Tell HN: Please bring back comment scores
I miss comment scores.
Perhaps we could strike middle ground by displaying a range, eg. 1, 2-5, 5-10, 10+
Perhaps we could strike middle ground by displaying a range, eg. 1, 2-5, 5-10, 10+
"that is not how mysql works" with 2 Points And "that is not how mysql works" with 102 points Are not the same piece of info.
I don't see any benefit to hiding this from people. It also helps newbies understand the customs here (they can get feedback on other people's comments)
edit now that I'm not on my phone:
When you google things, you probably skip over results like daniweb, about.com, or expertsexchange.com, and hope for a stackexchange page.
The little URL at the bottom of the description tells you "hey, this is from $foo source" and you're a smart human being that can put this information to good use.
Of course, you could make a case that this is bad, because you should read each one of the results and judge it based on its merits. Maybe we should even strip all of the identifying information away from the page, and just let it stand on its own (this would be a neat experiment, actually [and that's the experiment that I think we're performing here]).
The point counts on the comments act just like the URL does on google results. It's not saying "this is definitely 100% accurate", but it is useful piece of information that we can put to good use. Depriving us of this information doesn't break the comments, but I have certainly found myself reading comments a bit less lately as a result of it (instead of actually reading comments, I'm usually just skimming them now). With comment scores, things seem to have a bit of order to them, without, it just feels like a lot of people shouting at one another.
(Maybe this was the point?)
Naturally, I'm never going to stop reading HN; it is by far my favorite website on the internet. Complaining about the lack of comment points here is like complaining that my favorite bar switched to a new, very slightly different glass. I can see the difference, but it's not really going to change my habits.
While technically you're right, that seems like taking a really simple idea and making it really complicated.
Comment scores act as a "trend" of what people are thinking - it's not like a football game where the highest score wins. If I see a discussion it's useful to see a) how many people care enough to bother to vote either way and b) what sort of proportions are involved.
I think most people here can accept that not every comment in a thread gets the same number of eyeballs, and therefore the "scores" are not necessarily directly comparable.
The problem is that scores aren't like trends; they're like integrating a 'popularity' measure over time. As a consequence, you get a strong bias towards the stuff with the most points being comments from near the beginning of the discussion, not what's most correct or interesting or insightful.
When this thread is referenced to a week from now there is no way to see what sort of agreement occurred with either one of us. There are a bunch of "for" comments, and a bunch of "against" comments, but ultimately no resolution (that either of us can see.) So a plethora of "me too" postings occur to fill the void.
Of course the agreement in, and of itself, means nothing - as noted elsewhere popularity does not equal correctness, however it's a data point which exists, and in some cases is useful.
That just means that people will have to read the comments and decide for themselves based on the merits of the comments.
One might argue that that is better than deciding based on points because some people interpret points as assent to the beliefs expressed in the comment while others interpret points as a indicating the technical quality of the comment, e.g relevant supporting rationales and examples assembled logically.
The number can be used to gauge whether you should stop and read something carefully. Maybe an expert comes along and makes an extremely informative comment and maybe I just scan it. Seeing a bunch of upvotes indicates that I should probably slow down and try to learn something from it.
Or maybe a well-reasoned debate has a large voting discrepancy. I value the collective judgement of the HN community and this information helps me learn from the arguments.
The value of this exceeds the more fair voting that we might have now.
In your argument, you are using yourself as the measure of value, rather than using the community, i.e. saving you as an individual time by helping you ignore the less popular side of an argument doesn't demonstrably improve the quality of comments of the community in general (or even your comments in particular - though it may help you choose which side to comment on in order to maximize the upvotes your comments receive).
In the case where a "well-reasoned" debate has a large discrepancy then using points as a measure of value indicates an interpretation of votes as agreement with ideas rather than an objective indication of a comment's quality.
You imply that the sole benefit of the score would be to allow me to form my own opinion. Whereas I hypothosize that I already know my own opinion - the score allows me to measure the opinion of others. Those who have no opinion of their own are easily swayed, with or without a score, however for those with an opinion of their own, it is useful to know the relative popularity of that opinion.
More data is almost always better - especially for those who respect the value, and limitations, of raw data.
I think part of the reason for hiding the points is to affect the process of forming opinions and better facilitate changing opinions already formed - I doubt that the use of scores to confirm one's own opinions is seen as beneficial to the community. For people who believe that scores should reflect the quality of a comment rather than the popularity of the sentiment it expresses your argument tends to support the hiding of scores.
