Lately, the total hash rate on the Bitcoin network has been the topic of significant conversation in the crypto community.
The prevailing theory about why the hash rate dropped suddenly around April 17th/18th, was a major (regular) mine explosion within a certain province in China where several large mining entities are allegedly located. It has been claimed that these same miners in that province had their operations thrown offline as a result of the ensuing electrical outage that occurred in the province following the explosion.
Above is a link to a 'CryptoBriefing' article titled, 'Bitcoin Crashes After Accident in Xinjiang, China Halts Mining Operation'.
The article's subtitle parrots the incorrect thesis, stating, "A coal mine explosion in China has allegedly caused a massive drop in Bitcoin's total mining power."
Before even evaluating the article, the first thing that we need to do is find an independent (non-crypto) resource that can shed some light on the nature of the explosion for us.
A report by 'GlobalTimes' (Chinese publication), published on April 12th, 2021 sheds a bit more light on what exactly occurred in this province on that day.
Major Takeaways From the Article
- This event occurred on April 10th, 2021
- Flooding took place around 6 p.m. (assuming this is in local time for the province)
- The explosion is the cause for troubles in that region, but the actual 'effect' that led to the alleged outage of several miners was the "coal mine flooding" that ensued.
Fortune Reports Differently
For some reason, Fortune magazine's version of events takes a major break from facts of the case that we knew to be real (at the time).
The article (linked above), begins with the following:
"One of the great Bitcoin unknowns has long been the amounts being produced, or 'mined', in what's believed to be the top locale for mining the signature cryptocurrency: China's remote Xinjiang region. We got the answer when an immense coal mine in Xinjiang flooded and shut down over the weekend of April 17-18."
Major Problem Here - the coal mines flooded on April 10th, not "over the weekend of April 17-18".
So why is Fortune magazine telling us otherwise?
Examining the Veracity of Claims Correlating Bitcoin Hash Rate With the Coal Mine Explosion
Regardless of what Fortune magazine has told us, we can begin to construct at least some of the 'known' variables in this universe and use that to come to a reasonable conclusion on whether the claim that the coal mine flooded could have precipitated such a massive drop in hash rate.
Step Number One - Looking at the Actual Hash Rate For the Network
There aren't many websites that provide this type of information (in a meaningful manner), but there are at least some.
And fortunately, 'BitcoinVisuals' is one of those websites.
Specifically, historical hash rate information obtained via their full node (running on the Bitcoin network), provides us with comprehensive information on what changes (if any) to the hashrate can be observed from the beginning of April to the time of writing (April 29th, 2021).
Readers may remember that this article isolated and emphasized the finding that the coal mine troubles in that province began on April 10th, 2021.
Referring back to the Chinese publication that covered the issue as it was occurring (originally), "The flooding took place around 6:10 pm Saturday in Hutubi County, Changji Prefecture of Xinjiang, when the coal mine was undergoing technical transformation, resulting in underground power outage and communication interruption."
- The article, published April 12th, 2021, captures the events of April 10th, 2021.
- It confirms that, at the time the article was published, there were observed electrical outages in the region as a result of the coal mine flooding.
- Thus, according to the logic put forth by Fortune Magazine, Crypto Briefing, and numerous other publications, we should expect the hashrate chart to show us some sort of dropoff in the hashrate from April 10th-12th moving forward (well before April 17th).
Evaluating the Raw Hash Rate Changes From April 10th - April 18th 2021
Below is a screenshot from the site, 'BitcoinVisuals', showing us the estimated hash rate for the network on April 10th, 2021:
In the chart above , we can see that the hash rate was approx. 181 Eh/s.
Shortly after that, there is a moderate decline in the hash rate - down to 167 Eh/s.
While this is a definite drop in hash rate, this ~7% (positive or negative) fluctuation falls well within normal variation for Bitcoin's hash rate these days.
Hash Rate Actually Spikes Post April 10th
Now let's compare the hash rate that day (April 14th) to the following day (April 15th):
As we can see in the photo above, the hash rate spiked to 198 Eh/s.
To be more specific with the exact peak hash rate, we decided to use a second source (fork.lol; that's the entire name of the website [fork.lol]).
'fork.lol' shows us that the peak hashrate on April 15th, 2021 actually hit 218 Eh/s (which would be a hashrate record for Bitcoin):
Curiously, the data that fork.lol has for April 14th's hashrate low is almost identical to what we observed on 'BitcoinVisuals' (this is why you have to use more than one source).
