An alternative way to use the number of likes to find new games to play (list of games included)

[MrSilverLust]

The number of likes a game gets is one of the ways we can sort games in the Latest Updates section of this site. This metric is useful if you want to find the most popular games ever. But it’s also a heavily biased measure of a games quality: the older a game is, the more time it has had to get all those likes. So much so that it becomes a kind of metric to find classic games that stood the test of time. Which is fine. But what if we could also use it to find newer games worth playing?

I liked this suggestion by nokey69 that asked for a way to see how many likes a game got in a given period of time. Like, for example, see which games got more likes only during the last month to see what’s trending right now. Unfortunately, I don’t know how to get that data. But it did gave me a similar idea: find games with good likes/view ratios.

I’m not (yet) the best programmer, but I’ve always wanted to learn how to do some simple web scraping. So, I figured this was a good way to learn as any other. After a bit of research and trial and error, I got myself a list of the ~1500 more liked games in F95 (this obviously does not include all games on F95!), check how many views their threads had, and did the math to find out how many likes a given game thread had per 100k views.
I’m not sure if other people will find this useful, but I’ll share my findings with you.

A quick note: This way of looking at games is unfavorable for bigger games that have players constantly posting on their thread: this increases the view numbers a lot but not the likes, since a player can only like once, but can visit and comment in the thread many, many times.

So, the top 5 games with better likes/view ratio are:
Summer Heat [v0.1] [Naughty Fox Games]
Love Next Door [v0.1] [EroBro]
Real Dreams [v0.2] [ANW Creations]
In a Scent [v1.0] [Domiek]
Hard to Love [v0.01 Bugfix] [Qori Gaming]

As you can see, this metric finds games in early stages of development that got a very good number of likes at launch.
Check them out if you like to play what this community thinks are amazing first versions.
Disclaimer: I haven't played any of them, this recommendation is just based on the data.

But a lot of players might prefer games more established/that have had a couple updates under their belt. So, we can remove game threads with less than x views to exclude those still in early phases of development.

Here is the top 5 games with better likes/view ratio with at least 250K views:
The Bite: Revenant [v0.41] [madmate.games]
Eternum [v0.2] [Caribdis]
Heir Apparent [v.0.015b] [SilkandMilkProductions]
RSSU - Retro Style Soviet Undies - Case#1 Ithaca Project [v1.3.1] [ErosAmor]
Driven Affairs [v0.02a] [TEKUxMANITU]

And here with at least 500K views:
Eternum [v0.2] [Caribdis]
Apocalust [v0.03] [Psychodelusional]
FreshWomen [v0.3.0] [Oppai-Man]
A Wife And Mother Fan Game [Final] [Pixil]
Zeno's Anthology [v0.2.6] [Chemical Fire]

I attached my excel file to this post if you want to play with the list yourself.

I think it would be interesting if F95 implemented these kind of filters to help find new games. Doing something like this while also filtering for tags could be interesting.
Also, nokey69 filters suggestion would be really nice if possible to implement.

 

[Meaning Less]

I would start saying that number of likes is a bad metric, people like threads to thank uploaders, it hardly has any relation to liking the game.

Views just signal a thread is popular, again games that get pumped frequently to the first page are more prone to have more views regardless of how good they are.

Ratings also don't help much because plenty of good games get showered with low ratings just because they are sandbox, or have anything controversial whatsoever.

In the end all metrics really just help already popular games, the only thing that actually helped me find many hidden gems was using tags and filters. Because let's be fair, the current patreon and development model really doesn't help hidden gems, so people have to manually look for them unfortunately.

 

[jamdan]

Likes and views aren't good metrics. Unless it's in the extreme, a thread with 2k likes is probably pretty good. Or a game with 10m views is at least popular (or controversial). However, both of them are heavily influenced by longevity of the game being on here.

However, I think filtering by posted date could be useful. You can already sort via date for the latest update, but I'm talking about when the game itself was uploaded here (the date in top left of the OP). So, let's say you just want games posted in 2022 to be there. You put that in, and it filters out all the older games. Then you can apply whatever other filters you want. This is the most direct way to find new games to play, simply filter out all the older games.

