1) The 105 members in your sample received 10.75 merits each. This shows that there are enough merits for everyone, provided you are ready to make some efforts.
I just took a look at the outliers (should have earlier, I know) and it seems they got merits for dumb reasons (one guy used alt accounts, the other was selling ICO picks for merits). So a better stat would be that 103 legitimate (I assume) members have received 5.2 merits each.
2) Majority has no merits. This is something that needs to be looked after. The promotional offers (bounties and campaigns) should not expect members without any merits.
Suggestion: If it is possible for you to include data regarding the average posts made by the user, please include it. Especially for those 105 members.
Based on what I've seen in Gambling discussion and Bitcoin Discussion, those people without merits are probably just posting in spam megathreads or commenting stuff like "nice ICO!".
It's hard to automate a way to find the post quality of the user. I thought about finding the average character per post, but I noticed many have posts like:
Twitter Bounty Week X
Tweets:
<insert 10 long twitter links here>
Retweets
<insert 10 more long twitter links here>
They would inflate the average character count, while not being close to a quality post.
This does, however, bring up a very interesting question: how is merit distributed across different sections of the forum? It would be interesting to know which sections tend to produce the most quality content in the forums. This would probably be very hard to do considering that you are getting rate limited yourself with just 500 users in the sample size in a single section of the forum, but there might be an alternative way such as scraping the last x number of posts from the section and adding up the merit from there. However, with the number of sections in the forum I can see this being extremely slow. I do not have much experience with Sheets/Excel when it comes to scraping, but I might eventually write a quick script to do it and post the results myself.
I could write a program in some other language (nodeJS perhaps) to do it, but this data was a result of me getting the formula set up for the ChipMixer campaign, and deciding to reuse it.
Without looking though, I'm guessing that the 3 most merited sections are:
- Meta
- Development and Technical Discussion
- Technical Support
Some points might be:
Lack of merit sources in Altcoin sections?
More spam in Altcoin sections?
Members are ignoring merits or they are too lazy to send merit. Can you research how many bounty hunters sent merits and include number to spreadsheet? I guess you will find out that 90% members who are posting in altcoin section have never sent merits.
Is something wrong with my posts? Am I spamming? I don't see reason why nobody sent me merit while some guys here received 40 merits for couple of topics and posts in meta.
Should I become regular in meta now? Should I write more than it is necessary to write? Should I write essay over half page? Should I send merit to someone if i didn't receive any?
There you go, reasons and small rant. Decide which is which
Nice stats btw.
It's hard to scrape how many have sent merit, as you must be logged in to see those statistics. It would be too complicated to do it in Google Sheets, and probably still complicated in any of the programming languages I'm familiar with.
I took a look at your posts, and while they aren't spam, none of them really stand out to me as being worth merit. Most/all of your posts are very buried, and are on the 4th page of a thread, or more.
Wow what a nice statistic you got there, i guess this is not a hard job if it is done by an experienced manager.
This statistic really negate all the whiners who said that merits not given in some section and only meta's people getting a lot of merits. I dont know how does people's brain works if they are still denying the facts while we got some nice statistics here.
People in Altcoin campaigns likely still post in Bitcoin sections - after all, it's incredibility easy to make a post talking about why Bitcoin is better then Gold by reading the thousands of replies already made. I'm guessing that my Bitcoin campaigner data will show a much higher average too, so it's still possible to argue that merits aren't really being distributed in certain sections.
The account price estimator was for sure able to distinguish between written and quoted text, how it gave a list of addresses posted and addresses in quote seperate..
I think the estimater even had a post quality algorithm that also may have excluded quoted text..
You may be able to use a modified piece of the code used in that for your purpose?