This represents a +53.5% jump in the hash rate after the coal mining flooding. With that being said, the only plausible explanation that preserves the theory that the coal mining flooding was correlated to the drop in hash rate must include an additional explanation that identifies a sudden loss of electricity as a result of the April 10th coal mine explosion that didn't take effect until April 16th (at the very earliest).
If the above stipulation sounds like an impossible requirement to fulfill, that's because it is.
Evaluating the Source of These Claims
Fortunately, many search engines possess features to curate results from a restricted date range (i.e., some point in the past through to another point just a bit later than the first point, but not present; i.e., April 9th - 18th).
Based on the hash rate charts that we observed in the previous section, there is no possible way that any publications could have drawn correlations between the hash rate and the coal mine flooding in that province until April 16th (at the very earliest).
After some cursory research, this hypothesis was confirmed.
Below is a link to an article in CoinTelegraph making the claim that the Bitcoin hash rate dropped specifically due to the outages in China.
The article states:
"According to an article on Chinese media outlet Wu Talk, the region of Xinjiang is currently experiencing a 'comprehensive power outage safety inspection'."
What's even more curious is the statement made in the final paragraph of the article:
"According to ycharts, the outages appear to have driven a roughly 2.2% drop in the Bitcoin network's combined hash rate in the past 24 hours, which has slid from 169.4 million terahashes per second, or TH/s, to 165.8 TH/s as of this writing."
Cross Referencing Another Source
On April 16th, 2021, the Nasdaq also published the report as a re-post originally sourced from 'Bitcoin Magazine' (by writer Oluwapelumi Adejumo).
This source also uses the same reference as the CoinTelegraph article, stating:
"Regional blackouts instituted in Northwest China may be the cause of a drop in hash rate from several China-based bitcoin mining operations, per local media outlet Wu Talk."
The article also references the same exact tweet by 'Wu Blockchain' as well in its statement, "According to a BTC.com screenshot shared by Wu Blockchain, the hash rate of several major bitcoin mining pools has dropped significantly."
Wu Blockchain / Wu-Talk is a Chinese Media Propaganda Account
Since this site has served as the primary (and only) source of intel on why the hash rate dropped as much as it did ~April 16th, 2021, the next logical move was to look into the legitimacy of this source.
The specific tweet that's continually referenced by major news outlets is the following one:
What isn't spoken about is the '3 Days Lucky' category, which provides some outrageous values for OKex as well as SlushPool and ViaBTC (but that's aside the point).
Also, the consistent promotion and advertisement of various product offerings and mining initiatives seems to be within this entity's purview as well (examples below):
Focusing on the content of the tweet, it asserts that there was a drop in hash rate because, "Northwest China is undergoing a complete blackout for safety inspections."
Picking Apart the Simple Logic in This Claim
If China were instituting a "complete blackout" for 'safety inspections', then we should've expected the impact on mining operations to be immediate while occurring all at once.
However, this is not what we saw. Instead, there was a descent in hash rate (gradual, before accelerating rapidly).
This makes no sense if there was a "complete blackout" where these miners are allegedly located. For this narrative to be true, the drop in hash rate we see in the previous tweet should be reflective of the full extent of damage / disruption caused by whatever was occurring in China.
Checking Bitcoin Cash's Blockchain
No other news outlet thought to do this, but one of the things that should've been checked is Bitcoin Cash's hash rate.
If you're reading this and wondering why, the reason is because Bitcoin Cash uses the exact same mining algorithm as Bitcoin. So often, miners will hop between the two chains (depending on their profitability at any given point in time), in order to maximize the overall profit of their mining operations.
Below is proof of this claim:
This picture here depicts 15 blocks on both Bitcoin (left) and Bitcoin Cash (right).
You'll notice that 6/7 of the identified mining pools successfully found a block on both chains during that time spam (some more than once).
Overall, there were only 4/15 blocks that were found on Bitcoin Cash, that were not found on Bitcoin.
There were only 5/15 blocks on Bitcoin that were not found by the same mining pool on Bitcoin Cash.
However, out of those blocks, those were all mined by firms that are known to mine on both blockchains. Consistently.
Fortunately, for us checking the hashrate shouldn't be too difficult.
For this calculation, we're going to refer to the website 'fork.lol' (longstanding website, recommended by Jameson Lopp on his website for any and all that are curious about its legitimacy).
Below is a chart showing the total hash rate for Bitcoin Cash on April 10th, 2021:
As we can see from the chart above, the overall hash rate on Bitcoin Cash on April 10th was approximately 1.65 Eh/s.
Now let's check out what it was on April 16th (the day the hash rate fell so dramatically).
