What every user should know about the upcoming change in voting

in #steem8 years ago (edited)

There has been a significant amount of misinformation and propaganda being spread about the upcoming change from a 40 vote target to a 5 vote target.

The gist of this propaganda is that the change from 40 votes to 5 votes changes nothing, because all votes will be worth 8 times as much. So for example, if I could cast 40 100% power votes worth 10 cents each before the change, I would now be able to cast 40 12.5% power votes also worth 10cents each... or 5 votes worth 80 cents each. This is entirely false.

To explain why, I first have to explain a little about how voting weight works. There are 3 factors that affect your voting weight

1.your SP.
2.The percentage you choose on the slider
3.Your voting power. This goes down the more votes you cast. You can look at it on steemstats.com.

If 2 and 3 are expressed as decimals, you can use SP* percent*power to get a number that represents the complete power of your vote. Let's call this "P"

The maximum P you can cast per day is by voting the target number (let's call it T)of times at 100% on the slider. At this equilibrium, your vote power will stay at approximately 80%. Currently, the target is 40. So the max P you can cast per day is SP* .81 * 1 * 40

  • If you cast more than T, your lose out because your vote becomes too weak. For example, if you cast 80 100% votes per da the rapidly declining power of your vote would cause you to actually cast less P than 40 100% votes. (the decline in your voting power is not linear

  • If you cast 20 100% votes, you will lose out based on the number of votes you cast (even though your votes will be slightly more powerful)

The reward pool, and how your voting distributes it.

Note -- the math here is not exact, for a variety of reasons (for one thing, it's based on vests not SP). It does demonstrate how voting awards a share of the total reward pool.

Lets say I vote on a post at 100% on the slider and 80% power with 1000SP. That would make my P 1.810000=8000. How much is that vote worth?

To find the answer to this question, we need two more pieces of information, the total P cast that day by all users (lets call that P (sub A) and the total reward pool (lets call that R). V is the value of my vote.

P/P(sub a)=v/R
PR=V*P(sub a)
PR/P(sub a)=V

For the sake of example, let's say a total of 10,000,000,000P is cast each day, and the reward pool is $50,000.
(8000P*$50,000)/10,000,000,000P=$.04

So under the current 40 vote system, I could cast about 40 votes a day worth about $.04 each, for a total influence of 320,000P or $1.60 a day. Note that there are many other ways I can vote, but none of them will give me a greater daily influence than $1.60.

Note that each individual P cast is worth about 4 cents in this system. We can determine this by dividing the reward pool by P(sub a).

The reward pool is constant.

The new system is not changing that $50,000. The pie that is being divided up is still the same size.
Or 40,000P total. The $ value of this daily influence will do down from $1.60 to $.20

Well, what if we get 8x more P for our 100% vote.

It doesn't matter, if we're getting 40,000Px8 now, so is everyone else who casts a vote. P(sub A) will increase 8x just like P, and the two increases will cancel out. With a constant R and every P would be worth 1/8 as much as it was before in dollar value.

So how come @biophil thinks 100% votes are going to be worth 8x as much?

The reasoning is that because everyone can only vote 1/8 as much as they could before, then the total amount of P cast should go down to 1/8 of what it was before. SO our equation to figure out the $ value of our 80% power, 100%slider vote would be

(8000P*$50,000)/1,250,000,000P=$.32

why biophil is wrong.

This math only works out if every current user is maxed out and casting the optimal number of votes already. That is to say, if the current P(sub a) is 50,000M it will only be reduced to 12,500M by this change if every user is currently casting their optimal 40 votes a day.

So will our votes increase in value at all?

Maybe. They could increase in value, decrease in value or stay the same. Since whales control most of the SP, they also have the biggest effect on the total daily P cast, and thus the most significant impact on the effect of this change. There are three ways this will effect different whales.

  1. The Donkeypong effect -- For active whale curators like donkeypong and a few others, this will drastically decrease the amount of P they cast. Voters like this, for the most part, stay maxed out, so they will decrease by about the 7/8 @biophil imagines.

  2. The Ned effect -- For infrequent voters, like Ned, who already don't come close to their 40 vote a day quota, this will have very little effect. As an example, let's say ned currently casts 5 votes a day. The change will lower his voting power slightly by keeping it at an 80% equilibrium (instead of near 100% like ti would be now) but nowhere near the 7/8ths it would lower others.

  3. The Dan effect: Not to single dan out. Many whales have sockpuppet accounts, but do not use them to vote currently, because they do not need to. They vote their full 40 (or close) every day. However, if they were limited by a 5 vote max limit, they could just as easily bring in their sock puppets to cast the full 40 they're used to (or at least get closer to it). For example, dan typically uses @dantheman to vote, but he has at least one other account @dan, which is twice as strong and which typically hardly votes at all that I can see. Obviously depending on the extent of the sockpuppet accounts, and whether the whales actually bring them in or simply cut down their voting, this could go from having no effect at all to cutting down their voting by the full 7/8.

