1^st click!!!!! woot!

:😢: It's already so hard!!

You more veteran folks probably know better.

nah ... we don't really, arrendek ... we just have thicker skin (from the previous lashings) lol

Comment hidden because you are ignoring dag. (show it anyway)

I'm curious - although there is not a sure-fire way to measure it, how much harder is to reach gold star status now than say six months ago do you think?

I vehemently disagree with this move.

Comment hidden because you are ignoring dag. (show it anyway)

Care to elaborate DD?

I think you guys make too much over the queue issue. This Sift is bigger entity now. You're not going to have it be the way it used to be. As much time as it takes to find videos, vote and generally participate, I find this move extremely discouraging.

As someone else mentioned before, if/when the Sift becomes as large as other hugely popular Web2.0 sites, its basic function will have to drastically change accordingly. E.g., if there are 10,000+ videos sitting in the queue on a given day, how many should make it to the front page? I personally have imagined that when that day comes there'll be a limit of 1 queued video at a time and many votes to publish; either that or the basic functionality will be much different than it is today. I guess we'll just have to wait and see.

It's simply a matter of survival of the fittest. With more diversity, there must be more challenges to trim the herd and preserve only the best.

Comment hidden because you are ignoring dag. (show it anyway)

It's not the queue issue DD - and we haven't changed the number of posts allowed. This is just to slow the flow to the front page.

I can measure it from a purely anecdotal and singular observational point of view. I joined the Sift on April 14, 2006. I watched gold stars go up right and left as I believe you needed 4 or 5 votes to get a video published (semi-dynamic queue escape value) and postings were unlimited. Then, postings went to 2 in a 24 hour period and then to 3 in a 36 hour period. Gold stars were pretty easy to be had.

After those good ol' days, the queue escape level increase to 6 (230 days ago) and then, later, to 8 (215 days ago). I think that is when I began posting when it was at 8. (My first post was approximately 130 days ago.) There was also a 6 day expiration in the queue which went down to 5 and is now at 4.

It was about 190 days ago that this happened:

Posts that have less than 2 votes after 24 hours will be discarded.

Posts that have less than 3 votes after 48 hours will be discarded.

This will happen once a day at the same time we do our usual daily queue discard.

We've also raised the queue escape level again, as the front page is still going through a complete refresh every few hours it seems. The new peak-time escape level is 9 off-peak is 8.

Nine days later the submission quota of three in the queue at any one time kicked in.

All of these changes seemed to resolve the increasing submission glut for a while. Since then there have been some minor tweaks as well as some cool stuff: promote, save, collectives, new categories, et al. However, gold stars are harder to come by and it seems that diamonds are going to be nigh impossible.

I'm wondering if by having a collective that the videos submitted could somehow earn points for the collective owner toward diamond status. I'm just thinking as I type, but maybe a percentage of the votes earned by the original submitter would also count toward the collective owner. For instance, dag submits a video to the Rock and Roll Collective that garners him 20 votes - well, then MLX would get 5 votes as well (or 10 or something along those lines). Then, for those who were able to attain gold stars (which is becoming increasingly difficult) would at least have some kind of a shot at a diamond.

Long way of saying - yea, it's gotten tougher, but, yea, the Sift is cooler, too.

Have a great night,

Sc

agree! (with dotdude)

no one should expect to be able to watch all the videos published -- this place is too big for that now.

raising the bar too high makes it hard for more obscure videos to get published. so everyone tries really hard to guess what other people want to see, and ends up restricting their choices.

the Sift is at it's all-time best (most interesting, most exciting) recently. partly, it's because of the collectives. people are chasing after what they really like, and so we're getting more old rock music, more italian horror movies, more experimental animation, more of all sorts of interesting stuff.

so there's 60 videos published a day, even 100 or 200? use the channels, read the tags and descriptions, join a collective or two -- find what you want to see. Re-Sift the Sift.

excellent post, plastiquemonkey ... i have a full time job and a family and i still find time to sift through the sift ...

but i do think that this move makes it harder for newer members, which in turn causes more people to post and then leave (with little results seen).

