Astats imbalance in points awarded.

[AlexTheRussian]

I am brand new to Astats, just joined today for the hell of it. I am sure this topic have been covered, but I couldn't find anything relevant so I thought I'll ask.

Even from brief browsing of the site it is clear that point formula is heavily weighted toward free to play games, where entry level is minimal, first achievement is likely earned almost immediately but number of people 100% the game is abnormally small, so all the rarest achievements, regardless of actual objective or subjective difficulty are going to free to play games.

Examples are all over, from Card Hunter to Team Fortress 2 etc.

And it also seem obvious that people who are actively competing in Astats are forced to play those free to play games to earn points (never mind Zup style of games farming etc).

So my question is: Is Astats actually being developed/looked at since 2011? Or everyone using it is so vested in points they already earned that adapting the formula toward better balance is simply not going to happen?

Likewise, what's with Zup games? They clearly game the formulate to prey on people who would exchange money directly for achievement number, but while I see some discussion, not much is actually being done?

And finally, what's with 1500 hours played in the last two weeks stats? Clearly a person is running multiple games (for those who don't know, 336 is the hard codded limit of real hours :) on multiple machines but is that the spirit of Astats?

[lilcoffeebean]

It really depends on the F2P game. Some are worth a lot of points as you said, but the reason for this is because they frankly a lot of people start them on Astats but few complete them. There are others reason as well, such as many of the mmo's as an example are some of the least efficient games from an hour/point ratio. So the top players don't touch such games unless the gameplay can be largely idled, automated, or cheesed. Exceptions exist ofc.

You pointed out card hunter, but there's only been 1 person whose even 100%'d it, out of the 3887 that have started it. Regardless of what formula Astats chooses to use, that game is going to be worth a lot of points. The formula currently is 3/square root (x/y))-2.9
x= people with the achievement you're looking at
y= people with at least one achievement in the game
How many people own the game isn't used in the formula at all, just people who have an achievement and have a profile on Astats.

Another reason, is the fact that while these games can take many 100's of hours to complete, eventually they go offline and when they do, that game is now worth 0 points as Astats doesn't count games where achievements can no longer be earned by everyone. This scares a lot of people away from f2p mmo's that can take 100's of hours to complete, especially if they're not from a large, well known developer.

Game's like Zup, similarly are an issue on every site and it's something that's been and is being debated. The previous formula the game was worth 2200 points, a couple of weeks ago the formula was changed and now it's worth a tenth of that. It wasn't a perfect solution, but the debates showed that it was very difficult to come up with a solution that a) everyone would be happy with, b) addressed the problem in a way that was easily automated or c) addressed it in a way that couldn't be gamed. The Astats forums tend to be where this occurs over the Steam forums, for the most part.

The site is still being developed though and suggestions are listened to as well as debated. As far as the hours played in the last two weeks, there's nothing Astats can really do about that. The number is pulled straight from Steam and as people can play (or idle for cards) in multiple games and the time is counted for each it's going to go over the technical maximum. Even if this was arbitrarily capped at the maximum, it wouldn't accomplish anything. Unfortunately, it's a rather meaningless metric.

Also, weclome to Astats. :p It's not a perfect site by any stretch of the imagination, but of the sites that assign points to Steam achievements, it's one of the better and more active (development and playerbase) one's.

[AlexTheRussian]

Thank you for comprehensive reply, I am glad the site is still being developed and watched.

I think it was more of a newcomer shock to see unexpected point results, when I joined Astats.
Achievement in the free to play game that takes couple of hours (and 10 times more people have by percentage) to complete is worth more points than something 10 times less people have (again by percentage) and takes 30-40 hours to complete.

I expected the rarity and difficulty of the achievement to take more prominent role and was unplesantly suprised to see everyone's rarest achievements dominated by free to play games.

But understanding limitations of Astats that you mentioned certanly helps, I think I will use it as more of alternative view on achievements than actual competetive environment. It doesn't seem worth putting effort into climbing the ladder where fastest advancement is through gaming the system in place.

[twitch.tv/Hikikomori523]

Points aren't rewarded for difficulty, its rewarded by rarity.

Rarity can be insentivised towards certain platforms. If a lot of people play F2p but don't finish it, then the people who do finish it have a more rare achievement compared to the amount of players who played the game.

If you have a small game that very few people played, then its less rare compared to the amount of people who played it. It is weighted game by game, against the people who own and play the game.

