Qtum Mainnet Results October 23–29
The chart lineup is changing this week; we say “goodbye” to three of my favorite charts and introduce one new chart for Whale watching (wallets that win multiple blocks per day).
The educational feature for this week got out of control and was sent off to its own Medium story and tries to be a comprehensive introduction to Qtum Proof of Stake mining disguised as an Explain-Like-I’m-5 years old (ELI5 — which means a very simple, understandable explanation) story about racing, using the recent BMW Berlin Marathon as background. The graphic above gives a little taste for that story; check it out if you are interested.
I am an independent researcher, not affiliated with the Qtum Team, but I would like to acknowledge and have great appreciation for their advice and help along the way.
Charts and Graphs
This is my fourth week of reporting on the public deployment of Qtum Mainnet Ignition, examining the parameters and performance of the Qtum blockchain. Data sources for this article come from the Qtum Explorer and logging from the qtumd server application.
Mainnet activity began stabilizing this week with the exit of staking by Foundation wallets. Unique wallet addresses winning block rewards peaked at 434 per day on October 24th, and the network weight calculation peaked up briefly to almost 16 million — block 31,638 on October 22.
The number of Qtum ERC-20 tokens officially burned in the Ethereum smart contract reached 42%, up from 41% last week. This does not seem to be a useful number, and will no longer be reported, but you can check for yourself at the Qtum ERC20 smart contract (add the “burn addresses” 0x11111…, 0x2222…, etc.). These ERC-20 QTUM tokens could be taken out of circulation in other ways. The Ethereum smart contract represents a public accounting for the demise of these tokens.
Network weight increased to a brief peak of 15,907,217 for block 31,638 on October 22. Please keep in mind that network weight is a moving average of the weight of recent block reward winning addresses, not the total weight of all staking wallets (which would not be possible to know), which explains the block-by-block variation we see. This chart presents end of day (UTC) network weights.
Unique Reward Addresses
Unique reward addresses are a function of the number of large wallets that are active on the network. If the network were composed entirely of “little guys” then the number of unique addresses winning block rewards would approach the number of blocks per day, currently around 600. During the week, there was a peak of 434 unique addresses on October 24th, which may be a record for some time, since more big wallets are coming online. You can see a list of these big guys at the QTUM Explorer.io Rich List.
Taking the inverse of unique reward addresses (subtract unique addresses from blocks per day) shows the number of block rewards won by wallet addresses with multiple wins per day:
Active Transactions Per Day
Transactions per day peaked at 11,036 on October 20 as Coinoners withdrew QTUM from that exchange. This week transactions per day settled into the 3,000 range.
Keep in mind that the baseline for transactions is two per block — the blank coinbase transaction (the bitcoin block reward — not used for Qtum) and the coinstake transaction (the Qtum block reward), so the minimum transactions per day would be:
If we subtract these baseline transactions, we can see the transactions from activity above the baseline, here called Active Transactions per Day:
Block Spacing Variation
This week I continue obsessively tracking block spacing with the data set for October 22 through October 28 (blocks 31,067–35,261 inclusive, 4186 blocks total). In this week there were 6 blocks with more than 20-minute spacing (vs. 9 last week), with the greatest spacing for block 35,753 at 31:11. Last night (October 29) block 35,806 had a spacing of 54:06, the longest I have seen in the last month.
On the low end there was one block 35,245 logged at 1.6 seconds (but my logging is only += 2 seconds) — you might (correctly) guess these fast blocks rarely collect any transaction fees. Here is a look at the block spacing, for those spacings from 0 to 256 seconds, which clearly shows the 16-second granularity for wallets finding the kernel hashes:
I found an interesting result for the time synchronization on the network. By taking a moving average of the block arrival time compared to epoch time modulo 16 (modulo is the remainder of a division operation), the network is consistently tracking 4 to 5 seconds after time % 16 = 0. Sorry to be a little geeky here, but what this means is the network nodes, with a variety of time references, have a strong tendency to find new blocks at specific clock times, which doesn’t seem to drift.
For example, at 12 midnight UTC today (00:00:00 UTC October 30) epoch time was 1509321600 (You can convert from epoch time to local times with https://www.epochconverter.com/) with modulo 16 = 0, and the network wallets were likely to generate a new block in 16-second multiples starting 5 seconds later at 00:00:05 UTC, 00:00:21, 00:00:37, etc. Actually, block 35844 was published at 00:02:33 UTC about 4.4 seconds after the modulo 16 reference.
Regular readers of this report will expect a chart here, and won’t be disappointed. Below is the chart for October 27th, which shows that approximately 66% of the new blocks arrived in a four-second interval around the average time reference, with that reference shown here as the bucket -0.83 to +0.33 seconds.
There does seem to be an increase in the super-fast blocks at the very low end of the block spacing this week, for 3 seconds or less, with 9 last week and 30 this week. This may be due to the concentration of more wallets in close proximity within cloud data centers (e.g., Google Cloud) — the probability wouldn’t change for these wallets, but I think the pings between the wallets within that data center network would be faster.
We say “goodbye” to the annual return chart after this week, since there is an excellent real-time interactive resource for expected return available, the Qtum staking calculator by @calbert65.
Normalized Weighting for Foundation Wallets — And it’s Gone
We say “goodbye” to two more charts this week, the Foundation vs. Private chart and the Normalized Foundation Capacity chart, because as promised the Foundation discontinued staking from their wallets after the Coinone swap, and both these charts flatlined. The Foundation mined their last block 32,196 on October 23rd, at 9:14:56 PM UTC, and we thank them for their service launching and securing the network up to this point.
The methodology for calculating and charting the normalized Foundation capacity was given in a previous report.
Wishing you good luck and prosperity; may you win countless block rewards.
祝你幸運快樂, 財運亨通, 并赢得了许多無數的奖励
See my other reports on Medium.