16
\$\begingroup\$

When I google "can you counterspell a counterspell", google returns an answer snippet from the site:

enter image description here

Unfortunately, this snippet comes from this answer, which has a score of -17 at the moment, making it one of the top 10 lowest scoring answers on our entire site. This is obviously not a good "featured snippet". Normally, I would say "a score of -17 sends a strong signal to the reader about the quality of the information", however, the google snippet does not reflect this score, and if a reader does not click the link in the google snippet, they will never see that this answer is one of the worst performing answers on our site.

Deleting it should make it so that google's robots cannot fetch it for their snippet, so would it be appropriate to delete it?

Alternatively, if everyone reading this goes and provides "this is misleading" feedback to the snippet, it may just go away on its own.

\$\endgroup\$
8
  • \$\begingroup\$ Hmm... I'd guess this is because the other answers open on/focus on Defence against the strategy as opposed to answering the yes/no question and that confused the ML demon google is using. There's certainly a lesson here about not trusting the snippet google gives you, but follow the link and read the whole argument. \$\endgroup\$
    – Someone_Evil Mod
    Commented Jan 12, 2023 at 16:26
  • 2
    \$\begingroup\$ Amusingly when I now search for the same thing I get a snippet from the -10 scoring answer instead: i.sstatic.net/bxzxb.png \$\endgroup\$
    – Someone_Evil Mod
    Commented Jan 12, 2023 at 16:27
  • 1
    \$\begingroup\$ @Someone_Evil RIP, so there may be no winning against the machine. Just give feedback for misleading snippets and hope they listen. \$\endgroup\$ Commented Jan 12, 2023 at 16:31
  • 4
    \$\begingroup\$ Or we could reconsider our policy and delete bad unsupported answers. I don't think they actually teach anyone anything. \$\endgroup\$
    – Akixkisu
    Commented Jan 12, 2023 at 18:33
  • 2
    \$\begingroup\$ @Akixkisu Alternatively, we could add a "This post was marked as very low quality and thus was hidden. Click here to show it", a la reddit. No need to delete if those things are bad/wrong but still answers if we can instead hide them away \$\endgroup\$
    – T. Sar
    Commented Jan 12, 2023 at 19:44
  • 8
    \$\begingroup\$ There's a technical fix SE could implement by adding the data-nosnippet HTML attribute to very low score answers. I've suggested this on meta.SE. \$\endgroup\$ Commented Jan 13, 2023 at 2:13
  • 1
    \$\begingroup\$ I can't help but notice that both the "No" answers tend to focus on the fact that it shows up when searching. Whereas, I think the question here is "Should this answer be deleted" rather than "Prevent this answer from being shown in Google search". The latter is somewhat related but should not be used at all for justification. \$\endgroup\$
    – VLAZ
    Commented Jan 13, 2023 at 7:54
  • \$\begingroup\$ The question asks 2 distinct things and the title focuses on the second thing. It may indirectly fix the problem if we make the title reflect the question and/or split the question into 2 questions. \$\endgroup\$
    – NotThatGuy
    Commented Jan 17, 2023 at 14:22

4 Answers 4

15
\$\begingroup\$

Yes

The answer is wrong. Pure and simple. What a search engine chooses to show or not is a red herring in this case. At best it served as a catalyst to consider the answer but search engine previews should not really matter for the end result.

What does matter is the goal of the network providing useful information. And I cannot fathom how "wrong" can be "useful" given there is a correct alternative.

Were there a question (naive example) with answers:

Q: What is the result of 2+2?
A1: It is 4.
A2: It is 5.

Then I cannot see any reason to keep A2. Similarly, the answer discussed here is like A2.

Pyrotechnical says that the answer can be improved. I posit it cannot. It can only be entirely changed. That is not what on the network we typically mean by "improvement". Rewriting parts for clarity or correcting some supporting information - yes, but entirely changing the claim made in the answer - not really. At best, further development in time may require the answer to change drastically. If an answer was a "No" initially, it might become a "Yes" in ten years time. Yet that is not the case here.


