Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I think the concept people are reaching for is the idea of false positives and negatives. Yes, if you have a bad sentence there is a good chance it will highlight it. But, perhaps not (false positive). Similarly, it will flag many perfectly fine sentences not pitched to 7 graders (false negatives).

Here is text I more or less randomly chose from MOMA's site. Almost all of it is graded as "very hard to read".

Where is the cutting edge of the motion picture? Discover it first at MoMA. Building upon the Museum's long tradition of exploring cinematic experimentation, Modern Mondays is a showcase for innovation on screen. Engage with contemporary filmmakers and moving image artists, and rediscover landmark works that changed the way we experience film and media.

Any edit I make to that paragraph that makes the app happy seems to diminish the text.

In contrast, my first paragraph is graded better than the MOMA text, yet I think it is worse. The one thing it did complain about were the adverbs 'similarly' and 'perfectly'. The former is required to draw the comparison; the second is perhaps redundant, but I am emphasizing to make a point - redundancy is as much a tool in writing as it is a crutch or error.

I'm not saying the app is useless, just take the output with a huge grain of salt. Heck, if I paste text from Hemingway it is a sea of red and yellow.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: