mikheeeeeeeeeeeeeeeee <3
forum
Performance Points feedback and suggestions (Catch the Beat)
posted
Total Posts
775
Velps, illuminate us with your experience in that game on impossible/complex jumps. I think we can do, all together, a fair system that can detect complex jumps and compute it correctly 
Eh not sure if the ctb community considers it complex or not. But having a large non hyper jump right after a hyper jump can be more precise than if there was two regular large jumps. I have already mentioned that jump chaining is more difficult than regular standalone jumps, especially non-hyper chains. Hyper and non-hyper 1/2th node jumps that increase in spacing with each jump are quite tricky to time. Chains that suddenly become regular jumps after several hypers are slightly more difficult due to the first reason I presented. Jumps that start or end with horizontal mini-repeat sliders or wide wigglies are more precise, because you need to keep the ryuuta in a particular place to avoid missing a fruit when going for the jump.
Eh that's all that I can think of at the moment.
Eh that's all that I can think of at the moment.
Exactly. Few examples:TheVileOne wrote:
Eh not sure if the ctb community considers it complex or not. But having a large non hyper jump right after a hyper jump can be more precise than if there was two regular large jumps.
They are even so precise that you need huge amount of luck to hit them properly
Maps like (Sticks and Stones) became maybe even close to impossible to do because of the incredible accurate timing you need
So this kind of maps should give a good amount pp if you manage to FC them
I agree with mikhe.mikhe wrote:
then try to focus on jumps complexity instead of jumps distance, it's true that my map, romaji voice is not that hard, because the jumps are big but not complicated.
Now the problem is to know how to change the algorithm to detect "complex" jumps, it's hard point because the concept "complex" referring to jumps is not defined so good i guess.
Edit: I want to say it's hard to compute mathematically a complex jump because is hard to detect only with formulas.
so i think pp calculation should based from how "complexity" the jumps instead of distance.
but the main problem, to explain the "complexity" itself is more than hard.
also,
based from TheVileOne's quote. I also want to mention, there are many jumps are harder or chained jumps instead of distance jump.TheVileOne wrote:
But having a large non hyper jump right after a hyper jump can be more precise than if there was two regular large jumps. I have already mentioned that jump chaining is more difficult than regular standalone jumps, especially non-hyper chains. Hyper and non-hyper 1/2th node jumps that increase in spacing with each jump are quite tricky to time. Chains that suddenly become regular jumps after several hypers are slightly more difficult due to the first reason I presented. Jumps that start or end with horizontal mini-repeat sliders or wide wigglies are more precise, because you need to keep the ryuuta in a particular place to avoid missing a fruit when going for the jump.
also, based from my experience. there are many sides to see & to compare for the "complexity" jump itself .
so, my suggest. how about making a new pp convertion maps based from CTB maps(ranked/pending or even **** ctb map) or Some maps that "have complex jump". I think with that method, player can understand & give their feedback & suggest better to explain the "complexity" itself.
also. I think with this method will solve more problem than "pp calculations"
These types of patterns should be factored into the difficulty level of the map to begin with. The amount of pp a map is worth is based on this difficulty rating. I think that complex patterns are already considered towards the difficulty value. This value determines how much pp a map is worth. There is already calculations going on, but we don't know the exact values of maps yet.
I dont think that the difficulty is calculated for the CtB mode, i think the stars (the difficulty of a map) is calculated like an osu! standard map. Not totally sure, but it was how it works in the past. (If someone knows, correct me pls)TheVileOne wrote:
These types of patterns should be factored into the difficulty level of the map to begin with. The amount of pp a map is worth is based on this difficulty rating. I think that complex patterns are already considered towards the difficulty value. This value determines how much pp a map is worth. There is already calculations going on, but we don't know the exact values of maps yet.
