So it seems the conversation I was having in this thread is over. I'd still like to organize and sum up my thoughts on parrying in this game, so here we go.
First, I want to give a clear presentation of the concept of
hit consequence lag, because I'm going to be bringing it up later in this post, but also because
it applies to every attack in the game, not just parries.
There are three elements to a hit: visual confirmation (a visual effect), damage and, if poise is broken, reaction (flinch/stun/knockdown), which always comes last.
It's impossible for the game to have all three of these occur on the same frame, meaning that there will always be at least one frame between the actual hit and any reaction from the enemy, even if they've already taken damage and emitted some kind of hit effect. Visual confirmation and damage can come in either order, or can be simultaneous with each other or the reaction, but there is always at least one frame separating one of them from the reaction. For example:
Here's the frame where an attack connects. In this case, both the visual confirmation (the blood) and the damage appear on the same frame. Then the reaction happens on the next frame:
It is however possible for all of them to be spaced out, adding more than one extra frame to the whole process. Here's the contact frame of a hit where only damage is applied on time:
Then comes the visual confirmation on the next frame:
Then finally the reaction on the next frame, making it 2 frames between the hit connecting and the reaction:
It's even possible for any of these steps to take more than one frame each, creating even greater latency, but you probably get the point. This is a deep problem for a game where judging the timing of your actions is critical, and it's especially harmful to parrying, which needs serious precision by definition. Even if you don't care about parrying, you should care about this fundamental flaw. These observations may well have different causes; it seems plausible the ways the game calculates when damage taken from a hit is applied, when hit effects appear and when enemies should react to hits are all different processes, but I'm putting them under a single umbrella for clarity's sake and because there's no way for me to look under the hood.
Now that I've established that, I'll present my main thesis regarding parrying.
Tying the idea of parrying to guns was a catastrophic mistake. I think every flaw parrying has is tied to this, except those caused by hit consequence lag. Let's look at those flaws.
Parrying is slower than it should be
The fastest a parry in Bloodborne can possibly be executed from a neutral position (parrying directly after dodging has a shorter warmup period) is 11 frames: 9 for the warmup animation (using a spreadshot weapon; bizarrely pistols have 10 frames of warmup despite their description, they just have less cooldown so you can fire again faster), 1 for the bullet emission, and 1 frame to account for hit consequence lag in a best case scenario. Shooting at absolute point blank can eliminate travel time; the bullet makes contact on the same frame it's emitted. However, this is very rare and pretty much out of the player's hands, which makes this 11 frame ideal even less likely. A couple of frames of travel time is common even at what seems to be point blank. So if you're in an enemy's attack range and want to parry the attack, you need to input the parry at least 11 frames before their attack connects to have a total success in the best scenario. This is a very difficult task in itself for many attacks.
Let's say an enemy has an attack that has a 20 frame windup, two thirds of a second. That sounds reasonable, right? But take away that parry warmup period and you've got 9 frames, less than a third of a second to see the attack, distinguish it from other attacks that might be visually similar, and execute the parry. That can be doable with practice, but for most players it's going to be an intense pressure to rely on pure twitch reaction. Reacting that way is something the game (and any action game in a single player context) should pretty much never encourage for obvious reasons. And of course, there are tons of attacks that have more than 11 frames of windup but less than 20, making them possible but unfeasible to parry while in range. Making attacks too fast to parry may be a way of differentiating them as attacks that just aren't intended to be parried, but I'd be shocked if that was the intent. This might work if attacks were telegraphed very strongly from the very beginning of their animations, but this is very often not the case in Bloodborne, and it's probably just a bad idea regardless; just making parrying faster would maintain enemy attack diversity with no downsides. The long warmup time of parries is their biggest flaw, one that pushes the entire concept of parrying toward irrelevancy.
So how's this related to the fact that parrying is tied to guns? I suspect the reason parrying was given such a long windup time was to give other players time to react and dodge gunshots. If this is the case, I wonder why the concept of parrying wasn't heavily redesigned to compensate.
