Optimizing a crit fish blastlock?

[Vespa]

Hey there! Recently I've been having some fun building my backup character for a friend's campaign and was looking for some advice. The basic premise was to play a 5 raven queen tomelock/ 3 shadow sorcerer that would be crit based, as I haven't really ever seen a crit based caster build and I thought it might be fun. The basic idea would be to take find familiar(Book of ancient secrets) + the raven familiar(both taking the help action as needed) + the shadow sorcerer's discount darkness + devil's sight combo to have some reliable sources of advantage. I'd be taking elven accuracy, so if I was quickening an eldritch blast at the level i'd be starting at (8) that would be 4 rolls with what is essentially super advantage from elven accuracy.

This gets a bit ridiculous at warlock 6 because of a homebrew rule my friend has implemented. The basic idea is that when you reach the second rung of benefits from your subclass you can also take the first rung from another subclass. I was planning on taking hexblade's curse for not only the additional damage but the ability to crit on 19s. If my math is right (Which it probably isn't) that's a 10% chance to crit per roll, three rolls per EB beam at that level with advantage + elven accuracy, for what is essentially a 3̶0̶%̶ (about 27.1%, thank you divisiblebyzero for the correction)chance to crit per EB beam.

I'm wondering if there's anything I've got wrong here, or if I'm missing anything that could make this build work better? I'm especially lost on how I should distribute levels. After warlock 8 should I just keep going with warlock or should I drop it and start leveling sorcerer for more sorc points? Is this even viable???

[DivisibleByZero]

that's a 10% chance to crit per roll, three rolls per EB beam at that level with advantage + elven accuracy, for what is essentially a 30% chance to crit per EB beam.

That isn't how probabilities work.
It's close, but not 30% Closer to 25% if we're rounding, at about 27.1%
The easiest way to figure out the exact probability is to figure out what the odds are of it *not* happening, and multiply each result. So in this case, it's 0.9*0.9*0.9=0.729, leaving 0.271 left, which is your answer.

[Vespa]

That isn't how probabilities work.
It's close, but not 30% Closer to 25% if we're rounding, at about 27.1%

Ah, right. I'm still pretty happy with that probability though, I think.

[malachi]

3 dice on attack rolls = 14.3% chance to crit, if you crit on only a nat 20.

3 dice on attack rolls = 27.1% chance to crit on a 19-20.

[PeteNutButter]

I had a build idea that I wanted to try, which would be AL legal. It's just a kobold champion that dips warlock...
Kobold gives a fairly steady use of advantage when out of direct sunlight. Can cast darkness when needed.

Champion 4/Tomelock 4/Sorcerer X.

All it does is spam EB and quickened EB, probably with hex, and maybe Scorching Ray a little later. Never played it though, because it seemed boring.