Yeah at least KOF is looking better and better every time they showcase it. The trick is going to be getting a large enough player base to keep it going
what you are talking have NOTHING to do with what I said in my post.
160ms is 9.638F rounded to 10F not 8F… 16.6ms = 1F
11F is considered the generally speaking the lowest reaction time to a stimulus that you’ll find and that is usually just pressing a button not processing information and making a decision. The fastest ever recorded reaction time is 0.101s at a starting block(6F) and that may have even been anticipation for all we know. This extensive study showed that the mean average of male sprinters was 185ms ± 64ms in a 60m dash, that is 11.1~15F . There is no decision either, and you are prepping for one specific action. You hear bang - run. That is all you do. Keep in mind, even runners often have false starts so there is anticipation there not just reaction. In addition these are professional athletes not average people. Source
There is actually other studies that show that it can take up to 100ms just for your brain to recognize an object from a visual stimulus, audio is faster however, touch is fastest. That is BEFORE you even make a decision, before the time it takes for you to actually move your finger into position and press the button and/or move the lever. This is all assuming the monitor also has zero lag, there is no outside interference, and you are actually reacting. Adrenaline also affects your reaction time.
Sorry, /average/ human reaction to decision making is significantly more than 8F. 11F is on the lowest end you can expect to find with olympic sprinters outside of very specific cases. Computers found that on average the time it takes to recognize something for a computer (this means beyond just hitting a button when you see a dot) typically was around 50ms, 3F. Brake tests give as much as 2.5s leeway for a person to hit the brakes and the time for reaction to a situation + hitting brakes was ~0.7~1.47s. You’re probably going to see most people reacting in a game around 15~22F not 8~11F. People get hit by Chun’s OH in SFV and that thing is 26F startup, even removing the 8~9F delay that’s still 17~18F to react.
Fair points, though at some point players who practice for it also essentially remove decision making from reaction to setups they are familiar with. I.e. they’ll have trained a response to visual and/or audio cues.
Sort of. I get what you are trying to say.
Think about how whiffing jabs or mp works to bait stuff. Shimmying, jump/empty jump/dive kick/instant airflame/legs/dash up. You still need to make decisions because many of these things can be performed at similar distances and the reaction you need to make changes. Somebody throws out a crMK you are anticipating it and throwing out your own crLK / sMK / whatever you aren’t really reacting to the stimulus of the crMK animation. If you have already planned to do something then you’ll do it regardless of the action. People that punish a fireball after not being baited by a jab ARE reacting but their reactions aren’t clocking in at 11F, I guarantee that they are specifically waiting until they can recognize a specific animation then the time it takes to make that decision + input it.
I think we continue to play and just hope it gets better dude
I don’t know what else to do 
Well its too late for us to not buy it, but we can be purists and keep playing 3s or whatever game suits us. No reason to give capcom undeserved love for a subpar product. Enough people go this route they may take notice.
We could to start a petition to change the main game of the CPT to MKX, dunno.
I’m just trying to tell my friends about this and they’re keeping quiet or telling me “Thats the game then!” and they don’t care because they’re having fun. Meanwhile their “reactions” are just predictions in most cases, and victories are from unsafe moves that are too fast to punish.
Eternal has good points.
SRT (simple reaction time) to one known stimulus is not what people should be talking about; fighting games are ALWAYS more complex than that, and choice reaction time is much slower than SRT. When there are several possible things that the opponent can do and you have to react, you’re not going to react at the maximum SRT speed of around 180ms. Absolutely not! If that was even possible, we would literally never ever see overheads hit in tournaments (or in any sets). Taking average reaction times and straight up comparing them to overhead frames or dash speeds is completely retarded, and like I said, I think that people exaggerate how much they would have to adjust if they were to lower the lag.
That is quite sad to hear. The best players that I know have all been irritated by the lag since release and have slowly dropped the game almost completely. (Me included)
I played some GG Xrd today after hearing that the PC version has only 1 frame of delay with the right settings, (post-effects off, v-sync off, anti-aliasing default).
I went to training mode, set the input delay to 6-8 and jesus christ how huge difference it actually is.
I’d recommend everyone to play 3rd Strike on Fightcade while waiting for word from Capcom. It feels like heaven to be able to whiff-punish mediums with a super.
Needs a source with data to back that up. What is the framerate running at? If it’s unlocked then yes, you can get down to very small latencies if you start hitting 200+ fps. If the game is locked to 60 then there will always be about 3 to 4 frames of lag. You get 1 frame for the input, the next to process, and the next to render, then on frame 4 is what we start to see. With higher frame rates you get faster processing, rendering, and so on, so you could cut your input latency in half by playing at 120 fps compared to 60.
The only problem with this? All the desync that will occur compared to playing against people with different frame rates.
It sounds like it should be peer reviewed.