Still, if we really take it that the answer is improvable, then there is already a mechanism for that. The answer can still be deleted, then the author can edit and flag for a moderator to review and undelete the answer. Users with the privilege to undelete answers (20k reputation) can also vote to undelete if they deem the answer to be useful. Users with 10k reputation can see posts with undelete votes in the /tools section.

\$\endgroup\$
15
  • 5
    \$\begingroup\$ There are plenty of earnest answers on this site which are wrong and the solution for that is to down vote. Frankly, I consider the answer to have some measure of value by virtue that it introduces the notion of interactions between spellcasting mechanics for somatic and material components being in conflict. While the end result of that analysis is incorrect, it is an interesting thought exercise and ultimately one which at least 3 people thought was worthy of giving an upvote. Ultimately, the user has a net negative on their answer and they can choose to delete if they want. \$\endgroup\$ Commented Jan 13, 2023 at 15:11
  • 5
    \$\begingroup\$ Highly-downvoted answers are actually useful in some ways, as a way for the community to express disagreement with an argument or interpretation that may seem plausible. If some future reader was wondering "what about this reason" and there's already a downvoted answer, they don't have to take their own time writing it up. If they just want the correct answer, they should stop reading before getting to the negative answers. This is what makes it different from your multiple-choice answer example, where it's just wrong with no attempt at reasoning or justification. \$\endgroup\$ Commented Jan 15, 2023 at 16:53
  • 1
    \$\begingroup\$ @PeterCordes while I could agree for other cases, I do not see this one as being one. The answer simply wastes space on the page as the correct answers already have the situation covered. \$\endgroup\$
    – VLAZ
    Commented Jan 15, 2023 at 22:43
  • 3
    \$\begingroup\$ It's also worth noting that in this specific case, the author's last comment on their answer indicates that they actually agree with keithcurtis' corrections (the comment explicitly states that the original spell's effect successfully goes off) – meaning they disagree with their own answer. \$\endgroup\$
    – V2Blast
    Commented Jan 20, 2023 at 22:36
  • 1
    \$\begingroup\$ -1. Rules for a ttrpg are not simple math. Right and wrong is not trivial to ascertain, which is the entire point of a vote based system. \$\endgroup\$
    – user2754
    Commented Jan 24, 2023 at 15:55
  • 1
    \$\begingroup\$ @user2754 Right and wrong seem trivial to ascertain in this case. And do remember the context is for this case, not a general case. There is a direct quote from an official source which contradicts the answer. There is no huge mystery. It's not unfathomable, nor is it unclear. The answer is wrong based on pretty objective facts. That being that what it describes is directly counter to literally the same situation explained by official sources. There is no real need to vote on facts here. The question is unhelpful for being completely wrong. \$\endgroup\$
    – VLAZ
    Commented Jan 24, 2023 at 16:01
  • \$\begingroup\$ -1 : Just because an answer is down voted doesn't make it "wrong", just not popular. I've had quite a few answers that I back up with quotes, logic, and example. Yet, it still received multiple down votes as it wasn't the "fun" answer which used different quotes, different logic, and/or different examples. Both answers are valid, but one resonated more. \$\endgroup\$
    – MivaScott
    Commented Jan 25, 2023 at 19:08
  • 1
    \$\begingroup\$ @MivaScott you had it completely backwords. It's not wrong because it is downvoted. It is downvoted because it is wrong. Objectively and very provably wrong. Please keep up with where the discussion is. \$\endgroup\$
    – VLAZ
    Commented Jan 25, 2023 at 19:55
  • 1
