Ever thought about diminishing returns?
It's not hard to code such an algorithm for experienced coders i think.
However, it would really help to take away the fear, that converted maps could count too much.
Let's say you have a stream of 20 or 30 dons in a converted map above 200 bpm, which is still hard but not seen as "skilled" for taiko as you state.
Now the diminishing returns comes into effect:
The first don will give 100% of points (invisible ones, since it shouldn't affect the way of the scoring).
2nd: 50%
3rd: 25%
4th: 12.50%
5th: 6.25%
...
and so on.
If there is a Kat, it will give 100% points again and reset the multiplier for the next don to 100% again also.
So if there is a map with very little variation and much spamming, it would only give a little amount of points compared to taiko specific maps with much variation.
You could also let the system detect repetitive patterns and give less points for every same pattern as the previous.
This way "dkdkdkdkdkdkdkdkdk" wouldn't give 100% all the time because it is repetitive.
So that is why i said, if you have a good algorithm, you really don't need to fear, that converted maps could count too much, because if they'd do, the algorithm would be wrong.
Greetings,
NoYzE
It's not hard to code such an algorithm for experienced coders i think.
However, it would really help to take away the fear, that converted maps could count too much.
Let's say you have a stream of 20 or 30 dons in a converted map above 200 bpm, which is still hard but not seen as "skilled" for taiko as you state.
Now the diminishing returns comes into effect:
The first don will give 100% of points (invisible ones, since it shouldn't affect the way of the scoring).
2nd: 50%
3rd: 25%
4th: 12.50%
5th: 6.25%
...
and so on.
If there is a Kat, it will give 100% points again and reset the multiplier for the next don to 100% again also.
So if there is a map with very little variation and much spamming, it would only give a little amount of points compared to taiko specific maps with much variation.
You could also let the system detect repetitive patterns and give less points for every same pattern as the previous.
This way "dkdkdkdkdkdkdkdkdk" wouldn't give 100% all the time because it is repetitive.
So that is why i said, if you have a good algorithm, you really don't need to fear, that converted maps could count too much, because if they'd do, the algorithm would be wrong.
Greetings,
NoYzE