User blog comment:HeresyHorus/Critical Strike Chance/@comment-1330314-20140415041810

I think there are several reasons for why a constant AA damage multiplier would be a good update to crit, but here are my thoughts on your points:
 * There used to be an evasion mechanic called dodge, but it was removed because the random tankiness it gave was unreliable and frustrating to fight against. This is a similar mechanic, only directed towards a single class of champions. Not only would this not be a good idea due to its extreme unreliability (it's a chance to block damage from an attack that is itself based on chance), but it's also not a great mechanic in terms of balance: there is pretty much only one class of character to whom crits are a central mechanic, and what you're proposing thus only affects them. ADCs in the game are pretty alright as is, and this would warrant a buff for them, but assuming they're overpowered, having an item with an effect that is completely useless on four fifths of the enemy team is not conducive to intelligent build decisions. It also forces the issue where your item is either powerful enough to warrant purchase, and ends up being overpoweringly strong, or soft enough so as to not shut marksmen down completely and thus not a good purchase in any situation. Take Thornmail, for example: its armor and on-hit effect are super-strong against ADCs (though not just them), but it's still an extremely niche purchase because it was toned down until it wasn't as much of a hard counter to autoattack-based champions, and thus became a terribly inefficient item in all but a few situations.
 * Making crit chance more complicated than it already is (and it's already a mechanic with visibility problems) would probably create more problems than it would solve. From what you're suggesting you want crits to be more effective against tanks and bruisers (which makes sense, since these are the roles marksmen are designed to do well against) and less against squishies, but that sounds like a problem with the damage output of marksmen in general and not just crits. I do agree that marksmen can deal more damage to squishies than they're supposed to, but I think the problem comes from, which is super-effective against squishies despite being a counter to armor stacking. The effect you're proposing sounds like something you'd have on an item, though, so you could suggest that instead.
 * Likewise, having a separate system for champions and non-champions sounds needlessly complicated, with the added problem that this is just a straight nerf to crit with a few caveats. Junglers don't build crit, and crit isn't reliable enough to benefit last hitting, so you might as well retune the pseudorandom number generator for everyone in one go. I also think this is a case where you're focusing too much on the outliers (i.e. 3-4 crits in a row), when the lack of such outliers would be even more unlikely. With a crit chance of 50%, the chance of landing 3 crits in a row is 12.5%. With a PRD the chance would likely be lowered, but then the likelihood of that event happening also increases with the frequency of hits the marksman throws at you, so odds are you're going to take 3 crits in a row at some point.

While you may not like the idea, I think the idea of a damage enhancer to autoattacks would be a better idea, for multiple reasons:
 * It's actually possible to balance: One of the complaints I see when this idea gets thrown out is that turning X% crit chance into X% increased autoattack damage would make crit overpowered, but I think that's actually a non-issue: if a 1-to-1 conversion is too strong, then that just means crit itself is too strong, and its excess power is being obfuscated by unreliability. A constant autoattack multiplier with a 1:1 ratio to crit chance leads to exactly the same DPS, so marksmen would not end up dealing additional damage.
 * It can only affect builds in a healthy way: Another issue that gets thrown out is that turning crit into an AA amp would lead to AD assassins and bruisers picking as part of their build. I personally don't find this to be a bad thing either: first off, crit is far less valuable on most non-marksmen when compared to raw AD, since it doesn't interact with the AD ratios most bruisers and assassins have, but even if its reliability makes crit items more widely purchased it may actually be a good thing. Crit right now is a terrible stat in terms of itemization: unless you build a lot of it, you're not going to make much use of it. It's particularly terrible on AD caster assassins, who typically have only a short window to take down their target before they win or die, and so can't really rely on a random-chance mechanic to get out extra damage at just the right moment. This also ties into balancing: if making crit non-random makes all AD champions too powerful in general, then that would boil down to crit itself being too strong and requiring a nerf. However, AD casters also suffer from the problem where they're so bursty they can't rely much on their autoattacks, despite being AD, to the point where they're almost divorced from them save for one ability on their skillset. This feels like a missed opportunity, and giving burst champions the ability to deal more autoattack-based damage (as with  and  in their recent reworks) would make for healthier gameplay.
 * It's ultimately the simplest way of making all crit-based gameplay fair: This might sound stupidly obvious, but changing from chance-based critical hits to a constant multiplier changes nothing in terms of gameplay, and that's a good thing: it gives the same autoattack-based DPS minus the random factor, which makes it both better for the players itemizing for crit (more reliable autoattack damage -> better power evaluation and last hitting) and players itemizing against marksmen (again, more accurate power evaluation). It would also be the simplest way to clean up crit-based abilities and mechanics, which often fall in the scenario where they're either sub-par (i.e. not critting) or far too strong (i.e. an early game Parrrley crit), with nothing in-between. The biggest offender to me is, whose early game is one of the most unstable out there: either his passive doesn't give him the lucky crit he needs and he either dies or gets bullied out of his lane, either he cheeses his opponent with sheer luck and snowballs super-hard. The fact that his mobility is also tied to his critical strike chance makes his early game even rockier.