    \$\begingroup\$ @VLAZ, I'm completely up to date with what I said. Down votes != wrong. Something can be correct and down voted, something can be wrong and up voted. In this case, it was wrong and down voted. But down votes are not always an indication of right and wrong. Therefore a blanket statement of "If down voted, it's wrong and should be deleted" is not seeing the whole picture. Which is why I down voted this answer. Is your answer "wrong"? No, it's completely valid, but it's not the best solution for every time an answer is highly down voted. \$\endgroup\$
    – MivaScott
    Commented Jan 25, 2023 at 20:10
  • \$\begingroup\$ @MivaScott "Down votes != wrong" not my point "down votes are not always an indication of right and wrong." Not my point - we're talking about a particular answer here not whether or not in general votes indicate correctness. ""If down voted, it's wrong and should be deleted"" and "* it's not the best solution for every time an answer is highly down voted.*" are both things I've never claimed. AGAIN, my point is that the post is wrong. Nothing to do with the votes but the facts. Reciprocate and don't strawman. You clearly are not following the discussion. \$\endgroup\$
    – VLAZ
    Commented Jan 25, 2023 at 20:30
  • 1
    \$\begingroup\$ @VLAZ I think the point that Miva's attempting to make is the same one that I'm attempting to make, which is that just because an answer is wrong and downvoted does not mean that it crossed the threshold into obligating deletion via moderation. It was an earnest attempt to answer the question (4 people think so). For that reason, it should not be deleted via moderation tools, but only by the respondent if they choose to. \$\endgroup\$ Commented Jan 30, 2023 at 13:42
  • \$\begingroup\$ @Pyrotechnical the respondent who hasn't been active for 3 years? Also, why is it the respondent the only person who should do this? The model of SE is that the content is maintained by the entire userbase. Everybody can curate the content. Some of the actions require more reputation to unlock, yet the point is to give all regular users tools to moderate content. Because the content is shared. It's not a forum, where each person "owns" their posts. It's our combined effort that keeps the content relevant and useful for future visitors. Gating this away is a disservice to everybody. \$\endgroup\$
    – VLAZ
    Commented Jan 30, 2023 at 17:37
  • \$\begingroup\$ @VLAZ, to the contrary, each person does own their posts and they own the up votes and down votes that come with it. Furthermore, on RPG.SE, we take it extremely seriously that people posting questions are the sole authority on how their question is phrased. Similarly, people's answers are their own as well. This is especially true when an answer focuses upon an incorrect interpretation of the rules because it's very feasible that others might make a similar interpretation, so it's helpful for us all to see that the answer is wrong and why. \$\endgroup\$ Commented Jan 30, 2023 at 21:55
  • \$\begingroup\$ @VLAZ it's from this combined effort that an overall shared content is created that is useful for others in the future. And those future others may benefit from a wrong answer that incorrectly asserts that a caster can't perform the somatic components of two spells simultaneously. Which honestly isn't an unreasonable common sense assumption, I could see why someone would make it. It's only clearly wrong because WotC released a rules clarification document that definitively contradicts. And after all, WotC has never released rules and rulings that are contradictory /sarcasm. \$\endgroup\$ Commented Jan 30, 2023 at 22:08
  • \$\begingroup\$ @Pyrotechnical I guarantee you that the licensing employed by SE ensures that anybody can use, reuse, and even change the posts. That's how editing works - hence why edits are allowed on others' posts. That's baked into the SE model. If it were true that somebody could post something and then never return, then SE is no better then the forums of past it sought to replace. And it did replace them. I absolutely disagree on the autonomy and independence you claim posts have. They don't. Do feel free to seek clarification on this on MSE. I'm not really up to repeating this over and over. \$\endgroup\$
    – VLAZ
    Commented Jan 30, 2023 at 23:24
14
\$\begingroup\$

