Why not tie difficulty to number of transactions

[bit2.it]

sure, you might have to wait for more blocks for confirmation, but the length of time should stay the same as the old way.  Any attack on the network is still going to have to come up against a huge amount of hash.  Am I missing something?  Just make blocks complete faster.  It shouldn't be any riskier than before the difficulty was increased in the past.  The hashrate will still be over 51% and the chances of some lone double spender or hijacker still won't be able to overpower it.  Coins are found faster, limiting the power expended a little and we might reach the day the network moves to pure transaction fees faster but if it can't survive that we are just putting off the inevitable.

Please someone explain why this wouldn't work.

I've been thinking of network latency, people flooding transactions to mine blocks faster, and other stuff.  I don't know the reason why though, I guess I'm too nooblet.

[achow101]

First of all, that would mean that the inflation rate is not stable, it would depend on the number of transactions. Secondly, it becomes extremely difficult to predict roughly when a certain blockchain event will occur, like the halvings. Next, this is something that is very easily gamed. A miner could make and broadcast a ton of transactions so that they can lower the difficulty and thus mine more blocks and earn more money. Faster blocks also means more orphans.

Lastly, and most importantly, the number of unconfirmed transactions and what the unconfirmed transactions are varies from node to node. There is no way to fix the difficulty to the number of unconfirmed transactions because every node would then calculate the difficulty differently since not all nodes would have the exact same transactions in the mempool. If you were to tie it to the size of the block, then that is extremely easily gamed by a malicious miner because then they could risk no money of their own, and just make massive blocks to decrease the difficulty.

[DannyHamilton]

In addition to all the things achow101 has said, making the change would also require a hard fork. That means it it would require an overwhelming majority of all USERS (Users, not just miners. That's nodes, merchants, exchanges, wallet creators, etc) to ALL agree to run software with the new changes.  We can't even get a simple 50% majority to agree on a simple blocksize change.  How do you expect to get 80%, or 90%, or 95% to all agree to something that so drastically changes the nature of Bitcoin?

[bit2.it]

Thanks for taking the time to reply, explanations were quite succinct.

[BillyBobZorton]

I would say this is at the same level as wanting to change the 21 million limit... there will never be consensus.
Hopefully some things will never change, including the BUcoin disaster.