Comment hidden because you are ignoring dag. (show it anyway)

OK, I've added a poll, as there obviously some strong opinions on this.

If you are voting to keep it at 12, you are voting for fewer videos making it to the front page, and therefore less frequent updates of the "published" areas of the site. You are also probably voting for a larger queue, and some publish-worthy videos not making it out.

If you are voting to move it back to 10, you are voting for lots of videos making it out to the front page, but perhaps fewer comments per posts, as some threads are missed, and the possibility that you will miss stuff that has fallen off the front page before you got a chance to see it.

option #3? keep the escape number to 10, but cut the post-at-a-time number down or in half.

If we strangle the sift anymore we are gonna remain a niche.

If we let the sift evolve naturally with current standards it would only push users to explore the collectives system to find what they click with.

VS = MTV, Collectives + Playlist = MTV:2, VH1 et al.

I can't believe you just compared VS to MTV, Farhad. For shame. >.<
In any case, I'm for keeping the escape number to 10 and keeping the post-at-a-time the same as well. I personally would like to see more features to expose more videos in the queue to those interested, as the collectives and channels are attempting to do.

For starters, I suggest the default number of videos displayed per page be set to 30 or even 35. I just changed mine to 35, and I've been exposed to and voted on a lot more videos as a result.

well ... taking off from remy ... i have my pages set at 50 ... so much doesn't escape me in front, queued, or recent comment sections ....

i, too, agree with farhad that it is time for the sift to grow exponentially and evolve as it will ... the base that has been created is bar none ... but it's going to be impossible to keep the current standards and front page without sacrificing something!!

thumb-foot derrick, and long-path possessed.....



stay here, and don't get anxious,
let the collectives do their jobs,
and have a bi-weekly memeber cleaning, for all the join and get bored, wannaplays.....the folks a part of this site for the sake of joining once and creating a handle, are minions and expendable...let the numbers silverchord is fixated on, work for the holdouts.....the contributors and holdouts should have the best of the site.....if you guys got ants in yer pants anticipating a deluge you cannoot contain, then yer in the wrong post office blues business......cause cards and letters is what you really want,.....huh? Until ya have to pay

for lawyers...heheheheheheh

really dig the rembar twist, the visual impact kept simple, pluggs more synaptic paths in, and gets the juices flowing....dynamic, and simple.......like gloogles®™

Comment hidden because you are ignoring dag. (show it anyway)

I would say that our front page already updates more frequently than other sites like Digg, Reddit or Slashdot.

choggie's in da house!!!!!! *or is that westy??

Also, I'd like to suggest the idea of having a random section, where sifts are displayed at - surprise! - random. This would hopefully expose more middle-section videos to sifter as a whole, seeing as how sifts have a tendency to lie around for days after they move off the front page.

Can't we just adjust the limits when we need to do so? When traffic is high, we set higher limit. Slow days, set it lower.

I understand that there are clearly a lot of various complicated issues in this decision, and I would like to cleanly express why I support raising the queue escape level.

1. VideoSift will suffer from a reduction in quality due to obscure videos being overlooked or unposted in the frantic drive towards "star power," or ranking.

I think the collectives help devalue this claim, in the past when we have increased the queue escape level or the top 15 system we have caught similar flak, but currently the top 15 includes such titans as

"1. Polite Distances For Holding Open A Door

2. What is it with Kitties & Laser Pointers??

3. Incredible Shoot-out Goal (Voted best of 05-06 season)
[From YouTube: this is a video that was created by Karl Fisch, and modified by Scott McLeod. Ther]

4. Eye-Opening Stats: The Global Economy and Computers"

To me it is clear that this increase has not reduced our niche value; this is to say that I am interested in preserving the vibrant community that is at the heart of videosift, but I think we need to look at how we are going to grow.

I guess what I am trying to say is that essentially the front page is the life blood of videosift, long term members know how to find collectives, and playlists, but what about the first time visitor. When they come to the site I like to think that they see a representative of the sift ecology that we have all come to know and love. Currently I think the video turn over rate is too high, and in fact in contrast to promoting awesome videos, I think it suppresses them. If we promote 60 videos a day how can we expect the 59th one, a documentary about post Castro violence in Cuba, to receive the light of day it deserves.