No - That's Google's Problem

Google has a faulty algorithm that is picking up the wrong answer. Google has control over that algorithm - we don't. We might change our policy to do something that works well-ish with Google's algorithm currently. But what if the algorithm starts creating snips for answers that have a -1? Or even only a +1, when there's a +200 accepted bountied answer to the same question that's short enough to snip? Moreover, what if Bing starts doing that?

We don't control these sites, so we can't guarantee that we'll actually be able to fix the problem anyway. Trying to do so now as a matter of policy (even going against our policy to do it, as Pyrotechnical explains) will result in our policy becoming Sisyphus, rolling forever up a hill to perfect search engine integration everywhere but never being able to achieve it.

\$\endgroup\$
5
  • 7
    \$\begingroup\$ I'd like to see an argument why should the answer be kept, rather than focusing on Google. Since that is a red herring. I could strap on explosives to my car and say it's everybody else on the road should be careful not to hit me. Yet, I'd hope others will still say my car is dangerous and I shouldn't be doing that. \$\endgroup\$
    – VLAZ
    Commented Jan 13, 2023 at 7:19
  • 1
    \$\begingroup\$ @VLAZ Pyrotechnical and yourself already discuss that well in your answers - if we want to delete the answer for other reasons, I'm all for it. But I don't want to encourage a specific policy exception just because of something Google is doing poorly. \$\endgroup\$
    – ESCE
    Commented Jan 13, 2023 at 17:19
  • 4
    \$\begingroup\$ This is a problem Google can fix, but Stack Exchange is probably more capable than Google of figuring out what is and isn't appropriate Stack Exchange content to potentially be displayed in snippets (and there is apparently a way to tell Google this). It's about saying "this is bad, don't use this", whereas Google should figure out how to prioritise the not-bad content and what from that to put in a snippet. If Google picks a +1 answer above a +200 answer, that's their problem, but if they pick a -17 answer, that's SE's problem (too). \$\endgroup\$
    – NotThatGuy
    Commented Jan 17, 2023 at 14:34
  • 2
    \$\begingroup\$ @NotThatGuy I'm all for adding that tag to the answer! I wasn't aware Google gave us any control over it (aside from displaying content or not). I still don't think we should delete the answer (which goes against our current policies), but adding data-nosnippet to the divs is fine (and doesn't require us adjusting our policies). \$\endgroup\$
    – ESCE
    Commented Jan 17, 2023 at 16:09
  • 2
    \$\begingroup\$ It's worth noting that in this specific case, the author's last comment on their answer indicates that they actually agree with keithcurtis' corrections (the comment explicitly states that the original spell's effect successfully goes off) – meaning they disagree with their own answer. \$\endgroup\$
    – V2Blast
    Commented Jan 20, 2023 at 22:37
7
\$\begingroup\$

No

Stack Exchange sets the criteria for answer deletion as follows:

  • The answer is extremely low quality: There is little to no scope for improvement.
  • The answer doesn't attempt to answer the question; it may be a comment or a separate question altogether.

Personally, whenever I conduct a review regarding deletion, I take both of those criteria very seriously. There are plenty of bad answers on this site, but not very many actually make it into the review queue for deletion.

I think a good reason for this is because the bar for deletion is pretty high.

Is the subject Answer extremely low quality with no scope for improvement? I don't think so. The Answer has a score of -17, but the breakdown is +3 and -20. The respondent also provided multiple comment responses attempting to clarify their point of view. While the community at large does not agree, this doesn't make the answer have 'little to no scope for improvement.'

Is the subject Answer a legitimate attempt to answer the question? Unequivocally, my answer is yes. Do I agree with the answer? No. Does my opinion on the answer matter for this criteria? Also no.


Overall, our prerogative is to use moderation deletion tools in conformance with the site's stated criteria. Extending beyond that for things like search engine optimization is not something I can justify us doing barring a very good reason. From my review of the site's analytics, I can't see any startling data that justifies altering our current path (granted I'm not overly proficient in interpreting the analytics).