 

[MrSilverLust]

I would start saying that number of likes is a bad metric, people like threads to thank uploaders, it hardly has any relation to liking the game.

If that was true, every game would have similar number of like per view ration because everyone would thank uploaders the same way. That’s not true. Maybe people thank uploaders more often if they like the game?

Likes and views aren't good metrics. Unless it's in the extreme, a thread with 2k likes is probably pretty good. Or a game with 10m views is at least popular (or controversial). However, both of them are heavily influenced by longevity of the game being on here.

Exactly. I’d even say that they are useless unless you’re someone finding adult games for the first time and want to play the most popular ones ever.

But more experienced players already know all of the classics. How do they find new games to play? Right now, only by looking at the latest updated games and see if something fancies their eyes. Can we do something else than that?


The idea here is having more metrics to help find great games to play that we might not be aware of. This is nothing new,

You must be registered to see the links

and

You must be registered to see the links

have similar likes per view ratios to help figure out what videos are good and we might want to see next. Checking what the community thinks is good content is a great way to find good content.


For example:
The newest NLT, The Genesis Order game has 959 likes. Summer Heat has 224.
The Genesis Order has 1.8 million views. Summer Heat has 97 thousand.


One game is so much more popular that it’s impossible to compare them.
But we can normalize both to the same number of views, let’s say 100k, and better compare them:
The Genesis Order has 53 likes per 100K views.
Summer Heat has 230 likes per 100K views.


As you can see, the few people that clicked on Summer Heat during the 2 days that it was in the 1st page of the Latest Updated list thought it was a good game. The rest of us that were not online that day missed it and will never hear about it.

This metric would help players find these games. When was the last time you sorted games by likes or views and found something interesting? Now check the lists I posted above and see if you find something you like. That’s the advantage of using likes in other ways than just sort by the total number, which is not very useful.

 

[Meaning Less]

If that was true, every game would have similar number of like per view ration because everyone would thank uploaders the same way. That’s not true. Maybe people thank uploaders more often if they like the game?

Well, people need to see the game first before they can like, so games that are released as completed are bound to find less eyes regardless of how good they are, it is no coincidence that most popular threads are from 2019 or earlier. They had enough time and bumps to reach that spot.

This can be easily verified if you look at japanese games. The most popular ones have been bumped multiple times to the first page either due to updates or translation patches, meanwhile multiple measurably better games are hidden deep in the pile of completed games because people just won't stumble upon them that easily.

 

[anne O'nymous]

I would start saying that number of likes is a bad metric, people like threads to thank uploaders, it hardly has any relation to liking the game.

Without counting the fact that this number include facepalms.

I don't have names that cross my mind, but I already seen games so horrible that the thread received more than one hundred of them.

 

[anne O'nymous]

[Sorry for the double post, but I finally wanted to say more regarding this subject]

If that was true, every game would have similar number of like per view ration because everyone would thank uploaders the same way.

Sorry, but this is just pure none-sense.

Since you focus on "liked new games", I'll pass on the fact that people goes back to the thread to talk about the game. But it still left a big problem that you haven't took in count: The way people found that the game exist...

There's four way to access a game thread:

• From a link posted on a thread somewhere here ;
• From the latest update page ;
• From the game section summary page ;
• From the latest update list (bottom of forum main page).

Two of them don't permit you to have a preview of the game visual.
Let's say that I don't like 3D games. I'll give one view and one like to games that are in 2D, but I'll only give one view to games that are in 3D.

Three of them don't permit you to know what are the tags of the game.
Here, the difference will be even bigger, because it's not limited to me liking or not the visual aspect of the game. It's a gay game ? Well, sorry but the straight guy I am isn't really appealed by such game. It have NTR ? It have water sport ? It have sharing ? It's have a female protagonist ? It's a none nude game ? It...
There's so many way one can goes to the thread, and so add a view to the counter, without even reaching the line where the "like" button is because he already don't like the game and will not even try it.
And the opposite is also true. There's kinks that are so rarely found, that one will like any game that have it, even if it's the worse games ever.


You aren't the first one who had the idea to count the number of likes per million views and, like I said at this time, you can twist all the numbers you want, it will never give you a list of liked/interesting games.