Clearly this is an allusion to a false possibility, but what I am saying is that I have faith in the sift to promote these videos, regardless of the queue escape level to a point where more people might see them. In all honesty I think raising the bar to entry to such video goldmines is an ethical misstep, and would like to make sure that our community maintains its power to draw attention to such things. If we continue to allow queue turnover at this rate we risk driving attention from important issues to ones that merely garner x percentage of a niche support, and for me this is why we created collectives. Not everyone here agrees what should be on the sift, but if it dies an ignominious death the video shall live on in the collective. However, a video that should be seen by all might perish in the waste bucket that is continuous community creation. We owe it to our new users, and those who might not care to think, to provide bleeding edge access to all that is visually important, and I don't see how we can continue with our current queue escape.

This is not to say that I would not listen to suggestions, I merely ask those that are long term users to balance their desires for community growth, with what someone sees on their first visit to the site.

Ok, given what you've said, James, I understand the reasons for raising the escape number. However, I do have a large concern.

From what I've seen, the exposure rate to a video drops to nearly 0% upon reaching the third page, and does not increase again until it reaches the last page and makes its way onto the list of top 15 videos to be discarded soon. What I'm afraid of is that raising the escape number will just force more videos into that dead zone of the third page and on, and that will not raise the quality of the Sift, but rather simply cause more videos to die. If the escape number or anything is going to be changed, I would like to see some way that a video gets more exposure if it isn't published before falling off the planet after passing the second page.

Heck, I'd like to see some way of exposing more people to videos on the third page and on, period. Randomize the videos, display the first-page and third-page videos simultaneously, I don't know. But there's basically no point to the middle pages at present, and I think that issue needs to be addressed as soon as possible.

Don't agree with the idea and I think it'll just annoy people more than solve a problem. I'm not saying it's a bad idea as I can see your reasons but it's already becoming a lot harder to get some clips out of the queue and the sheer size of the bastard is quite daunting. I'm already seeing loads of sifts get about 5 votes in the first 10 minutes and then vanish for the next 3 days.

And it's only gonna get worse.

I honestly think the only way you're gonna get round this problem without frustrating everyone to the point of self-mutilation is to have more than one queue, maybe a different one for each channel (not as it is now with the filter but a forced seperation), but still only one front page. You could do this (already making the queue seem smaller to us infidels), stick the video expiry period back from 4 days to 7 and then knock up the escape bar. It'll have the same desired effect but the queue(s) will be easier to live with.

That's my thought.

The question then is, what is the first thing desired by concencus?, that is to be seen,,,,the Next visitors are the ones who will carry this, unison-style, with the firsties/holdouts.....otherwise, make it simple, and dynamic enough without perpetual fine-tuning, ...unless that will make it unique...that's the ultimate goal, holding its own,as a dynamic entity.....
perhaps a reasonable target is static fluidity....

deputydog has a point. Maybe the solution is to add another sifting queue.

First posting queue, second the well received queue, then finally the front page.

How about...One must begin from this point on, on the site, with 3 published in 3 weeks......or however the #'s work,cause you start small...always, but the dedicated prosper.....that's how real life is, innit?

wow that's cool.....queue(s) to doos.....

(to James Roe:)

What someone sees on their first visit to the site is 10 videos, no matter what the queue length or escape vote setting is. The only question is how quickly those 10 videos are going to cycle through. New visitors likely won't even notice how quick the turnover is -- since they've just arrived, they see whatever the 10 videos are when they get there.

Possibly the objection could be made that a lower queue escape will lead to those 10 videos on the front page being of lower quality. Of course, they will be less popular. But the less popular videos here (as long as they're still reasonably popular) are often very, very good, while the super-popular stuff is often (not always) quite predictable.

Have there been any complaints from newer users that the front page is too low-quality, or being updated too often? The complaints about turnover published on SiftTalk seem to be mostly from longtime users who are used to being able to keep up with the site in its entirety.