enter image description here

enter image description here

\$\endgroup\$
9
  • 6
    \$\begingroup\$ "Is the subject Answer extremely low quality with no scope for improvement? I don't think so." It seems to me that here the only way to improve the answer is to change it to a completely different answer. Because, as it stands, it is completely and provably wrong. Thus an improvement would be to make it correct by changing the entirety of what it claims. With this in mind I wouldn't call it an "improvement" since it will be a wholesale replacement. Furthermore, if the goal of a Q&A site is to provide correct useful answers, then what is the use keeping wrong answers? \$\endgroup\$
    – VLAZ
    Commented Jan 13, 2023 at 7:17
  • \$\begingroup\$ @VLAZ there are plenty of earnest answers on this site which are wrong and the solution for that is to down vote. Frankly, I consider the answer to have some measure of value by virtue that it introduces the notion of interactions between spellcasting mechanics for somatic and material components being in conflict. While the end result of that analysis is incorrect, it is an interesting thought exercise and ultimately one which at least 3 people thought was worthy of giving an upvote. Ultimately, the user has a net negative on their answer and they can choose to delete if they want. \$\endgroup\$ Commented Jan 13, 2023 at 15:04
  • \$\begingroup\$ @VLAZ to that end, is there a manner by which the respondent could structure a more fully fleshed out argument that affirms their stance and is mechanically correct? Maybe. Could they evaluate the situation more completely and perhaps reverse their stance, thereby changing their answer? Also maybe. But both of those avenues are present and the answer is earnest, so I don't think it's appropriate to delete just because the majority disagrees. \$\endgroup\$ Commented Jan 13, 2023 at 15:07
  • 3
    \$\begingroup\$ I said it's verifiably wrong. Adding more support is not going to change that. And changing the stance would not really be an "improvement" but a brand new answer. At any rate, deleted posts are still editable and can be undeleted, thus I don't see why it has to stay. Do also note that the answer hasn't been touched for 7 years. While that isn't a measure by which to decide whether to delete a post or not, it is a very strong indication that the author is not interested in improvement. Nor is anybody else as evidenced that no other user has stepped up to edit, either. \$\endgroup\$
    – VLAZ
    Commented Jan 13, 2023 at 15:33
  • \$\begingroup\$ @VLAZ One may also observe that the post's author has not logged in for over three years. \$\endgroup\$ Commented Jan 13, 2023 at 19:49
  • 1
    \$\begingroup\$ It's also worth noting that in this specific case, the author's last comment on their answer indicates that they actually agree with keithcurtis' corrections (the comment explicitly states that the original spell's effect successfully goes off) – meaning they disagree with their own answer. \$\endgroup\$
    – V2Blast
    Commented Jan 20, 2023 at 22:36
  • \$\begingroup\$ +1. No reason to look outside SE for the answer here. Surely we can do our own laundry. iirc if 5 people flag it as not an answer it will be deleted by the system. "This was posted as an answer, but it does not [...] answer the question. It should possibly be [...] deleted altogether." - no meta post required; no moderator intervention required. This 12y site is either self-correcting or the experiment has failed. - It's at -28... how's that not happened yet? Happens to me all the time. I must have my own FBI guy I guess.... \$\endgroup\$
    – Mazura
    Commented Jan 21, 2023 at 20:11
  • \$\begingroup\$ Except that's probably for being rude, not wrong. WHICH IS WHAT DVs ARE FOR. I blame all the users who didn't chide them into getting with the program and self-deleting. That's where it "failed". \$\endgroup\$
    – Mazura
    Commented Jan 21, 2023 at 20:18
  • \$\begingroup\$ @Mazura "iirc if 5 people flag it as not an answer it will be deleted by the system" no, if 5 people flagged it as NAA, it would have 5 NAA flags on it. That's it. A diamond moderator would still need to review tthis. NAA flags do not automatically delete content. Spam and Rude or Abusive flags will destroy a post however, they also come with a hefty penalty to the user account and are not appropriate in this case. \$\endgroup\$
    – VLAZ
    Commented Jan 28, 2023 at 15:28
7
\$\begingroup\$

Although we don't need to do anything on account of Google itself, this highlights a potential gap for us to address in our Q&A collection:

  • We don't have an actual question covering whether counterspelling a counterspell is actually permissible, or how it works.
  • We don't have a clear answer stating “yes it is, here's how.”