That’s not true. Maybe people thank uploaders more often if they like the game?

But is it really the game that they like, or is it the idea they have of it ?

It's while the game is downloading, therefore before you've played it, that you are facing the "like" button. Therefore, you can expect that at least half of the likes mean: "Wow, this game looks really promising, it even have this rare kink I like so much, thanks guy for sharing it, I'll try it once it will have finished to be downloaded."
And obviously, one would not go back to the forum, search for the game thread, then remove its "like" because he finally haven't liked the game.

Exactly. I’d even say that they are useless unless you’re someone finding adult games for the first time and want to play the most popular ones ever.

Even for that it's useless. The second game in terms of view it the now abandoned Milfy City, then come A Wife and Mother, and only after it come the game that have the more patrons, Summertime Saga.
One would say that a game who have more than 30 000 patrons is a very popular game, more popular than Being a Dick and its 14 xxx patrons. WVM, that is without doubt among the popular games here (despite it being overrated), isn't even in the first page in terms of view.

But more experienced players already know all of the classics.

Really ? I'm near to my five years of presence here, and I still discover a classic time to time.

How do they find new games to play? Right now, only by looking at the latest updated games and see if something fancies their eyes. Can we do something else than that?

The question is more: should we do something else than that ?
There's now more than 4.5 millions members on the forum, yet there's only one who time to time search another way to find games.
The others, especially the ones that are here since years, know perfectly that they are missing games, but don't care this much about that. Partly because they know that many games end abandoned, partly because it's in fact more pleasant to suddenly discover a gem you haven't found before.

What would be more interesting is to find a way to know that "this game" interest you, without having to follow the game thread and be notified every time there's a new post in it. Because I think that I'm far to be the only one among the oldest members, who time to time end on a game thread and think, "oh yeah, I remember this one, I really liked it... well, seem like I missed a lot of updates".

When was the last time you sorted games by likes or views and found something interesting?

I never sorted games by likes or views for another reason than to prove that they mean nothing. Precisely because I know that they mean nothing.
I prefer, by far, to browse by tags that effectively interest me, and find my gems that way. Perhaps it's less precise, but at least I'm not lured in loosing time with games that I'll never play again because I don't remember them enough. This just to discover, one year later, that they in fact have been abandoned in between.

 

[MrSilverLust]

This can be easily verified if you look at japanese games. The most popular ones have been bumped multiple times to the first page either due to updates or translation patches, meanwhile multiple measurably better games are hidden deep in the pile of completed games because people just won't stumble upon them that easily.

But then, aren’t you agreeing with me? I’m also saying that total number of likes or views is a metric heavily biased to older games. Do you think that, as is, the number of likes/views/rating/updated date are fair metrics for games released as completed? To me, it looks like that all of those sorting lists heavily benefit frequently updated games and punish, for example, games released as completed. Is this a good thing? Is "punishing" completed games what we want?

If, instead, we look at the likes/view ratio to compare a Japanese game that got a nice number of likes when released as completed vs the game with most likes on here:

Imperial Knightess Wein's Irregular Life: 119 likes, 212k views. Likes per 100k views = 56.
Being a DIK: 8821 likes, 38.9M views. Likes per 100k views = 22.6

The completed game would be doing better on this sorting list. Sure, as I explained in the OP, this metric would be unfair for popular games that go on and on forever, but they already have plenty of attention. This would benefit good games that are flying under the radar or were released as completed. Is it bad to have one more metric that helps finding good games that aren’t old/been updated many times? Or is it better to keep just using a system that popularizes games that are frequently updated and buries down everything else?

[Sorry for the double post, but I finally wanted to say more regarding this subject]

You make good points, even though you also strawman some of the arguments in favor of using likes as a metric.
But this is the core question:

The question is more: should we do something else than that ?

IMO, “the system” as is promotes games with lots of updates in short periods of time. Is this a good thing? Should games that release as completed or take their time to release only big updates be so discouraged? Can we use another metric that is independent of how old or how many updates a game has had? One that helps players find games that rely on providing good, substantial updates, instead of small, frequent ones?

