Wikipedia talk:Today's featured article/August 21, 2022

i am admittedly having a hard time reconciling the numbers reported for the japanese casualties. if the numbers in the infobox are correct, then if 774 to 777 soldiers had died out of 917, i would assume that 140 to 143 had survived the battle. however, the lead states that "[a]ll but 128 ... died" and the blurb states that "only 128 ... survived".

one of the footnotes explicitly states that the frank source reported that 777 died, while another footnote confusingly states that "Smith says 128 of the original 917 ... survived, meaning 774 were killed after subtracting the 15 captured from the total lost in the battle". i am wondering if the smith source considered a soldier to have "survived" if the soldier did not die and was not captured. if this is true, then i am assuming that the number of soldiers that did not die should include the 128 that "survived" and the 15 that were captured (assuming that they were not killed shortly after being captured), bringing the total to 143. would it be more appropriate to be more conservative in the blurb and replace "only 128" with "at most 143"?

notifying tfa/r nominator Z1720 by ping, and urfa/2020 reviewers Hog Farm by ping and SandyGeorgia by user talk page. (fac nominator Cla68 has not edited since 2016, and urfa/2020 reviewer AustralianRupert has retired.) dying (talk) 09:35, 19 August 2022 (UTC)[reply]

Unfamiliar with this myself and have to leave for work soon. Also pinging Nick-D to see if they have any familiarity with this subject. Hog Farm _Talk 12:18, 19 August 2022 (UTC)[reply]

Thanks for the post, Dying. Cla68 is no longer around, but it would be very unusual for him to make a mistake. I have three ideas: 1) it is possible that my husband has the sources and I can check with him when he is home later today (or poke around his extensive library myself); 2) I have an odd recollection of this having come up a decade ago and there being a logical answer (we need to review the talk page archives and the FAC); 3) we need to check old versions of the article to see if there was vandalism. If we come up with nothing, and Nick-D has no answer, we may have to figure out a rewrite. I'll be on this over the day, and watchlisting. (It's OK to ping me now, since I finally got a system that I get along with, where pings go only to my email, which I check twice a day ... but posting to my talk is faster.) SandyGeorgia (Talk) 14:10, 19 August 2022 (UTC)[reply]

Dying, my husband's library has nothing on this. I seem to recall something somewhere about those who died later, but unsure of the source of the discrepancy, and have not yet found time to examine archives. Pending Nick-D feedback, anyway, might we switch the blurb from discussing the small (confusing) number who survived to instead the sourced number killed ? The Marines counterattacked after daybreak, killing at least 774 of Ichiki's original 917 men. SandyGeorgia (Talk) 16:10, 19 August 2022 (UTC)[reply]

Ah heck, I guess that won't solve it, as the 128 is the sourced number, and we seem to have backed in to the 774. Back to the drawing board. SandyGeorgia (Talk) 16:12, 19 August 2022 (UTC)[reply]

@Dying and Nick-D: my memory worked; this was covered at the FAC, Wikipedia:Featured article candidates/Battle of the Tenaru. At the time, Cla68 said: I was rounding for aesthetic purposes in the intro, I've now changed it to read 917. I know that 917-15 (captured)-128 (killed) doesn't equal 777 but that's what the source states are the numbers. The source admits that the numbers don't completely add up but it's the best that can be done with the surviving primary source available. So that's all we've got. We should re-do both the blurb and the lead so that we aren't referring to specific numbers, and we should add an explanatory footnote somewhere. Cla indicated that the source admits the numbers don't add up; we need the wording from whichever source that is to add a footnote. SandyGeorgia (Talk) 18:00, 19 August 2022 (UTC)[reply]

SandyGeorgia, i'm impressed that you remember that this issue was mentioned at the fac review about 16 years ago. i like the idea behind the "killing at least 774" proposal, though the rewording suggests that those 774 were all killed during the counterattack. although 128 appears to be the number sourced from smith, it appears that the frank source explicitly reported that 777 had died. would the following be an acceptable rewrite?

The Marines then counterattacked after daybreak. Out of Ichiki's original 917 men, around 777 died.

alternatively, the following rewording avoids referencing either 774 or 777, though i am not sure if it violates wp:calc. (note that this option would push the blurb past the character limit, which can be resolved by replacing "subsequent" with "later" in the blurb's concluding sentence.)

