The Personal Talossan Activity Index

When Tamoran Dal Nava proposed to create a Talossan Activity Index, I had risen to the occasion and built and automatic calculator, as explained in last month’s issue of Preßeu Zespenat.

The index was wonderful to pin-point the differences in activity between the various days or weeks and more importantly, showed the relative quality of threads.

I remembers days (or rather nights) where Ron Rosalez or Andreas Lorentz would roam the empty halls of Wittenberg and post threads after thread with their random musings which would too often be forgotten the next day but would sync the WIx for several days.

And so, I wondered, is there a way to show why the TAI or WIx fluctuated? Is there a way to see who contributed the most?

Of course, being a man, the true question was obvious: it there a way to create an index to show who the strongest poster was and if it’s the case, is it me?

Don’t get me wrong, I did a lot of effort to create a neutral objective index and if I was so often ranked #1 was only because of my level of implication in the Kingdom (imagine the boosts to my number from my various SoS posts!) and my boredom at work…

Raw Factors

So, what are the factors?

Raw Values:
P: Posts by the user
R: Replies to posts by the user
T: New threads by the user
TF: Thread force: Number of replies to threads started by the user

Please note that Wittenberg back then allowed replied to specific posts and didn’t have a single thread view like today’s Wittenberg. The R factor would be much harder to calculated, if not impossible…

Perhaps Quotes and Likes could be counted, or many all posts in a thread with a higher number than the poster.  The problem is that in my calculations, the R value is truly pivotal, as we will be able to see further below, but it would be really hard to calculate a valid R using the current way forums are built.

Intermediary values

There are 3 intermediary values, two of which rely on R:

GA: Generated Activity: P + R
DGL: Distributed Generated Load: R / (GA+1) - between 0 and 1
TCF: Thread Creation Factor: TF/T+1

The Generated Activity simply counts the number of posts made by the user or as reaction to the user.  In short, it is the volume of talk generated by the fact that the user posted.

The DGL however, my favorite intermediary values of both the TAI and the PTAI, is a ratio calculating in short the percentage of that activity which is from other users, on the basis that multiple replies to your posts show a strong correlation with interest (even if your post simply caused an outrage) whereas multiple posts without replies show a lack of interest by the others.

The TCF on the other hand, calculated how strong your threads were. If you were lucky and one of your threads exploded, the TCF would be really high. If you posts at night a few random threads of no interest, it dropped down.

Interestingly, you could trick the system by replying to your own threads, which would raise your TCF, but lower your DGL…

Please note that both divisors have 1 added both to avoid dividing by zero and because I saw often that many indexes are calculated by adding 1 to the divisor (perhaps to avoid a division by zero).

The PAI

Interestingly, the main index of the PTAI calculations isn’t called the PTAI but rather just the PAI, the Personal Activity Index. The reason is simple and very logical, but to be honest, I forgot it. No, don’t insist…

So, the PAI calculations:

PAI: Personal Activity Index: (GA * DGL) + (TCF*DGL)

In short, the PAI was the sum of both the Generated Activity and the Thread Creation Factor, each weighted according to the Distributed Generated Load.

Retroactivity

The PTAI calculations, unlike the TAI calculated, surprised me in that they were constantly shifting retroactively as new posts were made in older threads.

This meant that a thread posted 3 days ago and which still generated activity today would still contribute to the TCF factor from 3 days ago.

As a result, the whole PTAI table shifted, moved, and no number was definitive as people often posted replies to really old threads.

At first, I tried to find solutions, but it dawned on me. It wasn’t a bug or a problem. It’s what made the PTAI beautiful. It’s chaotic nature, it’s never ending pulsations, it’s non-finality.

As new replies piled on, old indexes vibrated and soared, old mistakes could come back alive and bring success, but at the same time, those old successes contributed to the indexes of the past, not of the present. What have you done today is what counts for today, but all of the future comments on today’s work will retroactively count for today.

Almost poetic, at least, if you love Math.