Does getting a mechanical keyboard increase (damage) if I play an mmo?

[Heera]

Hello, I was wondering if mechanical (modern) gaming keyboards help with increasing damage when I play an mmo? I am currently on k120 (cheap as hell) and was wondering if I would lose dps (damage per second) because the keys are not registering as fast. So my question is: If I were to use a gaming mechanical keyboard (over $150), would it increase my dps in game? I understand that latency has a factor but by how much..? I understand it but I want to fully understand the process of which where on keypress, how fast is the speed processing and sending it to my desktop, then to the game, and if this would increase my performance on a mmo game. Thanks in advanced!

[Anakey]

the latency in your internet connection is by far the most limiting factor here. Unless you can hit the keys as fast as a hummingbird beats its wings then you will not see any improvement no matter what marketing bs they use

[matt3o]

depends how bad it is the membrane keyboard you are actually using. in an ideal world where your connection is blazing fast and ping few milliseconds.... yeah maybe a gaming keyboard can help you, but you don't need expensive ones. if it can scan at 1000hz you'll have a lag of approx 5ms (excluding other bottlenecks, like a borked USB host), which is more or less the best you can expect.

[purdobol]

Back in my days of gaming. Interrupt based protocols were considered better then packet based protocols (mouse is a different story).
So PS/2 keyboard was better than an USB one. With that said times have changed and adjusting USB polling rate is a thing. Many cheapo keyboards can handle higher polling rates with ease. Of course when we're talking latency all things have to be taken into consideration. To get instant response, monitor has to display said change with no delay. CRT's did that. Today panels probably are ok in that department but I doubt they come close.

All of this is moot though where we're talking online gaming, unless the game has godly network protocol, and you live nearby datacenter or have insane connection.
This kind of stuff is only applicable in LAN events and at a very high skill level. Where players look for any kind of improvement just to get an edge.

tldr: no

[vvp]

If the key registration on a network server (instead of registration on the local computer) is important then the internet latency is the biggest problem because it can vary a lot between players.
Here are some estimates of common latencies:

• internet: typically from 20 to 60 ms roundtrip
• keyboard with mechanical switches (e.g. Cherry MX): 5 ms
• USB protocol: half the keyboard polling rate; it is 0.5 ms at best; you can query the polling rate on linux with lsusb -v -s <busId>:<devId>; check the value for bInteval; e.g. for my keyboard it is 2; that means the average latency on the USB protocol is 1 ms
• display reaction time: about 8 ms (from refresh rate) + 5 ms (from pixel flip)
• human reaction time latency (eyes to finger movement) is about 200 ms; notice that eyes to movement is one of the worst cases, sound to movement is quicker; touch to movement is even more quicker

So your overall reaction time will be something like 200 ms + 5 ms + 8 ms + 1 ms + 5 ms + 40 ms to the server. If the speed of clicking (instead of the initial latency) is important then get a programmable keyboard where you can program it to generate key press/release events as quickly as possible given the selected USB polling rate. E.g. programs in my keyboard can generate a one keypress/keyrelease pair each 4 ms.

[andrewjoy]

I dont think it would help you in game in any real way unless the board you are using is truly terrible.

I would work on your keybinds first and optimise them not sure what MMO your playing but , main rotation on say Q12345E so you dont have to move from WASD have second and 3rd action bars with shift and control modifiers

7x3 keys should be enough for any MMO

[Muirium]

You're being way too objective here, fellas. The correct answer is in fact: yes! Because a good keyboard makes you feel better, whatever it is you're actually doing. Your fingers will appreciate better keys.

It's not just computer hardware that adds up to a good score. A lot of it is you.

[matt3o]

human reaction time latency (eyes to finger movement) is about 200 ms; notice that eyes to movement is one of the worst cases, sound to movement is quicker; touch to movement is even more quicker

human sensory delay is about 300ms but pro-gamers actually act before things are happening, almost instinctively, so basically negating such delay.

