Once I introduced my Son-in-Law to gridcoin and got him started, the question of whether to join the pool or go solo came up. I immediately encouraged him to go solo - even though he had no GRC. I planned to give him enough GRC right away to start staking.
Then I hatched a plan.
The Staking “Problem”
When it comes to staking as a solo miner the suggested numbers are all over the place. If you're going to mine GRC solo you must have at least 10,000 or is it 2,000 or perhaps that's "hundreds" of GRC. So which is it? It’s an important question in part simply because one goal of all this science being connected to Gridcoin is to encourage more research by helping to pay for it.
Staking regularly becomes even more important when you learn that if you don’t stake within 6 months, any of the gridcoin which you are owed will disappear because they never actually exist until you stake a block and mint them. Your GRC owed amount is like a conveyor in which the five month old coins are still ready to be minted, but there are no seven month old unclaimed gridcoins. So with a hat-tip to Chicago politics, the rule should be: stake early, and stake often!
How many gridcoin then does it take to stake in a reasonable amount of time (at least weekly)? Sure there are formulas out there which endeavor to answer the question. But who has time for math when we can do science? Besides, as my Lab Partner noted, the formulas all likely work off of the “worst case” scenarios. But these don’t really work perfectly for a few reasons.
- Not everyone who can stake is always staking. They might be offline on accident, or as I was for the duration of this experiment, on purpose.
- Some people put coins in reserve so that they don’t stake against them, which makes them available to spend. (This is done in the wallet SETTINGS > OPTIONS > MAIN > RESERVE).
The Staking Experiment
Now that my trustworthy Son-in-law was up and running as a solo mining boincer with not a single GRC to his CPID it was time to answer the question to the best of our ability.
With about 4,000GRC in my account I can stake reliably every three or four days. But what happens on the other end? Thus was born the experiment.
Every day I would send him 100 GRC, and we both would measure our Estimated Time to Stake (ETS). Once those coins had fully matured for 16 hours. In the interest of science he and I would both record these numbers to the best of our ability by recording the highest and lowest amount of ETS we observed. These we dubbed ETSH and ETSL respectively. Let me stress observed, because we both work for a living and couldn’t just stare at the screen all day, but we resolved ourselves to watch as frequently as we could and take note of the numbers on a spreadsheet.
It is likely that @jmarndt’s numbers are more accurate because I was trying to keep my wallet locked all the time so I wouldn’t stake, and only unlocked it for a few brief periods to check the staking potential.
Observing the Numbers
So here it is in all the glory, a chart that displays our activity:
Note that against the left axis are the total Gridcoins held by @tcblack and @jmarndt over the two week test. Against the right axis is a linear scale of how many days to stake are estimated by the Gridcoin wallet. Additionally the high and low forecast for the days to stake is given in light and dark blue lines for @jmarndt, and light and dark red lines for @tcblack.
For comparison, measure reds to reds or blues to blues.
Note that the dashed lines are for Gridcoin balance, so higher relative to the left axis means more grc. As for the solid lines, they represent estimated time to stake measured in days relative to the right axis, so the closer to the bottom the better.
We actually intended the experiment to go on longer, but I forgot and left my wallet unlocked and thus staked a large backlog of GRC on the 19th, additionally the recent forking issues cut it short. Nevertheless, I think we got some good data which demonstrates that you can reasonably go solo at current network levels for considerably less than the normally recommended 2,000 GRC. j
EDIT
An excellent point was brought up in the comments that I needed to carry in Net weight as a point of comparison. I didn't quite have that data for the full length, but I can add that here for most of the time.
So here is a chart for Estimated Stake Time -vs- NetWeight. In my second experiment, coming up in a few weeks I'll have more to add on the impact of this metric.
End EDIT
Points of Discovery
We had a few points of discovery along the way.
Coin maturation
It does take exactly 16 hours for coins to mature once they hit your wallet. This is no surprise to those who know the gridcoin blockchain of course, but it was nice to confirm it in action as a verifiable data point.
Staking is Dependant on Network Activity
Additionally on the sixteenth of January the price of GRC did drop significantly, and in @jmarndt’s words, “So instead of seeing my ETS drop consistently like it had done every other 100 coin increase, it actually went up slightly. ...I believe that due to the price drop there were more GRC being purchased than normal, which means more wallets were open/staking, making it more difficult to stake.”
If there’s one lesson to come out of that, it is that your ability to stake is largely dependant on who else is competing for the stake.
Staking Thresholds
We also noted, and you can observe it in red on the chart above, that there were certain staking thresholds. Network Difficulty being taken into account there was an overall increase in stake time for me when my balance went from 3,800 coins to 3,700. Additionally in blue you can note that @jmarndt’s ETS also dropped dramatically in the early days from a high of 102 days (essentially never) to a low of 15 days on January 15th with a balance of 600 GRC.
There was also an anomalous low on the 17th for @jmarndt, but it did reflect a dramatic drop to single digit low ETS once his wallet hit 800 GRC.
The Stake
The moment we were all hoping for happened! He staked his newbie block on the 24th with about 1300 GRC in his wallet. And may I add that the improvements to the V9 blockchain and wallet meant that he did not temporarily lose this investment to the former bug.
Because that happened in the initial days of forkapalooza, days which are thankfully coming to an end now, the network difficulty was quite low which in a sense artificially inflated his true chances of staking. But that’s blockchain in my experience: a little bit of luck combined with math and science.
