nintendo, because he has made the Mona Lisa of Mario Maker Levels.
d is the best fam
Even the name is beautiful
Can someone link me to a video playthrough of that level? Can't seem to find it.
*WELL IT'S ACTUALLY SURPRISINGLY SIMPLE!*
Hooray
Bumping Speed for 12 hours?
222 likes :0
*PLOT TWIST*
Yeah it's actually not that simple.
Huuu-RAY
Everyone's favorite squidgirl: Squidgirl
As a Splatoon 2 fanboy, I couldn't stop laughing
Not to be confused with Squid Girl, from the Squid Girl mangal/anime
Epic_stuff squidgirl isn’t from splatoon 2
Which is probably why he's laughing, Bum... he's saying that the Splat 2 girl is his favorite, or at least moreso than the Splat 1 girl. Or maybe he likes Pearl more, or one of the Sisters?
Personally, my favorite is Callie, though I should note that Octolings aren't technically squid girls...
ShadwSonic I know
ShadwSonic you filthy Callie loyalist...
Of course, we could have thousands, even millions of alternative universes where the "best" Mario Maker level is good...
But in this one, it's just...bumpers.
yea like... bumpers + coins in the middle + saws at the bottom = best level ever...or it would have been..
In order to answer that we need to talk about parralel universes
Theres infinite universes. so yes theres some good levels out there, somewhere....
@@calebd512 citation needed.
@@ProjjectDelta. which you can't spell
Fell into a big fat TH-cam hole of your videos. Keep up the awesome work dude
same also hi :)
Hey dude im subbed
"...to not over-strain Nintendo's VERY fragile servers" lol. The way its put is awesome.
2:05 If I didn't know how long the video was I would have actually thought that was it! He ended it with the music and even the sacred *Hooray*!!!
NOT THE SACRED HOORAY
OOF level: 5 oofs of 5 oofs
@@momchan4917 ?
Wait, Devoun?
GORG PAG
The translation for the Japanese level at the end is basically "This is several variations of hell. See if you can get the best time"
You forgot "please".
And I'd go with "There are various ways in which this level is hell. Please try for the best time."
2:10 Well, that was actually suprisingly simple.
Jojo, I was here first, and I made that joke first!
Jk is good
Dich seh ich doch auch oft bei Domtendo!
Was machst du denn hier? Guckst du auf TH-cam nicht ausschließlich Domis Videos?😂
Hallo deutscher Justin Y.
d
Hey, data scientist here.
I truly enjoyed this video, the fact that you gathered this dataset is awesome. Would you make it openly available?
Regarding the grading, I believe you could have used some bayesian posterior mean formula to have a more significant result. For example, the one used originally by IMDB (according to wikipeida) could be slightly modified to get:
Weighted rating for a given level = (R*v + C*m)/(v+m)
With R = percentage of stars, v = number of players, m = threshold to be considered (what you set at 20k), C = mean of all percentages of stars across all dataset.
I wonder if you’d get the same results using this formula.
I came looking for the data after watching the video on my TV. Please share the dataset!?
Keep asking; maybe email
NEEEEEEEEEEEEEEEERRRRRRRRRRDDDDDDDD!!!
(I'm kidding, your math stuff confuses me but at the same time it's very cool. Good on you my dude!
The scraping of this dataset isn't horribly complicated (the most time consuming part is having to throttle reads, avoid duplicates and deal with UTF8 characters). If Ceave hasn't responded to you yet, let me know and we can discuss getting this dataset.
You wish you were
So... The key to the best stage is bouncy stuff? (bumpers and springs)
Might explain all the video game companies making pachinko machines.
It explains Sonic's popularity.
You cannot skip a stage while in mid-air, and trying to hit Restart when pissed off at being stuck will often make you star something on accident. That level traps Super Expert players until they beat it or lose every single one of their lives. Over such a long time, it is extremely likely to star it on accident.
can’t exit while in the air. on bumper = in air. L O G I C
@@Zorgdub it's annoying to me that you can't change your vote after the fact. At the least there should be a grace period where you can do so.
That fake ending made me laugh so hard.
d
_And we can see that this is the BEST rated Super Mario Maker Level. I hope you enjoyed this little video, and if y-_
YES it made me laugh way more than I should have
Congrats dean!
😂
maybe it got bumped to the top?...
Boo
stop yourself
Stop it
why
BADUM-TSHHHHH
This is the objectively the best SMM video
Lol
no
You are objectively wrong
How could you say that objectively
@@snekk9701 burden of proof
This is basically basic statistics mixed in with mario maker.