The Marines counterattacked after daybreak; only about a seventh of Ichiki's original 917 men survived.

dying (talk) 20:45, 19 August 2022 (UTC)[reply]

I have to go out for the rest of the evening, and we haven't heard from Nick-D (MILHIST), so the decision is in your competent hands. I hate the word subsequently if that helps. This is a hard call to make, but whichever option you choose, please also sync the article lead. Wish we had a better answer, but it is what it is. The around option may be safest at this late stage, but what do I know ... I am always shocked at what gets picked on vs. what passes on the main page. Best, SandyGeorgia (Talk) 21:33, 19 August 2022 (UTC)[reply]

@SandyGeorgia and Dying: - I just picked up Neptune's Inferno by James D. Hornfischer today for $2 at a used bookstore. Hornfischer is a recent academic and his book about the Battle off Samar was well-received. p. 111 of Hornfischer places Japanese dead at over 800, has the 15 prisoners, and says only "a few" Japanese escaped entirely. Also accessed Stanley Jersey's Hell's Islands (TAMU Press) via the wikipedia library, and p. 213 has the Japenese losses at "about 777" killed with the 15 captured. Jersey lists 30 Japenese who escaped the fighting and 40 who were in the rear burying dead from earlier action. Interestingly, both Hornfischer and Jersey give American losses as 34 killed and 75 wounded, so I wonder if some of the confusion is if the total reported casualties are including only the Aug. 21 action or all fighting beginning when Ichiki's men landed. Given the ambiguity, I don't think trying to work backwards to 128 is appropriate here. Hog Farm _Talk 22:23, 19 August 2022 (UTC)[reply]

I've also left a note pointing to this discussion at WT:MILHIST. Hog Farm _Talk 22:25, 19 August 2022 (UTC)[reply]

I'm out for the evening so can't help; great work, Hog Farm! SandyGeorgia (Talk) 22:28, 19 August 2022 (UTC)[reply]

I got ahold of Frank from the Internet Archive, and Frank has "at least 777" killed on p. 156. A close reading of p. 681 of Frank indicates that the 128 figure includes 11 people already there before Ichiki and not counted in in his 917, so backtracking 128 from 917 to get losses clearly isn't right. Two sentences of note p. 681 that I don't have time to fully parse right now: If it is assumed that 117 members of Ichiki's First Echelon reached Taivu, this would make losses an even 800, including prisoners. The 900 figure would include losses in Shibuya's patrol, whose bodies would not have been along Alligator Creek ("the Tenaru"). I'm unsure if Frank's reference to "the 900 figure" is a mistake for the even 800 estimated earlier, or a vague reference to Ichiki's roughly 900-man total force, but the review of Frank strengthens my view that we shouldn't be back-calculating numbers here and that in rough terms is probably the best way to present numbers here, because reading the sources suggests to me some ambiguity between what figures are only for the Aug 21 action at Alligator Creek and what figures cover the patrols/other skirmishing. Hog Farm _Talk 22:42, 19 August 2022 (UTC)[reply]

that's great work, Hog Farm. also, i feel like an idiot for not having thought about looking for the frank source; i had only looked for the smith one, albeit unsuccessfully. i agree with your analysis, and am also further confused by frank's description on page 681 of "777 'killed' (presumably including the fifteen prisoners of war)", as this makes me unsure about the accuracy of frank's earlier count of "at least 777". i am guessing that the "900 figure" frank referenced appears in the "S.S., Vol. 14" source mentioned on page 657, but i do not have access to that source to confirm. in any case, i agree that presenting a rough estimate would probably be best here. dying (talk) 19:42, 20 August 2022 (UTC)[reply]

The crack on the computer screen now reaches from top to bottom, which isn't good but does at least keep it from selecting a single point. @Dying and SandyGeorgia: - I've made changes to the blurb and article, with edit summaries on some of the potentially controversial ones to point to this discussion. Changes to article. I've made changes to the article to reflect this, including adding a brief statement citing Hornfischer, Jersey, and Frank p. 681 for casualties of about 800. I've also tweaked the infobox, removed a claim in a footnote that Frank states that casualties were 777 (because Frank gives somewhat different details in his notes on p. 681). Less on-topic, but I've also removed an external link that was broken because it relied on flash. Hog Farm _Talk 21:45, 20 August 2022 (UTC)[reply]

looks good, Hog Farm. thanks, both of you! dying (talk) 22:16, 20 August 2022 (UTC)[reply]

Hog Farm, you're a trooper. Hope your computer gets well soon. Dying, how ever so very nice it is to work with someone who addresses main page issues to solve problems rather than shoot gotcha arrows :) :) Bst, SandyGeorgia (Talk) 04:37, 21 August 2022 (UTC)[reply]

haha, thanks, SandyGeorgia. considering your impressive work with featured articles, i am similarly glad to be working with someone as knowledgeable as you. dying (talk) 03:18, 27 August 2022 (UTC)[reply]