[AlexTheRussian]

That would be logical in a level playing field, however, there is a huge difference between F2p games and normal games.

The fact that the game is free, entices a lot of people who have no intentions of playing it beyond few minutes or hours, because again, it's free.

Normal games require commitment of some kind, because you are paying for it, you already for the most part know that you will play it for some duration.

That creates a huge disparity in number of people who get 1 basic achievement in either games, in free to play that could measure hundred times more people because it cost them nothing to do so and they didn't stick around.

To give a comparable example:

In Astats world, climbing the Everest would be much less of an achievement than riding subway from end to end.

Very few people attempt to climb the everest and fewer yet succeed because of difficulty and skill.
However, millions of people ride at least through at least one station on subway daily, however, the number of people going from one end to another is insignificant in comparisson.

By Astats formula, those people REALLY did something amazing and will get the highest score for their achievement, while person who completed Everest will be graded much lower.

We can do the math if you prefer.

I don't expect Astats to change, because I don't know how you could have formula automatically account for difficulty of achievement. But that by no means that Astats formula is really viable to measure rarity because it doesn't account for level playing field.

Using it as competetive environment would be strikingly similar to real sports where doping is allowed. If you aren't doping, you aren't going to place high, so everyone at the top is doping.

It's not their fault, it's the only way to advance in the system that is imbalanced.

[lilcoffeebean]

I'm inclined to agree, the site actually does an extremely poor job of reflecting difficulty but you more or less touched on the main issue. It's very difficult mathematically to come up with something that would lead to more ideal results. The option for manual adjustment of course exists, but that brings up other issue's, largely with subjectivity, workload, and an overall lack of consensus on what qualifies as difficult.

However, I don't agree that you need to 'dope' to reach the top. As an example, look at Dirty Harry. He is currently ranked 43rd and has only played 4 free to play games. Gajusz at 9th has played 13 one of which isn't counted (TF2) towards his points. I won't split hairs though, it would be easier to farm the easier/fast f2p's to get there quicker, but for me as long as it can be done then I'm content. I would love to actually have access to the metrics, to see how many points from f2p were actually attributed to x persons account, but alas it's too much work to figure that out manually.

I certainly understand what you're getting at though and can recognize why you wouldn't want to use it as a competitive environment.

[Jaketwo]

By the way, the formula also heavily favours games with multiple DLC, as they almost always have a diminishing return of players. DLC costs money, so it's kind of the opposite of favouring free to play.

[AlexTheRussian]

Good point on DLCs, same issue with the formula. That said, since free to play games also have paid DLC (Card hunter for example) it affects them as well.

[lilcoffeebean]

I'll admit I'm a bit perplexed with the dlc issue. Both AchStats and TSA seperate dlc into a seperate entry which avoids the dlc inflation problem. That makes more sense to me, but perhaps there's an issue within Astats that prevents it whether it be technical or just workload related.

[Twistor]

About the >336 h game time:

I have this currently because I run Clicker Heroes and Idling to Rule the Gods basically 24/7 (today being an exception to the rule).

I run them on the same machine, and I keep running them while playing other games (lately e.g. both Killing Floor 2 and Logistical at the same time).

Steam counts each game separately, therefore 336 is easily exceeded by such measures. Besides, why should it be forbidden / discouraged to play multiple games at the same time? Actually, Steam already prevents me from running games simultaneously on multiple machines (logging the other sign-in off when attempted).

[Azatoth]

There is 1 game, and as far as I know, only 1 game ever on steam where the DLC has a different appid than the main game. This means that it requires someone to manually tag each and every achievement that requires DLC.

When I was first made a mod on Astats 4 or 5 years ago, I spent 3 days marking 95% of DLC achievements. The problem is that not all achievements added with DLC actually require you to own that particular DLC. Then there is the problem of working out which achievements even require any DLC to start with. There often is no way to tell unless someone plays it to find out.

That was back then and if you havent been around that long, you probably dont realise how big steam has grown since. Check out this site
http://kotaku.com/nearly-40-of-all-steam-games-were-released-in-2016-1789535450

Long story short, it is not feasible or accurate to seperate DLC achievements because Steam has no rules and too many achievements are added with no-one to verify if they are DLC or not.

[AlexTheRussian]

Thanks for the information. Honestly, aside from this two issues, the astats formula works and everyone agrees that manual input is not a solution.

However, after thinking about it a bit, I'd like to spitball some ideas, maybe collectively there is something we can do?

