Leaderboards issues

[def750]

So, while ago I found 2 issues and I completly forgot to report them here, so first is

1. Selected Mods Leaderboard
When looking at selected mods lb, only pb is shown. What does that mean
For example user a has 2 plays on Harumachi Clover, One with DT and one No Mod.
User a has personal best on No Mod score, when someone, including user a tries to look at selected mods lb with DT, he won't see his score because it's not his PB.

2. PP sorting, but also score at the same time
So, I don't know if it's on purpose or not but in mania this is very annoying, anyways. This got tested in mania, idk if it works the same in other modes but probably yes.
When user a submits score on map xyz with 850,000 score and 220 pp for example, his score will be first
Then user a submits another score but this time with DT, he managed to get 770k score and 320pp. His DT score will be displayed as first even though it's not PB, same thing happens on HT. Due to mania score cap being 1 milion, it's really annoying and if it's on purpose then it doesn't make sense in mania. Also I think that this may be related to first bug.

[minisbett]

About the first thing: it's not a bug I suppose because it's like this on a lot of osu servers including ripple based ones. But yeah it would be much better to check for status = 1 or 2 instead of just 2.
To the second one: mania uses PP override, therefore a score is considered the best score on a map based on pp, haven't checked but based on what you've described I suppose that it is like this

[tsunyoku]

minisbett is basically correct on the first one, however i don't really understand the issue you're trying to describe on the second point. plays override based on pp, as the point of pp is to evaluate what is the better score. what's the problem

[minisbett]

I think the issue is that despite being overrieed based on pp, the score is still what's displayed on the leaderboard I think

[def750]

minisbett is basically correct on the first one, however i don't really understand the issue you're trying to describe on the second point. plays override based on pp, as the point of pp is to evaluate what is the better score. what's the problem

So basically in mania scores don't override on leaderboards, highest score = pb but pp is counted from best play
Here are examples
Score 1: HT 220pp 449k score - PP is calculated from this one
Score 2: NoMod 100pp 600k - This one is displayed on LB

Score 1: DT 300pp 750k score - PP is calculated from this one
Score 2: NoMod 220pp, 920k score - This one is displayed on LB

[tsunyoku]

i’m really not understanding lol

scores always override by PP, so there’s no way the score it’s showing on leaderboards is different to the one the api/db says the best; the one it says is the best will also definitely be the highest pp one

also the 2nd and 3rd screenshot are from bancho so where is the relevance??

i do not understand, lol.

[minisbett]

No the problem is just, you see score in the leaderboard but it uses PP override in the background, it should display pp in the leaderboard instead because otherwise it's confusing. Same goes for vanilla, scores override by score and score is in the leaderboard. For relax, pp is the score override and it shows pp in the leaderboard

[minisbett]

Oh nvm now I understood

[minisbett]

Seems to be a bug I guess

[cmyui]

1st is planned yes - per-mod best scores is something i've wanted to implement for a long while
2nd - are you asking for your highest score to be displayed on map leaderboards, but highest pp to count towards global pp?

[cmyui]

also - each github issue should represent a single problem - if you have two issues, please create two issues on github, even if they are from the same category (e.g. leaderboards).

categories can be represented through issue tags.

[skrungly]

2nd - are you asking for your highest score to be displayed on map leaderboards, but highest pp to count towards global pp?

presumably, since this is how it works on bancho (at least for mania).

[tsunyoku]

first issue already exists at #560, second issue i don't really see happening because there's a lot of semantics and opinion around it so not really worth it