Whatever the base rate, the ~3f difference in lag is real. Delay 3 training mode on PC feels like PS3, and just the 3f less lag already makes an insane difference in game feel.
Displaylag notes PS3 Xrd as having 5f of lag, which would say PC Xrd sits at about 2f.
A camera at 60 fps can only give a quick and dirty idea, since you can’t sync the camera with the display’s refresh rate. We’ve already talked about the camera method before. The one frame or half frame variances he’s getting most likely aren’t from variable input timings but rather the camera not being in sync with his display.
Vsync was turned off for his PC test and he’s using a CRT. A CRT monitor can run at way higher refresh rates than 60 Hz. The camera at 60 fps wouldn’t be able to capture half or quarter “frame” differences for a number of reasons. Not that it matters, since once you disable vsync a card is free to tear all it wants to get the fastest refresh possible. Still, if somewhere his drivers have vsync on, and the refresh of the monitor is 120+ hz, then it’s going to be much faster. Usually the OS defaults to the “native” Hz of the monitor and usually that isn’t 60 when it comes to CRT monitors. Source, I have some. If the monitor is set to anything but 60, like say, 75 or 90, then you’re talking 1.25 to 1 or a 1.5 to 1 ratio of frames rendered to frames captured by the camera. Factor in the camera isn’t synced to the display and, well…
IMO, his equipment setup is off. There are a number of things I’m taking issue with, such as vsync disabled but he has a constant 60 fps. Does the game soft lock to 60 fps? If so, then it’s still rendering on a 16.7 ms frame window, and without being able to have someone look at the code, we can’t tell if it checks for inputs in half frames or only full. Considering it’s a fighting game built for 6o fps and keeping frame data in check is integral I have to wonder. If it doesn’t soft lock and scales the animations with the frames based on an algorithm like pretty much every made for PC game to date, then there are a number of new issues that crop up.
PS360+ 0 ms input latency? K. TE Kitty at .27 ms? K. Did he test this with logic circuit testers? Unless he has a camera that can do 1/1600th of a frame, how does he get these numbers without significantly expensive hardware? I’m literally balking at the 0 ms input latency of PS360+ on PS4. Come on, everything has latency, even if it’s small.
That test certainly needs a peer review, and definitely needs to be proven repeatable via someone else testing.
Wow, this is not true at all!
It’s very well possible to have under 3 to 4 frames of lag in 60fps, there’s no reason that all of the parts that you mentioned would require an entire displayed frame in the game’s code. Why would you think so? That would be absolutely nuts bad design; in fact it would be harder to implement that way (I mean handling those things separate code cycles or ‘frames’).
Lag doesn’t have anything to do with framerate aside from it never being able to go under 1f of whatever framerate the game is refreshing at.
In fact there are many games on MAME that have input lag of 1, and you can test it by playing the games frame by frame using the emulation pause. No camera required.
In the shoot 'em up genre, 3-4 frames is considered high amount of lag (…and the games never run above 60fps), which is why we have things like Shmupmame.
And I still consider 3-4 frames bad for fighting games too. There’s no reason to not have 1-2 as your priority if you’re the developer, unless it requires some serious sacrifices from graphics or gameplay.
Like someone already stated, even few frames of lag can have insane impact on game-feel.
EDIT: Eh I don’t know what you’re disagreeing with, when what you’re saying is provably false and what I’m saying here is provably true. Explain?
Your claim that inputs need to be handled via several phases on different frames and that because of it there will always be 3-4 frames of latency, are both claims bullshit.
I’m going to trust people who know what they’re talking about with experience and data to back it up over you, which contradicts what you think you know.
Which part of those almost 10 years old articles do you want to me to read? I don’t have time for them all, and it’s usually not good for the sake of discussion to just throw a bunch of links especially when you’re trying to prove a smaller, single claim.
I gimplsed at the first one, and hilariously it has it’s Killzone 2 input latency test wrong. The input lag in that video is 6 frames and not 12 like he counts; his camera was recording at double the framerate of the game, and he was counting the frames from his monitor (in the camera view).
Does this guy have ‘wonky hardware’ too? (Battle(non)sense)
These are milliseconds too, with the blue bar being average latency. Like Xrd, I’d guess that these games have 1 frame (~16.6ms) of lag too + some milliseconds from PC and monitor, averaging at almost 2 full frames but still not quite.
Still i kinda find it funny that this input lag has been in sf5 since prolly the first beta an now ppl complain or use that reason as to why they lost.
People actually did notice it since the first beta. But it was initially attributed to online server issues instead of something that was just in the game.
Literally every good player that I know have complained about it since release.
Now that it has become a mainstream subject, they’re more open about it as it doesn’t sound like a personal excuse anymore. Also people don’t use it as a ‘reason to why they lost’, no-one that I know at least.