One thought I have I perhaps using the weight of this group to draw Steam attention to the issues and get them to change their approach, GUI and/or API.
If someone with more influence than me creates a coherent post on steam suggestions and rallies the group to go to that post and put their "bump" in, we might get some love?

Additionally, and again, just brainstorming, how about this:

Write a simple algorithm that parses when achievements added to the game (if API allows).

If not, There is obviously a database behind astats site, so once a new game is added to Steam the exact achievements are known and anything added after gets point penalty? Don't know how to do it retroactively but smarter people than me can chime in.

Alternatively, pressure Steam to add quantified to achievements stating DLC?

With regards to F2P games, it shouldn't be hard to tag them in astats as they already get steam tag. Once tagged, a differently designed formula applies to them, bringing them in line with pay to play games.

Now, I know this is not going to be a popular suggestion for a lot of people, typically people who got to the top on the system that exists won't like the status quo change, but I hope most will at least participate in conversation and see it as fair balance change.

In the end, everyone agrees that fair system is healthier for the community in the long run.

Thoughts?

[Ossi]

Originally posted by AlexTheRussian:

Additionally, and again, just brainstorming, how about this:

Write a simple algorithm that parses when achievements added to the game (if API allows).

If not, There is obviously a database behind astats site, so once a new game is added to Steam the exact achievements are known and anything added after gets point penalty? Don't know how to do it retroactively but smarter people than me can chime in.

Why? I mean ok it`s your opinion that this should happen, but why should Rocket League (f.e.) DLC Achievements have LESS value, people paid for it, people played the game to get those. Most extreme Example would be Payday2, i remember when the game had 88 (or was it89?) achievements, maybe it had even less before tht but when i started playing PD2 it had around 90 achievements. Yes it sucks that every not and then it adds a few Achievements, more often than not tied to a 5/6/7 dollar DLC but why in the hell should an Achievement, obtained by 500people out of god know who many played this game get a point penalty?
They even add new Achievements or free DLC with Achievements to the game. Have fun editing and figuring out Which Achievements came out when and which of those don`t require to buy the DLC (which is even more complicated in PD2 since you don`t need to buy Heist DLC to play the heists... but you won`t get the Weapons/Mods contained in the Heist DLCs).
While an algorithm might be able to sort out when the Achievement was added, finding out if it "needs the DLC to be played" would be alot of manual work.

Originally posted by AlexTheRussian:

Alternatively, pressure Steam to add quantified to achievements stating DLC?

Won`t happen.
Even if you rally lets say a few hundret people to "pressure steam" You think they put labor into editing every single Achievement, or you think they would or could pressure every single Dev to edit their Games? It could maybe just maybe applied to games or DLC that come out afterwards but thats about it.

Originally posted by AlexTheRussian:

Now, I know this is not going to be a popular suggestion for a lot of people, typically people who got to the top on the system that exists won't like the status quo change, but I hope most will at least participate in conversation and see it as fair balance change.

In the end, everyone agrees that fair system is healthier for the community in the long run.

Thoughts?

it would not change a thing... Well maybe some one "jumps" from #99 astats to #95 or drops from #90 to #92 but it would not change much on the topspots. The last Change on astats, adding the 0.1 value (which i really, really liked because zup3 is shit just because it got 1700 achievements and there are a few games similar to this and it is completely against the point of hunting achievement...) but some people who got most of their achievements from Achievementshovelware dropped a few hundret positions and it was very funny to see them complain "why is my 1700 achievements in 30minutes game not worth 2500 astatspoints anymore" DELICIOUS!

Astats might be changing their point system in the future,tweaking it here and there but i don`t see why DLC Achievements should be handled in such a complicated manner. While i like the idea of getting a "100%" for each payday2 DLC or Zaccaria table (which is all seperate tables and would be handeled as single DLCs) i don`t think that this system would be benificial.

[Azatoth]

Originally posted by AlexTheRussian:

If not, There is obviously a database behind astats site, so once a new game is added to Steam the exact achievements are known and anything added after gets point penalty? Don't know how to do it retroactively but smarter people than me can chime in.

As I already stated, not all new achievements are DLC and not all new DLC achievements require the DLC to earn. Some DLC is even free so is more like an update, sometimes added automatically, sometimes needing you to select it. An example would be Payday 2 where most paid DLC is not required except by the host, and the others just gets added with a patch. Astats tags achievements that require purchasing DLC, not achievements released with DLC. The aim being to inform people which, if any, DLC they would need to buy if they plan to 100% a game.