Parrying at range isn't parrying
At its core, parrying is an escalation of the standard of attack avoidance. Instead of dodging to avoid damage and gain advantages that aren't intrinsically connected to the act of dodging itself, you take on greater risk of being hit and gain a big direct advantage (the visceral attack) if you succeed. But what if you could avoid this risk and still get the benefit? This is what parrying at range allows. It's very possible to parry an attack while being completely outside of its range, meaning the player is at no risk of damage whatsoever. The balance of risk and reward that defines parrying is shattered in this scenario. It's a scenario that has no reason to exist with one big exception as far as I can see: when an opponent heals.
Allowing healing to be parried, and remotely, is a very clever way of stopping players from being able to break off from the fight and recover at their leisure, and it would likely work well if AI enemies were capable of healing too. It also counterbalances the speed healing takes place at in Bloodborne. It's the one case where parrying at range really makes sense, but it would've been better to make it a special case than to always have bullets be the parry delivery method.
Anyway, parrying healing is fundamentally different from parrying attacks, and there's no excuse for allowing attacks to be parried at range. Instead of the simple, great system of intense risk and reward in Demon's and Dark Souls we've got a weird wishy-washy thing that rewards players for backing off and staying out of danger. Also, parrying at range eliminates positioning as a factor; where before you had to be right in front of the enemy now you can be in an arbitrary position, shoot for the parry and run in for the riposte. That's lame too.
Parries resulting in trades is a failure of design
Since the parry delivery method is physically separated from the player themselves, and because of the issues with hit consequence lag, a scenario that didn't exist in any Souls game becomes possible when attempting to parry: a trade. This is when a parry succeeds but the parried attack also succeeds. This is only possible when a player attempts to parry while in the range of the attack. In the Souls games, a parry that was timed slightly off or came from the wrong position resulted in the parrying party taking partial damage and both sides maintaining poise. This scenario has been eliminated in Bloodborne, and now near misses result in trades. This was probably once again necessary to keep the idea of using a gun intact; something between total success and total failure doesn't make sense when the delivery method is disconnected. Also, in a bizarre twist, parrying perfectly, having a parry become active on the exact frame the enemy attack makes contact, will always result in a trade because of hit consequence lag.
Trading is a bad thing for a few reasons. First, it's confusing. Having players suffer the full consequences of a hit despite parrying that hit successfully is some seriously mixed messaging. I traded and took damage, so I failed, but I also succeeded? Huh? But even weirder is how trading can also develop into a total success or failure state because of circumstances. If the player trades with a hit that doesn't kill them outright or stun them for long enough for the parried party to recover, the regain system will allow them to take all the health they just lost back with the visceral attack, meaning the trade is without consequence unless the attack also caused a status effect or something. On the other hand, if one of the aforementioned consequences was attached to the attack, they can lose out, sometimes brutally. Trading is incredibly generous and ridiculously harsh by turns, and above all it simply feels unfair. I find it unbelievable that parrying was designed with these bizarre outcomes in mind, and again I wonder how there weren't big changes to avoid them.
---
These are the major problems with parrying that I think exist, and they're all possible because the idea of parrying is tied to the use of guns, in some way or another. There are probably even more details I'm glossing over. I was really excited by the idea of parrying being tied to guns before the game was released, since the change in metaphor seemed to give the designers freer license to make parrying useful against more enemies and types of attacks, but it seems to have backfired terribly. If you enjoyed parrying in the Souls games, that paradigm is as good as dead in Bloodborne, and the thing that replaces it is broadly inferior. The new parrying scheme can work well for you if you're willing to constantly retreat from the enemy and poke at them from outside their attack range, but I feel that's in direct conflict with the fundamental ideals of this game, where fast paced but controlled close combat is the goal. I'm very disappointed with the end result.
If anyone has data that disproves any of my claims, please go ahead and post it. Dissecting this stuff is both enjoyable and enlightening, it's the kind of discussion I really like to see.
