How many errors per page volume is typically "okay" in a book?
Suppose I'm in charge of getting a three hundred page book published. I'll do all reasonable things to get it proofread, but still there is a chance that some errors will persist.
Now the book goes to the bookstores and readers start reporting errors. I'd like to know whether the rate of errors is "acceptable" or if I haven't really done my job.
Is there any standard, or any informal rule, for how many errors are "acceptable"?
This post was sourced from https://writers.stackexchange.com/q/6750. It is licensed under CC BY-SA 3.0.
4 answers
Aim for as few errors as possible. Editors and publishers may overlook the very occasional error, but not if they see five a page. Get your manuscript proofread as much as possible before you send it.
This post was sourced from https://writers.stackexchange.com/a/27639. It is licensed under CC BY-SA 3.0.
0 comment threads
If it's okay with Stephen King, it's okay with me. I have proofread my book no less than ten times and I still find errors. I'm not perfect and neither is anyone else. I feel confident that most readers don't judge a book by its typos.
This post was sourced from https://writers.stackexchange.com/a/26856. It is licensed under CC BY-SA 3.0.
0 comment threads
I have to edit my own books frequently. But I publish ebooks -- which allows you to update the ebook file if you find errors.
With printed stuff, you really need a second pair of eyes to find typos. The risks are too great -- especially if you are paying for advance copies.
On the other hand, print-on-demand services (like Createspace) do allow you to update the manuscript after publication.
If you are not confident of your quality control process, you might want to look into a distribution method which allows you to update the manuscript after publication date.
Frankly I worry more about typos than about inconsistent punctuation because one typo lowers a reader's confidence about the book's quality.
Also, technical books and manuals have to undergo additional vetting from an editor who does technical review-- because a mistake can result in bad information. At the same time, all technical publishers know that errata will occur and often have a web page on the company site listing issues and corrections.
A typical developer buys a programming book with the near certainty that it will usually contain technical errors. But they don't sweat it as much because it will help them with general concepts, and generally they can spot the error -- even though it may not run properly. The typical developer wouldn't mind an occasional misspelled word as long as the code works or the technical knowledge is basically right. Also, many IT books are out-of-date by the time they are published, making it all the more important to check for errata.
This post was sourced from https://writers.stackexchange.com/a/26862. It is licensed under CC BY-SA 3.0.
0 comment threads
Whether such a threshold exists or not is irrelevant, since you can never know whether you have met it or not.
If you detected an error, you would fix it, so the number of known errors is always zero. The number of unknown errors is unknown because you can't count unknowns. Therefore there is no way to know if you have an error rate below any given threshold.
Of course, over time, readers may discover errors in your work. There is no way to know if any one reader discovers all the errors, or if all the readers collectively have discovered all the errors, since a) they don't report them systematically, and b) you can never know how many undiscovered errors remain.
This means that you can know if a work has passed the acceptable error threshold for an individual reader if they write to you and say, "I discovered 10 errors in the first 73 pages so I stopped reading." But that does not mean any other reader will ever notice all those errors, or that if they do it will provoke them to stop reading and write you a nasty letter.
Finally, even if no reader has found enough errors to stop them reading yet, that is no proof that the next reader may not be more sharp eyed or more irritable than than all the others may not find enough errors to trigger their refusal to read on.
A much better measurement might be this: the rate of error discovery. That is, how many errors do you find and fix in each pass through your work. Presumably this number will decline each time through. How many errors do other readers discover each time they read it? You can then set an acceptable error discovery rate per pass for a certain number of readers as a measurable quality metric.
Of course, the Web changes all of this, because now errors can be fixed after publication, something that even extends to published ebooks which can, like software, download periodic updates.
0 comment threads
0 comment threads