# Statistical tables in old books purposefully wrong?

I remember having read a while ago that in old (pre-computer days) books, the last digits of the theoretical quantiles shown in the appendices were inaccurate in order to discourage plagiarism (the idea would be that if another book has a table in the appendix where the last digits are identical to the ones found in yours, then that author must have plagiarized your table).

I am trying to find back the source for this story, or a link to a mention of it, but after hours of searching I couldn't.

• It was the case with mathematical tables more generally, rather than statistical tables specifically. (If I pull up a reference I'll offer an answer) Jul 20, 2016 at 22:56
• I wonder if this might possibly do better at hsm.stackexchange.com ? Jul 21, 2016 at 3:53
• A similar effect (careful insertion of small fictions among many facts) occurs in mapping, see for example en.wikipedia.org/wiki/Trap_street and wiki.openstreetmap.org/wiki/Copyright_Easter_Eggs. It's said that old tables could deliberately put a round off value to the 'wrong' side in those case, which would have minimal effect on computations, but a larger effect on copyright! Jul 21, 2016 at 12:43
• As anecdote, I TA'ed for a stats professor who presented the same homework for each quarter and distributed the answer key after the homework was returned (even though this was the same homework assigned last quarter). While the steps were correct, the actual math in the answer key was riddled with errors. Like, 2+2=5 level of errors. We would regularly see students hand in homework in which the wording had changed, yet we still say the exact same 2+2=5 errors. It was depressingly amazing. Sep 23, 2016 at 21:52

The Wikipedia article "Fictitious entry", which is on the more general subject of "deliberately incorrect entries in reference works", cites one example of something close to this:

By including a trivial piece of false information in a larger work, it is easier to demonstrate subsequent plagiarism if the fictitious entry is copied along with other material. An admission of this motive appears in the preface to Chambers's 1964 mathematical tables: "those [errors] that are known to exist form an uncomfortable trap for any would-be plagiarist".

The citation is to page vi of:

Comrie, L. J. (1964). Chambers's shorter six-figure mathematical tables. Edinburgh: W. & R. Chambers.

• +1 As an aside, this always seemed like a bizarre gambit. "Hah! You have egg on your face because I published incorrect results!" As if publishing incorrect results is not, itself, an own-goal.
– Sycorax
Jul 20, 2016 at 20:53
• @GeneralAbrial Perhaps there's an ulterior motive of being able to claim that any mistakes one has made were just clever ruses. :P Jul 20, 2016 at 20:58
• Right, but even then, you're saying "Ah! You fell into my trap of assuming that my publication was accurate, when, in fact, it is deliberately inaccurate!" Like, the endgame of the ruse is publicly admitting that one's own publication is faulty, and deliberately so.
– Sycorax
Jul 20, 2016 at 22:03
• Something like this (coarse language warning): i.imgur.com/n8umjWj.png Jul 20, 2016 at 22:37
• The important thing is that the inaccurate information must be benign. In a publication where statistics are incorrect, it would clearly be the case that the author has been an idiot in publishing incorrect statistics, but in the case where someone is commenting on statistics, for instance, they might include the name of a town where the statistics were not gathered, when it is obvious but unimportant from the source that this is incorrect. This would satisfactorily prove that someone has not researched their works sufficiently to justify credit for them. It's hard to plagiarise statistics. Jul 20, 2016 at 23:19