It's almost as though the future reader would need to form an opinion... for themselves.
"likely ... positive", granted, but strong correlation, no.
Keep in mind the change was an experiment to mitigate effects of HN growth and the corresponding trend towards average of audience. The "tend" you mention had been unquestionably moving back to the middle.
That said, I don't have a problem with points switching on after the comments page goes to archive mode. I can ignore them then, while those who want to see them then can neither vote based on bias nor vote to correct perceived bias.
Discussion and reading should be thoughtful instead based on clicks
One's own comments can be self-appraised by seeing one's own score in real time, influencing one's own future thoughtfulness.
Edit: of course, this argument would be satisfied by no more than making comment scores visible on discussions older than n days.
I agree that this might work better for determining correctness of informative comments than just plain point scores, however point scores work better than nothing at all.
I miss the community-provided context of comment points.
I think having the scores hidden is a good decision. It means that we concentrate on content rather than a 'karma game'.
""that is not how mysql works" with 2 Points And "that is not how mysql works" with 102 points Are not the same piece of info."
In my opinion, both of those comments are bad comments unless they further qualify the statement. I don't think karma should be used to infer whether a statement is correct.
Except if it's nested.
Sometimes the most useful comments are replies to less-interesting top-level comments. In this case the lower-scored item is positioned higher up the page than the higher-scored item.
Also I love the discussion in this post since, when with Georgify (http://news.ycombinator.com/item?id=2410195) I had tried first to seperate the upvotes with points of a comment, everyone (and yes a blanket term) had said it was a wrong thing to the extent that I had to make changes.
Of course it is not my site to make decisions for the users of the site, but still almost a dual reaction when the same decision was taken on HN by two different people in a very nearby time span, one being me (for Georgify) and the other being PG himself (for HN).
Completely agree. In my limited tenure here at HN, I've learned that karma scores have never been about being correct, factually or otherwise. Rather, they should be given to comments that help increase and move the discussion around the given topic even if that commenters opinion or assumption on something isn't factually accurate.
I wonder how these people ever managed to make it through a newspaper without scores telling them which articles are worth it.
Learn to survey, skim, and dive in. All your endeavors will be the better for it.
PS. Your method (don't read if don't like score) also ensures you won't be soiled by accidentally reading an unpopular but truthful point of view.
Then why sort comments at all? just order them in date order (or the reverse) and be done with it. Let the reader decide what is useful and what isn't.
Seems like having comments sorted in an implicit order of 'goodness' (even if there is no score) would handicap people in the development of their ability to 'survey, skim and dive in'.
You make an excellent point, applicable here: newspaper brand = web brand.
Articles not flagged [dead] on news.ycombinator.com already ensure some quality, and both similarly share a "front page".
My point was comparing newspapers brands/articles to HN comments, not articles, since you don't have much guidance anymore over what comments are worthwhile or not.
This comment is near the top of the page. Does that make it agreeable?
There's a lot of faith in the wisdom of crowds with these systems and that's not a faith I find particularly convincing. The lowest common denominator shouldn't be encouraged. These systems are good for getting rid of trolls and other cruft, but not magical quality or truth machines.
You are absolutely right though about source links on Google. They give us context for Google's ordering. While Google has done the work to return the best ordering of results possible, which might be the best results in the average case (though maybe not) site have earned a personal reputation for us. Personally, I know that I don't want to check the expertsexchange.com sites, so I skip over them.
I think that what HN would benefit greatly from is a more codified personal reputation system. Right now, the only way we have that is by learning to recognize the names of HN posters which, for me, has only happened with a precious few of the most prolific and respected. If HN would only give me some indication that I had up (or down) voted this commenter n number of times in the past, that would be far more useful than displaying the current vote count. It would allow me to discover whose voices I like, and who I might ignore.
I recognize that not having this is part of the design of HN. Instead we are encouraged to consider every comment fresh, without any context. It's a meritocracy of comments. I don't have much answer to that, other than to say that I would really like to see people be able to earn a good reputation. It would make the community feel a bit smaller and closer.
If anything i'd suggest that comment scores should also have minimal impact in karma, except for questions, where an answer is a valuable contribution. Otherwise, everyone's entitled to an opinion.
Unpopular opinions are exactly what make the conversation interesting. While not every unpopular opinion is necessarily valid, it's fair to say that on many occasions the "unpopular" viewpoint is both valid, and often correct.
