Measuring voting participation

Hi, I thought a bit about motion voting and condition probabilities. From a statistical point of view it would be very interesting to have a rough estimate about the number of blocks that actively participate in voting. It would also be useful to know up to which block new motions where considered by the creator of a new block.

Both things are of course not directly measurable from the blockchain. What we could do is to create a recurring motion, say every month or 4 months, which basically contains the following content:

[current date or block hash] This motion has no effect beyond providing statistical information. Please add this motion to your votes if you are actively participating in the voting process and read and understood the current proposals.

If people who follow the discussions cooperate and add this motion every time it is updated, then we could get much more information about how healthy the voting system is.

What do you think?

Conceptually, I like it. I’m going to think about potential negatives, but on the surface, I don’t see any.

Hm. You want to filter the blockchain for every block without votes?

Have an algorithm that starts with block 0 and counts boolean if votes are set or not.
Let the algorithm wait/stop, when the current blockcount is reached.
Count the blocks with vote “true” and et voila… :sunny:

That shouldn’t be a problem, or am I missing something?

Well, if a block doesn’t contain any motion it doesn’t necessarily mean that the shareholder isn’t voting, maybe the shareholder really wants to vote “no” on all current motions.

Secondly, and more importantly, there might be many addresses who are still staking but where the owners don’t pay any attention anymore to this forum. These people have some ancient motions in their voting and would be recognized as active voters by your method.

So there has to be an recurring event that requires an action of the shareholders and which is measurable on the blockchain.

I agree with you, it could just mean “No.”

Concerning your second point:
Whether motions had passed at the time the block was found, should be tracable through the API.
I think same applies for custodial grants.
Sorting should be possible.

I think I’ll script something tomorrow…

slightly off-topic idea: Maybe some kind of notification service would be suitable … Custodians register with a webapp that monitors their address and notifies them (e.g. via Bitmessage), that they are voting on passed grants/motions.
Shareholder gets poked. Blockchain stays slim.

It’s important to know how many shares are voting. NXT experiences show that seemingly overwhelming opinion in the forum could represent only a tiny fraction of all stake holders.

However if the highest "yes’ count of any recent motion is an indicator of the lower limit of how many shares are voting. So I think currently at least 50% are voting. One can use the speed of ‘yes’ votes increase relative to how fast blocks are found after the vote started, a more accurate measure can be made.

What do you think about active “NO” voting?

1 Like

The ternary voting system has been also proposed earlier by @Ben . I think that to answer such a question we should really dig deep into the literature on the topic. I am sure there are tons of books, conferences, research on the matter . It goes beyond the blockchain domain and into the governance models.

I finally took a free moment and published those thoughts in a topic. Thank you for the reminder, @desrever!

1 Like