One last thing, we don't need to know the algorithm, the exact pp of a map, we ONLY should know how it works, the future should be: "I see that pretty complex jumps in that map, i'm sure that this map give me pp"... *after some play* ... "I got X pp, nice
". There are some differences between to know the EXACT value and to know that SOME pp will be added to your profile.Now, i want to say something about the algorithm. I think is wrong, because it judges the pp amount of a map in base of the quantity of jumps and the distance between this jumps. What is wrong? Easy, the jumps are so overrated that some player can do HR+HT or EZ and get one of his top ranks, some examples:
- RAMPAGE88 with Team Nekokan - Can't Defeat Airman (hid+HR+HT)
- Dark Diego with the same (hid+HR+HT)
- ExGon with IOSYS - Cirno's Perfect Math Class (EZ+FL)
- ExGon with Yousei Teikoku - Kokou no Sousei (HR+HT)
There are 3 pro players, i'm sure there are so pro, but they has a huge valoration of his skills in those maps with HT and EZ!! And they have so many records even better than these.
Solution, in my opinion, first of all we have to nerf HalfTime and Eazy mods in the pp calculation. Maybe a solution could be that if you use Eazy you will get 0.5*pp and if you use HT 0.3*pp (like the nerf in the score when you use these mods).
Secondly we should open our mind, a map is not neccesary harder if we have 100 jumps, even if +50% of jumps are with hyperdash. Why is a map harder than other one? I think, always is my opinion, if you do a map with insane streams+jumps it's soooo hard, for example 44teru-k remixed wa. - Higanbana Daini -Shuusei no Lycoris-, even if you don't have "jumps" properly, but your map have streams and can't get the fruits simply stand quiet and you have to move so fast to get or to be lucky to get without any movement, for example Shounen Radio - neu.
So, i guess you don't have to do (i'm talking with you Tom) an algorithm that try to count jumps, distance and calculate pp (ofc i suppose that you count the mods, DT, hid...etc in the algorithm). I think we should create an algorithm which can detect patterns (using a data base for example). Now the point is... how many pp are fair to each pattern? I really don't know at the moment, but all the community of CtB can help us in that, i think we can work from this point all together
I'm waiting some response
(positive or negative but always with respect!)Edit: I'm sure we can find more examples of complex patterns in some CtB maps, please help me people, say some!
mikhe is right about everything there
Example
It looks like you can just stay in the middle to catch the fruits, but you just can't. It's like in between two movement frames, so that you can't quite position yourself in the correct place to catch all the fruits, you are forced to move back and forth for each fruit.
It looks like you can just stay in the middle to catch the fruits, but you just can't. It's like in between two movement frames, so that you can't quite position yourself in the correct place to catch all the fruits, you are forced to move back and forth for each fruit.
You can stay in the middle actually, you just have to be precise, I've hit that part many times by just standing still.
That's a better example of you want to sayTheVileOne wrote:
Example
It looks like you can just stay in the middle to catch the fruits, but you just can't. It's like in between two movement frames, so that you can't quite position yourself in the correct place to catch all the fruits, you are forced to move back and forth for each fruit.
And IT IS impossible to pass only standing quiet. One thing more, the players who has Fc'ed, they don't have this map as a top rank... is it logical? i don't think so... they must have this like one of their top ranks. Conclusion: The pp system doesn't count this type of patterns as "insane".
It isn't the purpose of this threadDakkyChan wrote:
Is that so hard to fix the hyperjumps again ?
. We want to get feedback about pp system, not about hyperjump system.Be more clear with your idea, because i can understand it like to cope with this pp system to change it or to adapt to it.Seph wrote:
Is trying to cope with change that hard?
DeathXHunter
Yep, you are right, the rest of the map is pretty easy... i dont know how to weight that, but at least something i think.DeathXHunter wrote:
Is it worth to be in the high top performance just because they managed to do that 1 pattern? The rest of the map is pretty easy,imo.
Topic Starter
The algorithm isn't just counting the jumps. It'd mostly taking the hardest patterns, which are currently determined by jump speed, and I agree with you that it isn't very optimal. Thanks for all the input, also to all other people who posted here, and I hope to incorporate that into the algorithm at some point.mikhe wrote:
I dont think that the difficulty is calculated for the CtB mode, i think the stars (the difficulty of a map) is calculated like an osu! standard map. Not totally sure, but it was how it works in the past. (If someone knows, correct me pls)TheVileOne wrote:
These types of patterns should be factored into the difficulty level of the map to begin with. The amount of pp a map is worth is based on this difficulty rating. I think that complex patterns are already considered towards the difficulty value. This value determines how much pp a map is worth. There is already calculations going on, but we don't know the exact values of maps yet.
