Posting in gameplay and rankings because it fits here more than feature requests.

I have been wondering the reason for half time existing in its current state. I've messaged peppy recently about halftime and its current standing within the game however I feel that his solution (simply adjusting pp weightings) is not enough. I find that halftime is great on its own; as a mod which reduces play speed of a song to allow a larger player base to enjoy maps, however the issue for me (and I think a lot of people will agree) is that halftime affects the ranks of others in higher level play and generally dominates the rankings when maps reach a certain difficulty. Here is the original message sent which conveys a few of the ideas of halftime within the pp metric:

jesse1412 wrote:

Just wanted to ask how pp works with half time. As it is, half time allows people to turn otherwise the highest pp rewarding maps (or at least will be highest) and most certainly the hardest into an average, long insane difficulty. For pp to be skill based a few problems would need to be addressed should halftime only give a minor reduction to pp reward; with halftime in the state that it currently is not only does it push 'better' ( and I use that tearm loosely ) players back in the pp race but it also allows people who would otherwise be unable to pass a map normally to beat people rank wise who can achieve hundreds of combo with a high accuracy simply because they managed to keep a higher combo.

I would also like to mention that half time can generally only be used to rank maps which are of this extreme difficulty simply because they are the only maps where the majority of players cannot hold a high combo. I understand that halftime cannot simply be removed entirely but maybe just throwing some way to punish this mod further than a simple score multiplier.

My final subject would be whether or not using halftime penalizes accuracy to a degree that a normal score with 95% would beat a halftime 99% in pp rewarded accuracy on one of these said impossible maps. Not sure if this may require a little too much disclosure of the pp system and weightings but I'm also wondering at this point; would it be possible to have the associated pp weightings per mod made public?

Thanks for reading.

example maps:

http://osu.ppy.sh/s/40344 (Atama no taisu - Lucifer x 0108 - 0108 - nogard) <-- difficulties in opinionated difficulty order
http://osu.ppy.sh/s/27752 (Chipscape - ragnarok
http://osu.ppy.sh/s/41823 (The Big Black - WHO'S AFRAID OF THE BIG BLACK)
http://osu.ppy.sh/s/39804 (FREEDOM DiVE - FOUR DIMENSIONS)

and of course the response:

peppy wrote:

Thanks for your input. I have reviewed cases with HalfTime and adjusted weighting accordingly.
I will not be releasing weightings etc. publicly as I believe it would be counter-productive.

I thank peppy for actually changing the reward of halftime within the pp weightings however the issue still remains that scoreboards are still dominated by those who choose to use halftime. The idea that people with no-mod, accurate scores are pushed out of the rankings is daunting for some players, why pass a map and lose out on accuracy while gaining no pp and failing to top rank it when you can just halftime S top rank?

The issue now is the fact that people who have played and ranked such maps would have scores they potentially put a huge amount of effort into nullified. You could always say that if this is the way it's meant to be, why wouldn't we do this? The reason I can see this being a valid argument is the fact that this mod only ever effects higher level play which is the place people prove themselves skilled. Of course this doesn't rule out the fact someone put a lot of effort into it, but since when was effort valued as much as results? If someone tried their hardest should they get top marks on an exam? The answer is no - they still lack the ability to attain a high grade. Achievements are measured in results not effort.

And the final thing is what do we do with halftime if it is changed? We could unrank halftime completely, however I feel this is unjust, people were led to believe their score would remain until someone beats them, how can we just remove halftime if people genuinely believed their score would remain forever. Another idea would be to lower the rewards of halftime however this would yet again create issues because we're essentially calling valid scores invalid without removing the problem at all. So in the end what do we do... Well how about removing the scores from their pedestals? Hide them by default; Make halftime scores still exist but force players to show them manually for each map while removing them from pp/toprank calculations for other scores. Do not get this wrong, halftime scores will still appear as the rank they are on a users profile and will still give the user the accuracy and score they are entitled to - the only difference is they will be hidden as alternative scores. Hopefully this would be an acceptable approach to the 'issue'.

