This is with regards to a recently closed issue that created a hard fork in the official project.

Recently there was a commit related to the above issue that changed how your post frequency affects the award you get from that post. Before, posting itself increased total bandwidth, and exceeding a certain threshold would mean you would not be able to make any transactions (for huge spamming of the network), but other than that, there was no penalty on the award you received if you “got through.”

Recently, though, there was an addition of an individual tracking parameter called the “post_bandwidth” for each account. The goal was to penalize people who post excessively in order to get lucky and “hit it big.” The consequence to the network of that behavior was that we would be flooded with lower quality posts. Thus, this was a change in the network in order to limit that incentive to post frequently.

This only applies to top-level posts (not replies/comments).

This is how it works (Examples below so do not worry)

• If you post no more than once every 6 hours, you will not receive any penalty.
• • Every time you post, 10,000 is added to your post_bandwidth. If you have never posted before, this will start at 0. The math for calculating the penalty for the new post includes the addition of the 10,000.
• • Your post_bandwidth, once you post and add 10,000, will decay with a half-life of 12 hours. This means that if you post at noon for the first time ever, at midnight, your post_bandwidth will be at 5,000.
• • Every time you post, 10,000 is added.
• • If you post every 6 hours, these are the values for post_bandwidth (rounded or truncated to the (nearest) whole):
• • • Hour 0 (first post) = 10,000
• • • Hour 6 = 17,500 (it would decay to 7,500 because 6 hours = half of a half-life (which is 12 hours), so you go down by 25%, or in other words, 75% remains. But once you post, this goes to 17,500 because you add 10,000 – this is the same math going forward).
• • • Hour 12 = 23,125
• • • Hour 18 = 27,344
• • • Hour 24 = 30,508
• • • Hour 30 = 32,881
• • • Hour 36 = 34,660
• • • Hour 42 = 35,995
• • • Hour 48 = 36,996
• • • Hour 54 = 37,747
• • • Hour 60 = 38,310
• • As you can see, this number is approaching 40,000. Normally it takes 4-5 “half-lives” for a number in this kind of process to approach the steady-state. This sounds about right.

But why do I care about this number? Sigh!!!

• Because your reward for the new post is equal to 1 or (40,000^2)/(post_bandwidth^2) [ whichever is less ] times the normal award you would get (remember you pre-add the 10,000 for the hypothetical new post onto your post_bandwidth at the bottom).
• This means that as long as after adding 10,000, if you never exceed 40,000, then 1 will always be the smaller number, and you would be multiplying by 1, which would be the same as not having a penalty.
• This also means, that if you do not post for a full day, and submit 4 posts at once, you will still not be penalized, as long as you wait 24 hours again before posting. This allows for a “burst posting” option.

Ok that is a bunch of math, damnit! This sucks!

Have no worries. The good thing is the system keeps track of the post_bandwidth. You can find it at www.steemd.com/@your_account_name_goes_here on the left-hand side with a bunch of your account information (near the bottom).

Ok… well this still sucks. I am lost about how I can use this to really help me figure out if my new post will be penalized or not (practicality).

This is where it gets practical

• I am about to post, and I want to know if I will be penalized.
• Go find your current post_bandwidth (see above)
• Add 10,000
• If this number is 40,000 or less, then there will be no penalty
• If this number is greater than 40,000, then do the following
  -- 1,600,000,000 / (post_bandwidth ^2). So if your post_bandwidth was 35,000 when you looked it up, you add 10,000 and get 45,000 and then you square it which makes it 2,025,000,000. Now do the division (1,600,000,000) / (2,025,000,000) = 0.79. The penalty to your payout will be 21%

I have precalculated a chart of pre-post post_bandwidth, and the penalty.

Finally

Based on my testing, edits to the original post do not count.