But I really, really like it. This guy knows what he is doing. This is honestly a rarity in these sort of videos.
The reason I subscribed.
U'RE RIGHT!
At 3:19, three of my levels showed up on your bookmarks page.
bruhhh
@bruhhh SMM /NiatsuFTW_Ros can you give me the ID, please?
Koopaling Island is great btw
Jesus that's amazing
Id i play them
If you would show the Level without any context, I would think it´s the worst Level in Super Mario Maker XD
Eh, there were worse. Way worse.
Yes, but it also don´t really looks like beeing the best Level
It was probably created by a streamer with a sizable following and gets a laugh out of most people in 100 Mario challenge but it is still beatable
For spot of best stage in creativity, I would nominate "Pause for Effect." It uses puzzles that you have to solve by pausing the game at the right time.
Agreed on "Pause for Effect." This is why when it comes to judging, subjectivity/quality has to play a part over objectivity/quantity. That said, what would be the highest liked over views video for Mario Maker videos on TH-cam? That would give a different source of data all together (which would include non-Mario Maker players like myself :P) (that would also increase the demographic pool)
Dude I laughed so hard when you said "Either I just don't get the level, or we just officially found the limits of statistics"
LOLOL You're the man dude.
Man i remember back when Mission Impossible was at the top, that is my favourite stage
It's actually still the most starred stage, so it still reigns supreme.
I love the way you say hooray, it seems a bit happier than the regular way of saying it. :)
Can you share the Database so other people can look through it?
you can just scrape their website yourself
@@nope4180 that's hella work though. You could make it a lot faster by running multiple instances through different vpns or tor exits, but youd still have to write the script
He said he's going to make more content using the statistics and he linked to the levels without actually linking the database so I doubt he's going to just drop this database, at least not now. Maybe after he's made the videos he wants to make using the data lolol.
You could probably do it yourself in a day or two if you wanted. It's trivial to loop through the different search results with curl, but it would take a moderately long script to parse though the html for the statistics and then output them to a database. I don't even know much about html but I could look through the curl results and pick out how it writes the liked numbers and so forth (typography-). Also it seems Nintendo are blocking a lot of tor exits because they refused my connection when I used tor. So you would probably have to limit the rate of requests to avoid getting blocked. Which means actually scraping the site might take longer. If anyone feels like pitching in to help get a community script/database of SMM levels that is updated every so often let me know I might be willing to help write the script, but I'm too busy to do the whole thing myself.
@@koolaidjerk a quick search finds this comment:
www.reddit.com/r/MarioMaker/comments/8y4ylq/is_there_a_database_of_mario_maker_statistics/e2c3p67
If any1 needed the data they could probably pm that person and ask for it. Sucks that his script isn't OSS afaik tho :/
Ceave Gaming keeps Mario Maker alive! It's so great to watch videos dedicated to the game.
Ceave keeps mario maker alive, J-Bizzle keeps mario maker poppin!
And BTG
@@Flare0227 yes agreed
The fact that you guys created a script like that just for this video really shows how passionate you are!
Just discovered your channel but already loving it!
Keep up with this awesome work man!
I subscribe because of the sacarstic "Hooray".
It's not always sarcastic, but I find it just as enjoyable either way... except for the time he went "HOOO cares" instead, the subversion was hilarious!
Hoo... *(A Googolplex years pass by)* ...ray
*horayyy
Sarcastic?
"There are three types of lies : Lies, damned lies, and statistics."
- Graham McNeil
(... the fourth one is the cake) ;)
No. The cake is truth
Empire player eh?
"Lies, damned lies, and statistics."
HA
there's also true lies
Lol great video Ceave! :D My buddy and famous hacker Jaku has a lot of data on levels, btw. Let me know if you need anything!
FYI I plan to post a video this weekend showing people how to get some modded items uploadably onto their Wii U :) (Psycrow)
I'd love to get my hands onto more data! :D
I'm looking forward to your video, that sounds amazing!
I'd really like to see a video where you show your code and explain how it works. I am a bit of a Rookie at Javascript so I would really appreciate seeing it being used in such cool scenarios.
@@CeaveGaming lol no
@@CeaveGaming It's easy! Sort by highest star-to-play ratio BUT exclude the 100%s.
No crow give us the database!!!!
I love your intro, it's so iconic to me and I started watching you like a week ago
Huge huge fan. It's been a while since a channel has me this enthralled on TH-cam and the incredible amount of work and detail you put into your videos really shine
Trials of Time is the best level ever made.
Please no XD
Get yt out of your name if i wanted to check out your channel i would click on your name. The yt makes your name look dumb.
i agree
Why are y'all roasting this guy's name? Leave him alone.