I think we should address that by having a Q&A that, you know, states this. “Can I counter a counterspell? How can I do that?” Then people can learn from that, Google can pick it up, and we can reference it in the question we're discussing.

I considered suggesting that this question should be edited to do that. It comes pretty close: the question asks how to respond to it when this happens and none of the upvoted answers challenge the legitimacy of the events, and the answers that do are downvoted, which goes pretty far to imply things. But proving you can do that is a lot of additional work and a different topic, so that's a major scope change we shouldn't just slide in here.

\$\endgroup\$
9
  • \$\begingroup\$ I agree that a direct question asking "Can I counterspell a conterspell?" is useful, but it is not clear how we should proceed. Editing the question does not seem fair, wrt the original OP, and a new question could be closed as duplicate. \$\endgroup\$
    – Eddymage
    Commented Jan 16, 2023 at 21:52
  • \$\begingroup\$ @Eddymage I don't think it ought to be closed as a duplicate — they're different questions with different answers, and we've got good reason to keep them both separate. We do have one answer from Rowdy saying you can counter a counterspell, but it's not actually answering the question that was asked by Nicholas at all (which was not can/how do I do this but what do I do to protect myself against this), so it would just get removed and posted as an answer to the new question. \$\endgroup\$ Commented Jan 16, 2023 at 22:49
  • \$\begingroup\$ @Eddymage if the dupe is a worry, why not make the question doppelgreener proposes, and then make the old question a dupe of that? (I'm not 100% convinced it'd be a dupe, but I'm also not sure if there are some other questions or answers that would be). \$\endgroup\$
    – ESCE
    Commented Jan 17, 2023 at 16:11
  • 1
    \$\begingroup\$ @ESCE and doppelgreener: maybe you're both right. I am preparing one, maybe with a self answer (since it is already answered in SAC), with maybe some discussion about casting times. \$\endgroup\$
    – Eddymage
    Commented Jan 17, 2023 at 16:51
  • 1
    \$\begingroup\$ "seeding" questions seems off to me. We don't actually have someone who has a real question on this - and it hasn't actually been asked (or answered) yet. Just creating a question to fill the gap without an actual need for an answer seems weird. It may be just fine, it just feels wrong. Our questions on here should come from honest sources of problems. \$\endgroup\$
    – NotArch
    Commented Jan 18, 2023 at 15:33
  • 2
    \$\begingroup\$ @NautArch I'm looking at this as a potential to create a canonical question on a reasonable topic we haven't tackled, and this very meta Q was prompted by someone googling this very question. (Maybe it's made obvious in the text in fact?) \$\endgroup\$ Commented Jan 18, 2023 at 15:37
  • \$\begingroup\$ @doppelgreener, "seeding" questions is something I brought up a while back. There was an influx of questions all centered around the same theme with only minor variations. It becomes difficult to say which should be the example, and which should be the duplicate as the first isn't always answered the best nor is the best explained question. (1/2) \$\endgroup\$
    – MivaScott
    Commented Jan 25, 2023 at 19:49
  • \$\begingroup\$ Which then sparks the questions, at what point do we create a "seed" question and how do we pull the "best" answers from all the questions that came before? How do we mark the question as a "seed" so people don't mark IT as the duplicate? Will everyone have to rewrite their answers? How do we choose the "right" answer when there were 3, 4, 5, 10 different versions of the same question all with "right" answers? In the end, it may just muddy things up more than fix them. (2/2) \$\endgroup\$
    – MivaScott
    Commented Jan 25, 2023 at 19:49
  • \$\begingroup\$ (1) This isn't about question seeding; that concept is unrelated. What I'm describing is the process of creating a canonical question. (2) If it's a decent canonical question that stands on its own, this will be self-evident. (3) Nobody will have to recreate their answers because it will be a different question. \$\endgroup\$ Commented Jan 25, 2023 at 21:14

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .