You need to know these "little-known facts" about mining.

As a journalist who has been following the mining industry for a long time, I have found that many people like to express their opinions and inferences about Bitcoin mining when the market is extreme and the industry is turbulent, causing investors to spread and panic.

But in fact, a considerable part of this is derived from wrong concepts. Many people who think they understand the principles of mining or have had contact with the mining industry actually have only a superficial understanding of some basic concepts of mining, or even have misunderstandings.

For example, on March 12, when BTC plummeted, the Bitcoin network did not produce a block for an hour. Many people (including me and some friends in the mining circle) thought that it was caused by the shutdown of mining machines and the sharp drop in computing power, and therefore worried that a "mining accident" was on the way.

But in fact, it has often happened before that the network has not produced blocks for a long time. This is purely a matter of luck under coincidence. No mining machine collided with a hash result (that meets the difficulty) within 1 hour.

There are many similar examples, from miners to coin holders, who have made many over-interpretations or even distorted interpretations due to improper understanding of the meaning of many data. If they use this to guide transactions, it is obviously very risky. In the opinion of professionals, this can be completely avoided.

Therefore, Odaily Planet Daily and Poolin Coin Printing Pool have carefully compiled a list of important but often misunderstood mining "cold knowledge" for everyone. Come and see how many of them you know.

Is “network-wide computing difficulty” a real data?

On the morning of March 18, four days after the Bitcoin crash, everyone was afraid of the coming of a "mining disaster".

A piece of news that "Bitcoin computing power dropped by about 40% this month, falling below 100E" scared many people.

The statistic was collected by Skew Markets, a well-known data analysis agency and Twitter celebrity. The data source used was BitinfoCharts, a professional data website. It seems that there should be nothing wrong with it. However, experts can tell at a glance that the data is outrageous.

Going back to the data source BitinfoCharts, by subtracting and dividing the high point (133.29) and low point (95.96) of the month, we get the maximum drop of 28% in the month. Therefore, the figure of 40% is "far from the truth".

If Skew did not make a mistake in his calculations and concluded that Bitcoin has fallen 28% this month, does that represent the real situation? In other words, is “the total network computing power on a certain day” a reliable data?

The fact is - you can’t believe it, because this “total network computing power” is derived from a formula, it is a theoretical value, not real-time monitoring data.

The "difficulty calculation formula" commonly used in the industry is as follows:

In fact, there is no record of how many mining machines are running on the entire network, and miners cannot find any authoritative statistical method.

According to the formula, the length of the block time determines the theoretical computing power displayed by the block browser. But we must know that Bitcoin's hash collision block generation is a probabilistic problem. The same computing power may produce a result in 3 minutes or half an hour. Therefore, the short-term theoretical computing power cannot represent the actual increase or decrease in computing power.

For example, normally 144 blocks should be produced every day, but in fact the number of blocks produced in a day is less than 144. Whether it is caused by the rapid decline of computing power, we need to look at it over a longer period of time. If the number of blocks produced every day is decreasing over a long period of time, we can further determine that it is due to the decline of computing power.

However, people cannot make instant decisions based on data from a month or even half a year ago. Therefore, mining practitioners generally use the average computing power of the past seven days as the computing power for a certain period of time, taking into account both accuracy and efficiency.

Knowing this, you will just laugh it off when you see headlines such as "sudden drop" or "flash crash" in Bitcoin computing power.

Bitcoin hashrate has plummeted too many times

Isn’t the difficulty adjustment cycle once every 14 days? Why can it be extended?

On the evening of March 12, regarding the impact of the Bitcoin crash on the mining industry, Pan Zhibiao, founder of Biyin, made a bold guess: There are still 11 days until the next difficulty adjustment, but if the computing power drops by 30% during this period, the difficulty adjustment cycle will be postponed by 5 days to 16 days. Therefore, miners must generate at least half a month's cash flow (to wait for the difficulty of the entire network to drop to a level suitable for mining).

Many people may be confused when seeing this. As we all know, Bitcoin adjusts its difficulty every 14 days, so how can it be postponed?

The reason is actually simple. The purpose of adjusting the difficulty of the Bitcoin network is to adjust the speed of block production to an average of 1 block every 10 minutes, with every 2016 blocks as a cycle.

The 14-day adjustment that we are familiar with is based on the fact that the computing power of the entire network is relatively stable in the short term and the block time of 10 minutes does not change much.

As mentioned above, if the mining computing power really drops rapidly and the computing power of the mining machine is not enough to find the target hash value within the average time, then the block time will also be extended, which will affect the difficulty adjustment cycle.

The price has fallen below the shutdown price of some mining machines, so why has the computing power not decreased?

To understand this problem, we need to know how the shutdown coin price is calculated.

The shutdown price refers to the point when the coin price drops to the point where the income of a certain mining machine equals the cost (that is, the net income is 0).