At the end of the day, some decrease in P(sub a) is very likely (in fact, some decrease in unavoidable). But it will most likely be quite a bit smaller than the 87.5% required to increase voting power by as much as indicated. Personally, i would anticipate a vote value increase per 100% power vote of around 25% to 50% . But thats just me spitballing.

Voters should watch the effects of their votes very carefully after this change (which, i believe, is on the 13th) because if they are significantly less powerful than 8x, then a lot of people have been selling you a bill of goods.

Sort:  

Bloody hell this is complicated !

FOr the record, note how you can hear crickets every time you ask for documentation about this mythical 8x vote

I don't know how to read code, but it only makes sense that if the target number of 100% votes is changed from 40 to 5, each 100% vote would have to be 8x more powerful. I'm not the only one who thinks this:

https://steemit.com/steem/@steemitblog/announcing-steem-0-14-0-release-candidate

In this link @smooth said: "it is 5 more powerful votes per day. Also you can vote more often if you voluntarily reduce power per-vote using the slider. 10%-power votes after the change will be about the same as regular votes now."

Also, in another post, @arhag agreed with my analysis.

These are 2 of the top 3 witnesses who have said something similar and agreed with my analysis, respectively. These are the people who know more about Steem/Steemit than just about anyone.​

I am researching and posting more in comments later on tonight. I am no longer sure i am 100% right.

I am no longer sure i am 100% right.

Didn't see that coming...

sorry I couldn't resist :)

Didn't see that coming...

Is it surprising that i could be wrong, or that i'm admitting it?

coz if its the first one, im right there with you

My take away from all this is you have to take into account how everyone else votes, what that total P(sub a) ends up being. For example, if everyone only voted 1 time per day at 100% power, then your relative share of P(sub a) would be unchanged from what it would have been before the fork under the same circumstances. But if a lot of whales start voting less because they are already maxing out their optimal quota today, then your relative share of P(sub a) could change, affecting your vote's value. So to really be sure of the effect, we would have to know the current voting habits of every user in the system. In short, it's hard to predict the exact effect this change is going to have. I guess we'll just have to wait for next week and see what happens. Oh the suspense!

I think you've done a lot of calculations for the post above, I have to admit all those formulas there are a bit hard to get unless you really focus on them for a minute.

What I think will happen is highlighted in the graphs below. The amount of recommended used voting power remains the same overall and if I will have a recommended voting number per day of 5 upvotes at 100% but those 5 votes will eat up the same amount of VP as the 40 did before = 20%.
Hence since I dump more VP with each vote, I also dump more weight e.g. more reward.

Imgur

PS. I think that is great for me, as my upvotes (as an evolved minnow) will weigh more that they currently are and I'd love to reward people more. Otherwise, I'll just use the slider.

I don't have a slider yet so my vote isn't worth much to start with

I think the threshold is at 1MV which currently sits around 319 Steem.

1 MV = 1M VESTS = 319.881 STEEM = $258.144

Source here: https://steemd.com/distribution

So does that mean I'm close or not. Thank you

You're close. After you get over the 319 Steem Power you will start seeing the slider.
I see you have 306 SP and 68 SBD in your wallet. Use the Market to change some of that SBD to steem, and then Power up the STEEM. You should be good with about 14 more steem.
That is off course if you want the slider that bad :)

I think I'll just try working harder see what happens. Thank you

really it looks like you should. I heard the cutoff was 300 sp.

When I vote nothing shows up. Am I missing out

im not sure what the cutoff is... you would definitely see it if you had it.. i remember hearing that it was supposed to be 300 by the time they rolled out the new hardfork

Suppose I will have to keep trying and hopefully I will get a slider too

It's calculated in vests, which is appx 300 sp.

So I'm not good enough . ouch that hurts lol

So how come @biophil thinks 100% votes are going to be worth 8x as much?

The reasoning is that because everyone can only vote 1/8 as much as they could before, then the total amount of P cast should go down to 1/8 of what it was before.

