[Suggestion] Fix/Improve incorrect statistics.

[Floriiss]

For the past couple of weeks I've noticed a lot of errors regarding trophy statistics. I'm curious why it takes that long to adjust the statistics (Or even worse, sometimes it doesn't adjust at all!). I'll be demonstrating this issue with a few examples:

 

Badland/Dragon Fin Soup
We all know both of these games have unobtainable platinums. At some point there have been hackers who obtained those platinums. Later on they got removed from the leaderboards. This happened weeks, if not months ago. Yet still when you visit those games the statistics show 2 platinum achievers, when it should show 0 platinum achievers. Even when clicking on the platinum trophies you don't see any achievers on the list, but it displays 2 achievers beside the trophy rarity.

 

 

 

Even though this doesn't affect the rarity of these games, because there are 0 players on the list, there are a lot of other games which have incorrect statistics due to this issue.

 

Cloudberry Kingdom
As we can see in the list there are 17 people who obtained the Shenanigans! trophy. This equals 17/22147 = ... x100 = 0.07675983203%. So basically this trophy should be 0.08%, yet it shows 0.09% with 19 achievers? What causes this problem? Are there 2 people who hid those trophies but still count towards the percentage? Or were there hackers who obtained it but got removed (just like badlands/dragon fin soup) but the statistics haven't adjusted correctly?

 

 

This doesn't only count for Cloudberry Kingdom, but for a lot of games.. I'm curious what causes this issue and if anyone else noticed it.

 

Edited January 24, 2018 by Floriiss

[Darkette]

Pretty sure it’s cheaters that get removed, then the rarity and the numbers take forever to adjust. Hackers always seem to be on CK apparently, just look at the psn rarity in comparison to the one here. Luckily there’s in game leaderboards too. I think the stats adjust quicker when someone hides the game than it does when a cheater gets removed, but I can’t confirm that. 

[Zolkovo]

While we're on this topic, why are these two figures calculated differently?

 

[Stevieboy]

27 minutes ago, Zolkovo said:

While we're on this topic, why are these two figures calculated differently?

 

 

I know for a fact that those 2 numbers are different calculations. The top number is the completion percentage of the trophy points you have earned against the total trophy points of all the games on your profile (excluding the ones at 0%). Whereas the bottom number is the average of all the individual games' completion percentages on your profile. I'm not sure exactly why the numbers are different if you don't have any 0% games on your profile, but maybe someone else would be able to explain that.

Edited January 24, 2018 by Stevieboy

[Sergen]

1 hour ago, Dark said:

Pretty sure it’s cheaters that get removed, then the rarity and the numbers take forever to adjust. Hackers always seem to be on CK apparently, just look at the psn rarity in comparison to the one here. Luckily there’s in game leaderboards too. I think the stats adjust quicker when someone hides the game than it does when a cheater gets removed, but I can’t confirm that. 

 

Unless part of the requirements for Cloudberry Kingdom require you to be online and submit a leaderboard score to unlock trophies, I think using the leaderboard as an indicator of whether the person did the trophy legit shouldn't be a reason to flag them. Some people prefer to play games while not being signed into PSN, I don't play offline games while signed into PSN often because I get messaged while I'm playing that game and I'd rather focus on the game. Also, the 0.09% for Cloudberry Kingdom is probably accurate as it makes it equal to the 0.09% Injustice USA platinum on PS4. Surely a 100% game can't be harder than Injustice, it's the most difficult fighting game plat on PSN and the rarity shows that... It's an Injustice to think it's not a very insane platinum. 

[Zolkovo]

41 minutes ago, Stevieboy said:

I know for a fact that those 2 numbers are different calculations. The top number is the completion percentage of the trophy points you have earned against the total trophy points of all the games on your profile (excluding the ones at 0%). Whereas the bottom number is the average of all the individual games' completion percentages on your profile. I'm not sure exactly why the numbers are different if you don't have any 0% games on your profile, but maybe someone else would be able to explain that.

 

Yeah I have nothing at 0% and my profile is a bad example since the discrepancy is quite small. Every profile I've checked has had a difference of 1-2% which is a lot more distinguishable, but I can't recall if they had any 0% games.

 

Edit: If we take the above poster's profile as an example (@Sergen Apologies for using it without your explicit permission), he has a much larger sample size without any 0%s either and also has a noticeable difference of .33% at the time of posting.

Edited January 24, 2018 by Zolkovo

[Stevieboy]

1 minute ago, Zolkovo said:

 

Yeah I have nothing at 0% and my profile is a bad example since the discrepancy is quite small. Every profile I've checked has had a difference of 1-2% which is a lot more distinguishable, but I can't recall if they had any 0% games.

 

You should check my profile then. My top number is currently at 92.38%, while my bottom number is at 59.56%. This is because I have 217 games on my profile at 0%.

[Zolkovo]

1 minute ago, Stevieboy said:

You should check my profile then. My top number is currently at 92.38%, while my bottom number is at 59.56%. This is because I have 217 games on my profile at 0%.

