#Difficulty filter

58 messages · Page 1 of 1 (latest)

sly blaze
#

this has flat brackets, so simply what it does is divide scores by downloads of each map and put them in a difficulty bracket based on the result of that division

with Bob's idea what i would instead do is sort the maps by scores/downloads, cut the list into five even pieces, then sort each piece back by IDs. its an idea for sure, and it would make things completely hands off, but i would have to try it

#

i dont have that much time during the week anymore but if there is any kind of interest (even non committal) im willing to experiment with different ideas and maybe add cw4 as well to the browser

waxen plaza
#

You've had this for a number of years, right?

Did you ever mention it to knucracker? Did you get feedback from him?

Did you get feedback from others or some consensus opinion? are any folks using this to select maps? Is it published somewhere where folks can use it to evaluate it?

Lots of questions, I know.

sly blaze
#

i did not mention it directly to KC. figured those things go through you so as to not be a bother. i had mentioned this link in your presence before.

i dunno about consensus opinion, but i figure those who engaged in the automated difficulty conversation might be into it. users who mention SMM as well. doubt anyone's using it to select maps as it is not integrated with the game after all.

not sure what to make of the last question, its on github so you can both use it and read the source if that's what you're asking

#

luckily for me KC does allow cross-origin access to the API's so it simply works through the web from a basic github site

deft finch
#

I do like that I have 3 pages dedicated to my maps for "Expert" difficulty.

sly blaze
#

yea, some well known insanely hard map authors feature often there too

#

which i find neat

deft finch
#

well, its a neat tool. will you do CW4 too?

sly blaze
#

i should.. back then i didnt know how, i do now

#

though on second thought i can't actually, since cw4 doesnt have scores or downloads 😅

#

just remembered that minor detail haha

waxen plaza
#

users who mention SMM as well.
? que

sly blaze
#

super mario maker

waxen plaza
#

oh, that

#

Since I don't have or play any of those games, all the references to that jsut pass me by.

prime ocean
#
  1. Yes, I was referring to difficulty.
  2. I don't think a user driven difficulty system (rating) is great and I was suggesting an objective metric for difficulty.
  3. In Mario Maker, this is done by attempts/completions. The levels are short and you cannot save.
  4. In CW (perhaps IXE) the closest equivalent seems to be downloads/ (unique?) scores
  5. I am not suggesting a strict order by difficulty that is evident to the player - instead difficulty tiers or buckets each made of 20% of maps.
  6. There is no specific reason I picked 20 (and they need not be the same size either) but it seems straight forward enough without complications. 3 buckets is probably the minimum and 5 is probably the maximum.
  7. If you just did an order by difficulty, you would have people 'seeking' out the hardest and easiest maps cementing their position - this is indirectly the problem we currently have with thumbs up in CW4
#

From a user perspective, they would select which tier of maps they want to play.

From a server perspective, the server orders them by (unique?)scores / downloads and then takes or skips as appropriate to get the proper tier

#

As I mentioned above, the server may also apply a secondary ordering (by id?) so they are not strictly ordered by difficulty

sly blaze
#

server needs not do any work. as in prev games, entering the custom map browser downloads the metadata file with all maps. the game then can do all the sorting/categorizing it wants, the same way i can do in my web map browser

#

the only challenge on the server is adding a unique counter for scores/downloads. in cw3/pf these counters are not unique, whch is ok because there is no incentive to artificially manipulate them. but if this system entered the fray, an incentive would appear

sly blaze
waxen plaza
#

So a few points on consideration. "difficulty" has never been a prime consideration for me personally, "fun" is more of a factor, as is replayability.

I guess that is perhaps why I've never discussed or mentioned your scheme to knucracker. I will rectify that at a future opportunity.

I do believe I'm not the only player that replays a number of maps many times over, to look for alternate means of completion, to get a better time, or to just have fun once I have figured out a map. This may result in multiple scores - or not.