Also I'd quickly like to mention that I'm going to assume that an accuracy of 80% with no mods on a song like freedom dive will not lower an accuracy of say 98% and will raise an accuracy of 90% (potentially not a good example because it is more about speed than accuracy but I'm using this map because of the severe lack of passes and good scores in general.) This would safely allow halftime to be discarded while allowing users who would want to keep a higher accuracy to play no mod.




tl;dr - make halftime scores hidden by default and make them not effect scores without it in pp calculation and map rank position.

once upon a time, renard banned forever was filled with half time scores

they will -eventually- slowly fade away, as players catch up with skill.

As a big fan of the HalfTime mod here is what I have to say on the matter: I'll just put it bluntly:

Osu's scoring system is broken

This means exactly what it reads. The scoring system does not function. You're supposed to get higher scores for better plays - in reality, better plays are not rewarded by a higher score. Because of this, the scoring system is broken.

Instead of writing a TLDR post that nobody will go over about how to create a far more accurate scoring system, I'll just outline the obvious things everybody already knows and a simple fix:

1. Quadratic scores do not represent skill level correctly. Simple example: If a map has [easy,hard,easy] parts, there's is no reason the last part of the map would give the most score per correct hit, since it's not even the hardest part of it.

2. The idea behind quadratic scoring is that gaining long combos would be worth more score instead of getting random hits, so that players who actually play well but do a random mistake every once in a while would get rewarded more than somebody who mashes and hopes to be lucky for every hit.
The idea is good, the implementation is bad. There needs to be a cap. In fact this is already done in taiko, where scoring works much much better and better plays actually get better score almost always - even with HalfTime taken into account.

WELL THAT'S IT!

Put a cap on the quadratic increase for hits. For example, starting from 20th combo hit, subsequent hits give the same score as 20th hit and the factor stops increasing, if you miss it resets to 0 and you only have to regain it up to 20 to start getting max per hit again.

The problem with this that MAT/BAT/Osudev is about to point out: adjusting old scores.
Easiest solution in the world: just keep the old scores. Add a tag, header, flag, whatever you want to call it, to map files, SCORE_SYSTEM = int (even bool would be enough, but maybe int for alpha/beta/future iteration stages? etc), which osu will read and use to decide which scoring system to use for the map - the new one or old one.

Maps without this tag, i.e. all the old maps, will just keep using the old scoring system, so all of their scores stay and there's basically no problem.
New maps starting from the time when new was introduced would use the new system, and their entire scoreboards would be made of the new style of scores. And then people who play with HalfTime and get decrease to 30% would actually fall below people who played without HalfTime and had lots of miss but played a much harder map due to speed, and then everything works fine.

fml it's still tldr.

Tanzklaue wrote:

once upon a time, renard banned forever was filled with half time scores

they will -eventually- slowly fade away, as players catch up with skill.

the problem is that the halftime scores are potentially made without any where near the skill of none halftime scores and yet still beat them on the scoring. These ideas are to remove the fact that "people should struggle to beat them because halftime is legitimately broken." A pro player shouldn't have to struggle to overcome people playing the 'impossible' maps at half speed and this brings up the fact that if someone managed to doubletime said maps or hd hr them they still wouldn't even compare on the rankings. But all of this is completely fine because pros will get even more pro. Why should people who want to play a map as it was intended have to get incredible songs to beat people who played it with a mod which makes the map ALOT easier.

easy mod doesn't really fit the category of making the map easier for 99% of players and should probably be reviewed entirely - but that's a different topic.

Aqo wrote:

stuff

such a huge overhaul would be great for standard players. Simple as. Sad thing is it probably wont happen for a long time if at all, I still however have hope that one day it will be changed. Maybe even allow both ranking charts with an available toggle? I suppose such a thing would be so so sudden with a toggle that rankings would be a little chaotic if old scores weren't converted to the new system. So many problems...

osu!'s score system has been broken since it's creation.

Score should be calculated using accuracy as a multiplier (thus making every hit worth the same). Actually score is worth shit since all that matters is how you do during the last part of the map since that's what's really worth a lot.

Factor used on score calculation should be accuracy and combo imo. Making the last hits worth more than the rest is stupid.

Then again Half-Time is completely imbalanced. This can be seen in every extremely hard map, where pro's cant beat Half Timers by playing nomod (or they just have a hard time doing it) which is, in my opinion, completely retarded.

After this you got Easy mod which is broken since it actually makes the map harder for almost everyone (how many people can play fluently insane maps + easy?).

