Let’s dive into the details of our ranking system, because we want to hear your opinion on how we can our babes.
Our current ranking system is pretty simple: once a babe has 25 valid votes, she can rank in our top lists. The result is that a newcomer may shoot up to a top spot in our top 100. Many of them, especially new number ones, get voted down very quickly. If a handful of people decide she’s not worthy by voting under 5/10, she will probably never reach the top 25 again.
Since we started tracking the #1 babe in a separate list in August 2020, we have seen 252 models reach the #1 spot. Around 200 of them only stayed in the #1 spot for a single day.
If you aim to have a top list where newcomers regularly get a shot at a high listing, but a small minority can shoot them down, then this is a good system. If you aim for more stability where only proven, famous babes with many votes can list high, the system can be improved.
If we look at IMDb’s Top 250 Movies, they only list the all-time classics. Only very highly rated new movies have a chance of ranking there, and even then, it require a large number of votes. They use a custom ranking algorithm based on weighted averages. The big difference with our system is that newcomers don’t enter high in a weighted average ranking.
If we were to apply this weighted average to our ranking, we would almost never see a sudden new #1. Getting to the top spot would take a perfect score of 25 times a 10/10 (and no lower votes). Maybe one or two 8’s or 9’s will still allow her to reach #1. Anything more, and she’ll drop quickly. So strong newcomers will likely enter much lower.
Below is a screenshot showing a recent top ranking on the left and a simulated ranking based on the weighted averages on the right.
Babes with more votes can rank higher. Relative newcomers like Francielli Fontana, Emily Feld, and Giusy Meloni are pushed out of the top 50 and can be found between 100 and 200. Dayane Hofer is around #450 and Bella Oelmann is even lower. Besides these newcomers with fewer votes, the top region would be close to identical.
So to summarize: if we switch to this weighted average ranking algorithm, we’ll have a similar ranking with much more stability and very little sudden surprises. Since Babepedia is mostly about discovering beautiful women, we are not sure if such a stable system is desirable. We can also increase the number of votes required for a ranking. Or do nothing at all 😊
What do you think? Let us know in this poll:
UPDATE 01/08/24: The poll outcome shows a slight majority for the weighted average ranking. This shows us that, while there is interest in doing something against the sudden new number one’s, many users also don’t want it to be too stable. We have enabled the new system as a test and it will become active tomorrow, on 02/08. We have opted for a tweaked version of IMDb’s algorithm, where newcomers should still be able to rank high with it, but it’s unlikely they will reach #1. A babe that reached 25 votes and would’ve otherwise landed in the top 5, will now be somewhere in the top 100.
We will evaluate this again in the coming days. If you have feedback, feel free to leave a comment or reach out to us via email.
I’m all for a system where we can create our own rankings. Be it a top 100, top 10, top 50 and so on. This ranking would be based on the user’s own preferences, and numbers would appear to indicate the position of that woman for us. Would such a system work?
Another idea that would be good too, and doesn’t have much to do with this ranking business, but which I suggested was:
When visiting a user’s profile, you could show the number of babies we have in common with the same user. In other words, the same women that we add to our favorites in that standard list that comes from the beginning, without having to create other lists, for example. On the site, I have Angelina Jolie, Salma Hayek and Ellen Rocche added to my favorites. If I were to visit the profile of a user who also has them, a number would appear, telling me that we have 3 in common. Which in this case are the 3 I mentioned. But if we had more than that, it could also appear. So? What do you think of an idea like this for the site? Would it be interesting? I look forward to your reply.
These are good ideas, but outside of the scope of this discussion. I suggest finding the first idea on our ideas page and voting for it, plus adding your comments.
The second idea is not there yet, I think. I invite you to submit it as a new idea with these comments.
https://feedback.userreport.com/bc265223-cb9d-4f67-bef3-9d57dc999860/#ideas/popular
I know with IMDB the old/simplified rank formula for Top 250 used to be a a Metacritic score of +80 and a weighted user score of +8 (with at minimum 1000 votes).
I think Babepedia needs to find it’s own formula. There has to be an optimized way for the same profiles to be rated equally by all users without relying on list to dig through all the pages of babes.
I do find it weird when we these crazy pushes from random profiles happen.
IMDb’s current formula is well-known. From their FAQ page: The following formula is used to calculate the Top Rated 250 titles. This formula provides a true ‘Bayesian estimate’, which takes into account the number of votes each title has received, minimum votes required to be on the list, and the mean vote for all titles:
weighted rating (WR) = (v ÷ (v+m)) × R + (m ÷ (v+m)) × C
Where:
R = average for the movie (mean) = (rating)
v = number of votes for the movie = (votes)
m = minimum votes required to be listed in the Top Rated 250 list (currently 25,000)
C = the mean vote across the whole report
We would use the same formula, with m being 25 and C being around 8.37. The table in this blog post uses this formula.
I’m a big fan of “data fidelity”. I’m not sure if that’s the right term, but if people are able to easily game the ranks by downvoting, then it defeats the rank in the first place and you aren’t receiving honest votes. I expect that honest votes, i.e. honest and accurate data, is what you want.
If discovering new models is what is most important to you, then as a site operator, I’d say create a separate “top 10” list of newcomers and pin that at the top of the homepage with the “all-time top 10” list appearing just below that. Maybe models that pass the 25-vote mark are only eligible for the “newcomer top 10” list for 6 months after that?
PS: I’m probably getting into the weeds here: what do you think of the idea of providing definitions for your star ranking system? I am guessing that the majority of people either rank a model a 10 or a 1 with little regard for gradient and nuanced voting. Maybe a little pop-up window can appear every time someone votes in the extreme, saying something like, “are you _sure_ you want to vote her a 10? Is she so perfect in every way that all other models are a 9 or less?” 😆
“I’d say create a separate “top 10” list of newcomers and pin that at the top of the homepage with the “all-time top 10” list appearing just below that. ”
We already have such a newcomer list with the “Rising Stars” page.
Oh. In that case, I’d say pin that to replace or be on top of the existing “top 25” on the home page. Making better use of that list may help with the data fidelity issue.
“Since Babepedia is mostly about discovering beautiful women, we are not sure if such a stable system is desirable.” Completely agree. I would normally support the stable system with a consistent ranking, but the idea for discovering new women only creates more opportunity for them to have a broader, more thorough ranking as they will be exposed to more people.
I dont care. My preferences: More pics, height and weight, bio, etc. Just my point, dont hate me.
I recently discovered this site and liked very much. I think besides the ongoing ranking system why not do a bracket (s) tournament (s) as an another way to add points and change the ranking too.
I don’t know if its possible, but I world love tô see it.
We’re playing with this idea, but haven’t found a good way to put it into practice yet.
https://feedback.userreport.com/bc265223-cb9d-4f67-bef3-9d57dc999860/#idea/394447 – This user suggestion could be interesting here.
Unlikely for a new girl to get #1? Todays #1 has 55 votes and wasn’t ranked before.
We’re still tweaking and testing the system. She got an exceptionally high score when she got her first rank. 9.38/10 is higher than most previous #1’s. After the weighted average calculation, she is still #1 by a small margin.
Do you think this is problematic?