Even multiple restarts may be indicative of a fun/challenging map. To this end I started the prior thread trying to determine if there was a metric that could capture that element of maps and provide insight to players.

For all metrics, difficulty, fun, replayability, etc. there are two ways to approximate it. One is on the back end via collected data, the other is via direct player feedback.

Now we get to the difficult part. it is my belief that knucracker, based on his predilection for creating games, is not really interested in the back-end issues and he does only what is necessary. Now of course, one can make the argument that a better map rating algorithm is necessary, but folks have been playing all the games since CW1 without that and it does not seem to be a major factor in considering acquiring or playing any of the games.

So I don't know that making a case for an extensive ranking system in-game is going to be a success. As I said, this will certainly be a topic we'll have a future conversation on.

sly blaze
#

does difficulty not correlate with fun in some way? if a map's too hard for you, you wont have fun. conversely, if youre looking for a teeth gritting experience, you're not going to enjoy a trivial map. obviously, it's not perfect, but a basic difficulty categorization based on collected data could break up the monolith that is colonies and make it a bit more approachable to the average player

#

historically, filters have been there to help tailor your search, but they are pretty unwieldy to use - you have to already know what you're looking for. what the games do not have is a few simple buttons for "we think these maps are going to be easy, and those will be hard"

#

i am merely trying to prove, successfully or otherwise, that it can be done without too much work. whether such addition is seen as a positive is another matter

waxen plaza
#

does difficulty not correlate with fun in some way?
This is the crux of the matter.

Is Sorrontis' "The Wave" fun? It sure is difficult. But is it fun in the way the LPAC maps are fun? (I've not checked Thumbs rating on "The Wave" for a while, so I don't know.

Point is that difficulty does not directly translate into fun/enjoyable. At least not for the majority of the player base.

And I think that is the holy grail, to find a metric that is adequately representational for all players. Which, on reflection, is perhaps as simple as the current Thumbs rating?

Again, on a personal note that is probably not representational of the larger player base, I have mostly gone back to procedurally generated CW2 maps. Where I choose specific maps that I think are hard, but do-able. Anecdotally I've heard from Bilbo that he has returned to CW3 DMD maps.

Reading the forums, hard CW3 maps generate a fair amount of discussion, if that equates to many players having fun, I've not checked.

Using your filter system, I noticed that a lot of the CSM maps come up in "Easy" and "medium" categories, which probably is accurate based on player feedback.

sly blaze
#

i think we can agree that fun is a complex problem that comes down to a variety of things - difficulty, design, gamemode, maybe even story. and it is not the same for everyone

perhaps a good way of putting it is that this system would not necessarily give you fun maps, instead it may improve one's chances of finding a fun map for them by attempting to solve one of those factors

#

what i've made is nothing more but a glorified filter using map metadata like all the other filters, but more in your face, with bigger and shinier buttons. yet i feel like that may make all the difference. it establishes a baseline format for breaking up colonies into a few logically sorted pieces as a starting point for the player when looking for what to play.

as a side note, a manual star rating system for difficulty could totally be employed under the same UI concept as well to attempt to achieve the same purpose. i personally feel like an autmatic solution based on data is going to be more reliable than user input, but nobody knows for sure

prime ocean
#

I think objectively (opposed to subjectively: ie rating system) is superior for difficulty ratings

sly blaze
#

i know not everything can be taken into account, e.g. you restarting a map a few times because you enjoy it would not factor into the unique scores/unique downloads ratio. but what factors in is the cumulative experience of hundreds of people, and the result comes out to a number of people who enjoyed the map enough that they took it to completion as opposed to leaving it behind

prime ocean
#

I think there is a player desire to play certain difficulty maps - reactions to Farbor or Wallis may be a good indicator here

