LowEndSpirit got some traction: numbers and figures

alwyzon

When I first stumbled over LowEndSpirit through some signature below a comment on LET, I was curious whether there is actually some activity on this new platform or whether its just a nice idea which isn't well accepted by the community. Yeah, there were some posts on the front page, but I prefer hard figures over gut feelings. So I did my homework and wanted to share the (surprising) results with you folks. I hope this is as interesting to you as it was to me and that you don't mind me sharing the result.

Here it is:

There are still less new discussions on LowEndSpirit than on its older brother, but the difference is much smaller than I thought. Seems like we got quite some discussions going on here too, great job @AnthonySmith! Definitely worth keeping an eye on LowEndSpirit.

EDIT: Just replaced the figure with one counting discussions as logged in user (see discussion below). The original chart that just counted disucssions visible to non-logged in users can still be found here.

InceptionHosting

@alwyzon said: There are still less new discussions on LowEndSpirit than on its older brother, but the difference is much smaller than I thought. Seems like we got quite some discussions going on here too, great job @AnthonySmith! Definitely worth keeping an eye on LowEndSpirit.

There is growth month on month in terms of unique views, number of comments and new users as well as new hosts, the google organic hits are ale growing substantially and now account for around 30% of traffic.

I think your figures are a little bit out, Jan 2020 for example:

That said if you applied the same method to both sites it probably evens itself out, I suspect whatever is doing the counting is not a logged-in user and perhaps you cant see that rants/Les exclusives and other topics that are not indexed.

LES is now just gone 9 months old, I would say it has a more diverse offer range with a slightly amended ruleset to allow exclusives.

Anyway, I appreciate the feedback and the post very much, I hope to have LES at least at the level of LET within 12 - 18 months but with better quality discussions as many of the posts on LET from memory are "I want VPS with 20 IP's, budget $2"

As a general and personal observation, the maturity level is higher, people conduct themselves better, the willingness to help others is higher here and the shitposting is WAY lower.

That is a deliberate direction/decision by Jon though, I probably still have the email from him from when i was admin there whereby he actively wants/likes the shitposts and drama to be unmoderated because its great for view numbers etc. (paraphrased)

Would be interesting to do a wider comparison and include WHT, VPSB and Hostedtalk, the last 2 being offshoots/forks of LET and the former being the thing that forced LET to exist because if its refusal to evolve.

I am a little bit concerned here that the cesspit is being used a lot for things that really should be new posts in their own right (guilty myself) although I have not decided how to deal with that yet.

Examples:

https://talk.lowendspirit.com/discussion/comment/35885/#Comment_35885

https://talk.lowendspirit.com/discussion/comment/35835/#Comment_35835

https://talk.lowendspirit.com/discussion/comment/35559/#Comment_35559

https://talk.lowendspirit.com/discussion/comment/35647/#Comment_35647

And some big discussion about think pads I can't pick a link for.

suggestions are welcome on how to bring those sorts of posts back to the main forum, I have an interim idea to possibly change mindsets immediately.

I do enjoy "inside baseball" so these discussions are always very welcome.

cybertech

Now getting used to the modern vanilla interface, not to mention its much faster.

with real exclusive LES offers its really enticing to remain active here.

Not to mention some fun discussions and OCD bench threads that providers themselves have taken interest to as well.

alwyzon

@AnthonySmith said:
I think your figures are a little bit out, Jan 2020 for example:

Thanks for sharing the numbers as reported by Vanilla. I guess the difference might arise from me just counting new discussions, whereas Vanilla seems to be counting active discussions. Sadly, we can't compare these Vanilla numbers without knowing the same ones from LET.

As a big proponent of transparency and reproducibility, here is the source code to reproduce the above figures: https://gist.github.com/hohl/532b7edb85f09a7758172865aaf7dc94

InceptionHosting

@alwyzon said: whereas Vanilla seems to be counting active discussions.

Ah very interesting, I had not realised it was active rather than new that vanilla reports as every other metric is for new (whatever item), where did you find that out?

Does the script run as a logged-in user? as if not it won't see at least 3 of the categories that people actively post in whereas LET only has 1 which is/was only generally used for the cestpit

May 2020 for example in vanilla shows 193 discussions.