Then the HR / HD discussion, both mods giving the same bonus. HR being a damn hard mod, since on insane maps you get AR 10, an OD close to 10 and a smaller circle size, meaning you need to have very good reflexes, a damn good accuracy since OD turns from being damn easy (like 7) to almost 10 which is really hard, and your aim needs to be way better (plus distance from hit to hit is increased a little by this). While Hidden just makes a visual change where snapping gets a little harder, and all you need to do is remember a few random hits that are hard to see using this mod. Only way this balances out nowadays is that most HR players can, for DAMN obvious reasons, do HR HD (there may be a few exceptions), but it still dumb that HR alone gives a lower bonus than another mod which requires you to be way better to FC. And remember, even if you're very good, getting 98% on HR is harder than getting 98% on HD, because of OD.

And on the last corner you got the competition between FL and DT, both mods giving the same bonus, but requiring completely different skills. FL being possible for every map if you play it enough times (it's just logic, play any map you can FC no mod for 2 weeks several hours a day, you will eventually FL FC) and DT requiring in many cases insane (many times impossible) skills to get the FC. Still many maps on DT are stupidly easy and FL requires a big amount of time to master/FC any map. So what should be done here is to change DT completely. DT bonus should be based on the base AR the map has (AR 9 DTs are just too hard for almost EVERYONE, even if the map itself is easy) and the base BPM of the map. Why the BPM thing? Many insane low BPM maps are just too easy, even using DT (example: http://osu.ppy.sh/s/30237), and those are cases where FL should be rewarded higher than DT.

But wait! There are some cases where low BPM maps are impossible on DT, if there was some accurate star rating system to see how hard each map is, I'd say to base DT scores on both BPM and difficulty (a very hard map should be rewarded A LOT MORE@DT than @FL).

tl;dr: Everything should be reworked/checked.

Wishy22 wrote:

osu!'s score system has been broken since it's creation.

Score should be calculated using accuracy as a multiplier (thus making every hit worth the same). Actually score is worth shit since all that matters is how you do during the last part of the map since that's what's really worth a lot.

Factor used on score calculation should be accuracy and combo imo. Making the last hits worth more than the rest is stupid.

Then again Half-Time is completely imbalanced. This can be seen in every extremely hard map, where pro's cant beat Half Timers by playing nomod (or they just have a hard time doing it) which is, in my opinion, completely retarded.

After this you got Easy mod which is broken since it actually makes the map harder for almost everyone (how many people can play fluently insane maps + easy?).

Then the HR / HD discussion, both mods giving the same bonus. HR being a damn hard mod, since on insane maps you get AR 10, an OD close to 10 and a smaller circle size, meaning you need to have very good reflexes, a damn good accuracy since OD turns from being damn easy (like 7) to almost 10 which is really hard, and your aim needs to be way better (plus distance from hit to hit is increased a little by this). While Hidden just makes a visual change where snapping gets a little harder, and all you need to do is remember a few random hits that are hard to see using this mod. Only way this balances out nowadays is that most HR players can, for DAMN obvious reasons, do HR HD (there may be a few exceptions), but it still dumb that HR alone gives a lower bonus than another mod which requires you to be way better to FC. And remember, even if you're very good, getting 98% on HR is harder than getting 98% on HD, because of OD.

And on the last corner you got the competition between FL and DT, both mods giving the same bonus, but requiring completely different skills. FL being possible for every map if you play it enough times (it's just logic, play any map you can FC no mod for 2 weeks several hours a day, you will eventually FL FC) and DT requiring in many cases insane (many times impossible) skills to get the FC. Still many maps on DT are stupidly easy and FL requires a big amount of time to master/FC any map. So what should be done here is to change DT completely. DT bonus should be based on the base AR the map has (AR 9 DTs are just too hard for almost EVERYONE, even if the map itself is easy) and the base BPM of the map. Why the BPM thing? Many insane low BPM maps are just too easy, even using DT (example: http://osu.ppy.sh/s/30237), and those are cases where FL should be rewarded higher than DT.

But wait! There are some cases where low BPM maps are impossible on DT, if there was some accurate star rating system to see how hard each map is, I'd say to base DT scores on both BPM and difficulty (a very hard map should be rewarded A LOT MORE@DT than @FL).

tl;dr: Everything should be reworked/checked.

Everything here needs a lot of work, but currently If I know right just 3 osu!Developer is working which is not enough to do something like this. So just be happy with this system.