I mean, you’re not wrong in saying that just me and a couple others seemed to look for alternatives... it's well possible that I'm alone on this.

I prefer, by far, to browse by tags that effectively interest me, and find my gems that way.

You still need to sort them somehow. Which, I suppose, will be by their latest update date. Which makes it much more likely that you find games that are often updated. Which might not be what you are not looking for, I don't know.

I think a way of sorting games that promoted good, long updates, good 1st versions, or even completed games would be of benefit for the community. What exactly this metric would be is difficult to tell. I made a simple suggestion of one that could be implemented relatively easily, I think. Might have problems, but, as the saying goes, the perfect is the enemy of the good.

 

[Meaning Less]

I think a way of sorting games that promoted good, long updates, good 1st versions, or even completed games would be of benefit for the community.

True, but the issue is that there isn't really any reliable metric to measure game quality, so changing how things are sorted would just redistribute bias to another metric making other games reach the top even if those aren't "better" either.

We probably would need a recommendation system first or something to know what should/n't be promoted to different users and help them find games that would interest them more precisely.

Afterall what is good or bad is even harder to measure when we are talking about adult games, because different users are looking for entirely different experiences, some want the quick-fap linear experience with minimum gameplay, others want the 50hour long run with as much gameplay as they can have.

tl;dr It's pretty hard to know what is/isn't quality when people are thinking with their genitals.

 

[anne O'nymous]

But then, aren’t you agreeing with me? I’m also saying that total number of likes or views is a metric heavily biased to older games.

No, because I affirm, and more or less demonstrate, that it's a metric totally biased and insignificant, period.
At no time I believe that it could be relevant under some condition ; this number is never relevant, whatever the way you look at it.

Do you think that, as is, the number of likes/views/rating/updated date are fair metrics for games released as completed?

I provided a link explaining my views... But since you clearly haven't read it, here's a TL;DR version:
Not a single one of the fucking metrics you can imagine is a fair metric, period.

Imperial Knightess Wein's Irregular Life: 119 likes, 212k views. Likes per 100k views = 56.

Oh, looks... Yet another one of these Japaneses RPG Maker games where you play a naive girl that will be abused and turned into a slut. And not a single combat in the screenshots...

I comment on it because it's symptomatic of the problem that your metric have.
It's the kind of "yet another game" that max its number of likes in one month. Those who are interested, and so will probably give a like, will not need more than this month to find it. And the others will find it by chance, what will increase the number of view, but not at all be interested, what will not increase the number of like.

The completed game would be doing better on this sorting list.

Doing better in regard to what exactly ?

Explain me, like I'm 4, how your metric will help completed games to to better than they already do with:

• The latest update page where you include the "completed" prefix ;
• The game section page filtered by first message and including the "completed" prefix.

What your metric offer that those two possibilities, that exist since age, do not offer ? And obviously, I mean "except yet another biased number that mean absolutely nothing".


If something is punishing a game, to use your words, then it would be his author itself, for failing to make it look interesting when presenting it. And nothing else.
Those who want to know what are the games released as completed can do it, and they are doing it. It's those who don't really care who, possibly, are missing those games... but like they don't really care...

IMO, “the system” as is promotes games with lots of updates in short periods of time.

But your opinion is wrong:
A game like Family affair, that is updated every 2 weeks, when not every week, only have 3 millions views after 18 months of existence, and a ratio around 215.
This while a game like Corrupted Kingdom, that is updated every 2 weeks, have 10 millions views for 32 months of existence, but a ratio around 195. Yet it's clearly the best of the two.

This compared to games like:
Being a Dick, updated every now and then, have a ~225 ratio.
WVM, updated every now and then, have a ~260 ratio.

But it's not really a question of seniority:
Corruption, a historical game, updated every month since years, have a ~128 ratio.
Nor a question of popularity:
Dating my Daughter, a historical and popular game, updated every now and then, have a ~150 ratio.

Can we use another metric that is independent of how old or how many updates a game has had? One that helps players find games that rely on providing good, substantial updates, instead of small, frequent ones?

And how your metric is helping exactly ?