#

I think that there is a strong correlation between fun and difficulty for many players

waxen plaza
#

a player desire to play certain difficulty maps
And this here is the crux of the whole issue. Some love Farbor, others hate it. If you can propose a system that recommends Farbor to those who love it, while not recommending it to the haters, you're golden. And I really don't think Grabz' (or any other system, for that matter) can do that

sly blaze
#

in super mario maker, they use a clear rate system. they divide the amount of times a level was cleared in total (non unique) by the amount of times a level was started from the beginning, through launching a level or dying (thus going back to the beginning). using this clear rate, they categorize their levels in several difficulty brackets. users can browse levels by those brackets as well as participate in endless modes with random levels from a chosen brackets.
i noticed that in that system, levels that are fairly new are sometimes slightly miscategorized due to variance. e.g. a level that is Super Expert tier could appear in Expert. people playing Expert pick it up, die a bunch of times on it, and the level eventually gets moved to Super Expert due to the shifting ratio from people expecting an Expert level. once the level gets enough tries, its status and ratio becomes solidified.

KC games do not have nearly as many players, heck, some SMM levels get millions of attempts. plus, SMM levels are short, few minutes at most in 99% cases. so a direct 1 to 1 translation makes no sense.

as i played cw3, i eventually noticed myself oogling scores/downloads as a gauge how hard a map might be. that was before i even considered making this suggestion. so once i thought this could be a filter, i set my idea into motion and made that github project.

sly blaze
prime ocean
#

Maybe you don't agree with those points

#

I grant that it is not a perfect system, and it certainly isn't as nuanced as is possible. But we are attempting to propose a minimum of functionality to get closer to where we want

sly blaze
#

(as a clarify, do note that scores/downloads in cw3 are not unique. a dataset using unique versions of these stats does not exist. the desire for uniqueness in these stats is mainly necessitated in my mind to prevent abuse of the system, due to lower overall playerbase. glorification of these statistics adds motivations for abuse, although systems in KC games have historically not really had issues with abuse despite many abuse avenues existing)

#

well, they are sudo unique in that most users dont have to download a map more than once unless they wipe it from their disk. and scores are unique if a user sticks to the same name. so they are both abusable, but with no reason to abuse either, most users have no reason to care, so for all intents and purposes theyre likely mostly unique

waxen plaza
#

Huh, I don't understand your position.
I posit that there is a difference in the population of players who enjoy the difficulty or challenge posed by Wallis and The wave in CW4. Or Farbor and that map Sorrontis made that's almost random in CW3.

#

And again, speaking strictly from personal experience, there were some hard maps in CW3 that I enjoyed tremendously and replayed a few times. One I even ported into CW4.

waxen plaza
#

Just came across this for Tucana in CW1. 🙂 Aligns with my previous post re relative difficulty/fun.

image.png
sly blaze
#

i must admit i'm not sure what point you are making. having a difficulty categorization benefits both those who enjoy the challenge posed by hard maps, as well as those who are looking for a chill experience

#

campaign is a bit of a different discussion, i mean normally games offer an option for adjustable difficulty, KC game campaigns generally offer a mash of easier maps with one or two harder maps thrown in so some clashing with certain players is hardly surprising

#

at the end of the day i'm just looking for a more comfortable experience selecting maps in colonies

waxen plaza
#

all I'm saying is that not all difficulties are the same

sly blaze
#

but an average can always be drawn

#

and it works in other games

waxen plaza
#

ok

sly blaze
tawny thunder
#

Fun is subjective. I'd like better ways to filter for maps that might be fun to me. More metrics, maybe hidden behind an advanced search button, would be nice.

sly blaze
#

maybe i should turn this around and instead ask you to find me a custom map in cw3 which resides in hard/expert bracket in my map browser yet is trivial to beat

#

or vice versa

#

cuz you did basically just claim objectively that my system would not help avoid recommending Farbor to people who don't want to play it and yet i want to say had Farbor been a custom map it'd easily have fallen into hard or expert and people not looking for a challenge would thus not find it

sly blaze