Talk:digitgrade

Too rare for a misspelling entry; cf. (1,021) with (39). Delete. †  ﴾(u):Raifʻhār (t):Doremítzwr﴿ 18:09, 28 September 2009 (UTC)


 * What's our cutoff? 3.7% seems like plenty for a misspelling, especially in print works that have (mostly) undergone some degree of editing.  I wouldn't have gone out of my way to create it, but I'm inclined to think that anything over 1-2% is keepable.  Further, turning to the web -- which is what most previus misspelling discussions have been based on -- I get 7530 for  vs. 44100 for .  At 14.6%, that's a higher error rate than accomodation (which is 10.9% by my Google).  I'm OK with it if we want to adopt a 5% b.g.c. cutoff or similar, but unless we do adopt a strict numeric cutoff, this looks keepish to me. -- Visviva 00:36, 29 September 2009 (UTC)


 * I rechecked: the number of hits for *undefined: became thirty-two by clicking on the last screenful of hits; of those, these:, , , , , , , , , , were invisible and this one is a scanno of the correct  (see the centre of its “(d)” paragraph). So, only twenty book hits for *undefined: are confirmed, or <1.96% of the number of hits for undefined: (before similar adjustments are made to the raw b.g.c. hits for it). Google Fight gives undefined: (11,800) vs. undefined: (583), making undefined: <4.95% as common as undefined:. Bear in mind as well the fact that we’ve had an entry for undefined: since 15:40, 10 May 2006 , so that’ll inflate the number of hits Google yields for undefined:. Compare  (217) with  (87,100): making undefined: <0.25% as common as undefined:. This, I think, is a very good reason not to listen to Google Web Search hits when considering such things in future. I conclude again that this is too rare to be kept as a common misspelling. †  ﴾(u):Raifʻhār (t):Doremítzwr﴿ 02:06, 29 September 2009 (UTC)


 * Good point. But if we apply the same correction to the b.g.c. hits for '+digitigrade', the total number of hits becomes 336: .  So the actual percentage would seem to be around 8.7% [32/(336+32)], which is again quite high, particularly given that we are dealing (primarily) with edited works.  I hadn't heard of any "common misspelling" criterion that would place the cutoff above 5%.  -- Visviva 02:29, 29 September 2009 (UTC)


 * OK. I really didn’t expect 660 hits to be fake. Yeah, I suppose that’s common enough to be retained. However, I think the point about the especial unreliability of Google Web Search still stands. †  ﴾(u):Raifʻhār (t):Doremítzwr﴿ 02:38, 29 September 2009 (UTC)


 * Google Web counts are fiendishly unreliable (and often impossible to double-check), yes. Perhaps we could use a cross-section of the smaller Googles instead? -- Visviva 02:42, 29 September 2009 (UTC)


 * Yeah. Book Search seems OK with some corrections; Groups Search has inspired suspicion; I haven’t tested the others enough to comment. †  ﴾(u):Raifʻhār (t):Doremítzwr﴿ 02:46, 29 September 2009 (UTC)

Consensus was Keep, deletion template removed --Kakurady 03:10, 17 December 2009 (UTC)