But regardless of the correctness of the opinion, it is the unpopular viewpoint which makes us stop, think about, and then articulate our own viewpoint.
Now wouldn't it be ironical if this very comment is deemed unpopular?
I read HN primarily for the informative comments, not the opinion pieces. A high comment score on an informative comment that contains a factual statement is a relatively good way how to easily tell that the information has been validated by the community and is likely to be correct.
However, this is more of a problem when I'm trying to assess information in areas that I'm not already familiar with. If I'm searching for information on which DNS providers are best and I find an Ask HN from 4 months ago, I can no longer tell what the true community consensus on it is. I expect the current masking of comments will provide less biased voting, so I think displaying comment scores on stories that are more than two months old would eventually provide the best of both worlds.
Comment scores in the midst of an active discussion narrow thinking. Comment scores on historical artifacts are useful for research.
Are we going for more informative or more participatory?
Now, there is little to no incentive to one-up someone, and I don't consider people refuted based on their score, but rather based on what I think of their comment. That last part has nurtured my curiosity, I find myself exploring thoughts I didn't on the 'old' HN
I feeling that I am contributing differently, assessing comments more on merit and adding my own thoughts less. I'm certainly voting on comments less (both up and down).
But I find I lack context in many of those discussions - when there are two different views on technical matters, I can't tell the difference in the group's opinion (especially if one is a reply to the other, so not comparable by which is 'higher' in the tree). Group think is bad, by blhack's mysql example is positive group interaction.
I also notice a similar lack of feedback on what is and isn't acceptable by the group. If a snide one-liner sits at the same level as a well-reasoned response, there's nothing to tell us all that one has 15pts in support and one has 1pt (an extreme example - other, more subtle ones, are now only noticeable if someone is greyed out for a negative score). There's no opportunity for us to learn from other's experience.
All in all, I want them back.
Let's say you say "mysql won't help with foo" and get no upvotes. Ten minutes later, in the same thread, someone says the same thing but adds a link and get 20 upvotes.
You can't tell that their behaviour - adding the link - was valuable to the group. They're above you, which could be a time thing only, so you learn nothing from their actions.
And next time, you make the same reply. If you could see their votes, you would learn from that (for most HN instances of "you").
A more likely case is a person who posts "mysql won't help you with foo" with no link and for no upvotes and "this is the ninja lisp trick" with a link and notices that they get far more upvoates for it.
But the most likely case is a person who notices that the first person to refer to Balmer as "monkey boy" in a hit thread gets massive upvotes.
Yes, at times I'm curious what a comment's score is, but that's my lizard brain talking.
In fact, I think the general quality of comments has improved since the comment score mechanism has been removed.
Of course, that's the wonderful aspect of comments - we are all free to disagree.
But today I did think of a different improvement I'd like to see, which for lack of better place to put it, will say here:
When I click on "reply" to a comment, it takes you to a page with just that comment and a text box. I'd like to see that comment's parents all the way to the OP, to give me some more context as I frame my reply.
I think it would improve discussion as you'll see the context in which the person you're replying to replied, and might interpret their words a little differently.
Let's take the oft-repeated example in other comments: two comments on MySQL, one with a handful of points and the other with a lot of points. There are a number of indicators of comment quality:
>> comment sorting works very well
>> if a low-vote comment is controversial, you will surely find spirited discussion below it--the volume of discussion would be an indicator of community disagreement
>> hit up Stack Overflow and find out if MySQL works as stated in the low-vote comment or if it works like the high-vote comment describes.
The arguments for visible points boil down to "I don't know how to think about this statement, so I need external confirmation." Indicators of comment quality still exist, and karma gamesmanship looks like it has decreased a lot. Finally, you really want to avoid groupthink--hiding scores accomplishes that pretty well.
If some people want to treat it as a "who can get more points" game, then so be it. I find that I can learn a lot from looking at which way public opinion is leaning.
That seems like cognitive dissonance to me and a rationalization. If you truly make up your own mind about something then why would you care so granularly about how much/little HN likes a comment. The ordering of the comments allows you to know where the comments all compare, so I don't see what you're really missing other than the exactness.
Also, comments sort within threads. Sometimes the best reply is to a no-so-great parent comment.
What does that even mean? That by not showing the comments we're somehow disrespecting those opinions?
|Sometimes the best reply is to a no-so-great parent comment.
