[XB] Year Recap

[XB] Year Recap [Paid] 1.1.3 Patch Level 2

No permission to buy ($25.00)
Overview Updates (7) History Discussion
Not a biggie - but consider next time you're doing changes to the addon to make it a permission to see past years statistics.

I think it would be cool to see all the past year statistics.... but probably only want my users to see the current past year to keep it simple.
 
smallwheels said:
Isn't this a normal statistical effect as possibly all threadviews since creation up until today are calculated? So a thread created back in 2022 will often have many more views than one created in 2025 simply because it had a couple more years to gain views?
If that's the case it makes a bit of a unfortunate picture as thread views for the current year will be always way lower than for the year before (at least if you have a lot of long-running threads or a lot of older threads that are still viewed a year or more later).

However, if I compare 2024 and 2025 in my forum thread views are dramtically lower as well - but the numbers of the monthly view aggregated for 2024 are equal to the total. So are these really monthly views for 2024 or the views of the threads in 2024 and 2025 added per month?
View attachment 331739

View attachment 331740


View attachment 331741
View attachment 331742

I am wondering b/c I did not see such a dramatic loss in visits, rather the opposite: My forum grew over the years.

So, depending from the calculation maybe the wording should be adjusted to something like "total views of the threads created in $year until today" and possibly the monthly numbers split into years if possible.
Click to expand...

You are right - the current behavior is not intuitive and retroactive rebuilds leads to distorted and confusing results.

This logic will be changed in the next version to better match real yearly activity and expectations.


smallwheels said:
Indeed thread views seem to be way off. My forum was created in Sept 2022, so only three month in 2022 with very limited users. Statistic says:

• 200 new Users (correct)
• 16 active users (which is clearly incorrect by miles)
• 233 new threads (correct)
• 741.000 thread views (which seems way too high)

in comparison, for 2025 it shows:
• 1000 new registrations (correct)
• 1843 active users (possible)
• 1000 new threads
• 545.000 thread views (which seems way too low - in comparison to 2022 as well as to 2024 where the stat shows 1.100.000 thread views)

Would it possibly be easier to count thread views for all threads viewed in a given year (instead of views of threads created within this year)? This seems possibly a more relevant number, easier to understand for the audience and hopefully a more correct number as well.
Click to expand...
smallwheels said:
Active users seems indeed also a bit off for the current year: According to the stats I would have had the most active users by far in December which does not at all correlate with the usual pattern of my forum (winter is calm in comparison to spring, autumn and summer) and also not to the posted messages stats and contradicts the stats available in the acp:

View attachment 331760


View attachment 331763
View attachment 331762

Does not fit - something seems to be off.

Same goes for the number of posts and for the number of post reactions (though the difference to the numbers in ACP is smaller):

View attachment 331764
View attachment 331765

View attachment 331766

The question remains which statistic is correct and which is wrong (or if they simply show different things for whatever reason).
Click to expand...
All discrepancies come down to how, when, and what data is counted and displayed.

Registered / active users:
ACP statistics include invalid users (e.g. unconfirmed, rejected, etc.), while the recap add-on counts only valid users, matching the logic used in the forum’s public statistics.

Reactions:
The reaction counter you checked in ACP applies only to posts.
In my add-on, reactions are counted across all content types where the reaction is marked as counted.

So the numbers are not necessarily "wrong" — they are simply based on different scopes and counting rules.
 
Do I need to run a rebuild to get the "Better than X% of users" to update, or do users need to log in first?

Mine is saying I am better than 100% of users, but when you click to view the leader board I'm nowhere near (for example my 19 threads, whereas the top 5 isn't lower than 400!)
 
Feature request:

Year recap previous value difference display formatting

Add percentage as an option (show the percentage difference of the previous year rather than the actual amount)
 
benFF said:
Do I need to run a rebuild to get the "Better than X% of users" to update, or do users need to log in first?

Mine is saying I am better than 100% of users, but when you click to view the leader board I'm nowhere near (for example my 19 threads, whereas the top 5 isn't lower than 400!)
Click to expand...
I see the same in various rankings of the addon for myself. Probably a question of how you display the value. It your forum has let's say 20.000 members and you are regarding the number of threads placed at rank 25 this would make you better than 99,875% of all members. Rounded this would be better than 100% - mathematically correct. Just that this is a bit of a misleading message: With "better than 100%" one would clearly expect to be on the very top of the list at rank 1.