The Deluca Family, is a popular game that have big updates every now and then, and a ratio around 390...
Heavy Five, is a less known game that have updates bigger than most finished games (It is 7GB and still at it's 4th update), and a ratio around 218.

 

[MrSilverLust]

No, because I affirm, and more or less demonstrate, that it's a metric totally biased and insignificant, period.

I know you don’t agree, I wrote that quoting and answering Meaning Less, not you.

I provided a link explaining my views... But since you clearly haven't read it, here's a TL;DR version:
Not a single one of the fucking metrics you can imagine is a fair metric, period.

Same thing here. I don’t know why you are answering me like if I haven't read what you wrote when I was quoting and talking to Meaning Less not you. I answered you down in that post, and appreciate your contribution to the thopic and what you have to say.

Doing better in regard to what exactly ?

Doing better as in showing closer to the top of that sorting list which would, theoretically, make it more easily found.

I comment on it because it's symptomatic of the problem that your metric have.
It's the kind of "yet another game" that max its number of likes in one month. Those who are interested, and so will probably give a like, will not need more than this month to find it. And the others will find it by chance, what will increase the number of view, but not at all be interested, what will not increase the number of like.

If it increases the number of views without increasing the number of likes it lowers it in the sorting list, promoting games to the top that the community thinks are more interesting. It’s working as intended.

But your opinion is wrong:
Family affair, a ratio around 215.
Corrupted Kingdom, a ratio around 195.
Being a Dick, have a ~225 ratio.
WVM, uhave a ~260 ratio.
Corruption, have a ~128 ratio.
Dating my Daughter, have a ~150 ratio.
The Deluca Family ~ 390.
Heavy Five ~218.

Those games are all already popular with millions of views!! This metric punishes those games by not showing them at the top. In fact, the most popular might be pushed down the list because they will get more people commenting and adding views all the time, but only liking it once. They would show up at page 50 or whatever.

On top of the list would be new games like Summer Heat with a ratio of ~2320 or In a scent with ~2020. This is an order of magnitude bigger than the numbers you showed. These are the games showing on top of this list and being promoted by this metric. Using it in those popular games it's useless and gives weird results. That's not the point at all.

Explain me, like I'm 4, how your metric will help completed games to to better than they already do with:

• The latest update page where you include the "completed" prefix ;
• The game section page filtered by first message and including the "completed" prefix.

What your metric offer that those two possibilities, that exist since age, do not offer ? And obviously, I mean "except yet another biased number that mean absolutely nothing".

That’s the point. Create "another biased number that mean absolutely nothing", but that is biased in a different direction. Older games already have likes, views, weighted rating in their favor. Is it such a big problem to have something that favors great newer games for those that want to find them?

Besides, those sorting lists show games in order of the most recently completed/updated. How is that relevant to know how their quality was perceived by this community?

At no time I believe that it could be relevant under some condition ; this number is never relevant, whatever the way you look at it.

The difference of this metric would be to present a list where the top is composed by good but yet unknown games that are still at early steps of development or were released completed (aka had less exposure).

If this metric works or not is testable. Using it, I gave a list of games that shows on top of it right now:

Summer Heat [v0.1] [Naughty Fox Games]
Love Next Door [v0.1] [EroBro]
Real Dreams [v0.2] [ANW Creations]
In a Scent [v1.0] [Domiek]
Hard to Love [v0.01 Bugfix] [Qori Gaming]

We can come back here in one year and check if any of these games become popular. Yes, this is a popularity contest, it’s impossible to objectively decide which games are better than others. But, if implemented, we could use tags to better refine our search to find games at early release stages that will become popular within the genre. In essence, find the good games, early. All the examples you mentioned have millions of views. These have less than 250k... This metric is focused on finding these games.


I might be completely wrong, but, by using this metric, I predict that these games are good and will have a bright future, without playing them or opening their thread.

If you’re so sure that this metric is useless, you’re free to use your prefered sorting method and write here a list of 5 games with less than 250k views that you think will be popular in one year, without playing or looking at any of their threads (because that’s what I did).

Then we can compare if this metric, as I expect, is good at identifying and promoting great games at yearly phases of development, or if you can use any other currently available metric to make a better prediction (or if any sorting method is crap and useless for this, and I’m wrong, and this task is hopeless).