One last thing, we don't need to know the algorithm, the exact pp of a map, we ONLY should know how it works, the future should be: "I see that pretty complex jumps in that map, i'm sure that this map give me pp"... *after some play* ... "I got X pp, nice
Now, i want to say something about the algorithm. I think is wrong, because it judges the pp amount of a map in base of the quantity of jumps and the distance between this jumps. What is wrong? Easy, the jumps are so overrated that some player can do HR+HT or EZ and get one of his top ranks, some examples:
- RAMPAGE88 with Team Nekokan - Can't Defeat Airman (hid+HR+HT)
- Dark Diego with the same (hid+HR+HT)
- ExGon with IOSYS - Cirno's Perfect Math Class (EZ+FL)
- ExGon with Yousei Teikoku - Kokou no Sousei (HR+HT)
There are 3 pro players, i'm sure there are so pro, but they has a huge valoration of his skills in those maps with HT and EZ!! And they have so many records even better than these.
Solution, in my opinion, first of all we have to nerf HalfTime and Eazy mods in the pp calculation. Maybe a solution could be that if you use Eazy you will get 0.5*pp and if you use HT 0.3*pp (like the nerf in the score when you use these mods).
Secondly we should open our mind, a map is not neccesary harder if we have 100 jumps, even if +50% of jumps are with hyperdash. Why is a map harder than other one? I think, always is my opinion, if you do a map with insane streams+jumps it's soooo hard, for example 44teru-k remixed wa. - Higanbana Daini -Shuusei no Lycoris-, even if you don't have "jumps" properly, but your map have streams and can't get the fruits simply stand quiet and you have to move so fast to get or to be lucky to get without any movement, for example Shounen Radio - neu.
So, i guess you don't have to do (i'm talking with you Tom) an algorithm that try to count jumps, distance and calculate pp (ofc i suppose that you count the mods, DT, hid...etc in the algorithm). I think we should create an algorithm which can detect patterns (using a data base for example). Now the point is... how many pp are fair to each pattern? I really don't know at the moment, but all the community of CtB can help us in that, i think we can work from this point all together
I'm waiting some response
Edit: I'm sure we can find more examples of complex patterns in some CtB maps, please help me people, say some!
When can we expect a more expansive list of difficulty values?
I would like to evaluate difficulty further, but it's very difficult when I cannot confirm how hard a map is compared to another map. It's pretty clear what some of the hardest CTB maps are, but it's not so clear where the maps that fall under them are. Maybe provide a difficulty calculator for us to play with? We can then manually choose CTB maps and have the difficulty value be calculated via an option.
Also I hope that you realize that mods can change hyper jumps and the difficulty of certain patterns. Hopefully mods like Hard rock give pp based on the extra difficulty given to the map, which can vary very greatly from map to map. It is not as simple as considering that the difficulty settings get increased, the random placement of fruits can add hyper jumps and very unusual and awkward patterns.
I would like to evaluate difficulty further, but it's very difficult when I cannot confirm how hard a map is compared to another map. It's pretty clear what some of the hardest CTB maps are, but it's not so clear where the maps that fall under them are. Maybe provide a difficulty calculator for us to play with? We can then manually choose CTB maps and have the difficulty value be calculated via an option.
You have to do that pattern twice in that map. Well at least you had to do it in the stream compilation twice and in that case it's more than halfway through a 10 minute drain map (12 minute long song). Anyways should be classify as a very hard pattern. Other patterns like it need to be considered as well. They fall under chain jumps.DeathXHunter wrote:
Is it worth to be in the high top performance just because they managed to do that 1 pattern? The rest of the map is pretty easy,imo.
Also I hope that you realize that mods can change hyper jumps and the difficulty of certain patterns. Hopefully mods like Hard rock give pp based on the extra difficulty given to the map, which can vary very greatly from map to map. It is not as simple as considering that the difficulty settings get increased, the random placement of fruits can add hyper jumps and very unusual and awkward patterns.