As the userbase grows, that's going to be impossible, unless the Sifting ratio (videos submitted : videos posted) gets higher and higher. If the target is a daily publish rate of 30-40 (for example), as the daily submission rate tops 100, or 200, or 300 -- it's going to be very difficult to get anything except the most popular stuff published. And that means the more unusual (or just less immediate) stuff will be choked out.

Even then, you could argue that the collectives and playlists will let experienced users find what they want. But most users (I think) don't bother posting things they think are likely to die in the queue (choggie and rickegee are exceptions, there may be a few others). So you lose that content right upfront -- eventually, you may lose that user entirely.

The argument isn't really about 12 versus 10, but what the motivation is for the change. "Too many videos a day for one person to watch" is a bad reason, because no one person should expect to be able to keep up with the aggregate interests of hundreds or thousands of others.

Well said PM

...here here

I'd also suggest turning the Top 15 on the front page into maybe the Top 30 if you have room. It would go some way towards stopping newly published vids vanishing as quickly and give sifters more hope / satisfaction.

BTW, well done with the collectives. Whilst there are minor negatives (as there are with everything) I think the positives are substantial, especially in the long run. I've personally seen some benefits already e.g. many videos I normally wouldn't have noticed in the queue have been presented to me as a result of collectives. It can only be a good thing.

Ditto what PM said, adding that I voted against the increase. Keep the publish rate at 10, but lengthen the 'top 15' and 'expiring soon' lists so that the more worthy videos get a little longer exposure.

It's hard enough needing ten votes - 12 is too many! Plus it will change what type of videos gets through - those appealing to the lowest common denominator?

Comment hidden because you are ignoring dag. (show it anyway)

You know, call me psychic, but I'm kind of getting the idea - that maybe you guys were happier with the 10 vote graduation limit?

Is the front page the life blood of VS anymore? It is an old way of looking at a radically transformed Sift.

I never, ever look to the front pages to find videos mainly because it is a glad bag or hand rag(cue Office UK theme). I want and have now received a more precise set of Sifts.

My worry is that the 12 votes will only help the Sifters that are already established, who can cleverly seed the collectives with their choices. There are many new great participating members and I do not wish to discourage their submissions because they may not feel comfortable yet asking me (or anyone else) to join a collective. And a collective is a massive advantage for vote-hogging.

More eyes are already diverted from the main queue by the collectives. It is simply going to hurt new submitters (the ones you and I don't vote on by submitter name value alone).

My more immediate and selfish worry is that I will never again have a published video and the catastrophic growth of my Damned list will break the VS server.

And I don't really care that the front page is turning over so quickly. I am still able to easily find things that I love.

I think another consideration is the promotion of the queue in the first place. I've noticed that it seems like the same 15-20 (maybe that's an underestimate) people are the only ones who actively sift through the queue on a daily basis... pretty much that is all I do anymore, rarely do I look at the front page because I've already seen it all in the queue. Anyway, when I look at queued videos, it seems like its always the same people that bother to vote on them, therefore, the queue audience is pretty limited. If we could figure out some way to get more members to actually look through the queue, the audience would be bigger. Just a thought, as I'm totally against raising the vote count... especially since I've had very few vids get published in the last few weeks... most have been stalling out after reaching the dreaded page 3!

>> My worry is that the 12 votes will only help the Sifters that are already established, who can cleverly seed the collectives with their choices.

This is exactly what I worry about... that vids will be promoted not just on the merits of the vid. If a vid is submitted to a collective, and an email blast goes out to its members, who are more likely to vote for it, that has a huge leg up over that same vid just plain submitted into the queue. This makes it harder for new members to break in unless they learn the politics of a collective and 'pledge' that collective. Discouraging new blood by having structural advantages for some members is bad for the Sift.

I totally agree with PM that we shouldn't increase the posting limit to satisfy Ant's ability to downvote more of the videos that we post.

Ant honestly, get a freakin' life, am sick and tired of you downvoting videos just because you don't understand them. Hypocrite. I label the videos for your convinence, I always LABEL HIP HOP or RAP and you still click it and you STILL downvote it. If you see something you don't like question why you clicked it. And not just DOWNVOTE it for dubious reasons.

