Global achievements stats (% of players who unlocked)

[Vic]

ChatGPT says it's calculated based on the total number of owners.

I asked:
wouldn't be more accurate to use a player base considering only who have opened the game at least once or played a minimum of hours?

It replied:
Yes, using a player base that considers only those who have opened the game at least once or played a minimum number of hours can provide a more accurate representation of active players. This approach can help filter out players who may have purchased the game but haven't actually engaged with it, thus giving a better indication of the achievement's rarity among the active player community.

Thoughts? Anyone knows if other platforms do the same math?

[Aachen]

AFAIR the statistics include only those who have earned at least one achievement in the relevant game.

ChatGPT, fwiw, is not an authority.

[zirkustier]

Originally posted by Aachen:

AFAIR the statistics include only those who have earned at least one achievement in the relevant game.

Are you sure? In a game I'm playing right now, 50.8% have the “Begin the game” achievement.

[BJWyler]

Originally posted by Vic:

ChatGPT says

Thoughts? Anyone knows if other platforms do the same math?

AI nonsense like ChatGPT is simply a novelty and mostly a complete waste of time.

Not engaging with a game is a form of engagement.

[Gwarsbane]

Originally posted by Vic:

ChatGPT says it's calculated based on the total number of owners.

ChatGPT and other AIs are well known to also "hallucinate" as in, it tells you the wrong information. Recently the one from google said an 1/8 of a cup of Elmers glue should be used if you are having problems with your cheese sticking to your pizza.... So don't believe anything that an "AI" tells you...

[Aachen]

Originally posted by zirkustier:

Are you sure? In a game I'm playing right now, 50.8% have the “Begin the game” achievement.

Nope, I am not.

[Sleepy Yoshi]

Seems to me Chatgpt is correct in it's conclusion at least, but I'll profess I don't know for certain how it works. If it required getting one achievement though, there are linear games with a set progression. You'd expect to see in these games have the achievement % being at or near 100%.

But that's not the behavior you see on Steam. Then you take for example games with a single achievement. How would these games have anything but 100%? But we see that's not the case either. For example Dustforce has a single achievement, but it only has a .4% unlock rate.

ACG on the other hand for the user is based on having a single achievement and that can be demonstrated more easily. I think from this people assumed global achievement percent worked the same way. So I think that at least can be ruled out.

I'm not sure how you'd really figure out whether it's owner or requires play time though.

[cinedine]

Originally posted by zirkustier:

Originally posted by Aachen:

AFAIR the statistics include only those who have earned at least one achievement in the relevant game.

Are you sure? In a game I'm playing right now, 50.8% have the “Begin the game” achievement.

It's already the way you want it, only players with playtime count against it. There is no way in hell 99 % of all *owners* even started the game.
The 50 % is simply players who idled the game for cards, started the application without actually playing or where the cheevo bugged out.

[nullable]

Originally posted by Vic:

ChatGPT says it's calculated based on the total number of owners.

I asked:
wouldn't be more accurate to use a player base considering only who have opened the game at least once or played a minimum of hours?

How accurate does the statistic need to be for a mostly worthless, little bit of fun? Anywhere in the ballpark is good enough.

Originally posted by Vic:

It replied:
Yes, using a player base that considers only those who have opened the game at least once or played a minimum number of hours can provide a more accurate representation of active players. This approach can help filter out players who may have purchased the game but haven't actually engaged with it, thus giving a better indication of the achievement's rarity among the active player community.

Thoughts? Anyone knows if other platforms do the same math?

When ChatGPT owns and operates Valve, then its "opinions" will matter. Using ChatGPT to bolster your confirmation bias isn't going to cause Valve to change their systems, or convince other users the subject matters in any way, shape or form.

[IFIYGD]

Originally posted by cinedine:

It's already the way you want it, only players with playtime count against it. There is no way in hell 99 % of all *owners* even started the game.
The 50 % is simply players who idled the game for cards, started the application without actually playing or where the cheevo bugged out.

Hahaha!
Guess what? Some games (at least one of them) unlock an achievement when you buy and install the game.
Secrets of Magic 5: Back to School.
As soon as you install the game, before ever launching it- you get the "Gifted Witch- Master " achievement.
Did a double-take when it happened, checked with other people who also have the game, and yep. Happened for them too. It shows 99.1% have unlocked it. Bugged? Misconfigured? Intentional by the devs? No clue- it's the first time I've ever seen it happen, though the devs have a known history of having bugged achievements in their games, so...

OP- Ask ChatGPT how that worked- I'd love to see what the AI makes up to explain it

[Ben Lubar]

The denominator for the fraction is the number of players who have opened the game.

In the Steam Sales & Activation Reports site, this is called "Lifetime unique users"

Source: the number given in https://arstechnica.com/gaming/2018/07/steam-data-leak-reveals-precise-player-count-for-thousands-of-games/ for my game (approximately) matches that number for the day the leak occurred.

[KalGimpa]

Originally posted by Vic:

ChatGPT says it's calculated based on the total number of owners.

I asked:
wouldn't be more accurate to use a player base considering only who have opened the game at least once or played a minimum of hours?

It replied:
Yes, using a player base that considers only those who have opened the game at least once or played a minimum number of hours can provide a more accurate representation of active players. This approach can help filter out players who may have purchased the game but haven't actually engaged with it, thus giving a better indication of the achievement's rarity among the active player community.

Thoughts? Anyone knows if other platforms do the same math?

that is a leading question

you should have asked if there was a better way to calculate it instead

[Dan5000]

Its not about getting the first achievement to be counted. It counts everyone to the achievement % that has only opened the game once. So you go into the main menu and close the game and you count towards the total already.

Also if a game didn't own achievements in the past and got them later on, it'll count everyone who played in the past aswell, even though they never opened the game again after achievements dropped.

[rawWwRrr]

Originally posted by Vic:

ChatGPT says

ChatGPT is not Steam. Stop depending on it to give you accurate answers.