Specifically, the hash rate for Bitcoin Cash on April 16th, 2021 was 4.08 Eh/s. This represents a hash rate increase of 245% on BItcoin Cash's network in less than a week - at a time when miners should've been impacted the most by the supposed power outages that were occurring throughout that week & weekend.
Evaluating the Total # of Blocks Produced During This Time Period
The 'average number of blocks found' on different time frames can yield a lot of information about what was really going on in light of the hashrate drop that we witnessed.
The reason for this will be explained thoroughly in the next section.
For reference, here are a few facts about Bitcoin as it pertains to mining:
- Bitcoin is designed to target 10-minute blocks
- This time frame is targeted by increasing / decreasing the difficulty for the entire network of finding a single block
- Mining itself is an overall probabilistic process, so the target is targeting a maximum % chance of finding the block at 10 minutes (63.5% specifically; the probabilistic targeting looks like a bell curve when the % rates are plotted out from 0 minutes to 20+ minutes)
- Based on this 10-minute target there are supposed to be 6 blocks/hour. Or 144 blocks produced per day. Or 1008 produced per week.
- The targeting algorithm bases its calculations on how long it took the protocol to find 2016 blocks. It simply takes the timestamp of that final block and the timestamp of the first block in that 2016 epoch and calculates how many seconds have transpired in that timeframe. This time is the only determinant used in calculating how the difficulty will be adjusted. In theory, one could produce 2015 blocks in less than a minute,then wait for two weeks before producing the final block - and the difficulty will remain unchanged (there is also a maximum to how much the difficulty will increase for a given epoch [400%+; announced by Satoshi on the forums )
The point made in the last item in the above list is verified by a statement made directly by Satoshi Nakamoto (re-published below for convenience):
Block Production Post-Coal Mine Flooding Incident
Once again, to reiterate the date of this coal mining explosion / flooding / incident was April 10th, 2021.
Average Number of Blocks Produced Per Hour on April 11th, 2021
Moving forward, things get interesting on April 13th, 2021.
On the chart above, we can see that the overall rate sunk down to a low of 2.5 blocks/hour.
Knowing that Bitcoin is targeting 10 minute blocks, the expected average is 6 blocks / hour. That means that Bitcoin was producing blocks at a ~60% reduced speed.
Hashrate is a Vague Estimation
There is no code in Bitcoin's source that allows for the calculation of distinct hashes (i.e., there is no recognition / acknowledgment of the concept of "hashes per second", this is a secondary calculation that's made independently by miners since they are mining these blocks non-interactively, only to provide a solution to the network later if they find the correct nonce value for the block that would extend the chain in a way that gives it the greatest Proof of Work).
Thus, a better approximation for the effective hash rate would be to perform the following calculation:
- Look at the overall difficulty target (this is a finite value; presented in hexadecimal form on the protocol)
- Calculate the exact number of hashes necessary to meet the difficulty's targeted time of 10-minutes (given the difficulty set for it; this will always be an approximation)
- Strictly go by the deviations in block time (for each block produced ) from that block to the one produced previously, then use that calculation to produce a closer approximate to the total hash rate '
- Plot this approximation using a 3-block moving average vs. using the hourly block count (or whatever algorithm these other sites are using). Make sure to be transparent about the values that are being obtained by your site's plotted chart.
The author has not yet computed the hash rate using this specific formula, so no affirmative comments can be made suggesting that this method of calculating the hashrate would provide a more insightful approximation of the hashrate than what's provided by whatever the current standardized method is (outlined in the Bitcoin Wiki for now).
Difficulty Adjustment Nearly Coincided With the Change in Hash Rate
One thing that has not been considered by any outlet covering this issue is the fact that Bitcoin's difficulty was due for an adjustment right around the same time the hash rate dropped so dramatically.
For those curious, 'btc.com' provide a chart showing all of the difficulty adjustments for Bitcoin (as well the exact time of each adjustment).
That chart is shown below:
As shown on the chart above, Bitcoin's difficulty adjusted upward on April 15th, 2021.
Let's revisit the tweet by 'Wu Blockchain' once more:
- This tweet was sent out at 12:41 a.m. UTC on April 16th (just a few short hours after the difficulty had adjusted).
- What's curious is that the author of the tweet is extremely familiar with mining operations (at least to the extent that they're able to provide exclusive late breaking news from the head of Poolin, Li Tianzhao, for example).
- Yet despite this familiarity, the fact that Bitcoin's difficulty had adjusted right before their tweet commenting on the drop in hashrate was published, the possibility that this macro-event may have had an impact on the hashrate (which its literally designed to do), wasn't even entertained or mentioned by the 'Wu Blockchain' account.
