01-13-2013, 08:42 AM | #1 |
FFR Player
Join Date: Aug 2005
Age: 29
Posts: 3,996
|
on the current SM difficulty system (also relevant to FFRCP4)
so I'm going through the 90+ files in FFR Community Pack 4 so far and I'm thinking of ways to set the difficulty...at first, I noticed that a lot of the files were rated according to (roughly) NPS, yet a LOT of them are much harder/easier than that number suggests.
for instance: Abandon by Valor. 1266 notes in around 3 minutes should be, in theory, a low 7 (the actual file is rated a 9). yet the file's chockful of incredibly unforgiving dodgeball sections, 32nd/24th bursts, and some really rough jumpjacks. I sightread it and got an A, to my surprise. compare and contrast to Guitar Hero by, once again, Valor. ~1850 notes (with a few tweaks by me) in roughly 3.5 minutes, right in 9 territory (again, the actual file is rated a 9, right where its NPS is). the actual file is super easy, though. it should be a sightread AAA for most proficient players barring like two mines. obviously there has to be a system in place so that we can fudge numbers up/down where they belong, but how would one standardize the system? I was thinking of the way FFR handles its difficulty ratings and got a rough sketch of how we might do things. divide things into around 6 tiers of difficulty as such: NPS +/- up to 5 difficulty points on this scale depending on external factors 1-3: beginner files. 4ths/8ths galore, consecutive 16ths are incredibly rare or nonexistent. patterns made to be as friendly as possible 4-6: lower-tier intermediate files. more complex rhythms or patterns might appear, but these should still be manageable. 7-9: edging into advanced territory with nothing really off-limits. you might see, say, 150BPM 16th jumpstream with some liberal use of jack theory and a few mines in this range. struggling on how to separate these files from files in the 10-12 range, though. anybody got any ideas? 10-12: either difficult/draining files or files designed to trip people up. AAAs should not come easily, even to advanced players. 13-14: unrelenting, intense files. AAs should not come easily by this point. 15+: forget scoring on these files since you'll be lucky to even pass them. if anybody has ideas, I'm all ears. |
01-13-2013, 09:00 AM | #2 |
⠀
Join Date: Dec 2010
Location: Singapore, SG
Age: 26
Posts: 6,858
|
Re: on the current SM difficulty system (also relevant to FFRCP4)
I definitely wouldn't mind seeing this. Though, I wouldn't use a 1-15+ scale though.
I feel that a tier (13-14) that's only 2 difficulty points wouldn't be all that accurate. A difficulty scale should be used to be as precise as possible, and giving a very narrow range for a particular "tier" of a difficulty would defeat the purpose of the difficulty scale. For example, let's say you're trying to compare say, Rave7 Oni. Using that is a baseline 14 (example, like I said, but it fits the description of being difficult to AA, and it's harder than most files that are difficult to AA). Then using Robotomy as an example; seeing that the highest tier is 15+ (nearly impossible to score on), Robotomy isn't nearly impossible to score on per se, but it's still much harder than Rave7. Would you put it as 15+ or put it as 14, albeit a very high one? Basically I think the difficulty tiers are fine, maybe except for the 6th one. Just that the ranges need to be expanded a bit more, probably 1-20+. Something like: 1-3: beginner files. 4ths/8ths galore, consecutive 16ths are incredibly rare or nonexistent. patterns made to be as friendly as possible 4-7: lower-tier intermediate files. more complex rhythms or patterns might appear, but these should still be manageable. 8-11: edging into advanced territory with nothing really off-limits. you might see, say, 150BPM 16th jumpstream with some liberal use of jack theory and a few mines in this range. 12-15: either difficult/draining files or files designed to trip people up. AAAs should not come easily, even to advanced players. 16-19: unrelenting, intense files. AAs should not come easily by this point. 20+: AAs should be almost impossible, even for the most seasoned Stepmania players. Advanced players should have difficulties trying to A files in this area by this point, and most average players would be lucky to pass them. Sorry if I sound confusing, I'm just really tired. I'll write up a longer/clearer post tomorrow. Last edited by EzExZeRo7497; 01-13-2013 at 09:26 AM.. |
01-13-2013, 09:32 AM | #3 |
Enjoy life!
Join Date: Oct 2007
Age: 31
Posts: 4,182
|
Re: on the current SM difficulty system (also relevant to FFRCP4)
i use a 10+ scale for my packs
basically every file should be at least difficulty 10, and I just bullshit some number thus not giving the file an actual consistent difficulty :3 That's just me though, I just look at the bpm and listen to the preview for like a second and that usually tells me what i'm in for haha The best actual grading system I think is NPS with fudge factor based on if the file contains anything difficult. So a 20nps file that has jacks could be a difficulty 23 or something because it has jacks and 20nps
__________________
http://www.flashflashrevolution.com/...d.php?t=149106 |
01-13-2013, 04:55 PM | #4 |
FFR Player
Join Date: Aug 2005
Age: 29
Posts: 3,996
|
Re: on the current SM difficulty system (also relevant to FFRCP4)
bumping. great suggestions so far.