Just another thought, could be bollocks.....

Whilst a video is in the dreaded queue, why not hide the number of votes already recieved and/or the submitter's identity to the sifters? Result - More excitement (I think) and less 'favouritism' (most people are guilty to some extent).

I'm sure this idea tied in with the 'queue limit debate' but I've forgotten my reasoning.

Wank.

>> Whilst a video is in the dreaded queue, why not hide the number of votes already recieved and/or the submitter's identity to the sifters?
I dunno.. the # of votes in queue relative to its peers is a key signal. Like if you walk into a restaurant at lunchtime and it's empty..that tells you something.

>> we shouldn't increase the posting limit to satisfy Ant's ability to downvote more.
Here here! Very annoying. Did the same thing to me, against Chris Hedges before he was for him.

>> Whilst a video is in the dreaded queue, why not hide the number of votes already recieved and/or the submitter's identity to the sifters?
I dunno.. the # of votes in queue relative to its peers is a key signal. Like if you walk into a restaurant at lunchtime and it's empty..that tells you something.

Maybe that's the wrong approach. Everyone has different taste.

Raven is correct. Every one of my videos that escapes the queue has an array of the same 20-25 voters comprising the formerly necessary 10 votes. There is just not enough traffic in the queue to support a raised vote level. It should go back to ten votes, most definitely. The front page "life blood" will not suffer because it's important to keep fresh videos on the front page for new visiters. Rembar brings it up a lot in these posts about how once a queued video gets to to page 3 it stops recieving votes, this is the case a lot of times. With the introduction of the collectives we are only now correcting this problem and by up the number the publish votes you just set the problem back at step one. I voted for 10 votes.

this siftacular makes my nipples hard and i *nominate this user for *banination.
I voted for 12 because whats another 2 votes? i always figured it would go up again, but what about 11, then 12 again later as a ratio relative to the number of users or vids in the queue.

Secondly, i think DD's idea about the queue containing anonymous post information a really really cool idea. I've mentioned before that diamond and 100 gold star posters generally get their stuff unfairly voted up by default, so why not level the playing field?

lucky760 cannot be banned because starred accounts are immune from banination. If you feel this account requires banination, please contact the administrators (ignoring banination request by gold star member ren)

i didn't seriously think that would work

Something else I thought that will happen a lot more with a higher number of queue graduation votes. A video could now with eleven votes. Getting eleven votes in the first place definitely means it is a pretty good video, and when it expires with eleven that just means someone else will post it the next week, and that will cause more sift talk posts about discarded video rights.

12?! No, no, keep it at 10! There has to be a better solution that to keep raising the escape level. What's next, 14? 20? Is our user base really that large? While checking the the lists of users, I see PAGES of users with 0 votes, 0 comments. Another increase is, like DD said, discouraging.
If you're concerend about refresh rates, why not have gradual promotions. If several vids hit the "magic number" all at once, put them on the front page based on order of submission. Also, I think the default number of vids on the front page could be increased, say to 20.
I have a feeling once the novelty of the collectives wears off, things may settle down a bit and refresh rates will decrease.

>> Maybe that's the wrong approach. Everyone has different taste.

Well, you're right. My priority/goal when reviewing the queue (or, back when I had time to do so), was to maximize my utility. The more filtered/upvoted a video is, the more likely it is to be good. "top 15 queued sift videos expiring soon" is a good example..those are bound to be good.

I'm totally down with hiding the posters (See Blind orchestra auditions better for women, study finds). But I contend that the bias from gold star posts is nothing compared to the (potential) impact of collective voting blocks.

just to point out, nothing in the argument i made has anything to do with ant particularly, or downvotes from anyone.

also, it's not "bias" for someone to watch everything MLX (one great example) posts by default. it's just recognition of her track record. people whose posts here do well again and again are (usually) worth trusting.

the same principle is why people have bookmarks for favorite websites, rather than reading the entire internet at random, with the source names hidden.

anyway, if you anonymize the queue, people will just make personal-taste playlists (like All Kinds of Awesome, This One Made Me Happy Today, Raven's Favorites, etc) with all their own videos on them. which will defeat the purpose.