- Additionally, there was never any evidence given by the account attesting to specific outages in the 'North-Western' part of China as a result of "safety inspections" being conducted in that region specifically, on April 15th / 16th (any and all evidence of this claim must fall within these dates to be considered legitimate proofs substantiating the claims of 'Wu Blockchain').
More Than Likely the Exodus of Miners Triggered Others to Exit
This portion of our analysis requires to take a step back from the analytics and begin looking at the situation from the perspective of rational actors (as we have no reason to believe that there are any 'irrational' actors in this situation, so far).
The increase in difficulty for Bitcoin that week, albeit small, was still an accurate reflection of the network - thus, making it an appropriate difficulty target.
However, when miners began withdrawing hashrate from the network in droves, those miners that did remain found the task of mining Bitcoin to be significantly more difficult than they otherwise would have been (since the network was operating with reduced hashrate).
As a result, it is likely that the mass exodus that occurred right after the difficulty for Bitcoin adjusted caused even more miners to leave (in response to the rapidly increasing relative difficulty of mining blocks for any miners that decided to stay on the network).
No Justification For Correlating Sell-Off With Decreased Hash Rate
In fact, this more so an instance of 'chicken and egg', except in this case, its relatively obvious what came first.
As the posts, news articles, references, etc., in this article have shown, the drop in hash rate occurred before there was ANY noticeable change in the price.
Also, in the days leading up to this massive drop in hash rate, the price had notched a new all-time high (for Bitcoin).
However, over the course of the next four days (from that period), we saw the price of Bitcoin decrease by ~20-25% (from the top of the candle wick on the 13th to the bottom of the wick on the 17th).
The screenshot below confirms this:
Tying it All Together
When we put the bigger picture together, it is clearly the following happened:
- The Bitcoin difficulty adjusted (April 15th, 2021)
- This led to an initial exodus of miners from Bitcoin (April 16th-17th)
- Price of Bitcoin undergoes a nasty sell-off amid a larger, industry-wide sell off that wipes >$500B (half trillion) in value off of the books (alongside several billion dollars in cascading liquidations)
Are we surprised that the hashrate fell as low as it did after this confluence of events transpired?
Considering this logically, its not far-fetched to suggest that the miners found it hard to justify their operations (financially) in the immediate term when considering:
- Miners that remained on the protocol would have to spend more money on deployed mining equipment (hashrate) to make up for the lost hash rate on the network to ensure their block discovery estimations would remain consistent.
- Bitcoin was worth 15-20% less than it was right before the difficulty adjustment
Is it possible that this difficulty adjustment was what was ultimately responsible for the downward adjustment in hashrate for Bitcoin? (its worth noting that the network provided its highest average hashrate ever [168.79 Eh/s] during this period [April 1st - 15th, 2021]).
Discrepancy in the Information Being Published by "Wu Blockchain"
Going back to the Twitter profile of the original "source", we can see that the account published this post recently:
The post (published April 29th, 2021), states:
"Security inspections in northwestern China have basically ended, and the bit coin hashrate has begun to recover. This inspection is not aimed at Bitcoin mining, but on the overall safety of electricity."
Before even probing into the graphic attached to the tweet, its worth mentioning that no country with any amount of resource should ever find themselves in a situation where they need to cut off power just to perform garden variety (safety inspections; worth noting that there's no proof / evidence that these inspections ever took place).
Debunking This Narrative With On-Chain Metrics
The accompanying picture that was published with the tweet by 'Wu Blockchain' gives the impression that the recovery of Bitcoin's hashrate occurred on April 20th / 21st.
Here it is re-published below for convenience:
In reality, the hashrate had recovered right after the massive dip that occurred on April 17th, 2021.
As a refresher, the hashrate for Bitcoin fell down as low as 80 Eh/s:
However, by the very next day, the hashrate spiked up to 175 Eh/s (really):
This indicates a hash rate increase of approximately +118.75% in under 24 hours.
This fact completely eviscerates arguments made that the coal mine flooding led to widespread outages (for a sustained, lengthy period of time). In reality, there was only a period of 24-48 hours where we observed a noticeable decrease in hash rate and the immediate bounce back was just as impressive (if not even more so).
This entire analysis was published with the intent of shedding truth on what's going on in the blockchain markets (behind the scenes).
Since there were no other publications or resources on planet earth willing to conduct the same level of due diligence / examination into this issue it felt incumbent upon Librehash to be the one organization that was willing to do so.
With that being said, the author(s) of this publication are far from infallible. So, if there's any information / details that you believe we've missed, please do not hesitate to reach out to us on social media (Twitter / Telegram).