Hi.
I have lately noticed that the maturity rewards are suspiciously low. I made some screenshots and in the past 7+ days, I have received 0.127% yield. And this is an 8Y neuron with 2.5y age! What is going on? I am following various neurons just to make sure I have fallbacks.
The only thing that has changed in the last months is that I merged a couple of neurons into this. I’m not sure how long I have received lower yield, but it’s been awhile I suspect. For months.
I suspect that your neuron is not actually voting on all proposal because of the complicated following setup you have. In particular, I think ELNA AI is not voting on most topics, which would lead to your neuron not voting if DFINITY votes yes.
Could you look up your neuron id on the dashboard and look at the past proposals? I suspect that you see a lot of “none” votes, which would confirm my theory.
Why would my neuron not vote if there is someone voting in my selection? I have selected a few neurons for fallbacks. This should really be built into NNS frontend so everybody could see their participation.
__
I checked the dashboard. The estimated ICP reward is roughly 10x bigger than what I get. The results are peculiar. I have roughly 10 votes in total for October. And no “none” votes ever. This seems odd, there should be many more votes in total. Dfinity has more than a hundred in October.
Btw, another one of my neuron hasn’t voted since July. But no “none” votes too. Just nothing. I suspect this could be related to merging neurons? This is the only thing I’ve done in the past months.
All neurons vote according to the consensus of all Followees that are configured. There is no such thing as a fallback neuron.
Consensus is defined as:
Adopt if greater than 50% of all Followees vote to adopt;
Reject if greater than or equal to 50% of Followees vote to reject;
No vote is cast if neither condition is met.
This is how neuron voting has worked since genesis, so it’s not a change.
Several things you might want to know based on your configuration which is not readily available information yet (since neuron configurations for known neurons are not yet public):
Synapse always votes on all proposals including the technical proposals, but it is done by following CodeGov on many topics and DFINITY on the rest (more info at synapse.vote)
Arthur’s neuron seems to follow CodeGov on all topics
CodeGov also votes on all proposals, many of which are actually reviewed by CodeGov team members and others by following DFINITY (more info on codegov.org).
Since you likely have many proposals where DFINITY voted Yes and ELNA didn’t vote, I wonder if your neuron is missing the None classification. Perhaps it is interpreting a vote by Dfinity as something instead of none. If this is the case, then I’m sure that part is a glitch.
It would be good for you to spot check some proposals that are not in your list that DFINITY voted to adopt and see if ELNA voted or not. I suspect they didn’t vote like @Manu said. Whether or not there is a dashboard glitch, if ELNA didn’t vote when Dfinity voted to adopt then your neuron didn’t vote since it didn’t reach consensus of your Followees.
I think this confirms that your neuron is hardly voting, which explains why you get lower rewards.
As @wpb mentioned, this is not actually how it works, you can read the precise details in the docs.
Concretely, I would recommend just following one neuron that always votes, like CodeGov or DFINITY (on “all topics” and “governance” and “SNS & NF”). That should ensure you always vote and get full voting rewards.
I agree, I think there are plans to improve that in the future, but i’ll pass on the feedback.
Oh. I see. For some reason I thought many followers fallback, not try to find consensus. In that case it makes sense. Just using Dfinity and CodeGov without overlap now.
Oh boy, I wonder how many rewards I have missed in the last 3 years. 
The ICP Dashboard only supports “None” in the Voting History table of known neurons, though we have plans to support this for non-known neurons in the future.
The reason for this is that the NNS has no concept of a “None” vote. In the recent_ballots array of a neuron, which contains the most recent 100 votes cast, there are only “Yes” and “No” votes. If a neuron hasn’t voted on a proposal, then it is never part of the recent_ballots array. The Voting History table of non-known neurons is simply displaying the votes from this recent_ballots array, hence there are no “None” votes. You can interpret the absence of a proposal in this table as meaning the neuron did not vote on that proposal within the past 100 votes cast.
When I wrote that, I wasn’t thinking of the changes for Request for Comments: API Changes for Public & Private Neurons. Since NNS neurons are now private by default, and private neurons do not expose their voting history, we can only add support for “None” in the Voting History table of public neurons.