'anyway, if you anonymize the queue, people will just make personal-taste playlists (like All Kinds of Awesome, This One Made Me Happy Today, Raven's Favorites, etc) with all their own videos on them. which will defeat the purpose'

Disable the option to add to playlist until it's published or expired?

"Disable the option to add to playlist until it's published or expired? "

No no no no no. I like looking through people's playlists and finding videos that have expired but which are still definitely worth watching! It makes playlists more interesting and personal.

no one has time to watch all the submissions. the username is a major piece of information predicting the quality of the post. why would you hide it?

also, this is a social website. having a history of being a good member of the community is important. giving people the option of giving votes to members they like, or withholding votes from members they dislike, is a powerful way of regulating how people behave.

about the collectives. most people are already spread across too many collectives to vote as a block. right now, there are 12 queued videos in the Arts collective that have 9 or fewer votes, even though there are over 30 Arts collective members. not much of a voting block! i doubt anyone has the political power to make it much more organized than that. and with people in multiple collectives, there will be no unity of the "let's all stop the cats and dogs people" variety -- since most people in the other collectives are also part of the cats & dogs collective.

Oh well. Just a thought.

I agree with PM, I don't particularly like the idea of anonymizing the queue. Sure, there is the possibility of a voting block, but let's also consider the possibility that higher-level members are at that rank due to their taste in sifts and their amount of time spent sifting, not merely because they're pseudo-famous. Also the fact that it is enjoyable to be able to pick out other peoples' sifts and make personalized comments.

Also, as Raven and WildmanBill have pointed out, I have a raging clue for the queue progression. Something needs to be done about the visibility of queue videos, because it's ridiculous. I can categorize all videos as one of two types of sifts: the Daphne and the Velma.

You see, the Daphne is the quick-riser, the one that gets huge amounts of initial attention and gets published within the first few hours because it's quick, easy, and a looker to boot. (Cute sifts, I'm looking at YOU). But it's sometimes harder to sustain conversation with the Daphne, because often (but nowhere near always) it's a bit of a ditz. The Velma, on the other hand, tends to be overlooked in the initial phase, and it takes a while to come into the spotlight in its own right. For some reason, the slightly bookish, thick-glasses-type sifts fit this category. (Documentaries, especially, have this effect.) Sure, it may get some votes initially, but it takes time for enough people to notice it, and realize it's smart AND sexy, so it takes the full four days and its appearance on the soon-to-be-discarded list to get voted to the top, and sometimes even that won't be enough to promote it, and that's dangerous, because without Velma's priceless intellect, sleuthing skills, and constant supply of doggie treats, how will Scooby and the Gang ever manage to crack the mystery of the Egyptian Mummy and save their professors from the mummy's stone curse?

...wait, what? Crap, I forgot what I was talking about.

Anyways, all this talking about Daphne and Velma has given me another big, stiff clue for queue visibility. I think randomized videos should be added mandatorily to the queue list. For example, underneath each time-placed video, you could include one small-sized video randomly from the queue list. It's probably a bad idea, but seriously, this whole deal of letting sifts sit on their butts after page 3 until they're just about to be discarded is annoying and will become even more so as it becomes harder to get a sift published if the escape number is raised to 12, and the only way I can see this being dealt with is by raising middle-page visibility so queued videos can be seen and voted on as they pass through each page of the queue, and the only way I can see that happening is by creating some sort of randomized display of videos to accompany videos in chronological order.

I don't think videos sit in the middle of the queue as much as you think they do. People tend to sort by most or least votes, at least I do, and that cycle tends to consistently promote videos that are doing well or should be doing well out of the queue. It is certainly true that the occasional nugget gets lost, and it HURTS when it's your nugget, but if you link me to a video that has died that "shouldn't have" I am willing to bet I can tell you why it did.

case and point

http://www.videosift.com/queue?page=6

has a bunch of videos ranging from 5 - 8 votes. The ones with 8 votes are almost certainly a shoe in to escape the queue at some point in the next 4 days.

