No, that's not what's happening

Something I wrote on this matter in another thread:

Actually, in addition to potential "latency" as usually considered as internet problem there is the additional constrict of flash rendering speed. But barring all those, that is to say you got a good machine that can render optimally and internet that have no delay, there is still a problem: chase. Specifically, the server side does not wait on a chase attack more than it does if the attack have no chase. So as soon as a reasonably long chase occur, client side fall behind by the amount of time it took for the chase to resolve. This would require some kind of code fix to resolve and certainly should be a priority. For the moment however, you can somewhat play around this by knowing the fact that the delay is not random but rather deterministic. That said, no amount of software can account for actual latency nor should a player be expected to wait on another player because the other player lags. So as far as delay caused by actual lag goes, aside from optimization to reduce lag, nothing should be done.

This really need to go into a FAQ somewhere... This is like the 10th time I'm explaining this.

I mean I agree there is a problem, but can we please stop assuming what is actually happening?

No it didn't pop up just a while ago, the problem have been around since at least 2.0. I discovered it on 2.0 due to the new talent which requires me to use some mysteries mid-turn. Before that, most of my cast are up-front, so I can't be sure. So with the newly discovered lag, I tried a number of things to figure out where it actually came from, and it was discovered that chase is the root cause. Similar thing might have happened to you, your previous team or opponent might have had no need to have mid-turn cast such that even thou it's there, you didn't notice it.

Don't assume a problem didn't exist because you didn't personally notice it.

Thou to be fair, I have minimal amount of experience using ONLY animated mystery. Because most of my mystery and that of my opponent leads to chase afterwards. And if you intersect with the need to cast mystery in middle of turn, the amount of experience is small on that end. The server may or may not wait an appropriate amount of time for animated mystery w/o chase trigger.
I don't know if the problem exists in CN/DE/RU, but that's hardly relevant.

On the other hand, If you have random lags w/o chase, then that's probably you(again, not sure if animated mystery might be a cause, but if you have neither that nor chase...), either due to internet or flash.

And then there is the interesting bit about CPU spike.

Guess what, you SHOULD have CPU spike in combat. Because the battle is hardly "even" in terms of resource consumption. Normal attack that leads nowhere are the cheapest while flashy mystery should cause spikes compared to that. So if you are not spiking, that might mean you are maxing out your CPU utilization which means that during the time when you should have spike, you are lagging due to insufficient CPU horsepower.

And no, not having 100% CPU utilization does not mean you are not maxed out. CPU utilization that you see means the amount utilized across all cores. But this game only uses one. This means you could have one core spinning at 100% while the others are at the nominal 10-20% so over all, for a 4 core CPU for example, you'd only see 30-40% CPU utilization. Feel free to verify this by observing the per-core performance of your CPU.

And that is, of course, assuming that your bottleneck is not your flash player or browser, which might have a lower resource cap than your hardware.

If there is a fix, I'd like it ASAP too.

But it would depend on the devs being competent for us to get this quickly, as opposed to waiting however long until the update cycle reaches us. Due to modularity and dependencies, sometimes you can't mess with update orders much. That means chances are we can't get the fix patch until we get all the earlier patches, and you know how they love to drag their feet on that.

I personally don't have a lot of faith in the devs, so I don't expect it to come anytime soon.

On a side note, as for the difficulty of deploying updates in general. Code is code, that's not really the problem. Translating text also isn't particularly difficult, considering they are hardly taking particular care to ensure correctness (like ensuring every reference to the same thing are translated exactly the same, don't you love when an area have different name from wanted mission to chase mission to on the map to quest to whatever else?) The most potentially difficult part about patching a game into another language, specifically Chinese to Latin-root language, is user interface. Specifically, the length of text will greatly expand, which means they have to change the size of the *on which might lead to change to lay outs and so on. It's not exactly a difficult thing to do, but is tedious and time consuming as they have to patch the text in and then keep trying different things to see if they fit nicely.

Granted, their update pace is slow even considering those difficulties, but it does mean that they can't just patch every new thing in within a month or two.

Chase delay have been around forever, and that's what I, and most others, talk about when delay comes up. I am still seeing this so I know it's still here.

I cannot say, however, whether a new delay exist since 4.0 due to whatever reason. That might be what you and the OP are on about, but that does not discredit the long term existence of chase delay. Claiming delay is a new thing will invalidate your statements as it's know to be not true. On the other hand, saying there is ANOTHER delay that is new could potentially be true. So you need to be clear with your statements. Because "delay only happened after 4.0 update" is most certainly not true whereas "this non-chase-based delay only happened after 4.0 update" could be true. So which is it? Personally, i use mei, which have an animated mystery which OP mentioned. But since I'm accounting for chase delay already, I might not notice a separate new delay.

And you need to understand the context of the thread. The OP says the game is pre-calculating. But pre-calculation would imply it is doing so intentionally, that's not true. The real cause of delay, whatever the source, is due to server state is ahead of client state. And the cause of that is the server not waiting long enough for the client state to catch up. Since, in reality, server side calculations are almost instantaneous. And there are two reasons for why this happens:

1. Server not waiting as long as it should: such as chase delay, and perhaps a new delay in mystery or dodging or whatever (For example, I can see, with its rather long animation, how minato's attack could cause a problem)

2. Client not going as fast as it should: such as flash rendering delay or network delay.

Problem #1 is the fault of the game, and need to be fixed. Problem #2.... depends. If it's causing problem even in premium software, it's a lack of optimization that is the fault of the game. On the other hand, it's the player's problem. (For the record, your hardware is above the "player fault" par, but the OP's might not)

No one is saying there is no problem. But correctly identifying the problem is important. First and foremost, identifying that problem #1 exists and not some "pre-calculation" conjecture that have no basis whatsoever. And secondly, identifying where exactly #1 is coming from--chase is a known case and have been around forever. Are we getting something else that falls in the category? If you are experiencing it, you should try to pin it down. Is it an excessively long animation on a normal attack (ex: minato), extra attacks not accounted for (a number of edo tensei), animated mystery(mei, tendo etc) or something else? Problems do not exist in a vacuum nor do they randomly pop up just because. There is ALWAYS a definite source. While it's not our job to identify the source, it would help fix the problem much more to do so than just reporting. And it is leagues better than counter-productive claims of conjectures that are wrong--because then they would check it, find no problem in the specifics provided and dismiss it.

I mean if you take the OP's report at face value. The devs certainly didn't make any sort of pre-calculation. So they know that problem does not exist. What do you think THEY would think? They would think it's the player's problem, that the player is lagging. And they would proceed to dismiss any OTHER reports of lag because they would assume it's the same thing. So the fact of the matter is, mis-identifying lag source is counter-productive.

And no, you obviously don't know as much as you think. Because you are comparing a CPU bound game versus a GPU bound game. Granted, 2600K should handle this game fine, but you are still comparing the wrong things.