I was thinking about how to calculate a player's improvement rate. I want to present this idea called the Improvement Rate Index (IRI), much like the Body Mass Index (BMI), it would measure a player's improvement rate. Maybe the weight counterpart would be the total play time and the height counterpart would be the performance points? There is still the IRI classification to know if you are playing in a healthy rate or not. I have no clue about how to calculate a player's improvement rate but it would be helpful for everyone to know if the rate they are playing is healthy or not.
forum
Calculating improvement rate? Improvement Rate Index (IRI)
posted
Total Posts
11
What
There’s no way you can accurately measure a user’s rate of improvement. Everybody is different and we all have some days where we either make some insane plays or completely play in a slump.
this is obsessive af and i feel like if it became a thing it would do more harm than good to be honest with you
https://www.reddit.com/r/osugame/comments/l05i9i/osustandard_statistical_relationship_between/
Someone already did this with playcount and PP (see link above). If you look at the bottom of the thread you can see someone asking for what you're looking for and OP responded. The problem now is that the data he provided with those mirror links is no longer available.
Also this russian guy made a post documenting some statistics for players in his country:
community/forums/topics/397055?n=1
The information is out there, the big problem is compiling the data. You will either have to manually gather it (which could take a LONG TIME depending on how big you want the sample size to be), find and use an already existing data collection tool or code one yourself.
Someone already did this with playcount and PP (see link above). If you look at the bottom of the thread you can see someone asking for what you're looking for and OP responded. The problem now is that the data he provided with those mirror links is no longer available.
Also this russian guy made a post documenting some statistics for players in his country:
community/forums/topics/397055?n=1
The information is out there, the big problem is compiling the data. You will either have to manually gather it (which could take a LONG TIME depending on how big you want the sample size to be), find and use an already existing data collection tool or code one yourself.
That is outdated and measures against playcount instead of hours. A better dataset is https://www.reddit.com/r/osugame/comments/l09a7l/modeling_performance_as_a_function_of_hours/ which says the average player's pp = 37 * (hours ^ 0.75)Simon12 wrote:
https://www.reddit.com/r/osugame/comments/l05i9i/osustandard_statistical_relationship_between/
If we say IRI is a ratio of your pp to the average player's pp with your hours, an equation would be IRI = pp / (37 * (hours ^ 0.75)), where 1 is average. My IRI is 1.121.
This equation also accounts for improvement rate changing over time; Mrekk's current IRI is 2.240 because he hasn't gained much pp recently but when he had 608 hours and 13313 pp his IRI was 2.939.
Topic Starter
thank youQuantumKatana44 wrote:
That is outdated and measures against playcount instead of hours. A better dataset is https://www.reddit.com/r/osugame/comments/l09a7l/modeling_performance_as_a_function_of_hours/ which says the average player's pp = 37 * (hours ^ 0.75)Simon12 wrote:
https://www.reddit.com/r/osugame/comments/l05i9i/osustandard_statistical_relationship_between/
If we say IRI is a ratio of your pp to the average player's pp with your hours, an equation would be IRI = pp / (37 * (hours ^ 0.75)), where 1 is average. My IRI is 1.121.
This equation also accounts for improvement rate changing over time; Mrekk's current IRI is 2.240 because he hasn't gained much pp recently but when he had 608 hours and 13313 pp his IRI was 2.939.
Also this https://ameobea.me/osutrack/ is an interesting website to see your statistics
thats not a way to check improvement rate with play hours thoughashura_amane wrote:
Also this https://ameobea.me/osutrack/ is an interesting website to see your statistics
Well we can look at correlations like play time and ranks, sleep time and ranks, average playcount per day vs improvement rate etc.. This will give us some idea whether its impossible after all.Stomiks wrote:
There’s no way you can accurately measure a user’s rate of improvement. Everybody is different and we all have some days where we either make some insane plays or completely play in a slump.
Considering BMI can be pretty inaccurate, u can assume the same for IRI. Despite the many numbers and constants that exist, it is often rare that smth (data,etc) can be assumed as constant, consistent and comparable unless there are control variables (which in this case, you'd be limited by the number of variables u could measure). Any constant to account for such a high level of variability is usually complex and takes incredibly advanced scientists and/or mathematicians (both isn't rare depending on what he testing for, tho numbers entails more mathematicians) and a long period of insanely dedicated time (yes, even more than the average osu! player).
Damn, this turned out longer than expected, mb (just finished my final exams 💀).
Damn, this turned out longer than expected, mb (just finished my final exams 💀).