Kurokami wrote:

Wishy22 wrote:

more stuff.

Everything here needs a lot of work, but currently If I know right just 3 osu!Developer is working which is not enough to do something like this. So just be happy with this system.

This shouldn't actually be that impossible for them to implement as it stands, I think the main issue is the fact that old scores will be changed which is unjust in many ways but as stated this could be solved by splitting maps into 2 different ranking styles - old and new.

also @ wishy. I actually asked this in a follow up pm to peppy, could mods ever be based on the difficulty of maps. That was about a week ago and I still haven't got a reply but I can already see a good reason why: we simply don't have a way to attribute difficulty settings to maps thus far. The star system hardly reflects map difficulty and this is where the issue arises, we just need someone to make a difficulty detector.

Until there is an actual code to calculate map difficulty there is no point to try and calculate true variable added difficulty from mods.

Could write pseudo code for an algorithm for this but I'm sure it would be a waste of time, rather just pick up a book to learn the C# syntax and then write actual code for it.

Why couldn't osu be coded in C++ or Java >_>

You could just make a new user-group with experienced players whose work would be to judge how hard a map is from 1 to 100, but that won't ever happen since, if everything keeps working the same, users with privileges will still be just users that don't really play the game (most BAT/MATs don't play the game seriously, just a few are good and even if they are most of them are not really active).

I agree with Wishy22 in regards to his tl;dr post. Remember my "gay" play? That beatmap actually serves quite a challenge with the mod combinations I used (it should be obvious how hard it is to SS Insane maps with Easy+Hidden mods). When I SS'd that, I actually received some amount of pp, but the problem is that I placed #447 there, and I may lose pp when that record is beaten out from Top500 (as implied on osu!Wiki). Although, this is the reason why I started playing some "easier Insane" maps without "Difficulty Decrease" mods.

Some random thoughts about possible solution: Perhaps a separated Leaderboard based on mods, like... HalfTime and/or EasyMode Leaderboard? (derived from No-mod Leaderboard)

While I don't take pp seriously, I somehow started to have interest in keeping my Performance above 5,000pp.

this should be considered in Taiko/CtB too. I've seen top ranks in Taiko (in easy/normal diffs tho) that HT people get higher scores than people without them. Same goes for CtB, some maps don't convert properly to CtB (i.e. pixel jumps appearing in maps, totally weird patterns), no mod plays should give higher pp gain than those with difficulty reducing mods.

@sephwin: That also happens at osu! standard as well, on certain "Easy" beatmaps like this.

lol~HT.

Wishy22 wrote:

You could just make a new user-group with experienced players whose work would be to judge how hard a map is from 1 to 100, but that won't ever happen since, if everything keeps working the same, users with privileges will still be just users that don't really play the game (most BAT/MATs don't play the game seriously, just a few are good and even if they are most of them are not really active).

problem with that is, you can't give a map a general difficulty. some people might not be able to play one insane that others can play wothout problem, but those people suck at other maps that the other people can clear easily.
and there are also mods. two factors here: some maps play easier (or harder) with specific mods than others. for example, map A is generally harder than map B, but map A is a lot easier to clear with HD HR than map B, so which one is harder? and of course every player has different modpreferences, which means that for example Flute would HR a map that no one else could HR, but then fail to do a hidden run on the very same map which every decent player could achieve.

short version: any difficulty rating of a map given by a user would never be objective.

If you get 10 opinions you will get a somewhat accurate average score.

you can make this exact same post about normal compared to DT

silmarilen wrote:

you can make this exact same post about normal compared to DT

Well apart from the fact you can't doubltime atama no taisu and beat all the no mod players with half the skill. Double time doesn't even apply to higher level maps, that would be a separate discussion completely. mods in their current standings can be completely impossible for 1 map while being incredible easy (if not helpful) for others. Not gonna go too off the topic of halftime but as an example: there is no way anyone can DT airman while some people can happily clear it hd hr for the same multiplier, or let's say there is an insane difficulty which is incredibly slow, some people may find it easier to DT it than no mod.

Also has anyone tried to easy mod the big black? I find easy to be relatively broken for the maps which people need it - the opposite end of the spectrum of halftime which makes it too easy. Food for thought.

I think that the argument that DT and normal are analogous to normal and halftime is actually quite good!

