The main idea is to introduce mod multipliers. You can adjust the difficulty of EZ/HR/HT/DT mod. Higher mod multipliers make the effect of the mods more notable. By default, EZ, HR, HT, DT would have multipliers 1.0x. I suppose the multiplier to cap at 10x, but reserve the possibility to scale it over 10x.

Let's talk about DT first. Currently, DT makes the speed of the gameplay 50% higher, with mod multiplier at 10x, it would become 500% higher maximally. But what about the score? Faster speed ought to lead to higher score, but as the speed increases, the bonus score should increase slower and slower. I think using the arctan function is a good idea here:



Now there are 2 requirements I want to fulfill:
1. When the mod multiplier is 1.0x, the score multiplier should be 1.12x.
2. When the mod multiplier is almost infinity, the score multiplier should be 1.24x.

And I'm going to scale the function linearly. I got this function:



The x-axis is the mod multiplier, and the y-axis is the score multiplier.
If you click the link above, you can see the formula I use on the left hand side.

When it comes to HR, we can just set the CS multiplier to 1+0.3*(mod multiplier), and AR/HP/OD to 1+0.4*(mod multiplier), all of which multiplied with the original value capped at 10. I'm going to use the same strategy, that is, when the mod multiplier is 1.0x, the score multiplier should be 1.06x, and when the mod multiplier is almost infinity, the score multiplier should be 1.12x. Here's the graph:



There are also those (arguably) difficulty-reducing mods, namely EZ and HT. EZ would multiply CS/AR/HP/OD with 1/(1+(mod multiplier)), and we can do the same thing as HR to the score multiplier for EZ:



For HT, we can multiply the speed with 1/(1+(mod multiiplier/3)). Unfortunately, the score multiplier would quickly become negative, I suggest simply to make them unrankable when the mod multiplier is over 2:

p sure theres already a request for different types of DT and such multipliers (1.1x and 1.2x speeds instead of only 1.5x and such)
also as long as that HT thing goes negative this probably wont happen at all.
too bored to check your functions and if your math is right but nice suggestion. id prefer it working in the arctan function instead of just a linear function (since the differences in speeds becomes relatively lower as we go higher up)

-Jukke- wrote:

p sure theres already a request for different types of DT and such multipliers (1.1x and 1.2x speeds instead of only 1.5x and such)
also as long as that HT thing goes negative this probably wont happen at all.

banning super slow time warping isn't actually a very bad thing either, or else there would be S plays on Mazzerin's crazy maps orz. That said, in fact I can change the function to converge towards 0, but that would make the algorithm of HT different than the others'.

UsaTewi wrote:

Faster speed ought to lead to higher score, but as the speed increases, the bonus score should increase slower and slower. I think using the arctan function is a good idea here:

Why slower? It gets exponentially harder. Shouldn't the score reflect that?

Also atan is not the function to use for this as it converges. That's bad if you are talking in terms of difficulty. Use ln instead as it does get slower and slower, but never converges. 1+0.05ln(10x+1) should get you around where you need it to be.

UsaTewi wrote:

For HT, we can multiply the speed with 1/(1+(mod multiiplier/3)). Unfortunately, the score multiplier would quickly become negative, I suggest simply to make them unrankable when the mod multiplier is over 2:

Use 1/(e^x)

Make CS/HP/AR/OD toggle-able in Mods
Controllable Song Rate

Please search around a bit when making a request next time