Question about how difficulty is calculated for pp, do streams lower the difficulty? Since that would make maps like freedom dive give very little due to it having so many long streams when it's still pretty difficult seeing as not that many people can FC it. Part of it might be the fact that it's so long since it doesn't really have that many hard parts, but the jumps that keep people from FC'ing it are still pretty challenging even if there's only a few.
Topic Starter
The difficulty algorithm is run seperately for hardrock, easy, doubletime and halftime. Don't worry, it's just just a flat difficulty bonus.TheVileOne wrote:
When can we expect a more expansive list of difficulty values?
I would like to evaluate difficulty further, but it's very difficult when I cannot confirm how hard a map is compared to another map. It's pretty clear what some of the hardest CTB maps are, but it's not so clear where the maps that fall under them are. Maybe provide a difficulty calculator for us to play with? We can then manually choose CTB maps and have the difficulty value be calculated via an option.You have to do that pattern twice in that map. Well at least you had to do it in the stream compilation twice and in that case it's more than halfway through a 10 minute drain map (12 minute long song). Anyways should be classify as a very hard pattern. Other patterns like it need to be considered as well. They fall under chain jumps.DeathXHunter wrote:
Is it worth to be in the high top performance just because they managed to do that 1 pattern? The rest of the map is pretty easy,imo.
Also I hope that you realize that mods can change hyper jumps and the difficulty of certain patterns. Hopefully mods like Hard rock give pp based on the extra difficulty given to the map, which can vary very greatly from map to map. It is not as simple as considering that the difficulty settings get increased, the random placement of fruits can add hyper jumps and very unusual and awkward patterns.
what really bothers me is the huge gap on pp right now..
I really do not understand the new system. I really cannot wait for a the wikipage or a topic for the CtB PP algorithms.
In general, I don't know if my PP doesn't update... or the map I just finished wasn't hard. I know I'm not that great yet, but it's deflating as a player to see no progression on completing maps that are beyond my actual capabilities at the moment.
In general, I don't know if my PP doesn't update... or the map I just finished wasn't hard. I know I'm not that great yet, but it's deflating as a player to see no progression on completing maps that are beyond my actual capabilities at the moment.
mikhe wrote:
That's a better example of you want to sayTheVileOne wrote:
Example
It looks like you can just stay in the middle to catch the fruits, but you just can't. It's like in between two movement frames, so that you can't quite position yourself in the correct place to catch all the fruits, you are forced to move back and forth for each fruit.
~
So I got bored and I decided to SS a bunch of extra songs and I got 0 pp for this, why?
http://gyazo.com/75ee6e12bf4d1761bf16ad14a3a4b338
why is top 10 weighed so low? and why is nomod not weighed at all?
http://gyazo.com/75ee6e12bf4d1761bf16ad14a3a4b338
why is top 10 weighed so low? and why is nomod not weighed at all?
Topic Starter
You already have a lot of scores of comparable quality. You will get rewarded more pp if you improve your hardest plays. The score rank you achieve on a particular map doesn't matter at all anymore.Sabi wrote:
So I got bored and I decided to SS a bunch of extra songs and I got 0 pp for this, why?
http://gyazo.com/75ee6e12bf4d1761bf16ad14a3a4b338
why is top 10 weighed so low? and why is nomod not weighed at all?