However... I think the real issue here is that it makes sense that normal scores might be able to beat out DT scores on a map that can only barely be DT'd, because the DT mod is supposed to imply difficulty. It does not make sense that halftime scores can beat out normal scores, since the mod implies ease. It really should not be capable of competing with normal mode on a difficult map that cannot be DT'd, since it's more of a fun mode to make maps accessible to lesser skilled players.

jesse1412 wrote:

there is no way anyone can DT airman while some people can happily clear it hd hr for the same multiplier,

Actually Cookiezi tried it with DT+HD. o.o

But just for fun.

Unrank all mods if you cry so much about this.
The Maps which you gave as example are special and to be playable for the people over the Top 100.
And please play the game for fun, not for Rank. Thanks.

No.

I rank for fun.

darkmiz wrote:

I rank for fun.

I agree with this. I find it fun to get high ranks and if all I could get was the same SS 100 other people had I would most likely just quit.

TheNutritiousGuy wrote:

Unrank all mods if you cry so much about this.
The Maps which you gave as example are special and to be playable for the people over the Top 100.
And please play the game for fun, not for Rank. Thanks.

For some (many) people fun equals ranks.

Also, the mod isn't broken for the maps anyone can FC. Of course he mentioned the extremely hard maps with high max combos. 1000 combo HT is worth way more than 300 combos nomod which is just wrong. That's where the problem exists in the first place.
Regarding your first sentence: Please re-read G0r's post, he explains it pretty well.

why does your avatar look like a stick figure with massive tits who's inside a circle

TheNutritiousGuy wrote:

Unrank all mods if you cry so much about this.
The Maps which you gave as example are special and to be playable for the people over the Top 100.
And please play the game for fun, not for Rank. Thanks.

Its actually kind of annoying that you would post an argument like this but I sort of agree, in fact I fully agree! I think the people who win the Olympics should be the people who have the most fun. The kind of stuff they do is only for the top 100 dedicated people so we should completely blow them off and propose everything caters for the average.

Basically no - I feel I may have been a little too sincere with this post as well.

I agree fully with easy maps, there should be a difficulty for every player but according to your argument we should only have maps which appeal to the average players with no challenge for the more 'skilled' players. In its current standing you are saying that it's fine to have thousands of easy maps for undedicated players but having a few hard maps where only a few of the most determained players can even pass should be completely limited to a half times win only (essentially - typical insane difficulty). People who play at this level tend to care about ranking songs alot more than others and that's not to show themselves off - it's to prove that they can pass. Essentially it's fine to ruin their fun because you disagree with it.

The worst part for me is that you came in blasting away about how they are special maps even though the thread is completely about the maps it applies to, HARD ONES. Seriously, you could have at least read a little bit of the thread.

thelewa wrote:

why does your avatar look like a stick figure with massive tits who's inside a circle

I swear you could bat off to anything if you can see a big titted women in that avatar.

TheNutritiousGuy is currently fortifying his house in fear of an Osu mob coming after him. It's funny how everyone reacted to this guy so fast.

G0r wrote:

TheNutritiousGuy is currently fortifying his house in fear of an Osu mob coming after him. It's funny how everyone reacted to this guy so fast.

Because he posted a ridiculous solution to the problem that was at hand.

Wondering if people's opinions are still the same as before. Also another peppy response on HT now would be nice.

Leave it ranked. Add a toggle amirite peppy?

there should be a negative multiplier for halftime

first one to play the map with halftime--> rank#1 -1120412 points

winber1 wrote:

there should be a negative multiplier for halftime

I actually like HT and I'm ok with this

Making halftime unrankable would be lovely imo.

it wouldnt in mine.
does it honestly hurt you so much that there are halftime plays in a top40?

It hurts my eyes to only see 1 HT record which is why I support t/88988
HT records make me wanna throw up.

silmarilen wrote:

it wouldnt in mine.
does it honestly hurt you so much that there are halftime plays in a top40?

Halftime is great is you want to practice specific things like slowstreaming at certain BPMs that lack a good fun streammap in normaltime or if you want to set different goals like going for an FC on a map where in normal time it just aint gonna happen, but using halftime for top scores is a no-no, it looks bad on the rankings and takes away from the quality of competition on the original map.

silmarilen wrote:

it wouldnt in mine.
does it honestly hurt you so much that there are halftime plays in a top40?

Yes.