@Tom94

Tom Long Dong

Ask @Tom94

Sort by:

LatestTop

Previous

Is it impossible for aspects of NoMod to be buffed without directly buffing DT as well? :(

Depends on how you look at it. It's possible to make slower patterns weight more relative to faster patterns and thus have an effect which you would likely perceive as a nomod buff.

Related users

I find it kind of dumb to have separate scoreboards for DT and NC when it's basically the same mod... Will you "fix" it or is it gonna stay like that ?

Kynan383’s Profile PhotoKynan
I personally think it makes more sense to throw them together rather than seperating them. We will see. :)

All the scores would be kept if a player plays: No-Mod, HD, HR, DT, FL, EZ, HT, NF, HD+HR, HD+DT, HD+FL, HD+EZ, HD+HT, HR+DT, HR+FL, HR+EZ, HR+HT, DT+FL, DT+EZ, FL+EZ, FL+HT, EZ+HT, HD+HR+DT, HD+HR+FL, HD+HR+EZ, HD+HR+HT, HD+DT+FL, HD+DT+EZ, HD+FL+EZ, HD+FL+HT, HD+EZ+HT, HR+DT+FL, HR+FL+EZ, etc...?

Yep, but only the highest pp giving score would count towards pp and the highest score would count towards ranked score.

Is there going to be implemented a system that prevents you from loosing pp? http://puu.sh/c0QNo.png I just made this score on Satellites, before i had 98.69% acc diff with DT, and I thought it would improve the pp to add HD. However, it just resulted in me loosing PP when i beat my own score :/

Thomas Surlykke
There won't be a system to completely eliminate losing pp, but scores will be kept per-mod, so you can't lose pp by making a higher score with different mods.

I was thinking it would be cool for people that specialize in one aspect pretty well can search maps by it, for example you can type stars=\<\>X , it could go for aim\speed\acc as well, people could search maps by aim levels for example 'aim>X' it would be easier for people to find maps they like.

Yes, I did plan things like that, however some changes need to be made on the client side for that to be possible in the first place, since currently the client only stores the final star ratings and not their components (aim, speed and so on).
Liked by: Eyal Fridja

For the performance point bonus "There is also bonus pp based on the number of ranked maps you have a score on. The bonus is: 416.6667 (1- 0.9994^Number_of_scores)" If I got a score on easy, normal, hard, and insane all of the same mapset, would that be 1 score or 4?

That's 4 scores.

do you, by chance, happen to have the source code for your old difficulty calculator client?

Yes, but I do not consider it in a releasable state right now, so I won't give it out, sorry. There might be some older iteration of it floating around in the depths of my puush account - you might be able to find something in one of my older answers.

So... OWC is coming! I know that you didn't watch it last year (you even streamed during one of the matches... like, wtf... WHO STREAMS WHILE OWC IS GOING ON?!), but will you watch it this year?

I'll participate this year in all likelihood.
Liked by: Cozzzy Faggod

Hey tom, if there's no ETA for small circle buff why not just do it now:P will satisfy most hr players and you won't get spam and slam daily. Better to do it now than never because we all are waiting

Because it's not trivial to do?!?
Liked by: thelewa

The new pp system is almost spot on, but there is still this map: http://osu.ppy.sh/b/118068&m=0 that doesn't get the stars it deserves, NO ONE except Cookiezi FCed and people are still far away from FC, so why is it less valued than Airman wich as been FCed multiple times?Same goes with BDsPony map

Because the algorithm isn't perfect. I am well aware of those maps and any attempt at giving them the pp they deserve distorted the results in other maps far too much. I think, that the algorithm needs some proper pattern difficulty formula to be able to properly deal with these maps.

What is the current speed of score processing? Like 10k scores per second? I'm writing a replay simulator and want to know whether it sucks or not.

The processing itself is at around 50k per second (on a relatively weak VM). When having to update the database values however the database becomes the bottleneck and it goes down to like 4k per second.

Will it ever be possible to calculate accuracy pp based on unstable rate?

For past non-top50 scores not at all and for all scores with replays stores maybe when our processing power is large enough and someone (maybe myself) finds the time to program a server replay simulator that can go as fast as possible.
This is basically a per-hitobject information thing. For future scores another possibility would be to just let the unstable rate be calculated on the client side and then sent to the server together with the score, but this would be very hard to reliably protect against cheating.

So you're biased towards DT players man it's guys like you that ruin the game once pp v2 was implemented there was a glimmer of hope but you just ruining it.

I don't know if I should laugh or cry. Please read my last answer again and realize that I was quoting examples...

Any plans to buff NoMod then? Or nerf the bonus received from DT?

I'm so tired of phrases like "nerf DT" or "buff HR". For the last time: the algorithm does NOT work at a mod-level. What I can buff or nerf is scaling of circle size, speed between notes and so on.
If it was as simply as dialing up a "NoMod number" I wouldn't be required for managing the pp system in the first place. x_x

Why is checking periodically for changes in difficulties required? Can't the BSS communicate with the difficulty processor every time there is a change?

Writing a specific communication protocol would be additional overhead that can go wrong and more work to program. Having a polling based solution like the current one (costs almost nothing to check for changes every 10 seconds or so) isn't really at a disadvantage and on top of that is more robust due to only depending on the database and being BSS agnostic.
Liked by: DroppedBass

Next

Language: English