Conclusion
As was noted by @Gregan, in his recent post, the days to stake is nothing more than a guideline, “by the way, the estimate never went below 10ish days, so be sure to treat it nothing more as a guideline.” That said this experiment might give you a better idea of what to expect in going solo with whatever balance you have. I certainly agree with @Gregan, 500 GRC is likely the bare minimum. 700-800 gets you a much better chance, and 1,000 plus should be quite comfortable, with 2,000 being indeed a virtual guarantee at current network loads.
Addenda
If you don’t happen to have a father-in-law who is interested in your success with Gridcoin the pool is a perfectly acceptable means of getting into Gridcoin. But if you want to go solo, you might need a quick refresher on how to obtain it. For now, you must buy a coin like bitcoin or litecoin from someone like coinbase or others; then using an exchange convert it to Gridcoin and send it on to your wallet.
Personally I started with Coinbase >> C-CEX. Afterward I learned that Coinbase >> GDAX >> Exchange skips out on Coinbase’s fees. Last month I discovered Flyp which makes the whole thing effortless. Just Coinbase>>GDAX >> Flyp and you’re good to go, and if you already have a Holy Transaction account I’m told you’ll find the fees there are even lower. Until they upgrade, I wouldn’t use C-CEX since they are still on the 3.6 wallet.
There are several other options out there. (see my recent post: Gridcoin - the time to buy is nearly now) And soon, very soon, the next stable wallet release will be out and all the exchanges will open again. GRC is vastly undervalued right now, so buying what you can will get you staking solo in no time.
Now let’s go do some science.
Love the article : your scientific approach is exemplary.
Also, may I suggest the use of bisq if you want to get GRCs, with fiat->BTC->GRC? There may even be an offer without intermediate step.
Interesting article!
My perception is that staking frequency is primarily influenced by network weight. That might means that some of the forking over the past month could affect that.
I also think that staking every 6 months is the minimum, but it is definitely very reassuring to stake more frequently! Plus you get a little more compound interest by staking more frequently.
Thanks for your post,
@dcj
Staking is absolutely based upon network weight. But please note that this experiment was run before the forking occurred, and was in fact terminated when the forking began.
Everyone's comfort level for staking frequency is going to differ. But please note that actually staking is not the same as estimated staking. There is some aspect of this which is somehow a random event.
Great article , I learned a lot of things !!! Without father-in-law I'll wait flype's return to go solo . Thanks for informations.
Keep it up @grc-son. I'm looking forward to Flyp's return as well.
Let me start by saying awesome effort and good methodical way of documenting the result.
However this is the worst time to do the experiment. The values are not at all reflecting general conditions.
Place network difficulty in the picture at each of those point and you will see a entirely different story.
I personally have stacking for the first time during this when i had balance ~ 10XX.
I applaud the effort but the values will be highly skewed and this would warrent a retry at some later stage when network at stablized.
I would have personally used this low time to actually gather more GRC's and then push them to wallets once they stabilise to get more stakability like you suggested.
Please note that I stated the experiment was terminated by the forking, so all of the "normal" network activity was happening during the experiment. In fact I state in the article that the forking terminated the experiment. So the numbers are still good.
Additionally, (@jmarndt would have to verify) when he finally staked the network weight was "normal" at 1.xxx on a non fork.
Network difficulty is always fluctuating. I've seen it close to 7.x and under normal conditions before the recent forks as low as 0.4x though that low was an extreme example of "normal".
OH, and you might be interested to note that a new experiment is in the works. "trust but verify".
agreed. and like i said full merits for doing it and full merits for having plan on doing it again. But next time also plot network weight in picture that will give more clear picture.
Thank you! I went back to my data and discovered that I did have Net weight data for most of the test. As a result, with your prompting I have added that chart. Take a look.
In my second comparison experiment I am including other metrics as well for a more complete picture.
As a matter fact, network weight will be included.
A very nice write up indeed.
I find it interesting that the chances grow seemingly exponentially in the higher bracket of the 500-1000 GRC range.
I concur with the above comments:
Great idea, almost scientific setup of a real-world experiment sending the coins in blocks regularly. Having read many many theoretical posts, this is one of the few that give me as a noob a solid expectation on how much GRC I need to stake the newbie block. I am buying BTC right now and also waiting for Flyp.me to re-open doors. :-)
Thanks for that, @tcblack!
That was the point of my experiment @xlm567. I'm glad you found it useful.
I'm running a second followup to verify and test a few other variables now.
I stake once every other week roughly with over 2000 but I dont solo at this point. I might once I'm in the top 10 on the pool tho!
Great write up! In your opening paragraph you say you recommend staking even though he had no GRC, are there advantages to staking and solo crunching over pool crunching? Or was this decided specifically for this experiment and write up?
Thanks for the article!
I recommended Solo to him because I was planning on kick-starting him. Benefits of being the Father-in -law. :-)
There are a few benefits to pool mining.
There are also some benefits to Solo mining.
Not a definitive list there, but it's a start.
Thanks for the reply, great info!
As a newbie in the Gridcoin World this article is an absolute life saver.
If I may I will refer to it in one of my upcoming Gridcoin Denmark videos.
I am glad to read your mathematical approach to the subject.
Maybe you could do a follow-up article in the near future?
You certainly may use it ... unless you're going to mercilessly make fun of me. LOL.
Happily I am working on a follow up right now with more data points, and more data in general.
Did you ever do the follow up article? Sorry have been off grid for a couple of months
Thank you for this! I am looking to get into gridcoin. I thought this would be a great idea.