 

Ah you beat me to my edit. There seems to be another factor in the equation then as profiles without any 0% games also have a difference, albeit small.

[Tankor_smash]

This has happened to Class of Heroes 2G as well.  It says there are 11 platinum achievers, but only lists 9.

[Stevieboy]

6 minutes ago, Zolkovo said:

 

Ah you beat me to my edit. There seems to be another factor in the equation then as profiles without any 0% games also have a difference, albeit small.

 

The only thing I can think of is the top number might include the trophy points for a platinum, while for the bottom number the completion percentage for the games don't include platinum trophies. TBH, I don't know what difference it would really make though.

Edited January 24, 2018 by Stevieboy

[Kristycism]

On 24/01/2018 at 11:43 AM, Zolkovo said:

While we're on this topic, why are these two figures calculated differently?

 

 

On 24/01/2018 at 1:02 PM, Stevieboy said:

 

The only thing I can think of is the top number might include the trophy points for a platinum, while for the bottom number the completion percentage for the games don't include platinum trophies. TBH, I don't know what difference it would really make though.

 

This was bothering me too so I did some calculations of my own because I'm weird like that.

 

What it seems to be, using a very simple example, is that the number on our main profile is the percentage of all your trophy points obtained.

So say you had two games, one with a plat available that had 1,200 points available for the game, and the other with no plat that had 300 points available. That'd be 1,500 points total available for you to earn.

 

So if you'd got 100% on the 300 points game, and 50% (600 points) on the game with the plat, you'd have 900/1,500 on your main profile, or 60%.

 

But the one on your stats page seems to be an actual average percentage completion. So using the same numbers above you'd have one 100% and one 50%, an average of 75%.

 

Does that make sense?

 

I could be wrong of course, since I didn't code this site or have anything to do with the numbers, but from my own calculations using my own page, that's what it seems to be.  So if you have a lot of games that don't have plats that you've got 100% on, it'll make your stats page average number higher than your main profile page completion percentage.

 

--

 

Now my turn for a question.  How on earth is DLC rarity calculated? I can't make heads nor tails of it at all.

 

Taking Witcher 3's GOTY from my profile for example.

 

There's 29,125 owners of the game, right. Now all of those actually have the DLC since it's included in the GOTY edition, but somehow the site calculates the Hearts of Stone DLC as having 10,737 owners.  I assume that's something to do with all users who have a trophy in that list?

But that aside.  Take the first trophy "I'm Not Kissing That".

 

It has 8,824 achievers, but listed rarity of 49.90%.  8,824/10,737 (DLC owners) is 92.18%.  8,824/29,215 (game owners) is 30.30%.  So where does 49.90% come from???

 

8,824 is 40.90% of 21,575 but what is this number? I don't get it???

 

And every DLC trophy rarity for every game is the same? Where do the numbers come from??????

[FawltyPowers]

This is how you calculate the Geometric Mean in Excel (Figures shown are for the Trophy - The Awakening in the DLC for SSF4 in Feb 2017).

 

X = Achievers (4452)

Y = Main Game Owners (247023)

Z = DLC Owners (38894)

 

Formula in Excel is:

 

=SQRT((X*X/(Y*Z))) = 0.04542 (4.54%)

[DEMONICRUBLE18]

You can use my profile to use for the stat's question. I have a lot of unplayed games. 

 

The way I think it works is the stat's in the stat's page looks at all games you own and calculates the average completion rate you get on games. So say I has 1000 games 500 at 100% and 500 at 0% the stat's would show 50%.

 

If that was on the main page it would show 100% because that one only includes games you have at least one trophy in. 

 

Link to my profile:

https://psnprofiles.com/DEMONICRUBLE18

Edited January 26, 2018 by DEMONICRUBLE18
Link to profile.

[Floriiss]

Can we please stay on topic? I'm not asking how calculations are made, but why it takes so long to adjust or doesn't adjust at all.
 

If you want to know how calculations are made, there are plenty of other threads around the forum that explain it in detail.

[Stevieboy]

3 hours ago, Floriiss said:

Can we please stay on topic? I'm not asking how calculations are made, but why it takes so long to adjust or doesn't adjust at all.

 

All I know is some stats are updated at different times. I'm guessing that individual achievers are removed immediately (either from being removed due to flagging, hiding the trophies, or making their profile private), but the total stats are only updated at certain times (such as once a week) to minimise server load. I think only Sly can confirm that though.

[Floriiss]

6 minutes ago, Stevieboy said:

 

All I know is some stats are updated at different times. I'm guessing that individual achievers are removed immediately (either from being removed due to flagging, hiding the trophies, or making their profile private), but the total stats are only updated at certain times (such as once a week) to minimise server load. I think only Sly can confirm that though.

 

I was thinking the exact same thing. It's just weird that certain stats take months to update. Or not update at all..

 

Lets hope it gets fixed eventually..

