The Powder Toy

Mrprocom

Powder Toy Statistics 2020

The following is a PDF file containing a bunch of statistics on Powder Toy saves, the forum, registrations, moderation, game development, TPTMP and the Minecraft server, collected throughout the year 2020. Some of the statistics also show the absolute/percent change from 2019 to 2020.

Powder Toy Statistics 2020

Possible Factors Contributing to Inaccuracy

The statistics are not completely accurate. The following are factors expected to affect the accuracy:

The raw data used to calculate the statistics come from my IRC logs. There might have been instances of me getting temporarily disconnected and losing some relevant logs that are supposed to contribute to the statistics.
Some statistics required me to manually categorise raw data into a small set of categories to be displayed as pie or bar charts. Examples of such statistics include "Percentage of Lock Reasons" and "Most Common Report Reasons". Such subjective categorisation is prone to bias. Additionally, some of the raw data getting categorised fell into grey areas, making them difficult to categorise.
Statistics dealing with reports are likely the least accurate due to how reports are announced on IRC. It is common for a moderator to deal with a report before it gets announced on IRC, preventing it from getting logged and used in the calculations.

Although these factors might seem to have a bad effect on the statistics, I expect the difference they make to be insignificant. In case you suspect that something is off with any of the statistics, please say so and I will try to verify it.

Are There Statistics for Previous Years? Will There Be Future Statistics?

The answer to both questions is yes. The first statistics I made were for the year 2019 (Powder Toy Statistics 2019). I will work on the 2021 statistics once the year is over. I plan to make these annually, though it has become a slight burden as I am becoming more busy with university and life in general.

Credits

I would like to thank @jacob1 and @LBPHacker for reviewing my drafts, spotting errors, suggesting things to add or change, and for generally helping me out during the process of making the statistics. I would especially like to thank @jacob1 for providing me with some of the statistics that I was not able to get hold of myself from my logs alone. Those include the statistics used to make the save and user leaderboards, the upvote to downvote pie chart, and the MC world size by the end of 2020.

Final Message

I am always open to new ideas and suggestions, so please post them here if you have any. I will not make the changes now for this year's statistics (unless they are extremely necessary), but I will consider them when making next year's statistics. Also, as mentioned previously, if you think that something is off, post about it here and I will try to redo the calculations.

I apologise for being very late with the 2020 statistics. I had a very busy semester and I kept delaying this project until the summer holiday.

I hope you have an enjoyable time going through the statistics and learning from them, whether it is by knowing the best times to upload a save, or by having a general idea of how things are in the community.

All Statistics

Powder Toy Statistics 2019
Powder Toy Statistics 2020

jacob1

Some people might be confused at the date, "didn't 2020 end 7 months ago?". But I'd like to dispel the rumors that these statistics are 7 months late.

If you consider the Chinese New Year, on February 12th, we are only 6 months late by that measure.

Better, if you consider the Islamic New Year, we are 25 days early.

So these statistics are not late, they are early. And we hope to be even earlier next year for the 2021 stats.

SpikeViper

@jacob1 (View Post)

I'm going to make sure to use this logic for Valour (my code project). Absolutely brilliant.

jacob1

@SpikeViper (View Post)
It is the best logic!

Also, as you can see, you contributed to a huge bump in users in late 2020 lol. I do remember that month, registration activity was through the roof and it was obvious to anyone who looked at it that it was way higher than average. So thanks for that.