{"I think randomized videos should be added mandatorily to the queue list."}whoever said that yeah
This would solve the dynamic of rotting on page 3(obit)

Hey.....to keep the direction of the little click vibe some folks think may be a dynamic, invite soem newbies to yer collectives, the ones that are potential quality contributors.

{"also, it's not "bias" for someone to watch everything MLX (one great example) posts by default. it's just recognition of her track record. people whose posts here do well again and again are (usually) worth trusting."}-pm

PMs' right, there are folks better at bringing a thematic vibe to us, dJ Style)...benjee(documentaries) farhad(eclectic blend of sounds from everywhere)
This dynamic of personal taste transmission thru submission, is one of the reasons I am even here.....

I do not like public masturbation, jack-ass, ambulance chasing,political agendas reaaaaaaaaaaaly piss me off, etc.)....but defer any who with these, thematically and skillfully, spin a series of otherwise boring or diversive viddies, and cause the synapse to wiggle.....

some folks love to engage, others remain detached...i.e. real world

I have always been for the queue being a completely RANDOM order weigh-station....this eliminates the first page hotter/colder/snowball's "chance in hell its posted" dynamic

Actually looking back over some of my past discards I think there might be some validity to queue rot. I should point out however that a randomized queue by default is a hairy coding problem. It involves creating a randomized list for each person on connection, and then maintaining that list during their queue browsing, not a trivial task. So as much as this may or may not be a silver bullet, I wouldn't hold your breath for it any time in the near future.

/maybe enjoys raining on parades
//maybe just a little

oooh, one more side note, a randomized queue by default limits the sifts capacity to publish "breaking stories"

Like the Takei clip that is climbing the top 15 at the moment.

James, I could give you a list of videos and argue from an anecdotal standpoint, but that's not really conclusive (and besides, wouldn't it be easier to just take a poll and ask the other sifters if they've noticed this pattern?). If our debate is of a factual nature, i.e. do videos sit in the middle of the queue as much as I think they do - and I think the type of videos one sifts has a lot to do with it - then the matter is simple to deal with. How's about we set up a data collection (if there isn't one already) that will record when votes are made in terms of a video's life in the queue? I'd be willing to bet that the two kinds of voting patterns will arise, and that there will be a dead zone of voting in the late part of day 1 all the way to the beginning of day 4.

Heck, such data would be invaluable anyways, because voting trends and timing would be great in figuring out all types of adjusting to the queue and such.

Did James just said he was getting burned out with the legistics of managing the site.....*ba.......wait let's give him one more chance!

how about this: change the "expiring soon" sidebar.

have 3 different ordered assortments, one each for day 2, day 3, and day 4. "top 15 videos submitted yesterday", "top 15 videos submitted monday" (or feb 19th, or "two days ago"), and of course "top 15 videos expiring soon". have one of the three sidebars show up randomly where the "expiring soon" sidebar usually shows up now.

this would increase the visibility of near-miss videos on days 2 and 3. sure, it would also make it a bit harder to find the about-to-expire stuff, but maybe more of it would be getting published earlier, or getting additional votes sooner, meaning there'd be less of a rush at the end.

much easier to code, i bet...

um, that should be "top 15 queued videos submitted yesterday", etc...

James, why does it have to be random per person? You could, for example, create a randomized-order list that refreshes every hour, and displays to everybody. That would be much easier to code and less stressful on the system. (Would it really be all that difficult to code your idea, anyways? It doesn't seem all that insane to me.) There are other ways, but my point is that I don't think a completely personalized random queue is necessary or even helpful.

Besides, my suggestion was for a mix of time-ordered and randomized videos. I think it'd be nice to have the combination, since the time element is nice (for breaking stories, as you pointed out).