[Stevieboy]

6 minutes ago, Floriiss said:

 

I was thinking the exact same thing. It's just weird that certain stats take months to update. Or not update at all..

 

Lets hope it gets fixed eventually..

 

Yeah, it shouldn't really take months to update. Unless it's the same 2 people that keep getting removed each time  .

[BlindMango]

The numbers on the stats page are the average of each "tier" shown in the circle graph. The number at the top of your profile is different and counts the average of all points on your profile (minus platinum trophies). We calculated the stat at the very top of your profile in a thread previously, but I forgot which one it was  

[RNumbers]

On 1/24/2018 at 5:01 AM, Tankor_smash said:

This has happened to Class of Heroes 2G as well.  It says there are 11 platinum achievers, but only lists 9.

 

Yeah, I pointed this out months ago in a similar topic. I could have sworn it was fixed at that time, but it's now wrong (again?). I'll do some digging.

 

EDIT: FilmFanatic linked the one I was thinking of below.

Edited January 26, 2018 by RNumbers

[FilmFanatic]

@Floriiss is this the same issue as mentioned in this thread?

 

Some other games with this happening are Bladestorm: Nightmare. 3 hackers have been reported for this game in the last month or so but the stats still show 22 achievers so 1 has been removed from the stats since then but not the 2 most recent ones. Tokyo Twilight Ghost Hunters that I mentioned in the above thread has the same issue again.

[Floriiss]

1 minute ago, FilmFanatic said:

@Floriiss is this the same issue as mentioned in this thread?

 

Some other games with this happening are Bladestorm: Nightmare. 3 hackers have been reported for this game in the last month or so but the stats still show 22 achievers so 1 has been removed from the stats since then but not the 2 most recent ones. Tokyo Twilight Ghost Hunters that I mentioned in the above thread has the same issue again.

Yes, although I'm talking about the whole system in general. Not only platinum trophies are affected by this, but also games with DLC, thus the 100% completed list.

[Zolkovo]

On 26/01/2018 at 6:23 PM, BlindMango said:

The numbers on the stats page are the average of each "tier" shown in the circle graph. The number at the top of your profile is different and counts the average of all points on your profile (minus platinum trophies). We calculated the stat at the very top of your profile in a thread previously, but I forgot which one it was  

 

Have I misunderstood the above? I have all my games in the same tier but the averages still appear different:

 

 

As you can see I have no 0% games, all games in the same tier and no hidden games.  

[FilmFanatic]

53 minutes ago, Zolkovo said:

 

Have I misunderstood the above? I have all my games in the same tier but the averages still appear different:

 

 

As you can see I have no 0% games, all games in the same tier and no hidden games.  

The number at the top of your profile is the completion rate for your entire profile, that is, you have 99.32% of all possible trophy points (not including platinums). The number in the graph is the average completion rate per game on your profile, that is, your average completion rate for any single game on your profile is 99.46%.

[Stardroid]

I am hijacking this thread to point out that both of the rarest platinums that you can achieve legitimately (I believe) are incorrect and I believe this has been for a long time too. Necomancer has five people legitimately 100% it. Yet the site tracks 10 legitimate achievers, that's double for what is considered the toughest platinum trophy out there. I don't know how many cheaters get removed on that game on a week to week basis, but I think this has not been updated in quite a while. 

Next game is Badland. I have reported quite a few people by checking that game once in a while. It's an easy way to filter out VITA hackers and remove them from the leaderboard. There are twelve legitimate achievers, the site tracks 22 completitionist. Even weirder is that is tracks 23 platinum achievers, which is the full 100%. Can we actually get a concrete answer here why this lags so far behind? When does it update if at all? I checked games mentioned int his topic and they are still not correct either. Some clarification please. The two rarest platinums out there legitimately shouldn't track double the amount of people having achieved it.

 Thanks for your time@Sly Ripper

Edited March 17, 2020 by Startropy X

[Floriiss]

4 hours ago, Startropy X said:

I am hijacking this thread to point out that both of the rarest platinums that you can achieve legitimately (I believe) are incorrect and I believe this has been for a long time too. Necomancer has five people legitimately 100% it. Yet the site tracks 10 legitimate achievers, that's double for what is considered the toughest platinum trophy out there. I don't know how many cheaters get removed on that game on a week to week basis, but I think this has not been updated in quite a while. 

Next game is Badland. I have reported quite a few people by checking that game once in a while. It's an easy way to filter out VITA hackers and remove them from the leaderboard. There are twelve legitimate achievers, the site tracks 22 completitionist. Even weirder is that is tracks 23 platinum achievers, which is the full 100%. Can we actually get a concrete answer here why this lags so far behind? When does it update if at all? I checked games mentioned int his topic and they are still not correct either. Some clarification please. The two rarest platinums out there legitimately shouldn't track double the amount of people having achieved it.

 Thanks for your time@Sly Ripper

As much as I appreciate your concern, I don't think this will get fixed soon enough, if at all. This thread was made two years ago and up until this point nothing has changed.