I think MEV is a confusing term.
The definition on https://t.co/6cEQ5IexH1[0] is:
> Maximal extractable value (MEV) refers to the maximum value that can be extracted from block production in excess of the standard block reward and gas fees by including, excluding, and changing the order of transactions in a block.
The problem with this definition is that it doesn't capture the thing we actually care about, the thing that makes MEV bad.
The thing that can make MEV bad, and, as far as I
understand, the only part of MEV we should care about
vis-à-vis Bitcoin's decentraliation, is MEV which requires
large fixed costs which don't scale with hashrate.
When a miner produces hashrate, fixed costs are those which don't scale with the number of units of hashrate, creating economies of scale, and thus encourage miners to merge. Let's say there's an MEV opportunity which generates 1 BTC per block, with a fixed cost of 1 BTC to take advantage of.
Looking at a period of 10 blocks and two miners each with 50% of the hash rate, each miner expects to make 4 BTC, i.e., 5 blocks x 1 BTC - 1 BTC = 4 BTC, for a total profit of 8 BTC across both miners.
If they merged, they would expect to make 9 BTC, i.e., 10 blocks * 1 BTC - 1 BTC = 9 BTC.
So, there is centralizaiton pressure, if the miners merge, they'll make more money.
But the definition doesn't capture this, so there can be huge "MEV opportunities" that we shouldn't really care about, because they don't introduce fixed costs, or the fixed costs are small enough that they don't matter.
For example, imagine the same example above, but the MEV opportunity takes a fixed cost of 0.1 BTC per block to take advantage of.
Looking at a period of 10 blocks, and two miners, each with 50% of the hash rate, each miner expects to make 4 BTC, i.e., 5 blocks x (1 BTC - 0.1) BTC = 4.5 BTC, for a total profit of 9 BTC across both miners.
If they merged, they would expect to make, again, 9 BTC, i.e., 10 blocks * (1 BTC - 0.1) BTC = 9 BTC.
Is this MEV? Yeah, I guess, under the definition of MEV, but it doesn't introduce centralization pressure. And note, this is an unrealistic example, almost everything is going to have fixed setup costs. But if those setup costs are very small, they shouldn't make much of a difference.
So the problem with the existing MEV definition is that you can hear "This is a huge MEV opportunity." without knowing whether or not you should actually care about it.
Is there a term or definition X which effectively captures the bad part of MEV? So if you hear "This creates a huge X opportunity.", you could know that it's a bad thing, instead of a possibly good thing that just increases miner profits without introducing centralization pressure.
[0] https://t.co/ViBJKocSTV