That is not his argument. He isn't talking totals but rather each vote individually. He never says that the total number of votes will decrease eight-fold, that's beside the point. The point is, people will now have the option to make a single vote worth more if they want. This higher-value vote will come at the expense of that individual's voting power.
https://steemit.com/steem-help/@biophil/announcement-all-votes-will-be-up-to-8-times-more-powerful-in-new-hard-fork

Oh, I might as well chime in. @sigmajin and I have been vigorously miscommunicating on that post you linked to. Our misunderstanding boils down to this, which I just realized today: we're measuring the "strength" of a vote by two different measuring sticks. I'm measuring a vote in absolute terms by how much it boosts the rshares of a post (the more rshares a post has, the more weight it gets in the reward pool), but he's measuring a vote by how much it boosts the payout of the post. His way of measuring is totally legitimate, and I'm guessing he's right - we probably won't see an 8x increase there. On the night of the hard fork, we almost certainly won't, because everybody's power will shoot up all at once and everybody will cancel each other out. A week later, we'll be back to some sort of equilibrium, and then we'll probably have a significant increase, but possibly not as high as 8x. It just depends on total voter participation; if participation goes up (a good thing), the increase won't be 8x. If it goes down (a bad thing), the increase will actually be more than 8x. It's very difficult to predict.

I really really really wanted to argue that his way of measuring is "wrong," but it's not. It's just unpredictable.

Here's the thing about "my way" of measuring it. A couple things actually.

The first is that the whole reason i was in your thread is @shenanigator sent me there. If you read his comments in neds thread, he is unambiguously talking about money, not about rshares. Which, as you note, is suppositious at best.

The second is that if people are saying this: "All votes will be up to 8 times more powerful in new hard fork!" most people are going to assume (just as shennanigator has) that what you're promising is a real money increase in the effect of a 100% vote. If you then go back and say "well rshares game theory hashtag quadratic equation" its going to sound like the owners pulled a fast one.

ANd yes, i realize that you said "up to"...

Yes, I was talking about money. As it turns out, it could be misleading to do so.

I'm more knowledgeable about it now, and I thank you guys for that. It will help me get the most accurate information out there. As I said in a reply above, I still think the money a vote is worth will increase by about 8x, but only after users' voting power is drained. It seems @biophil thinks this could take up to a week for things to run their course and reach the new normal. I believe that it will happen in a couple of days, 3-4 max.

If you then go back and say "well rshares game theory hashtag quadratic equation" its going to sound like the owners pulled a fast one.

Totally valid point, I'll definitely be more careful about this next time I write about it.

OK, so first things first. Keep in mind that im not a coder, so i can just look at documentation and discussion. Since other sources have confirmed it, ill take your word for it that this is doing what you say its doing. But heres my take and why I wasnt sure to begin with and still have my doubts.

So used.power is like fuel to assign rshares. And vote_regeneration_per_day is like how much gas gets put in your tank every day.

So in the current system, say a 100% vote is like using 1 gallon of vote.power fuel to assign 10 Rshares. What you're saying (which though it might be true, is not clear from documentaion or discussion) is this

under the new system, for each post you vote on, instead of consuming a maximum of 1 gallon of fuel to assign 10 rsahres, you will be able to consume 8 gallons of fuel to assign 80rshares

This might be true. I certainly can't prove it isnt. But based on what little information i can find in plain english on github, the following scenario could also be true.

under the new system, you will be able consume 8 gallons of fuel at once, but that 8 gallons will only give you the ability to assign 10 rshares.

Now maybe there is some essential link between consuming used.power that makes it so that consuming a specific amount inherently generates more rshares. But its not obvious to me. In the spirit of full disclosure my interpretation of the change in vote power consumed has been the second thing, so if its really the first at least part of my reasoning was wrong.

You know that I already believe this, but from my reading of the code, I'm 99% sure that it's the first of your two options. I'm sure enough that I'd unhesitatingly bet money on it.

After I had written some of these responses, I thought maybe I should let you speak for yourself. Or at least put a disclaimer that I believe I'm representing your argument accurately, but there's a chance I'm not because you're more knowledgeable about the changes. Please correct me if I misconstrued your argument in any responses.

Thanks for the clarifications here. I also thought that we wouldn't see an 8x boost immediately because most people's voting power will likely be close to 100 on the date of the change. For a few days, I think people will be placing a lot of high-slider-percentage votes which will cancel each other out. Several days after the change goes into effect, I think many people's voting power will be well below 100. At that time, it's my guess that a 100%-slider-vote will be worth about 8x what it was.

No, it was fun to show up here after arguing with sigmajin in my own thread, only to discover this thread and find you arguing my side over here as well. :) From what I can tell you've represented my standpoint well.

The point is, people will now have the option to make a single vote worth more if they want.

No they won't. Once again, this is a fantasy that you made up and that neither you nor BP has provided any documentation whatsoever for, I already explained in the post above why that wouldn't be possible. You can not increase the total amount of P and not decrease the value o f each P. Not with a fixed reward pool.

