A while back I was wondering about how the revenue is distributed among the models. What percentage of the models make what percent of the revenue? I thought that if revenue has a linear relationship to camscore, then I can figure it out. By linear, I mean that camscore is directly proportional to revenue. For example, someone with a camscore of 20000 makes 2x compared to someone with a score of 10000 and 4x compared to someone with a score of 5000. At the simplest level, it's like assuming that camscore means tokens per day. Not that that's the case, but it's assuming that sort of relationship.
One night back around Dec 5th I recorded everyone's camscore. I then added up all the camscores and figured out what each person's percentage was. The results were something like this:
her % Adding up all the camscore values equals 1052489. This column specifies what percentage that model contributes to the total. So for DevonAlexis, her value is 15282/1052489 = 1.45%. She is making 1.45% of the total revenue.
total % The sum of all the "her %" values up to this point. This gives the cumulative amount of revenue for all models at this rank and higher. So for 3rd row, it's saying that those three models are making 4.58% of all the revenue.
rank %Indicates the rank percentage of each model. There were 1116 total models. This value is rank/1116.
So what does this tell us? 20.23% of the revenue comes from 2.24% of the models (the top 25). 50% of the revenue comes from the top 13%. If a model has a camscore of 1500, she's right in the middle of the revenue. 50% of the models make more than her and 50% make less. 60% of the revenue comes from the top 20%.
This chart is just a guess. If the relationship is not linear, then the results would be incorrect. I'm assuming the formula is something like $=z*score. But if it's something like $=z*score^2, then the % would be different.
MFC may be artificially bumping the highest models score. If that's the case, then a higher % of models contribute to the top revenue. For example, MFC might be bumping the top 3 model's score higher making it look like they are bringing in a higher percent of the revenue.
There's also the 1000 score newbies. They have the effect of making the bottom models appear to bring in more revenue than they really do. If you take them out, the top models actually make an even higher percentage of the total revenue.
Of course, we don't know for sure how camscore relates to tokens. I have some ideas of how to figure out how they relate. I'll post them in the Camscore ??? thread.
One night back around Dec 5th I recorded everyone's camscore. I then added up all the camscores and figured out what each person's percentage was. The results were something like this:
Code:
rank name score her % total% rank %
---- ---- ----- ---- ----- ------
1 CrazySysy 16720 1.59 1.59 0.09
2 Tattgoddess 16253 1.54 3.13 0.18
3 DevonAlexis 15282 1.45 4.58 0.27
...
25 Laylie 4791 0.46 20.23 2.24
26 LacieJames 4759 0.45 20.68 2.32
...
151 SexxxyAndreea 1519 0.14 49.85 13.53
152 Imogen21 1510 0.14 49.99 13.62
153 Hottalicia 1504 0.14 50.14 13.71
154 YourHotAngel 1495 0.14 50.28 13.80
...
223 Briannaxx 1159 0.11 58.90 19.99
224 HelenAngel 1150 0.11 59.01 20.07
..
1113 LoveSunshine 63 0.01 99.98 99.73
1114 DivineAna 60 0.01 99.99 99.82
1115 PlayfulBoobs 51 0.00 100.00 99.91
1116 Anzelika 49 0.00 100.00 100
her % Adding up all the camscore values equals 1052489. This column specifies what percentage that model contributes to the total. So for DevonAlexis, her value is 15282/1052489 = 1.45%. She is making 1.45% of the total revenue.
total % The sum of all the "her %" values up to this point. This gives the cumulative amount of revenue for all models at this rank and higher. So for 3rd row, it's saying that those three models are making 4.58% of all the revenue.
rank %Indicates the rank percentage of each model. There were 1116 total models. This value is rank/1116.
So what does this tell us? 20.23% of the revenue comes from 2.24% of the models (the top 25). 50% of the revenue comes from the top 13%. If a model has a camscore of 1500, she's right in the middle of the revenue. 50% of the models make more than her and 50% make less. 60% of the revenue comes from the top 20%.
This chart is just a guess. If the relationship is not linear, then the results would be incorrect. I'm assuming the formula is something like $=z*score. But if it's something like $=z*score^2, then the % would be different.
MFC may be artificially bumping the highest models score. If that's the case, then a higher % of models contribute to the top revenue. For example, MFC might be bumping the top 3 model's score higher making it look like they are bringing in a higher percent of the revenue.
There's also the 1000 score newbies. They have the effect of making the bottom models appear to bring in more revenue than they really do. If you take them out, the top models actually make an even higher percentage of the total revenue.
Of course, we don't know for sure how camscore relates to tokens. I have some ideas of how to figure out how they relate. I'll post them in the Camscore ??? thread.