how fast can you repeatedly hit a key?

[alpslover]

this is a spin-off from this thread, where the discussion was which type of switches are 'better' for gaming, i.e. which ones perform better, particularly in games where repeatedly hitting a key as fast as possible is important.  the commonly accepted 'knowledge' being that non click tactile switches perform better/respond quicker than click tactiles.  i've always believed this to be a myth, at least partly because nobody has ever provided any evidence to support the claim.  a lot of hypothesizing, yes, but no real world data for corroboration.

so i proposed a simple test - open up a text editor, hit a key as fast as you can over a given time interval, and tally up the number of keypresses.  do a few runs (giving your hands a rest between them) and then repeat with a keyboard with different switches.  do you find there a significant difference between the keyboards?

i compared 7 keyboards:  a blue cherry filco, a brown cherry filco, a realforce 101, a model m, a focus 2001 (complicated white alps), an ibm thinkpad t42 laptop, and a cheap hp rubber dome.

i did 3 runs with each keyboard, each run lasting 15 seconds.  i used the index finger of my right hand (my strongest finger) to hit the 'j' key.  i bottom each keypress (that's just the way i type).


here are the results, in order: (i'm updating the results as i test more keyboards)

blue cherry filco:  138 132 134
brown cherry filco:  136 134 132
realforce 101:  130 133 133
focus 2001:  132 133 134
model m:  133 135 131
thinkpad t42:  133 131 133
(*)hp rubber dome:  138 133 135

conclusion:  for me, there was no significant difference between the keyboards.  i suspect i was a hair quicker on the filcos because my hand was not as fatigued.  bottom line, the limiting factor is not the keyswitch, it's me.

i'd be interested in seeing other people's results, particularly those of pro gamers, or those with cherry black keyboards.  i'm guessing those people who can press keys really really really quickly (200 keypresses in 15 seconds?), and/or 'dither' their keystrokes around the keyswitch actuation points rather than doing full strokes like i did, might have different results.


(*)  the hp rubber dome results were a tiny bit higher.  maybe not significantly so, but during the testing, one thing was immediately obvious - my hand was not fatigued as much after each run as on the other keyboards.  this hp keyboard has tactile rubber domes, and as far as rubber domes go, they're not bad.  i think what made this keyboard less fatiguing for repetitive fast key hitting is the softish way that the keys bottom out.  softer than topre capacitives, yet not mushy to the point of nastiness.  however, as far as fast normal typing and actually playing a rhythm game (where it's not a single key being hit repeatedly, but rather key sequences) goes, this keyboard is not a good choice as it has a slow key scan rate which would result in key sequence errors.

[DreymaR]

Presses per 15 s is what you measured, but I'd also report it in CPM or even WPM: For instance, 100 presses in 15 s is 400 CPM (over 15 s) or 80 WPM using the standard convention of 1 WPM = 5 CPM.

Malapropos that: I once managed 240 WPM pressing eight different keys over and over. Interesting to compare those 'perfect rolls' with your same-finger experiments where you get about 107 WPM.

[Rajagra]

My tests in order done:

Topre HHKB 101 100 101
Topre RF 87U 111 105 110
Cherry blue DAS III 112 100 108
Buckling spring Unicomp Customizer 110 102 108
Scissor switch Logitech diNovo 112 119 119
Cherry brown Compaq MX11800 118 116 120
Cherry black Cherry MX11900 117 120 120

Topre HHKB (again) 111 112 116

Looks like some of the increase in speed is down to warming up (moral: work out on the buckling springs!) but some key types have a slight edge for me.

Edit> One more: Belkin Nostromo N52te 120 111 120

[alpslover]

Looks like some of the increase in speed is down to warming up (moral: work out on the buckling springs!) but some key types have a slight edge for me.

thanks for doing the tests.  it does look like that you're slightly quicker with non click tactile switches.  which keyboard(s) do you use most on a day-to-day basis?

i fully expected to be slower on the m than the other keyboards, but was surprised to see that i was not.

[InSanCen]

I'll get on this later when I get some time.

I am typing on Cherry Blacks now, and also have BS (M13, 1391406), and Pink alps as well as something that pretends to be alps.

I am not, however a pro-gamer, though as I kid, I rocked at Daley Thompson's Decathlon as well as another game that for the life of me I can't remember (EDIT:- Track and Field I think it was). Both involved smashing one button (or 2) as fast as humanly possible.

[Rajagra]

it does look like that you're slightly quicker with non click tactile switches.  which keyboard(s) do you use most on a day-to-day basis?

The 87U is my usual board. Before I bought that the DAS was my norm. It's interesting how swapping around affects things though. I used to type relatively slowly on buckling springs, but after leaving them aside for a few months I tried them again and almost matched my normal typing speed.

When I game I use the Nostromo, so maybe I'm tuned into that for rapid key hits. Those keys are tactile, quiet, short travel.

[alpslover]

Malapropos that: I once managed 240 WPM pressing eight different keys over and over. Interesting to compare those 'perfect rolls' with your same-finger experiments where you get about 107 WPM.

not entirely malapropos, as the ability to do 'perfect rolls' are what's necessary in rhythm games.  which is why in the original thread, i mentioned that the das iii probably wouldn't be a very good choice, as its slow key scan rate will be a problem.

[alpslover]

When I game I use the Nostromo, so maybe I'm tuned into that for rapid key hits. Those keys are tactile, quiet, short travel.

i did have access to a cherry ml keyboard (3mm throw, i think?) when i did the tests, but it didn't occur to me to test it because it is not a keyboard i would willingly use unless i was under threat of imminent violent death.

i'll test it and some other keyboards later.  the closest thing i have to linear keyswitches (aside from the brown cherries) is the truly horrendous btc foam capacitive keyswitch keyboard.  i'll test it anyway, not that i would ever use it for typing or gaming.

while i was testing the realforce, it 'felt' slower to me, but the results showed otherwise.  i think it's because i use click tactile keyboards almost exclusively (and have done so for many years) and i've gotten so used to the aural feedback that it's something i expect when i'm typing, so a lack of aural feedback feels 'off'.