And even then those replies are sorted again... what is your concern? Time? If you were saying hey, I only have so much time, I could maybe understand. But... I think it would be better for HN as a community to train people for more thought out discussions, fewer flame wars, and more of a focus on actually reading the comments and evaluating them based on their own content and not just on what others have said.
As someone in this thread also pointed out, it also ends the perception that an argument can be won by popularity or upvotes-- I believe this has been especially beneficial at preventing/reducing typical fan-boy flaming. I think HN has already begun to benefit for the changes. Also, I would rather read fewer threads with better comments, than more threads with worse comments.
Note: Better subjectively being defined as more thought out, less antagonistic, more discussion based, more logic, and more citations/sources.
What I was trying to say is that I value the votes of the community. The culture at HN is that "+1" type comments are noise, but 73 upvotes gives me an indication that the comment resonates with this community (whose collective opinion means alot to me).
The opposite is true for others, I'm sure.
EDIT: this -> parent comment
I really miss the comment scores.
Sometimes I am not totally familiar with whatever the original post is talking about, often the top rated couple of comments give me some good insight or jumping off points to look into it further.
I respect the HN community and have learned a lot here. I generally trust their judgement and I have found if a comment is rated highly, it most likely adds a lot of value to the discussion.
Sometimes I disagree with the highest rated comment(s). I then see my opinion is in a minority and maybe I re-examine it or stand firm and make a comment to the contrary.
Unfortunately I don't think everyone has time to judge each comment on its own merits -- there are simply too many comments -- so we need a little crowd-sourced ranking. It does lead to some group think at times but that's a (relatively) small price to pay.
Skimming is a valuable skill, becoming even more important as the world generates information ever faster.
I'd prefer to focus on attention on filtered comments rather than waste time skimming hundreds of comments.
That, I believe is more correct. I miss the scores, personally, and found them rather useful.
Top comments on HN were becoming more "top 40" and something had to be done before people started posting links to Trollface, etc.
One approach would be to use category-based voting, which adds a lot of complexity.
One approach would be to implement some sort of vote weighting sytem (time based, reputation based, context based), but that's ad-hoc and may not fix the problem.
And one approach is to simply hide numerical comment scores from all but each user's own comments. This turns the quest for high karma into a personal battle against one's self, not a sport.
PG wisely chose to make Karma a personal Everest for each individual to care about (or not).
Conversely, I tend to upvote mostly what appear to be underrated comments that are low on karma. Not saying there's anything admirable about this approach -- what am I, some kind of Karma Robin Hood? -- but the new system definitely discourages it.
Some comments are so awesome, they deserve 50+ upvotes. Other comments are pretty good, and deserve maybe 8. I do not/did not personally try to upvote every single comment. I try to add upvotes to the comments that seem "best" in a particular thread discussion, and allocate votes in this way.
Perhaps my behavior is something that pg was attempting to fix with this change, but I have a feeling I'm not alone in this regard.
http://news.ycombinator.com/item?id=2403696
He wrote, "The problem has several components: comments that are (a) mean and/or (b) dumb that (c) get massively upvoted."
So the founder of HN thinks that before the recent experiment there was a comment voting problem: (a) mean comments were getting too many upvotes, and (b) dumb comments were getting too many upvotes, and (c) too many of the comments that got the most upvotes were either mean or dumb or both. Let's stop and think about what that means. That means that, according to pg posting as of that moment, comment karma scores were often NOT reliable signals of good comments, comments worth finding rapidly when skimming a thread.
With that condition of HN less than a month ago in mind, how do the highest-voted comments visible in the bestcomments list
http://news.ycombinator.com/bestcomments
look to all of you recently? Are there fewer mean comments than before? Are there fewer dumb comments than before? Are the comments that are "massively upvoted" since the experiment began mostly comments that are reasonably kind and well-informed, helpful comments on the whole? In most of the treads you visit, do helpful, thoughtful comments seem to rise to a position of prominence, while mean or dumb comments gray out?
A link and comment in another recent metadiscussion thread largely sums up the back-and-forth about visible comment scores as a signal on comments in active threads:
http://news.ycombinator.com/item?id=2465357
>> Please bring back the comment scores. It helps a lot in parsing the comments and assigning a proportional weight to each when reading them.
> I had to think about this a bit, and I disagree so far. I'm finding that I'm not pre-judging comments as much. It's nice to be able to read someone's comment without knowing first that 70 or 80 or 3 other people thought it was worthwhile.