@cybertech said: Now getting used to the modern vanilla interface, not to mention its much faster.

I am also working to improve that yet again fairly soon and its still running on a VPS you can buy for $7

alwyzon

@AnthonySmith said:

@alwyzon said: whereas Vanilla seems to be counting active discussions.

Ah very interesting, I had not realised it was active rather than new that vanilla reports as every other metric is for new (whatever item), where did you find that out?

Sorry, you are right - it's new discussions - according to Vaniallas official website, quote: "Number of new discussions that have been created for the time period"

@AnthonySmith said:
Does the script run as a logged-in user? as if not it won't see at least 3 of the categories that people actively post (...)

No, the script just runs as a guest and crawls through the /discussions-page. So, yeah, if there are more hidden forums on LES than on LET that might introduce a small bias in favour of LET.

@AnthonySmith said:
(...) I hope to have LES at least at the level of LET within 12 - 18 months but with better quality discussions as many of the posts on LET from memory are "I want VPS with 20 IP's, budget $2"

Indeed, counting posts isn't the best comparison anyhow - as long as there is no way to count "qualitative" posts only. And yes, I'm aware of the kinds of post you're talking about and they're no loss if they are missing here.

InceptionHosting

@alwyzon said: No, the script just runs as a guest and crawls through the /discussions-page. So, yeah, if there are more hidden forums on LES than on LET that might introduce a small bias in favour of LET.

You are welcome to make a dummy user to use with the script for authenticated counts, they seem to be out by around 35 - 40% otherwise

Ant.

alwyzon

@AnthonySmith said:
You are welcome to make a dummy user to use with the script for authenticated counts, they seem to be out by around 35 - 40% otherwise

I just gave it a try and made the scripts login on LET and LES as my user account before the crawling of the discussions. While there was only a minor increase for the LET numbers, there was an above 30% increase in the number of LES discussions:

(However, the script still only counts 174 discussions for May, not 193 discussions, but we are getting closer and I guess close enough. Maybe there were some discussions my user can't see or got deleted in the meanwhile.)

InceptionHosting

Not logged in:

Logged in:

Posting both together for comparison

good stats, appreciate you taking the time to do that.

Not_Oles

@alwyzon said: As a big proponent of transparency and reproducibility, here is the source code to reproduce the above figures: https://gist.github.com/hohl/532b7edb85f09a7758172865aaf7dc94

Thanks for posting this! I haven't yet done any of this sort of stuff, and it's intriguing to me. I might even try running it myself.

Just to add a joke to what I say above seriously: Maybe everybody could run this a few thousand times to see if we indeed do get reproducible results! $7 Would Cloudflare and @AnthonySmith even notice?

InceptionHosting

@Not_Oles said:

@alwyzon said: As a big proponent of transparency and reproducibility, here is the source code to reproduce the above figures: https://gist.github.com/hohl/532b7edb85f09a7758172865aaf7dc94

Thanks for posting this! I haven't yet done any of this sort of stuff, and it's intriguing to me. I might even try running it myself.

Just to add a joke to what I say above seriously: Maybe everybody could run this a few thousand times to see if we indeed do get reproducible results! $7 Would Cloudflare and @AnthonySmith even notice?

I am 99% sure you would at least have to do that concurrently for me to notice

imok

@AnthonySmith said: LES is now just gone 9 months old

What? I thought I registered last month.

vyas

We need more drama on LES to increase visitors
(Not really)

mikho

@alwyzon said: (However, the script still only counts 174 discussions for May, not 193 discussions, but we are getting closer and I guess close enough. Maybe there were some discussions my user can't see or got deleted in the meanwhile.)

Most likely as there are discussions among the different groups that are used.
I'm sure there are developer discussions somewhere, and the comments from the "main" site are hidden discussions here on the forum.
To get the most accurate numbers, one would need to be an administrator on each forum to have access and count all topics.

havoc

@Not_Oles said: Maybe everybody could run this a few thousand times to see if we indeed do get reproducible results! $7 Would Cloudflare and @AnthonySmith even notice?

Always wanted to build something like the LOIC.

Doesn't have many legitimate uses though so haven't done so. Maybe one day for load testing....