I fully understand that this won’t be useful for everyone, but I think that there are lots of players here that enjoy finding and following great games at early stages of development and would like a sorting method to help then find those good games in the middle of everything that’s published that’s not that good.

Also, this easy to implement feature would incentivize developers to take their time to work on their 1st releases (or even being benefited by releasing completed games!) because a great launch would put them on the first page of this sorting list for a while. I fail to see how this would be a bad thing for the community.

 

[anne O'nymous]

I know you don’t agree, I wrote that quoting and answering Meaning Less, not you.

Oops, my bad. Not being able to freely move my neck tend to make me misread the small parts, especially when they are dark on dark.

Doing better as in showing closer to the top of that sorting list which would, theoretically, make it more easily found.

So only for the persons who not only order differently than the default "last message date", but also order by number of likes. Are they really this numerous ?

Those games are all already popular with millions of views!!

3 millions view in 20 months (Family Affair, updated every two week at worse) or, worse, in 38 months (Heavy Five), on a forum with more than 4 millions members... How can one believe that those games are popular ?
It's not because half the games do not even reach one hundred of thousand views, that a thread with few millions views in years is popular. A popular game, like Being a Dick by example, have near to twenty times more views/month than Family Affair.

But anyway, I was just showing you that you are wrong when you claim that the actual system favorite frequents updates ; since the games with the most frequent updates have the lowest ratio. And that your metric favorite games with big updates ; the game with the biggest updates have one of the lowest ratio.
And in the same time pointing that your metric hardly punish interesting games with frequent update ; it's the game that have the lowest ratio.

On top of the list would be new games like Summer Heat with a ratio of ~2320 or In a scent with ~2020. This is an order of magnitude bigger than the numbers you showed.

So your metric is purely designed to favorite even more popular games....

In a Scent is awaited since a little less than a year and, after one and half month, the thread have more views (154K) than the ~3/5 of the games available here have had since the creation of their thread. The same apply more or less for Summer Heat, that have 97K views in just one week less ; more the ~2/5 of the available games.
And please, don't pretend that they aren't popular games. Above you claimed that, because it had 3 millions views in 38 months (78K views/month), a game like Heavy Five is a popular game that don't need help.

But well, congratulation I guess. You achieved to find an even less equitable way to order games ; a metric that favorite games that absolutely don't need it because they already are more popular than more or less half the available games.


No, seriously, thought about it for a minute. You just proved, without a single possibility of doubt, that people do not need your metric to find games that are, I guess, good ones.

 

[MrSilverLust]

So only for the persons who not only order differently than the default "last message date", but also order by number of likes. Are they really this numerous ?

I don’t understand what this means.

If I was not clear before, this way of sorting would be for anyone interested in finding new games that are good, but doesn’t want to go every day to the latest updated sorting list and check what’s new. Steam and itch both have a section called “New and Trending” or similar, where people can find good new games and not just a list of everything that has been released by date.

But anyway, I was just showing you that you are wrong when you claim that the actual system favorite frequents updates ;

Of course the actual system frequents updates! Game threads of new games get lots of views and downloads the first week after an update and are dead until the next update comes around. People only find new games they are not aware of when they are on the top of the updated list. Early on, the more updates a game has, the higher chance it has of being found for any given player.

But well, congratulation I guess. You achieved to find an even less equitable way to order games ; a metric that favorite games that absolutely don't need it because they already are more popular than more or less half the available games.

You have it the other way around. Because this metric found good games it means that it works as a means to find “New and Trending” games. Of course it'll find the ones that the community has shown to be good, otherwise it would be useless. It's doing what is intended to do.


But I reiterate my challenge.
This is the problem I face: I want to find the best games that have been released recently, (let’s say last 3 months). I don’t have time to check every day/week all the newest released games, then check all the threads, and then try to figure out what’s good or not. I'd apreciate some sort of list that is somewhat curated and already eliminated all the “bad games” that are released every week.

Please tell me how can I use the available sorting methods to get that list. Feel free to post the list here.
We can then compare our lists and see which one did a better job at finding "New and Trending" games.