A few months ago I got a cluster of complaints about the number of typos on PainScience.com. One of them was unusually rude, implying that errors are only slightly less common than words. That’s obvious hyperbole — this place isn’t perfect, but it’s polished to a high shine by internet standards. The unreasonable griping didn’t get under my skin much, but it did provoke me in a good way, and I decided it was time to go to war on whatever typos have eluded me to date.
I have been working with a good proofreader ever since to drive the error rate down to a “barely detectable” level. The 50 busiest pages on PainScience.com are now done, plus twenty other miscellaneous selections. Some fun facts:
- 70 articles, containing 339,345 words, with 527 mistakes. Lots of those mistakes were duplicated on other pages, so it was more like 700 typos from the reader perspective.
- So the overall error rate was about two tenths of a percent — that is, about 1 out of 500 words were broken, which is about what I expected to see.
- Two of the worst pages were more than twice the average, hovering right around the threshold where many readers might start to roll their eyes at all the mistakes.
- The numbers get dizzying if I estimate how often those errors have been seen: roughly several million error sightings per week, all now eliminated. (Very roughly. Page view counts are much higher than the number of people who actually read the whole page, and most of the errors will be missed by most of the readers.)
- The oldest typo found was probably about 18 years old, likely dating all the way back to the first version of one of my earliest articles.
There are undoubtedly still some typos left. I have already found a handful that my proofreader missed. It’s amazing how the damn things can hide in plain sight. Two readers will usually sport several different errors in a 5000-word article — I have seen that happen many times over the years. But I suspect the error rate is now down to something like one per 3000 words. In the busiest articles, anyway.
If you want to report some typos, here are some (very basic) guidelines here: Typos and other minor glitches: On proofreading and reporting errors to PainScience.com.