I really, really do get what youre saying. Youve been spamming the same cooked fantasy all over steemit. It simply ain't so. Basically what youre describing is a system thats the same as the current one, but you can cast 8 100% votes on the same thing.

Again, i think this would be an awesome idea, but its not whats happening. And repeating the lie again and again isnt proof, its just propaganda.

It doesn't matter, if we're getting 40,000Px8 now, so is everyone else who casts a vote. P(sub A) will increase 8x just like P, and the two increases will cancel out. With a constant R and every P would be worth 1/8 as much as it was before in dollar value.

Just going to keep peppering my documentation around. This is the code change:
https://github.com/steemit/steem/commit/6500bb65eb6282866f3c6f356cbcbe09fd03c7bf

One wouldn't be increasing the power of total votes. You would only be enhancing the strength of a single vote at the expense of your future votes

I understand the fictional protocol that you made up. the problem is its not whats on the table.

Even if it were, it wouldnt work, at least not like you think it would. but its not

You made up a fake protocol for voting that not supported by any documentation whatsoever. I can't prove a negative. Your vote at 100% will still be worth the exact same amount of absRshares it was before the change. There is no 8x power vote. Its a complete fantasy.

Produce some documentation for it if you have any (you dont because you made it up) but you can't expect me to prove a negative

You're arguing points neither one of us has made, so that makes me think you don't understand.​

You can not increase the total amount of P and not decrease the value o f each P. Not with a fixed reward pool.

You aren't increasing the total amount of P because the voting power decreases accordingly with a higher-powered vote in the new system. Therefore, subsequent votes would be worth less due to significantly reduced voting power.

Yes, once again. I get it. Its just like the current system except you can cast 8X power votes. Thats an awesome idea that exists only in your head and is not documented anywhere except in a never past bed time land. This is where your vote is worth 8X

@sigmajin already knows this, but this is the code, which you can see has now been merged into master: https://github.com/steemit/steem/commit/6500bb65eb6282866f3c6f356cbcbe09fd03c7bf

Not to belabor the point, but i do want to make it clear that the quoted text is precisely why @biophil thinks votes will be worth more (in money terms). We get there different ways (and his way is probably more correct, because he understands the mechanics of how steempower translates to rshares translates to money better), but the part of his theory that has people getting 8x more money for an individual vote absolutely, positively depends on 7/8 fewer votes being cast.

Do we know if this change is definitely going to happen? They said it was a proposal and they were looking for feedback. Given how much negative feedback there is, will it go ahead?

its up to the witnesses, but i suspect theyll go along with it for the most part.

SO does this mean the witnesses have approved, or does the change go out on the 13th and only become "real" if a majority of the witnesses accept it.

Huh. Interesting about the witnesses switching back. I wonder what's going on.

it was a glitch getting fixed... still hope the witnesses decide to quash this... a bunch of them seem to have voted for my other post, so who knows.

Some of the witnesses have approved, but not enough yet it looks like. You can see what version they're running here; 0.14.0 is the new version: https://steemd.com/witnesses

really interesting development ... it seems like all of the witnesses who had been running 14.0 went back to 13.....

Bookmarked for future reference

This was interesting, but where is this voting "slider" that you refer to?

I thinknits an old feature. Indont thinknthey have it anymore.

I can find it in the eSteem app created by @goodkarma actually, but I just never found it on the site.

Oh! Your first bullet point is wrong! If you cast 40 100% votes per day, your total voting output is exactly the same as if you cast 80 100% votes per day. This is on average, after you've hit equilibrium (which takes a couple days, I think).

This is because there is no penalty for voting too much; all voting too much does is spread your votes thinner.

Im not 100% , but my understanding is that voting power does not deplete linearly (though if the depletion is linear, you're correct). I think thats what the chart on page 19 of the white paper is about.

To stick with the fuel analogy, its like a car that burns more gas when the tank is more than half empty. You lose out if if you let your power get too low. I think.

I know - the whitepaper is very weird and misleading on that point. I had to reverse-engineer the algorithm myself from the code before I actually believed this, since it's not documented anywhere. Each time you vote, that gallon of gas you take out of the tank all goes completely to upvoting the post; none of it gets wasted on something else. Your car is just as efficient with an empty tank as with a full tank, it's just slower (might as well beat this analogy to death). It's exactly this:

power_after_vote = power_before_vote - used_power

where used_power gets multiplied by slider power and SP to calculate the rshares that go to the post.

That exact expression shows up in this line of code.

OK, i see what that graph is saying now. Wow... ive actually bee spacing them all this time.

Incidentally, based on this, I made a model... tell me if you think I'm on point in my analysis.

https://steemit.com/steem-help/@sigmajin/the-tale-of-the-5-brothers-a-voting-power-parable