[vvp]

I measured myself few years ago when I was playing games more and my response was around 220 ms (image change detected by eye to finger movement). This web site claims: "The average (median) reaction time is 215 milliseconds, according to the data collected so far." But when one looks at the data the maximum of the distribution seems to be around 260 ms. And the text is claims the median is at 273 ms. Based on the discussion later(*) it seems that the charts show raw data and the average delay from the system is not subtracted. The delay from the system will be about 8 ms from LCD refresh, about 5-15 ms from pixel flip and at least 5 ms from the input device itself. That would put it to around 245 ms which is nearer to 215 ms claimed at the front page.

If it is very easy for you to post reference to the 300 ms value then do it. If you do not remember it then don't bother. I do not care that much about this

If Heera can somehow modify his game or hardware so that the triggering sensory input is switched from eye to ears or even better to touch then he will get an advantage as well.

(*) They claim that the response time is getting worse over time which they attribute to the increasing system lag (e.g. mobile phones are slower than PCs, LCDs are slower than CRTs, ....). That indicates that the shown data charts include the system delay.

[matt3o]

I said about 300ms, our data basically aligns @vvp . my point is that you cannot apply a 200-300 sensory delay to a gamer who actually acts before things are happening.

[vvp]

Ok, that is true for the actions which can be anticipated. But if it can be anticipated about 200-300 ms in advance then it is probable it can be anticipated 5 - 10 ms more to compensate for keyboard delay. And that means that keyboard does not matter at all. Objectively. Of course, there is your argument that good keyboard makes players feel better and therefore it matters ... at least for the people who care about appearance instead of performance.

PS: I had an error in my previous post about average delay from LCD refresh rate. The claimed value was 30 ms and it should have been 8 ms. I'll edit it.

[matt3o]

unless keyboard is ghosting or jamming or you hit a keyroller limit... what keyboard you are using makes very little difference. it's more of a "feeling" thing. Having a higher actuation point or a linear vs tactile feedback. It's not that it makes you better, it just makes you feel better.

[Muirium]

And feeling better makes you better.

[vvp]

No ghosting and jamming is requirement for any keyboard regardless whetter is is a generic rubber dome or an expensive gaming keyboard. I doubt that rubber domes do jam or ghost more often. I never had this problem with the few rubber domes I had. I had some ghosting problems with a Cherry MX switches after about 10 years of use. I never used a rubber dome for such a long time though. I cannot claim any reliability difference from my experience.

Every modern keyboard will be 6KRO at least and that is enough for any game which dedicates one hand to mouse or joystick.

Curve of actuation force dependence on stem travel is an objective thing. Touch is our fastest sensory input and some people may need the tactile feedback to be aware about the actuation point better. Others may prefer quicker speed of a linear switch around the actuation point. There are significant differences between keyboards in this area.

The thread starter focused on speed. I think there is no significant difference between standard cheap keyboards and the expensive gaming keyboards.
Are typical debounce times of rubber domes bigger?
Are typical USB polling rates or rubber domes bigger?

We will not know debounce times of rubber domes without a look at the firmware which is typically proprietary. But people who have USB rubber dome keboard can report USB polling rate easily. So the answer on the second question is easy to find.

[vvp]

Kinesis Advantage 1 USB polling rate is 8 ms. This is Cherry MX Brown.
My custom K84CS keyboard polling rate is 2 ms. This one is Gateron Blue.
I have only PS/2 rubber dome now -> no rubber dome data from me.

[swampangel]

I'm sure this has been posted here before, but Dan Luu's experiment found that the keyboard controller can add latency between physical keystroke -> OS keystroke in the order of tens of milliseconds https://danluu.com/keyboard-latency/

[vvp]

Interesting. He adds key travel time to the latency and the key travel time is in the range of 16 - 32 ms. Cherry MX registers in the middle of the travel so that would make it 8 - 16 ms latency because of the the travel time.

Well, gamers should use microswitch keyboards with about 0.5 - 1 mm travel. Sure as hell I would not want to type on such a beast

[swampangel]

Here's a followup from another keyboard enthusiast http://thume.ca/2017/12/29/fixing-my-keyboards-latency/

In Dan Luu's test the OLKB Planck was middle-of-the-pack; it would be interesting to see if there are optimizations that can be applied to qmk.