If the price of the currency is lower than this point, it will be unprofitable to mine with this mining machine. Therefore, from an economic point of view, most miners should temporarily shut down their machines at this point and wait and see the market outlook.

The sharp drop on March 12-13 quickly broke through the shutdown prices of many old mining machines, and the high-performance new generation mining machines with 45-65W/T were also operating with tiny profits of a few cents to a few dollars.

However, we have observed that the computing power of Bitcoin has not dropped significantly in the past few days. According to estimates by many senior industry insiders, 50E of the 100E computing power at the end of 2019 came from old mining machines at the 15TH/s level. Obviously, these mining machines have not been completely removed from the shelves until now.

Are miners mining at a loss?

According to the calculation formula of the shutdown price, mining should be stopped when "income of the mining machine - cost (mainly electricity cost) = 0". At the same time and the same coin price, different electricity costs will affect the shutdown price. The lower the electricity cost, the more the miner can withstand the decline in coin price, which means the shutdown price is lower.

Therefore, when we say that the price has fallen below the shutdown price, we are using the general electricity price level in the industry, but it does not rule out that miners with cheap electricity resources can still run old mining machines.

If we take the electricity price of 0.24 yuan/kWh and the currency price of 5,000 US dollars as an example, turning on the Antminer S9 today will still bring in a few cents of profit.

Even if miners want to shut down their machines, it is not easy to do so. For example, many mining farms have reported their electricity consumption to relevant departments, so it is inconvenient to shut down suddenly. For customers who cannot afford the electricity bills, both parties may choose to temporarily rent the machines to the mining farms to continue operating. For mining farm owners, their electricity costs will be much cheaper than those of customers, and running old mining machines will not necessarily be a loss.

At the same time, many people in the circle of friends like to say that "if the price of BTC drops to xxxx, all S9s will be squeezed out", which is also a very untenable statement. As mentioned above, old S9-level mining machines still account for nearly half of the total network computing power. If these mining machines are shut down, the so-called "shutdown price" will also drop significantly, so everyone will not shut down easily, which becomes a "prisoner's dilemma". In reality, it is not so easy for all old mining machines to stop mining.

In addition, many miners now choose to use financial tools to survive the downturn in the industry. A short-term drop below the "shutdown price" does not necessarily affect the computing power.

Just like playing games and grabbing red envelopes, mining can also use cheats. Many mining machines are using cheats.

At the end of 2018, the mining community was very excited about a mining optimization algorithm called Asicboost. As the name suggests, the optimization algorithm can improve mining efficiency.

According to the calculation data of the Rawpool mining pool, the mining machine that adds this algorithm in the firmware can reduce power consumption by more than 12.5%. The logic of reducing power consumption is to use the characteristics of the block data format to cleverly reduce the number of times the mining machine receives data, thereby reducing operating power consumption.

According to public information, Asicboost was first proposed by two researchers, Timo Hanke and Sergio Lerner, in 2015 and subsequently patented. Perhaps due to patent restrictions, this technology was not widely used in the early days.

Until October 2018, when the price of coins was falling and many mining machines were on the verge of shutdown, Bitmain took the lead in announcing that it would add Asicboost to Ant S9, T9+ and other mining machine models.

Once the news was released, major mining pools followed suit and announced that they were "compatible with AsicBoost mining."

According to the latest data from the Asicboost.dance website, the usage rate of Asicboost has been rising steadily since then, from a penetration rate of less than 5% to nearly 70% now, which means that nearly 3/4 of the mining machines on the entire network are "using external plug-ins."

When transactions are congested, can the mining pool speed up a certain transaction in a targeted manner?

The sharp drop on March 12 and 13 caused many contract users who did not have time to cover their positions to be liquidated. At the same time, this was also another big test of the transaction speed of the Bitcoin network. Data shows that from March 12 to 14, the total number of unconfirmed Bitcoin transactions increased sharply.

Image from: Johoe's Bitcoin Mempool Statistics

Basically, this situation occurs every time the market changes.

At this point, if users do not want their transactions to be blocked for too long, they can only increase the handling fee to increase the probability of being packaged by miners first. Of course, the speed cannot be guaranteed.

In fact, directly "tipping" the mining pool can also speed up the process.

At present, many mining pools, including Biyin, have launched the "transaction acceleration" service. For users who purchase this service, the mining pool will first put the transaction initiated by them into the queue of the next block packaging transaction. If the next block is successfully produced by the mining pool, the transaction can be credited immediately. If the mining pool fails to successfully find the next block, it will increase the handling fee for the transaction, so that other mining pools will also prioritize the packaging of the transaction, thereby completing the transaction acceleration.

It is understood that for users who purchase this service, the average transaction arrival time is 25 minutes, which is equivalent to the time required for ordinary transactions when there is no congestion on the chain. When the transaction volume surges and the network is congested, this service also provides an additional option.