Alessia Uva

You are an old man on this site, maybe retire from this

Posted by kjbun3Reply0 02025-04-22 01:03  🛈  ⚠

I can't believe to my eyes.
The issue is NOT pushing some profile up.
It should require a max effort with low chance to succeed.
Top100 excellence requires 9.05+
So it takes only one 1.0 to annihilate nine 10.0 (it's not my opinion, it's math for 11yr old).
The real issue is how to avoid that few unlikely votes from trolls with multiple accounts, make disappeares some beautiful candidates in #5000 or even lower.

Real 1.0 given by trolls are too easily decisive.
A single asshole can overturn the global ranking for hundreds of babes.

THIS is what you should contrast.
You know who they are.
Alert them and, eventually erase their accounts.

We can't stand with this.
An overall ranking of 120.000 profiles controlled by few well known trolls.

Pushing up is very difficult.
Pushing down is too easy.

Posted by Omri_SboddereReply0 02025-04-23 00:41 (edited 2025-04-23 00:44)  🛈  ⚠

I agree, but it's hard to prevent... Lots of involuntary celibate users with parasocial relationships that only give ones and tens.

Posted by LeLoyon71Reply1 02025-04-23 01:52  🛈  ⚠

A simple weighted sistem based on a gaussian distribution centered on the average vote?
It would annihilate 1/2/3
Would give low importance to 4/5/6

(if you look at some voting distribution there are absurde bends in the graph just around 6)

And obviously, it would calm down strange, not proper amount of 10

Posted by Omri_SboddereReply0 02025-04-23 02:12 (edited 2025-04-23 02:23)  🛈  ⚠

"...users with parasocial relationships that only give ones and tens..."

Lol
I think I know one of them...

Posted by Omri_SboddereReply0 02025-04-23 02:14  🛈  ⚠

Times that by a couple dozen and you'll be closer to the number haha

Posted by LeLoyon71Reply0 02025-04-23 06:47  🛈  ⚠

Just found the second here Millie Tanaman

Posted by Omri_SboddereReply0 02025-04-24 21:45 (edited 2025-04-24 21:46)  🛈  ⚠

You're making many assumptions here, and most of them are wrong. The system is already taking several measures to make low votes less powerful. We are already applying weighted ranking.
However, the issue remains that a handful of votes, even in the 5 to 8/10 range, can push a relative newcomer far outside the top 100.
The current number one has a rating of 9.12/10. The current 10000 has a rating of around 8.77/10. That is a very small difference, making it very sensitive to votes that are not 9 or 10.

Rather than focusing only on the tweaking of the algorithm, we think it's better to focus on spotlighting girls that have dropped hard. We may add a list soon that focuses on this. We already have the "Most Days At Nr 1" list: https://www.babepedia.com/numberonetop100

Posted by Pryda (admin)Reply0 02025-04-23 11:56  🛈  ⚠

There are pros and cons to increasing the threshold. Increasing it reduces the number of high newcomers, but it doesn't eliminate it. If we increase it to 30, 40, or 50, you'll still have high newcomers. A bit less, but they're not gone.

You have to keep in mind that some models take a year or longer to gather 25 votes. The people keeping an eye on the Rising Stars list are the ones pushing these high newcomers. They tend to rate new models rather high.

Then you have a small group of people who don't check that list but do watch the top 25, and they also feel like most high newcomers don't deserve a top 1000 rating, rating them 6/10 or lower. This often results in them dropping to positions far outside the top 1000.

Increasing the threshold also reduces the variation. A model like Alessia Uva may never reach the top 25 if we require 50 votes to rank, because there may be one person who gives her a 3/10 or two persons giving her a 6/10. Some models survive the high-ranking start and stay in or near the top 100. Some people like this variation and the surprises. Would you rather have a ranking where there is less variation and almost no chance of someone new reaching the number 1 spot or even the top 25?

Posted by Pryda (admin)Reply0 02025-04-21 15:36  🛈  ⚠

For what its worth personally, seeing different people in the Top 25 or Top 100 is something I prefer over a mostly static ranking.

Posted by DarolisReply1 02025-04-21 17:43  🛈  ⚠

Or even a pop-up window - when voting - that alerts the user that a 3.0 given at babes just dropped out from RisingStars grid is statistically unlikely.
I mean : for 1.0 I can even take into account a typing error (it should have been a 10.0).
But for a 2.0 the algorithm should immediately ask the user : are you a troll or are you a retard? In both cases you are not welcome.

Posted by Omri_SboddereReply0 12025-04-21 19:42 (edited 2025-04-21 22:40)  🛈  ⚠

We shouldn't call people retarded because their taste differs from ours.
You have given over 100 models a 1/10 or a 2/10. Some of them deserve an 8+, in my opinion.

I don't think a popup will change people's minds too. It will definitely annoy them.


Posted by Pryda (admin)Reply2 02025-04-23 12:54  🛈  ⚠

I don't care one way or another... I was just explaining why it the rank threshold rather than foul play.

Posted by LeLoyon71Reply0 02025-04-22 02:37  🛈  ⚠