My impression too is that even with comment scores not visible, it is still convenient to browse threads to find thoughtful, informative comments, but now there is less anchoring bias
http://www.sciencedaily.com/articles/a/anchoring.htm
of most votes on a comment converging to one score level that shows up early in a thread's development, and more engagement by readers of HN in actively reading comments and upvoting (or downvoting) based on each comment's characteristics in light of the context of the thread. So far I can still find good comments quite readily here on HN. Indeed, I think that since the experiment began I am seeing more good comments more readily than before.
The main motivation stated by pg for the current experiment with making comment karma scores less visible is to "stave off decline of HN," and that is what will decide if the experiment was successful. If the previous visibility of comment karma scores led too many casual readers of HN to upvote mean or dumb comments, and too few readers to upvote thoughtful, informative comments and to downvote mean or dumb comments, the arguments on the side of reader convenience aren't going to be convincing. It isn't convenient for ANY reader of HN if the comment scores are a poor signal, and if bad comments become more prominent and good comments get skimmed right over by readers in a hurry. If a change of rules here makes every reader read comments more carefully and more thoughtfully, and vote based on comment inherent quality rather than on crowd appeal, that is a feature rather than a bug. For comment scores to be a good guide to every reader here, every reader can help by actively upvoting informative, helpful comments, and also by downvoting comments that are either mean or dumb--and especially comments that are both. As I recall, the experiment has also involved some changes in the effects of flagging, so flagging inappropriate comments is also helpful.
After edit: many comments in this thread ask about the karma rules and voting rules imposed by the software. We can all read the news.arc software ourselves
https://github.com/nex3/arc/blob/master/news.arc
if we would like to see what the rules do (except I think that maybe a few aspects of the current experiment are hidden from the current distribution of the source code), as previous HN threads have pointed out.
I can tell the difference between a stupid comment that got a lot of points for being stupid, and a useful comment that got a lot of points for being useful.
When I look at amazon and see that twilight has 3000 5 star ratings, I don't assume this means it's a great piece of literature, but it is useful information to me: it's a popular book.
Same goes for comments.
I'm not stupid, I can tell the difference between "popular because it's funny/pandering" and "popular because it's useful"
Let me make sure I understand correctly what you mean in light of what pg said about the reasons for the current experiment (quoted in my post to which you reply). Is it simply a mistake to say that high comment scores on comments that are mean or dumb is a problem? I ask, because pg has said that there has been a recent problem with comment scores.
One of the big assumptions that's being made here is that people are dumb calculators.
I think that there is a huge body of research showing that all human beings, without exception, high-IQ or low-IQ, are "cognitive misers"
http://www.psychwiki.com/wiki/Cognitive_Miser
http://blogs.discovermagazine.com/cosmicvariance/2009/11/04/...
(there is a good self-test at the link above)
http://lesswrong.com/lw/2ey/a_taxonomy_of_bias_the_cognitive...
in the sense of preferring less effort-demanding problem-solving techniques to more accurate, but more effort-demanding, techniques. Snap judgments spare mental effort for everything else that we are busy with.
Here on HN, the only thing most participants contribute to the community is good comments and good submissions and votes about other participants' submissions and comments. A few founders, pg and his core of curators (editors),
http://ycombinator.com/newsfaq.html
do actual work to keep the site running, and they get the best compensation for their volunteer efforts if all the signals about story quality and comment quality are true, thoughtful signals rather than noise.
With scores visible, most "discussions" end up as little more than opinion polls.
All in all, not showing comment points is a step backwards in helping people get the most out of the site in the most efficient way possible.
But also, please split upvotes from downvotes. There's a huge difference between a +25 / - 24 comment (apparently a controversial one) and +1 / -0 (probably a mediocre one).
Or perhaps, display only upvotes, and use some weighted form of (UPVOTE * U_WEIGHT - DOWNVOTE * D_WEIGHT) for positioning the comment among other ones.
The relative absence of these black holes of discussion is one of the things that brought me to HN in the first place, and I think that showing comment scores discouraged them on multiple levels. Public upvoting lets people express their view on the topic without posting points similar to those already expressed. When two comments have a lopsided point spread it lets one "side" of the debate feel more comfortable letting the other have the last word.
At the moment, a comment that's down-voted past zero becomes lighter. Perhaps very popular comments could be made more visible, or highlighted?
