10fastfingers.com

Update

Normal mode:

Traditional Chinese: 224 WPM (+5 keystrokes)
Indonesian: 173 WPM (+10 keystrokes)
Malaysian: 169 WPM (+21 keystrokes)

Advanced mode:

English: 194 WPM (+30 keystrokes)
Dutch: 162 WPM (+75 keystrokes)
Indonesian: 133 WPM (+5 keystrokes)
Czech: 126 WPM (+28 keystrokes)

I DID IT!!!

typing advanced indonesia 139 wpm, normal typing: 174 wpm

http://10fastfingers.com/user/321935/
134 wpm

OMG!
Here are competing robots!

Update

Normal mode:

Slovak: 178 WPM (+12 keystrokes)
Indonesian: 176 WPM (+13 keystrokes)

Advanced mode:

Spanish: 194 WPM (+28 keystrokes)
Dutch: 179 WPM (+87 keystrokes)
Vietnamese: 151 WPM (+1 keystrokes)
Hungarian: 142 WPM (+9 keystrokes)
Indonesian: 139 WPM (+30 keystrokes)
Czech: 126 WPM (+1 keystrokes)

The Korean language was adjusted down from five keystrokes per character to three keystrokes almost two years ago:

https://10fastfingers.uservoice.com/for ... e-accurate

This means that the record of 348 WPM, which is not registrable on 10fastfingers (in the stats and graph) with a limit of 255 WPM, should be reduced significantly:

https://www.youtube.com/watch?v=vpaKn-0NUwM

The score of 348 WPM is equivalent to 209 WPM after the adjustment:

348 WPM * 3/5 = 209 WPM

Yes ZappX, I'm already aware of that for Korean. Actually there have been several changes for that language in recent years, at least two or three. I think I'll watch the video in detail and accurately count the number of keystrokes needed to finally have a real estimate of this unrealistic score.
Anyway, looks like I have a lot to update on this thread.

Edit: Done! 859 CPM or 172 WPM for Andrew Kim.

It looks like my simple calculation of 209 WPM was wrong as it didn't take "space" into account. I figure you arrived at 859 CPM and 172 WPM by counting the actual keystrokes required for each correct word using the Hangul method?

Here's how the calculation of Korean happens on 10fastfingers:

Most other languages adds one keystroke per word because of "space", but for some reason each Korean word gives two extra keystrokes. I checked the video and counted a total of 299 characters and 133 words, where 295 characters and 132 words were correct. Here's the old calculation, which matches the numbers in the video:

https://www.youtube.com/watch?v=vpaKn-0NUwM

Gross keystrokes: 299 * 5 + 133 * 2 = 1761
Net keystrokes: 295 * 5 + 132 * 2 = 1739

WPM = Net keystrokes/5 = 1739/5 = 348 WPM

By the new calculation the numbers are:

Gross keystrokes: 299 * 3 + 133 * 2 = 1163
Net keystrokes: 295 * 3 + 132 * 2 = 1149

WPM = Net keystrokes/5 = 1149/5 = 230 WPM

I also found another Korean video from February 2016, and again the numbers are confirmed by the calculation:

https://www.youtube.com/watch?v=Gb3It6AVJaA

A total of 123 characters and 53 words were typed and 117 characters and 51 words were correct:

Gross keystrokes: 123 * 3 + 53 * 2 = 475
Net keystrokes: 117 * 3 + 51 * 2 = 453

WPM = Net keystrokes/5 = 453/5 = 91 WPM

With the current calculation system for Korean it looks like the known record is 230 WPM, as shown above. This is quite a way off the 172 WPM you calculated by presumably counting the actual keystrokes required.

Good find for dual spaces. A bit weird, maybe a mistake from Chris?

Indeed, even with 1 normal space after each word, the current calculation system is still too high. I guess 2.5 keystrokes per character would be more realistic. But instead of doing that, the ideal is still to know how many keystrokes requests every word of the Korean list.

I can totally do that, at least for Korean and Japanese, and send the files to Chris so he can finally fix it for these two languages. About Chinese, I still don't know very well the right input method. Besides, Chinese word banks are huge and it will certainly take more time.

Regarding the highest score in Japanese, I would be able to give the actual score, as I did in Korean (indeed based on Hangul), but unfortunately I can't do it from the graph only. I need a video (or at least the list of words that have been typed correctly).

Based on your count and calculation of 859 CPM we get the following average for each character, excluding "space":

(859-132)/295 = 2,46

2,5 keystrokes per character is very close to the actual number required, as was suggested:

https://10fastfingers.uservoice.com/for ... e-accurate

The final solution was rounding it up to three, which skews the score too much, as you mentioned. When in
addition adding two extra keystrokes per word the final result is way off. Indeed, that looks like a mistake.

The current Japanese calculation is also too high, and the best typists reach above 255 WPM. I'm thinking the same as you; the best way is to calculate the number of keystrokes for each word in the word banks. This is meticulous work, especially if you include the normal, advanced and top 1000 modes. Japanese currently has a normal and an advanced test, while Korean only has a normal test. None of them has a top 1000 mode.

I can think of two potential problems:

1) There might be a faster input method than you are using for these two languages. Even so, the most common input method for the country is preferable.

2) New word banks will have to be calculated in the same way to be comparable.

Video of 264 WPM from a Japanese competition:

https://www.youtube.com/watch?v=bkj8glLtSQk

Currently each Japanese character gives five keystrokes and as in Korean two extra keystrokes per word.
226 characters and 102 words were typed and 224 characters and 101 words were correct.

Here's the Japanese 10fastfingers calculation:

Gross keystrokes: 226 * 5 + 102 * 2 = 1334
Net keystrokes: 224 * 5 + 101 * 2 = 1322

WPM = Net keystrokes/5 = 1322/5 = 264 WPM

In the video it looks like a program is used to register the keystrokes. Most of the time both "space" and "enter" are hit for each word, and the words are underlined. The reason being that the user is converting the text:

https://www.coscom.co.jp/learnjapanese801/index.html