Also consider that some games have day 1 DLC which means right from the start, or even before release, there can already be DLC achievements.

[AlexTheRussian]

Thank you for your input, I am glad we are talking about it.

I want to clear up some misconceptions first.

We are talking about the fact that achievements added later (after game release, new paid or free dlc, etc) are given much higher point value than normal achievements. This is a fact, not an opinion.

The reason this happens is because when achievements added later, vast majority of players already abandoned the game and will not come back to it. So while the number of people getting first achievement is very large, the number of people who get dlc achievement is tiny.

The result is that an extremely trivial achievement in dlc is more rare than absolute hardest achievement in the base game. Example: Rise of Tomb raider, my most rarest achievements in it are from dlc that took less than 2 hours to get and literally no effort.

So, it's not about arbitrary punishing Dlc buying crowd, it's about bringing dlc achievement inline with normal achievement. Otherwise at this point, it's worth buying a game AND only playing DLC, you will get more points in 2 hours than playing the base game for 60 hours.

So to reiterate: it's irrelevant HOW the achievement is added to the game, what's important is WHEN. As long as achievement is added after the main release (make it 3 months or define more precisely) it needs to use separate formula to get points.

Again, none of this is opinion, all of it is facts, we can do the math on any game.

Hope this answers why the change is needed, it has nothing to do with DLC specifically.

To Ossi: Steam put achievements in BECAUSE they know it matters to players. Saying they aren't going to put effort in to support them is short-sighted and defeatist. If they see enough people are interested in the change, it's not a complicated business case to figure out benefit of supporting our community. They put out weekly patches with way more niche changes.

But this is not about forcing someone who doesn't want it. If you don't want the change, don't participate. You are not part of the community that wants improvements, and that's totally ok. Achievement hunting is not for everyone.

I hope I answered your questions.

[Ossi]

Originally posted by AlexTheRussian:

The reason this happens is because when achievements added later, vast majority of players already abandoned the game and will not come back to it. So while the number of people getting first achievement is very large, the number of people who get dlc achievement is tiny.

The result is that an extremely trivial achievement in dlc is more rare than absolute hardest achievement in the base game. Example: Rise of Tomb raider, my most rarest achievements in it are from dlc that took less than 2 hours to get and literally no effort.

It rewards player that come back to games and play it when new things come out. People who don`t bother 100% it won`t come back, why punish the people who actually do want to 100% the games? Balancing these things out is quite impossible, what would you base the pointsystem on?

Originally posted by AlexTheRussian:

But this is not about forcing someone who doesn't want it. If you don't want the change, don't participate. You are not part of the community that wants improvements, and that's totally ok. Achievement hunting is not for everyone.

I hope I answered your questions.

I like improvements, i don`t see what your suggestions would improve, other than punishing people who buy and play DLCs.

While we are at it, we should just put a difficulty rating to every game and every single achievement and make the score to the achievements that way. People have different opinions. I don`t see a point having zup3 giving 12 Achievements for finishing one single level, other people like it, that just difference in opinions. They change astats point system based on that and i very much enjoyed it.

ONE PIECE PIRATE WARRIOR 3 gives 77 astatspoints, while none of the Achievements is hard to get it will take alot of time get it 100% estimated 96hours slash it can be complited in about 2hours and gives 86points while also no having any particular hard Achievements.
Meanwhile games like dustforce only got 1 achievement and it is very hard to get and gives 0.1 point (+10points for 100% on astats) Estimated time 45+ hours, while VNs that take 5minutes to skip through give the same if not more points on astats. There are plenty of hard games most people won`t even touch so astats rewards low numbers of points based on their algorithm.
Is it perfect? no it isn`t, but it`s functional.

I see a point making the base game one 100% and the DLCs seperate 100%. this would automaticly lower the points in the DLCs but it would lead to another imbalance... it also would boost the Nr. of 100% quite abit. Just for me i would probably jump from 143 to 250 or something just because i finished plenty of games that got DLC, beeing it easy or hard to complete these.

If people want to hunt hard achievements, or high value achievements and go for certain games to obtain these, the numbers will eventually drop and balance things abit.

With a group this big you will never have a system that pleases everybody. Even getting 50% of the people happy can be hard.


If i could change one thing on astats i would probably try to boost the numbers for games like duskforce and other hard games to benefit those people that go through all the hard work to get those games done, even if they have just one achievement or very low astatspoints because so few people play those games.