Hi ytsunny
As someone who really loves spreadsheet stuff, leader boards and ranking data and Mario, this video is like a wet dream. Amazing video!
Also, it is all sooo worth it for that ending
This video made me realize that you're the panenkoek of Super Mario Maker
Except he isn't obsessive about quality to the point of mental self-harm (I love pannen, but his FAQ was a rather sad read).
Uh, DUH. They both rock.
@Doctor Whouse Holy shit you're right.
___
But in addition to feeling bad about myself, I also felt the need to punish myself for letting myself get into this situation at all. In other words, a masochistic rage ended up taking over my mind. I wanted to hurt myself physically, but that’s against my personal rules. So instead, I settled for psychological harm. So I went onto my Facebook account and unfriended about 80% of my friends, because I didn’t feel that I deserved to be friends with them. Later, I took a bath, but when it was over, I ended up staying in the tub cold and naked for hours, since that’s what I deserved. During this time, I also ended up going on a mental monologue detailing how worthless I was by going over all my past failures and shortcomings, and for each one going over my personal traits that caused those failures. Needless to say, this was a very dark period of my life.
This is the type of situation where under normal circumstances, I would channel some mental resources into making myself feel better, such as reassuring myself that I would get through this and that everything would be okay. But did I do this under the rolling rocks mindset? Nope. Remember, the rolling rocks mindset had been created specifically to get the ABC videos done, especially the rolling rocks video that had been eating away at me for months. So instead of channeling any resources into making myself feel better and combat the masochistic rage, the rolling rocks mindset ended up taking a more exploitative approach. The rolling rocks mindset managed to convince my masochistic rage that the perfect punishment for me would be to force me to work on the rolling rocks video. The logic was that I didn’t deserve to relax and enjoy my winter break. Instead, I deserved to slave away all day on the rolling rocks video, even though I hate commentating. While all my friends and family would be having fun over winter break, I deserved to be locked up in my room just working all day. And that was how the rolling rocks mindset (whose goal was getting ABC videos done) and my masochistic rage (whose goal was punishing me) ended up working together.
___
SonicFanChris 4 At least he doesn’t scuttlebug transport, beat a level in 0.5 a presses or go to parallel universes
I'll get to my opinion on that, but first we have to talk about parallel universes.
I love the way he says "hooray"
I really liked the objectively worst level, so I'm probably going to hate the objectively best level.
No, some Japanese person turned out to have it.
*hate
TripleJump *hate
...hmmm....
@@albond
I really have autocorrect.
...I MEAN HATE! DANGIT! 😂
I found your stuff like 3 months ago and now I'm binging the hell out of it, love the effort you put in them and the neat little trick you show in every vid, keep up the good work!
HOLY SHIT that joke at 2:06 made me laugh way more than I should have, GENIUS!
I'm always so fascinated and interested by the content you put out. Oh and charmed of course too!
Now we have the Ultimate Answer of Life, the Universe and Everything
Bumpers.
So 100bumpers=42
That's the same
what if thats the question?
Rubber baby buggy bumpers
The level that I hate the most is most popular?
Nice to know
Not most popular... Objectively best.
so be glad to know you have better taste than 24% of all SMM Players
5:44 it would take me about that long to make a video of this quality, well done again!
Your number of subscribers has exploded lately. I’ve been watching for years and can’t believe we came this far. Hands down one of the best channels on TH-cam
7:22 Everyone's favorite squid girl...
*Squid Girl*
CONGRATS ON HITTING 150K MAN, YOU'RE THE BEST TH-camR EVER.
You're awesome man, I really enjoy your videos!
I don't know much about super mario maker but i DO know that this is the objectively best youtube channel i've ever been recommended. Love your videos
"Non-ouching, but bouncing obsticles". 😂
THE OBJECTIVLEY BEST LEVEL IS UNSKIPPABLE
Great video, as always!
I remember encountering the top rated stage during a 100 Mario challenge. I was so annoyed that I had to reset the game since I couldn't just skip it..
We need more videos using this database, there is so much potential now that you have all this information !
Just watched your “completing Super Mario while getting 0 coins” and now every time you touch a coin I have a mini heart attack
Well that was anticlimactic.🙃
The objectively best level will be a "don't let the Goombas drop" level where Mario has to prevent Goombas from hitting the ground by placing a spring where it can bounce up the Goomba. That way Mario has to escort Goombas from the left tower to the right tower. Occasionally bob-ombs will be spawned instead, which will - instead of disappearing like Goombas will - be thrown back to the left tower and blow up part of it. 3 bombs and you win. After every bomb the spawn rate increases.
Contraptions required include:
- Don't let Goomba/bomb hit the ground (that then also prevents picking up). Seems hard, as bumpers seem to be the only SMM thing that enemies can pass and Mario cannot - but that's then be a weird hybrid with New Junk City :) and it would also require throwing the spring up and catching it to push the Goombas forward, which seems rather unforgiving.
- Distinguish Goomba from Bob-omb (a lazy solution could involve small Goombas and big bombs, or vice versa, making them easier to take apart by a contraption). Goombas fall offscreen, bombs get lit and thrown back by a spring.
- Spawn Goombas with an occasional Bob-omb in between.
- A note block loop playing "Italian Medley".
- Ideally spinjump-killing the Goombas should be forbidden.
I bet +JoCr2 aka Psycrow will _love_ this level. Too bad I absolutely can't think of how to make a "don't allow Goomba to drop to the ground" contraption. In SMW a first idea could be using switchable blocks for Mario to walk on, as enemies fall through these (and from there can trigger note blocks) - but then Mario would even have to throw the spring up to bounce the Goomba up which would be annoying. Not sure how to get the enemy spawning and the bomb/Goomba distinguisher in SMW though.
Basically at the end this would be a port of the Earthworm Jim level "Puppy Love". If that has been ported to SMM, SMM is complete.
The greatest of my congratulations to do this data base. You should share your excel file to us new comers to MM2 next June!!
Really? There isn't a single comment appreciating how deep he's gone into researches?
He didn't make this video, inside a mind did. Check his channel for the compliments. Inside a mind does a lot of research for channels like game theory and checkoutthis.
Thanks much!
Hmmm......
Hmmm......
Didn't kiavik come up with the concept of the 2nd best level of all time? The one where you have to search for the hidden blocks...?
2:05 I love how the music is getting louder and louder and Ceave is just doing his outro.
I love you and your madness, Ceave.
That is his signature theme song.
Cody Smith I know but it’s still funny 😄
Crackin video buddy! Keep it up
Wait a minute isn’t the time machine the most popular if it had 65076 plays and 12430 likes? At 11:27 btw for those wondering
19.1% play to like ratio.
P
@@SilverNitris the comment was from September 2018
@@SilverNitris He commented this on September 2018.
When examining the threshold for how many plays a level should have, you should view it the same way you determine sample size. For example:
I want a 95% confidence level.,
I want a 0.5% confidence interval (margin of error),
As of March 2018, 4 million copies of Super Mario maker have been sold,
Therefore I can say (based on those numbers) that a level with at least 38,051 (or just 38 thousand) plays is representative of what the entire community thinks of the game with 95% confidence and 0.5% margin of error.
20k is almost half of 38k, so I'd argue that's awfully low even though it sounds high. The point of finding a proper sample size is that you can say with confidence that it covers the opinions of everybody, or that it's representative. It's possible that levels with fewer plays really would be more objectively liked, but without the amount of plays to back that up it's kind of a stretch.
What you're missing is that not every play is random. People can manually find levels, and certain levels are going to draw different crowds of players. For example, that 38,051 players might not be a sample of the entire Mario Maker community, just a sample of the fanbase of an internet personality.
I was so afraid near the beginning that you wouldn't write a script/get someone to write a script to aggregate all the data you needed. Thank you for not killing my soul.
This conclusion feels like when the mice learned that 42 was the Meaning of Life in The Hitchhiker’s Guide to the Galaxy.
*H O O R R A Y*
Crowtem „who ray“
I want that level to be my screensaver
Big point that needs to be adressed. When players search for a level to play they search can search by catagory. What you are really measuring is "what is the level that best pleases the people who used search terms that would return it as a result?"
For example if the vast majority of players HATED levels of type x, and would never even begin to play them, they wouldn't count towards the stars per pay metric as they would never open the level to begin with. While players that do like level x may often star that level because it is exactly what they were looking for. In this way, a level that most players hate could have the best star to play ratio.
Make a couple videos looking at interesting levels, or really strange outliers in the data. Great video btw
20,000 plays isn't high enough, that's why it's messed up. Mario Maker has sold 4 million copies not counting 3DS sales. A more sensible threshold would be 40,000 - 1% of the WiiU playerbase. If a level is good than word is going to spread and it should have a lot of plays. Mario Maker has sold almost 7 million copies, accounting for people owning the game on both consoles and those who didn't really play the game, 40,000 should be representative of 1% of the people who actually played the game.
This would exclude all 3 levels. The first level is clearly not supposed to be there and ,as others have pointed out, the 2nd place level is a rip off of a more popular level (~100,000 plays) and isn't really a good mario maker level. It's a concept that is used in many games just created in Mario Maker. It's a good idea, not a good level.
11:23 Then the best level would be "The Time Machine".
@@maxfi878 Name sounds promising.
That is a good level. Wouldn't say it's the best, but definitely "pls star twice" worthy.
Except that's not how it work either, you cant find the objectively best levels, and increasing the required amount of plays won't help, there are hidden gems stuck at a few plays, especially now that the community size decreased. What you'll find will just be a list of the most popular levels
Daniel Payne All your opinion, m8.
Find a level with lots of plays but not many stars
So the objectively worst level depending on the star rate.
It would be interesting to see what level has the highest # of plays yet 0 stars.
Probably Pit of Panga or something similar.
The amount of work and effort you put into these videos is amazing.
such fabulous content, keep it up!
Best youtuber love you!
that level likely is only the best because it was made by a Japanese celebrity.
@@swagerdager6931 Nice Google Translate :p
ママミアパパピアの赤ちゃんは下痢を持っていた！
Look at all of those ouching spikes on your profile picture.
So glad to see someone still playing these games brings back soooooooo many memories
This game is 3 years old... are you like 12? You know he is rated 17+ right?
Lol so are you a 50 year old man looking for pics of my belly button
Are you ready, leeeets do this.
I love this channel!
me thinks this Macaron guy is just a japanese e-celeb.
I think there are a LOT of peapole here who would love having acces to the database!
That level is just sublime, tears are flowing down my face just looking at it.
Greetings from Styria, nice video btw ^^
I love how the second "objectively" best level ever is a complete plagiarized level of "Nikki s Puzzling Picture" by Kiavik.
I almost immediately went into the comments to post about this! I hadn't known which level was created first but I guess it's nice to know people appreciated the concept?
@@MasterPadoo Yeah, but it's plagiarism, and even if this person liked the concept, it's not very cool to take his concept, with the same ideas.
Sady Still doesn’t change the core design.
The best level is the levels made by ceave
I had a feeling there was some sort of twist at the end, but this one still made me laugh out loud
You guys do a ton of work just to entertain us (and make money). Thanks.
Can you make a new Super Mario Maker 2 prediction video?
I thought it would not happen but after they said in a Pokémon let's go they said "If you're on the *handheld* mode of the switch you can change the hairstyle." which means it wouldn't be a problem to use the touch screen!
@@billyjoestuff7297 splatoon2 slove that with a gyro camera
Martti let’s hope there is not smm2
That is incredibly awkward
Well the objectively best level would be the most creative
Like how the objectively worst would be the least creative/and or impossible level
It's not possible to have an objectively "best" video game level.
But there are objective measures and concepts we can try to implement into our levels that will generally improve them: Variety, scaling to the skill level of the player, etc.
B. Moser All of which can be subjective to a degree.
I already didn't like the idea if trying to find the objectively best level ever, fir I believed it still had tobe made or will never be made because there will alwayscome better ones.
But this, this truly convinced me that there will never be a better level.
I laughed so hard on when you said some trashy one letter stage was the best and the outro music began playing. What makes it so much funnier is that it's just called "d"
NSMB 2 without collecting ANY coins XD
I didn't understand at first and then remembered nsmb2 is all about collecting coins
@@markasmar1097 thats the deal you cant make it
Didn't Nintendo Unity already do that run?
@@HyLion i dont know
th-cam.com/video/y0CVz897BfU/w-d-xo.html it's actually been done
First
Hope no one has a blue shell
Trip lol
Blue shell
well I failed
rip.exe
No... i have two :3
Great video! Love it 👍
the explonation for the nr.1 best level is that the player can't restart a level when he is in the air. this also counts when he is stuck in between bumpers. most people don't know that you can't restart it and instinctivly try to do so, skipping the "start over" button and giving the level a star.
I love your voice.
Welp, at least I can oficially say that my level is better than the best one...
Sadly playing in 2DS, so nobody will ever see it a-a'
I know how you feel. ;-;I play 3ds version too and I think my levels are a little better than the ones on the front page...
Haha same
Same. I'm recreating Castle Crush from DKC2. The size limitations are more crushing than the ceilings, though...
I used to design SQL queries for work. This video made me nostalgic.
Haha nice video! So for the number one, I believe an actual translation of the name means "in many different sense, this is hell- please play this as a time attack challenge" hahah I guessed many people wanted to beat the best time maybe?