|
01-13-2013, 05:21 PM | #5 |
`~`
Join Date: Oct 2005
Posts: 7,327
|
Re: on the current SM difficulty system (also relevant to FFRCP4)
My personal NPS scale I've used roughly translated to this, although it's not all that accurate in the first place. Admittedly, many of the NPS ratings I gave to files are inaccurate and bullshit because I was strapped on time to meet deadlines and didn't have time to -thoroughly- play through everything, checking to see if my scores were consistent enough with the NPS rating I gave the files. I'd have to do this thoroughly with every file to make sure it was passable enough as a rating zzz
1–3: Sparsely used 4ths/8ths with very minimal jumps, holds, and 16ths or above. Should have a lot of spacing between notes to prepare the player to hit the note. Shouldn't have any hand or mine placements at all, though very rare exceptions may apply depending on the usage in itself. 4–6: Heavier usage of 8ths with 16ths sparsely thrown in which are often triplets or broken streams that are very spaced out to again allow the player to prepare. Heavier jump and hold placements, with very conservatively used mines and some hands. From my personal perspective as a player, I should be able to low SDP effortlessly or AAAA after a few tries. 6–8: This is pretty much where 16th streams and more complex notes come in. Slow and/or light jump-stream would be around the end of this spectrum. The "average" player in this community would typically fall here. From my personal perspective as a player, I should be able to SDP after a few tries or AAAA after several tries. 9–10: Denser jump-streams and light hand-stream. Overall dense files that are fairly easy to full combo for the "slightly above average" player in this community would typically fall here. From my personal perspective as a player, I should be able to get low teens or twenties on these kind of files, with some minor issues with FC'ing consistently. 11–12: Pretty much a 10 but has much more difficult scoring factors such as mine dodging, denser jump/hand-streams (such as 8th jump/hand-stream), dense hold streaming, jackhammers, trills, and the like. Would be much more difficult for the "slightly above average" player to full combo. From my personal perspective as a player, I should be having trouble FC'ing as well. 13+: I never really had a set criteria for anything above this, lol—this was purely based on scores I personally got from playing files. If I couldn't FC a file (keep in mind, I did actually have a decent amount of speed and stamina), it would fall around this range. The more trouble I had on a file, the higher the rating would be at this point.
__________________
Last edited by Gundam-Dude; 01-13-2013 at 05:30 PM.. |
01-14-2013, 02:58 AM | #6 |
FFR Player
Join Date: Aug 2008
Location: France
Posts: 235
|
Re: on the current SM difficulty system (also relevant to FFRCP4)
I used peak NPS or 16NPS for Sirius MP
Basically find the 16 measures that represents the most dense part of the chart Calculate NPS gg or use a program that does it alone xd It's more relevant than global NPS I used to do global NPS on Sirius MP and I couldn't get anything higher than 15/16 global NPS By using 16NPS difficulty increased overall but not in the same manner Some sub15 GNPS could had a higher 16NPS than some 15+ GNPS This happens because songs are not consistent and same goes to the chart Using GNPS on a song like Pink Nightmares will make it very low because of the quiet and empty part in the middle The idea behind using 16NPS is that the value represents the required skill to play the file Everything easier shouldn't matter to the player if u can do the hard part However just like GNPS this doesn't take in account holds, mines and patterns But the basic idea of using NPS for difficulties is to make unbiased difficulty ratings and there nothing better than that So u can do your idea but use 16NPS instead of GNPS u shoez
__________________
(heidy)(heidy)(heidy)(heidy)(heidy) (heidy)(heidy)(heidy)(heidy)(heidy) get mad h8ers |
01-14-2013, 03:12 AM | #7 | |
⠀
Join Date: Dec 2010
Location: Singapore, SG
Age: 26
Posts: 6,858
|
Re: on the current SM difficulty system (also relevant to FFRCP4)
Quote:
blargh i'm done Last edited by EzExZeRo7497; 01-17-2013 at 03:41 AM.. Reason: formatting consistencies |
|
01-14-2013, 05:41 AM | #8 |
FFR Player
Join Date: Aug 2005
Age: 29
Posts: 3,996
|
Re: on the current SM difficulty system (also relevant to FFRCP4)
lots of good ideas in here (especially Roar's idea, I like that a lot). I'll post back to reply to everything in a few days after I give this some more thought.
|
01-19-2013, 05:06 AM | #9 |
⠀
Join Date: Dec 2010
Location: Singapore, SG
Age: 26
Posts: 6,858
|
Re: on the current SM difficulty system (also relevant to FFRCP4)
|
01-19-2013, 06:10 AM | #10 |
FFR Player
Join Date: Aug 2005
Age: 29
Posts: 3,996
|
Re: on the current SM difficulty system (also relevant to FFRCP4)
sorry this is late! I was so busy with the graphics/nudging that I forgot about this hehe
honestly I think your descriptions + Roar's usage of peak NPS will prove very helpful. I'd adjust the tiers a little because there still has to be some tie to overall file density, so I'd use this scale 1-3 4-6 7-9 10-12 13-16 17+ I imagine we'll have a much clearer idea of whether this will work after we playtest, but the above scale should be fairly close to how the final scale will operate Last edited by moches; 01-19-2013 at 06:13 AM.. |
01-19-2013, 06:56 AM | #11 |
Custom User Title
|
Re: on the current SM difficulty system (also relevant to FFRCP4)
I'm all for using a more accurate scale
__________________
|
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
|
|