Suggested solution for @Painbaker : Display 100% only for rank one, else, if a rank is not top of the list but better than 99% display 99%.
 
smallwheels said:
I see the same in various rankings of the addon for myself. Probably a question of how you display the value. It your forum has let's say 20.000 members and you are regarding the number of threads placed at rank 25 this would make you better than 99,875% of all members. Rounded this would be better than 100% - mathematically correct. Just that this is a bit of a misleading message: With "better than 100%" one would clearly expect to be on the very top of the list at rank 1.

Suggested solution for @Painbaker : Display 100% only for rank one, else, if a rank is not top of the list but better than 99% display 99%.
Click to expand...
Yes things can get tricky here - because take posts for example, I wouldn't want to compare against all the years active users, only the users who actually made a post. It doesn't make sense to compare with the 20,000 who didn't post anything.
 
benFF said:
Do I need to run a rebuild to get the "Better than X% of users" to update, or do users need to log in first?

Mine is saying I am better than 100% of users, but when you click to view the leader board I'm nowhere near (for example my 19 threads, whereas the top 5 isn't lower than 400!)
Click to expand...
Currently, the add-on uses a high rounding threshold for value distribution statistics to avoid performance issues on large forums and to prevent generating excessive database records.

I will add an option to adjust this threshold. Using lower rounding values will result in more precise data with less deviation.


cdub said:
Can you explain what these two options do and how they are different?

View attachment 331839
Click to expand...
The “Rebuild community statistics” option also updates the overall statistics for the forum, not just the personal user statistics.
Retroactive statistics are user activity and thread views - it is best not to touch it for older years in order to preserve more accurate information.
 
Painbaker updated [XB] Year Recap with a new update entry:

1.1.3 Patch Level 1

1.1.3 Patch Level 1:​

  • Added option to change the value distribution rounding step used for percentile calculation. Larger values reduce precision but improve performance. Lower value = more precise results.
  • Added percentage display format option for previous year value difference
  • Fix: show 100% percentile only for top rank users

1.1.3:​

  • Changed counting logic for “Active users” and “Thread views” statistics
  • Changed...
Click to expand...

Read the rest of this update entry...
 
Thanks for the update, but there is a new bug - "Mostly in month" now just shows the yearly total (but if you click the chart, it shows the correct monthly breakdown)

PHP: 
            if ($valueFormatting == 'number')
            {
                $value = $this->app->templater()->filter($userRecap->value, [['number', '']]);
            }
            else if ($valueFormatting == 'number_short')
            {
                $value = $this->app->templater()->filter($userRecap->value, [['number_short', '']]);
            }

Should use $value instead of $userRecap->value

:)
 
Last edited:
Bug in the frontend: Missing Phrase in "active users": shows "pb_yearrecap_mostly_in" instead of the phrase.
Bildschirmfoto 2025-12-31 um 10.44.15.webp

Fixed itself after updating the caches for each year w/o checking the second option there:

Bildschirm­foto 2025-12-31 um 10.59.43.webp
The updating w/o the second option was necessary to show the newly calculated threads view counter.
Monthly distribution of the tread views is however missing (was there before), however, this is not a dealbraker.

This:

  • Fix: show 100% percentile only for top rank users
Click to expand...
does not hold true for me. Even if I am the top rank user it still shows "better than 99%" - but I can live with that, it is way better than the other way round as it was before).

Active users for former years is fixed as well: I started my forum only in 2022 and the numbers of registrations and active users did not fit:
Bildschirm­foto 2025-12-31 um 11.01.42.webp
After the cache rebuild they do now:

Bildschirm­foto 2025-12-31 um 11.17.59.webp


Thanks for the update and for the speed of it!
 
Last edited:
Recalculation for the new year 2026 seems partly delayed: 10,5 hours into the new year and sadly there's no way my forum had 4.000.000 thread views over these couple of hours. :p (in fact this number is - in opposite to the others - from 2025.

Bildschirm­foto 2026-01-01 um 10.34.25.webp
 
You must log in or register to reply here.
Back
Top Bottom