WPM Inaccurate?

[ztbrown]

I feel like WPM is inaccurate. This is just anecdotal, but I'm getting through a book much slower than expected.

How do we currently test the WPM?

[OnlyInAmerica]

You're probably right, and we don't have a test case for WPM yet.

The wrinkle in this problem is that certain words have increased delay for readability, and we display three blank "words" following each end-of-sentence punctuation.

Determining appropriate probabilities for these situations and scaling the displayed WPM could be a viable improvement. e.g: Say we determine a 600 WPM setpoint corresponds to a realized 500 WPM with average sentence and word length. Developing a test case to help determine this scaling value would be cool, but I don't have time atm.

[Matschile]

Had the same feeling, while reading a book.
I set up a unit-test (locally) with some article from a newspaper.

Results:
600 WPM in the setting was just about 240 measured.
500 WPM in the setting was just about 200 measured.

Not sure the test ist currect, yet. Going to look over it the next days.

[arpheno]

I can confirm this. There are tons of delays for everything.