The mysterious flash crash of bitcoin’s hashrate explained

The mysterious flash crash of bitcoin’s hashrate explained

Just a week ago we wrote about the hashrate of bitcoin setting new records. We were not alone because all the major online platforms widely reported the event. Later there was all sorts of press coverage regarding a flash crash of bitcoin’s hashrate. And indeed, on Monday, September 23rd, the hashrate did dip more than 30%. But what does this really tell us?


Image: the flash crash of Bitcoin’s hashrate (source:

The figure above shows how the hashrate significantly dropped for a brief moment on September 23rd. The “crash” caused panic and raised questions regarding the safety of the network and potential issues with the Proof-of-Work model. Why so? Because the graph above basically suggests that 30% of the mining rigs went offline and therefore the Bitcoin network’s security was in jeopardy and vulnerable to attacks! 

Even the Bitcoin whitepaper addressed this 30%. Emin Gün Sirer, a professor at Cornell University noted on Twitter that Satoshi suggested six confirmations per transaction for a reason. It was based on the assumption that a potential attacker would have 30% of the hashrate at most. 

The mystery explained

So what happened and is the dropping hashrate a cause for alarm? To cut right to the chase, the event is nothing to be worried about and the frightening news coverage can be considered FUD (Fear, Uncertainty, and Doubt). FUD that got even more ground due to the declining bitcoin (BTC) price. 

So, about hashrate. The hash is a alphanumeric string with a fixed length that is produced while mining blocks. It’s what miners do; they perform complex calculations to define the correct hash. This unique hash gets assigned to the next block on the blockchain. The amount of hashes a miner can spit out per second is what we call hashrate.

The hashrate is a performance indicator that is widely used. However, it is not an indicator that can be exactly determined. It can only be estimated by looking at historical hashrates or elapsed time between the creation of two consecutive blocks. So imagine a set of blocks that are mined a lot slower that the previous set of blocks. Both sets will result in a found block, but the first set will take longer to produce and will have a byproduct of hashes that did not result in finding a block. As Christopher Bendiksen, head of research at CoinShares, explains:

#bitcoin hashrate is not a known measure. It can only estimated from previous block intervals. Block times are Poisson distributed which sometimes cause large variance in block interval. If several slow blocks happen in a row, it makes these estimates spike down.

In other words, hashrate estimation graphs like the one above spike hard up and down, which is inherent to the Proof-of-Work (PoW) algorithm of Bitcoin. In fact, the hashrate dropped drastically before in november 2017. Nothing new under the sun and nothing to worry about!