Talk:Marion Parris Smith

Latest comment: 7 years ago by Beatley in topic infobox

infobox

edit
the m:Objective_Revision_Evaluation_Service indicates that the infobox adds to the quality of this article. if you have questions about the algorithm, ask there.
i presented objective evidence, and you presented a subjective personal essay. Beatley (talk) 18:40, 10 January 2017 (UTC)Reply
Please explain why you think this infobox is of benefit to this article. CassiantoTalk 18:41, 10 January 2017 (UTC)Reply
ORES says:
with infobox "B": 0.020803113987792523, "C": 0.03316255459156076, "FA": 0.0031200106133461377, "GA": 0.006058826715413562, "Start": 0.3985832693509729, "Stub": 0.5382722247409142 [1]
without infobox "B": 0.015729764197649678, "C": 0.02176572531599291, "FA": 0.0019667612672566977, "GA": 0.00430964431963926, "Start": 0.23811408793540903, "Stub": 0.7181140169640521
any objective facts to offer ? Beatley (talk) 23:14, 12 January 2017 (UTC)Reply
The objective facts are that you have been asked multiple times to provide understandable explanations, in English, for why you think an infobox will be an improvement, and all you ever do in response is to post meaningless and unintelligible strings of numbers. It's not a helpful contribution to the discussion. —David Eppstein (talk) 19:37, 14 January 2017 (UTC)Reply
So, to follow-up, I tried an experiment with ORES. This version of the article, that I created for the experiment, has the same effect as adding the infobox, to my mind: by purely duplicating what's already in the article, it adds no useful information, makes it harder to find the existing information, and has the effect of making the article appear longer and (to unintelligent software that can't tell the difference) better written. And, surprise surprise, ORES likes it better. ORES correctly judges all of these versions as on the low side of start class, giving the non-infobox version 0.24 chance of being start / 0.72 chance of being stub, while it gives the infobox verson 0.4 chance of being start and 0.54 chance of being stub. But it also gives my experimental duplicated-content version even better ratings, 0.49 chance of being start and 0.31 chance of being stub. I conclude that ORES is reasonably accurate at judging the quality of good-faith human-generated content (such as the original version of this article), but really bad at distinguishing useful new content from useless robotic duplicate content. Therefore, it is not an appropriate tool to be using in this context, where the debate is specifically whether the new content is useful and informative, not just whether it adds length to the article. —David Eppstein (talk) 21:21, 14 January 2017 (UTC)Reply
that's very nice. perhaps you would care to share your insights with the ORES team. your conclusion does not follow from your premises. you use aesthetic judgement to agree with it in one case, but not another. do you have any objective evidence that an infobox is not an improvement? are you going to stand upon your subjective judgement to prevent linked open data improvement for biographies? why are you having a policy discussion on a random article page? Beatley (talk) 21:42, 15 January 2017 (UTC)Reply
If your vaunted "objectivity" doesn't immediately tell you that my experimental version is much worse than the original, then something is wrong with you. And yet ORES ranks it as significantly better. What does that tell you about the value of an ORES score? And I'm not trying to change policy; I'm only trying to explain why I think your arguments here are worthless. —David Eppstein (talk) 22:22, 15 January 2017 (UTC)Reply
@Beatley: per WP:INFOBOXUSE the decision on whether to include an infobox is made through consensus on this talkpage. You're certainly welcome to argue that there should be one, but an ORES score doesn't automatically trump the en-WP consensus requirement. At present the consensus on this page opposes an infobox. Unless/until that changes, please don't restore it to the article. -- Euryalus (talk) 11:01, 16 January 2017 (UTC)Reply
gosh, wouldn;t be a nice thing, if an independent third party developed some objective metrics, as messy as they might be, to measure article quality? wouldn't it be something if it was the slightest consideration in discussions. what would we ever talk about? this person says two people are not a consensus. how many are? because i can get as many as you want. consensus is not shifting the burden on the other party, it requires shared values. Consensus_decision-making#Blocking is not required in the process, you may also stand aside. Beatley (talk) 19:03, 17 January 2017 (UTC)Reply