I think this kind of fuzzy indication of popularity might be a good compromise.
It would likely have to be normalized based on the score of the submission or the total number of comments on that submission, though, which might be more complexity than pg wants.
For my purposes, displaying actual scores within the range -4 -> 10 would be sufficient. the low end is already capped, and the high end could either just have a ceiling of 10 or a score of 10+.
I am occasionally a "someone is Wrong on the internet!"[1] type and my inclination to wade in is directly proportioanl to the perceived traction of the inaccuracy. Without such a heuristic the choices are: reply to all, none, or a random sample which result in "poor information", "needless pedantry", and " undefined behavior" respectively.
My gut is telling me that traffic is likely to be lower. Reality might be different of course.
If there was a comment that made me laugh (while sticking to the point) I will be more likely to upvote it, but if it already had 10+ upvotes a laugh on my part probably didn't justify another upvote.
I second what others have mentioned that it would be good to re-display comment scores on older posts. This way we could at least see the community consensus after some time has passed.
I don't think the solution is to bring back scores, though. A possible "simple" solution could be to color/star a comment above 50/100 points, etc. Comment scores could also be displayed as percentage or on a scale of 0 to 1, 0 to 10, etc. I'd be more likely to read a comment with a score of 95% than one with a score of 20%. This way you at least get an indication of the helpfulness of the comment other than just its position on the page.
Seconding this idea, I really love it. The biggest problem with a linear point count is the large dependency on a comment's visibility and number of impressions, which in turn leaves posters competing for that as much as for inherent comment quality. The implementation would be up for discussion, maybe something like show a comment's positive percentage once it garners a sample size of 5 or 10 votes, and maybe round off to the nearest 10% so you can't triangulate back to the actual score from the change inflicted by your single vote.
I liked someone's suggestion of basing the comment score on "upvotes per view" so older comments don't dominate, and I also like the idea of using a dark-to-light gradient (dot) instead of a concrete number.
Just sorting to the top (kinda) really just makes it hard to wade in, and makes me less likely to take a look at a topic I know little about and would like to see a couple of definitive words on it. It may be groupthink to some extent but this is a damn smart group.
That way only posts/comments get points/scores, not people.
With reflection, I think a good idea might be to show the score after you vote.
This way you get the feeling of making some difference i.e. immediate feedback, but also the knowledge that your vote wasn't subconsciously affected by a visible score beforehand.
The main downside of this would be people voting out of curiosity to see what a comments current score is. Perhaps displaying the score of a comment once it reaches a certain age (maybe three or four days old) would mitigate against this.
If voting reveals the score, then there would need to be a new way to abstain and reveal the score.
This helps because even if in hindsight, I would know if the comment was the general consensus or a popular one or not?
You would have to go out of your way to see the score but could still be used by us hackers who like to customize our experience.
I also feel like people are avoiding posting crappy comments with the intent of tapping into a popular vein for a high score.
This could be a placebo, or perhaps, if real, it's instead caused by an improvement in voting.
I'm reading a lot let now because there are no scores; I just skim the top and then move on.
Comment scoring allows the community to reveal quality.
My comment here explaining it has made me comment less has 40 points. I think a lot of people agree.
So the system could lead to more posts by undesirables (me) and fewer posts by those with more knowledge (you).
Also, the ordering of the comments does the same thing as having the comment scores!
I do wish there was a way to still search for the highest rated comments on searchyc.com, but I still think it's a small sacrifice for an overall better community/environment. I have definitely seen less iOS/Android/Windows flaming, so I think the site is already benefitting from the changes.
I understand your argument. Just keep in mind that there are multiple ways that HN is useful. Some of us want to go deep into a topic, will read all the comments. And some of us are in a hurry and want to a quick feel and extract a summary.
In the first case (I call it "researcher mode"), not having votes is a good thing. But it prevents me from browsing in the second mode (the "quick update mode").
I wish you could set this in my preferences, but it wouldn't fulfill pg's original goal of staving the decline of HN (that I don't see - but that's another discussion).
There is a tendency for short comments to sink down, regardless of how informative they are, simply because people spend less time on them, so they're less likely to hit that upvote button.
Also, like all forums, the first upvoted comments get more replies creating a positive feedback loop, not necesarily because they are the best, but because people know their replies will be more visible.
It would be interesting to have the statistics of number of upvotes vs position of the comment in the page.