what if all the top plays are SS's? :v assuming no spinners (it also looks like hidden is over rewarding pp despite the difficult barely going up)
edit: i also noticed an emphasis on FC to get pp, i rarely play insane's (almost all my ranks are hards, so im assuming my pp would be that of someone who plays hard and can barely beat easy insanes) despite this, i got a 300 combo on a 325 combo song and got <1 pp (1 rank), i would assume based on the system that you describe, that i would gain quite a bit of pp because i cant really beat insane's unless theyre really easy but i got so little i couldnt even see if i went up in pp or not
edit: i also noticed an emphasis on FC to get pp, i rarely play insane's (almost all my ranks are hards, so im assuming my pp would be that of someone who plays hard and can barely beat easy insanes) despite this, i got a 300 combo on a 325 combo song and got <1 pp (1 rank), i would assume based on the system that you describe, that i would gain quite a bit of pp because i cant really beat insane's unless theyre really easy but i got so little i couldnt even see if i went up in pp or not
How do I improve my top plays if they are SS's? Spins? or? HD? cuz HD is really gay, it has no factor in the game's difficulty at all, to me it's just a preference of style, and also I just SS"D over 30 extra's and got multiple top 8's and 10's and I still didn't get 1 pp from it, and also can you explain this? http://gyazo.com/dd1790888a59f3cbe4e52574e7bcc640 that guy is rank 255 with normals and easy's on top performance. watTom94 wrote:
You already have a lot of scores of comparable quality. You will get rewarded more pp if you improve your hardest plays. The score rank you achieve on a particular map doesn't matter at all anymore.Sabi wrote:
So I got bored and I decided to SS a bunch of extra songs and I got 0 pp for this, why?
http://gyazo.com/75ee6e12bf4d1761bf16ad14a3a4b338
why is top 10 weighed so low? and why is nomod not weighed at all?
edit: all this didn't give me 1 pp I think rin gave me like 1 or 2 http://gyazo.com/5b0ed8d80905584648bd0d7e0529b231
Topic Starter
Playing harder maps than your top plays also works. You don't need to improve on the exact same maps.Sabi wrote:
How do I improve my top plays if they are SS's? Spins? or? HD? cuz HD is really gay, it has no factor in the game's difficulty at all, to me it's just a preference of style, and also I just SS"D over 30 extra's and got multiple top 8's and 10's and I still didn't get 1 pp from it, and also can you explain this? http://gyazo.com/dd1790888a59f3cbe4e52574e7bcc640 that guy is rank 255 with normals and easy's on top performance. wat
edit: all this didn't give me 1 pp I think rin gave me like 1 or 2 http://gyazo.com/5b0ed8d80905584648bd0d7e0529b231
Topic Starter
Link to example, please (the user with the HDHRHT rank)? If that's the case then I'll adjust it. That's not supposed to happen.Seph wrote:
why does airman give more with HDHRHT than an actual nomod SS? please explain.
Seph
mine and Dark Diego
just checked and it seems a HDHRHT SS play gives more pp than the others, and it disregards the rank (or the spinner, w/e lol). RAMPAGE88 got 465 pp vs Dark Diego's 464 but not SS
just checked and it seems a HDHRHT SS play gives more pp than the others, and it disregards the rank (or the spinner, w/e lol). RAMPAGE88 got 465 pp vs Dark Diego's 464 but not SS
You dodged my question and all these maps are insanely hard, they are top tier maps, theres a strong reason why only 10 people have fc'd a certain map for instance why is kemu life game worth nothing even tho it's nearly impossible and you have to get lucky https://osu.ppy.sh/u/set44599 why isn't kemu life game ( https://osu.ppy.sh/b/187501?m=2 ) on his top performances if he holds the only FC and SS another case is nomod, why is HD weighted like a anchor?Tom94 wrote:
Playing harder maps than your top plays also works. You don't need to improve on the exact same maps.Sabi wrote:
How do I improve my top plays if they are SS's? Spins? or? HD? cuz HD is really gay, it has no factor in the game's difficulty at all, to me it's just a preference of style, and also I just SS"D over 30 extra's and got multiple top 8's and 10's and I still didn't get 1 pp from it, and also can you explain this? http://gyazo.com/dd1790888a59f3cbe4e52574e7bcc640 that guy is rank 255 with normals and easy's on top performance. wat
edit: all this didn't give me 1 pp I think rin gave me like 1 or 2 http://gyazo.com/5b0ed8d80905584648bd0d7e0529b231
PS: I'm not trying to hold torches and go after you, I know you don't play CtB but.. this ranking system needs a lot of work and you seem to think it's fine or something.
DeathXHunter
DeletedUser_500696
I agree with fabi, don't talk smack about my brethren please. Also, there is a huge range of ctb styles and preferences which cannot possibly be accounted for in a single algorithm (reliable player, HD player, HD+HR, spin pro, w/e). Don't judge others based on your style solelySabi wrote:
How do I improve my top plays if they are SS's? Spins? or? HD? cuz HD is really gay, it has no factor in the game's difficulty at all, to me it's just a preference of style, and also I just SS"D over 30 extra's and got multiple top 8's and 10's and I still didn't get 1 pp from it, and also can you explain this? http://gyazo.com/dd1790888a59f3cbe4e52574e7bcc640 that guy is rank 255 with normals and easy's on top performance. wat
edit: all this didn't give me 1 pp I think rin gave me like 1 or 2 http://gyazo.com/5b0ed8d80905584648bd0d7e0529b231
Shit talking other players? are you stupid or something? When did I shit talk anyone? are you retarded? or something holy shit. I was just wondering why there was normals and easy maps on top performances, I was using him as an example. I can care less about ranking I already know I'm good.DeathXHunter wrote:
Mad 'cause bad. Stop shit-talking about other players in public just because you think you deserve a higher rank and learn to play yourself.
AUUUUUUUUUU
DeathXHunter
Sabi wrote:
Shit talking other players? are you stupid or something? When did I shit talk anyone? are you retarded? or something holy shit. I was just wondering why there was normals and easy maps on top performances, I was using him as an example. I can care less about ranking I already know I'm good.DeathXHunter wrote:
Mad 'cause bad. Stop shit-talking about other players in public just because you think you deserve a higher rank and learn to play yourself.
I thought you cared, otherwise why were you so upset when pp was gone? You wouldn't care about all those "top tier" maps giving you 0pp either. Stop contradicting yourself, if you're so upset about the pp system then suggest ways for it to improve instead of complaining about it, that's all you've been doing lately.Sabi wrote:
I can care less about ranking
EDIT: I would also like to ask about what has been done when calculating the difficulty? I find it really odd that Go Berzerk still gives more than Calamity Fortune when it's worlds harder.
Also I think misses need to be penalized a little less, I like that droplets aren't penalized very much anymore (though they could be weighted slightly more idk) but on maps that are excessively hard to FC it's nearly impossible to get any pp because misses are completely destroying the pp they give.
Here's a good example:
I know that my score on Image Material is shit, but that's the very bottom of all my plays and I just don't see how that should be the 6th worst play I have out of nearly 3,500 maps that I have a score on.
"I thought you cared, otherwise why were you so upset when pp was gone? You wouldn't care about all those "top tier" maps giving you 0pp either. Stop contradicting yourself, if you're so upset about the pp system then suggest ways for it to improve instead of complaining about it, that's all you've been doing lately."
brb fapping
"I would also like to ask about what has been done when calculating the difficulty? I find it really odd that Go Berzerk still gives more than Calamity Fortune when it's worlds harder."
go berzerk is hardest song, it's worth a lot cuz its harder than even big blackums no jkoke that's why i'ts worth so mcuhz
brb fapping
"I would also like to ask about what has been done when calculating the difficulty? I find it really odd that Go Berzerk still gives more than Calamity Fortune when it's worlds harder."
go berzerk is hardest song, it's worth a lot cuz its harder than even big blackums no jkoke that's why i'ts worth so mcuhz
Oh damn, every time this fighting about ranking and pp and who is better and who is not. Just stop that already.
I completely agree with 119, it was way more pleasant to see the players profile without a rank.
I completely agree with 119, it was way more pleasant to see the players profile without a rank.
Zak wrote:
EDIT: I would also like to ask about what has been done when calculating the difficulty? I find it really odd that Go Berzerk still gives more than Calamity Fortune when it's worlds harder.
Also I think misses need to be penalized a little less, I like that droplets aren't penalized very much anymore (though they could be weighted slightly more idk) but on maps that are excessively hard to FC it's nearly impossible to get any pp because misses are completely destroying the pp they give.
Here's a good example:
I know that my score on Image Material is shit, but that's the very bottom of all my plays and I just don't see how that should be the 6th worst play I have out of nearly 3,500 maps that I have a score on.