BigHatPaul
New Member
Agreed on XP boosters. Any spam game really should be completely worthless. This new formula looks much better.
I have been experimenting with making some changes to the way EXP is calculated, to better deal with games that have a large amount of achievements. None of this is live yet, just have ran some tests to see how it would impact the totals.
This method is actually borrowed from SteamHunters, so thanks to them for coming up with this. Rather than counting the amount of EXP individually per achievement, the total EXP for each game is calculated from the rarest achievement using its unlock percentage - and then that value is distributed accordingly across all of the game's achievements (the rarest achievement will get the most points)
Spider-Man on PS4 as an example -
The rarest trophy has been earned by 8.96% of players on the site. We use the calculation below to get a total EXP value of 2,391.
(sqrt(100 / UNLOCK_PERCENTAGE) - 1) * 1000 + 50
To break it down, 100 / UNLOCK_PERCENTAGE calculates the worth of the rarest achievement, then the square root of that number is taken to avoid getting inflated values. 1 is subtracted from the result here to avoid very easy games from being worth 1000+ EXP. The final result is multiplied by 1000 + the base EXP value, which is 50.
As another example, Viki Spotter: The Farm on Steam - this game has a large amount of achievements (298) and all can be unlocked very quickly (under an hour).
With the new calculation, it only gets 68 EXP, versus the 2,980 it has now. Given that it can be completed so quickly this more accurately represents the worth of these achievements.
Third example, LOGistICAL on Steam - this game has a very large amount of achievements (9821).
With the new calculation, it only gets 2937 EXP, versus the 252,215 (!) it has now.
So I think this handles with the games that have a large amount of achievements much better, while not taking away from the potential EXP in games that have a more typical list. I think we won't have to devalue specific games with this approach either, aside from some isolated cases - I think the XP boosters on the Google Play should still be completely removed, since the base value of 50 EXP is still too much for those.
Any thoughts on this?
I really like the idea. However, I have a question.
How do you deal with games that have one or more 0% archievements. In this case you would end up with infinite amount of exp. Do you use the rarest archievements that does not have 0%?
EDIT: just realized that this should never happen, as it's based on the rarest archievement a player has earned, right? So UNLOCK_PERCENTAGE is different for "every" player?
Does this mean that games with over 1,000 achievements will now be counted? Either way, this sounds like a nice change and will definitely help keep the leaderboards more competitive.
If a game only has 1 achievement and it has 1% players have it, would the score be ~100,000? Or am I doing something wrong in the calculation?
I think this is a great idea...as for those clicker games i have played them and gotten many achievements from them (shane on me) so yes i dont think thise games should give xp but i think the achievement count should still be represented on the profilesI have been experimenting with making some changes to the way EXP is calculated, to better deal with games that have a large amount of achievements. None of this is live yet, just have ran some tests to see how it would impact the totals.
This method is actually borrowed from SteamHunters, so thanks to them for coming up with this. Rather than counting the amount of EXP individually per achievement, the total EXP for each game is calculated from the rarest achievement using its unlock percentage - and then that value is distributed accordingly across all of the game's achievements (the rarest achievement will get the most points)
Spider-Man on PS4 as an example -
The rarest trophy has been earned by 8.96% of players on the site. We use the calculation below to get a total EXP value of 2,391.
(sqrt(100 / UNLOCK_PERCENTAGE) - 1) * 1000 + 50
To break it down, 100 / UNLOCK_PERCENTAGE calculates the worth of the rarest achievement, then the square root of that number is taken to avoid getting inflated values. 1 is subtracted from the result here to avoid very easy games from being worth 1000+ EXP. The final result is multiplied by 1000 + the base EXP value, which is 50.
As another example, Viki Spotter: The Farm on Steam - this game has a large amount of achievements (298) and all can be unlocked very quickly (under an hour).
With the new calculation, it only gets 68 EXP, versus the 2,980 it has now. Given that it can be completed so quickly this more accurately represents the worth of these achievements.
Third example, LOGistICAL on Steam - this game has a very large amount of achievements (9821).
With the new calculation, it only gets 2937 EXP, versus the 252,215 (!) it has now.
So I think this handles with the games that have a large amount of achievements much better, while not taking away from the potential EXP in games that have a more typical list. I think we won't have to devalue specific games with this approach either, aside from some isolated cases - I think the XP boosters on the Google Play should still be completely removed, since the base value of 50 EXP is still too much for those.
Any thoughts on this?
I think this is a great idea...as for those clicker games i have played them and gotten many achievements from them (shane on me) so yes i dont think thise games should give xp but i think the achievement count should still be represented on the profiles
In addition to this point system change, what about adding something that allows for achievements/trophies to be flagged as unachievable or broken? Should these have no points allocated to them or should it be left as is?
I like the new point system except for one thing. If only one person plays the game he gets very few points. I think you need to give 10 points for an achievement as before. Is there any game in Steam that no one plays with easily obtained achievements?The point system changes I mentioned in this post have been pushed live today.
The EXP totals haven't changed yet, they will update tomorrow, so we'll see how it affects the leaderboards then. Most of the Steam games with a lot of achievements are worth significantly less under the new algorithm.
Civ 6 has essentially become worthless. I like the general idea of the new points system, but still feels a bit unbalanced.
View attachment 10537
I got 10 points.Before if only one person played the game, you got zero points.
In general, I think the new system sounds fairer.
However, some games seem to have taken a real hit, like Star Craft 2. This game has like 2500+ achievements apparently, but the vast majority of those will be "feats of strength" so ones that dont give points in the game, and often cant be earned after a certain point.
My guess is that those will get the lion's share of available score.
I like that on the PSN games you can now see both the % of players that have the achievement on this site, and on PSN.
People that link to a site like this are always going to be trophy hunters, so it's nice to see the difference.
Wonder if this will be possible for steam going forward?