Anyways, it's just my opinion that letting the queue stay defaulted as a time-ordered scheme is a bad idea.

having talked to lucky it seems that I overestimated the difficulty of randomizing the queue on a per user basis. He assures me that it is not so difficult for a shaolin code ninja like himself. I am going to start a separate thread for that discussion. So that this one can maybe, hopefully return to a discussion about the increased queue escape.

looking at the front page 5/10 videos were published in the last 24 hours with one being less than 4 hours old. Since the last time we raised the queue escape we have had 620 members join. If only one percent of these members are looking at the queue that's still an additional 6 votes. Is it really all that bad?

the random queue thread can be found here

The problem is too few video being shown on the front page. I have always disliked the fact that you can only see a small amount of videos on the FP. I would really like to see a thumbnail type of system where we can see essentially the same info such as title name, number of votes, tags, etc, but only thumbnails of the screenshot then once clicked or on a mouseover you get a lightbox or ajax style popup/popover to play the video in much like the way the login box works.

To me this would give us more videos to view on the front page and reduce the need to have more votes to escape the queue. You could also use the same format for the queue so we can see more videos at one shot.

I would love to see about 25 to 50 vids on the front page to sift through.

grspec, you can change the default page view in your profile already.

Also re: the ajax video system, digg does something like this currently, honestly i kind of prefer the set up that we have at the moment, but I don't know how others feel.

/prods conversation back to the new queue escape

yes but I thought dag said that it really can tank your system resources by doing this. I am thinking along the lines of something that won't kill your system and can be made default so that the problem of having videos move so fast home page would be fixed, which would then not require the increase in queue votes. I think this would also solve the queue rot problem too since they wont drift away so fast.

Poll results so far:

26% Keep it at 12
74% Return it to 10

So can we go back to 10 votes makes the front page now?

/prodding back

The peasants are getting restless...

It seems like that is the general consensus, can I get some feedback though on the fact that the front page is composed of 6/10 videos that are less than 24 hours old. It seems like the new queue escape is doing little to stifle graduation rates, grumbling aside.

Also it bears saying that if this doesn't happen now it will inevitably happen at some point. When would that point be? How do we determine it?

Comment hidden because you are ignoring dag. (show it anyway)

But wait - you can't go against my decision. I'm the Decider! I think we should listen to our generals on this one. Lets roll back the queue escape to 10.

(At the risk of letting you all in on the secret that we're just making this up as we go along)

Does that mean shitloads of vids which are on 10, 11 votes are now gonna flood the front page? OMG!! What the hell are we gonna do? We need a system to control the rollback - shall I start a new thread?

77 comments??????? WTF!

The problem is the bi-modal distribution of videos. You need to more quickly prune those crappy videos which will never escape, and leave room for good videos to be spread out instead of all them with 8 votes.

I would like to start seeing more statistics on queue videos, ie. Time based (# of submissions / hr), (# of total votes on all queue videos / hr), (% breakdown of queue submissions .. ie. % that get 1,2,3..10 votes after 24 hrs, 36 hrs, 48 hrs)

I have said from day one that scalability of this site will be a problem. You can't limit the number of front page videos per day. The queue just breaks down at some point (200 queued videos to vote on), So you need to firewall the front page with a "hot" videos or a secondary queue. If you consider the Categories page, then the number of new videos isn't overwhleming. Only when you look at ALL.

(BTW, The Channels page needs a Misc. for videos NOT in any category)

phew!!! joe is here! ")

Hmm, it looks like people want to go back to 10.

Did anyone notice this poll can be voted multiple times on multiple computers? Heh.

Joedirt, is right. If there is going to be some way we will solve this we need better statistical data from which we can ascertain trends on the sift.

Moar dataz please!

"Did anyone notice this poll can be voted multiple times on multiple computers? Heh."
Now we know where 8 of those 10 votes came from

I am happy that the three separate issues (Votes Needed to Graduate, Queue Rot, and Collective Pain) that have been raised about the Sift today have been broken out into their own playlists.

On the Votes Needed issue, the number needed to graduate should defined by a failure metric (or success metric if your sift is half full). How many videos do you hope fail?

I would hope for a success rate of 65-70% (or a failure rate of 30-35%) as to not totally discourage posting by members. If 1 out of every 2 of my videos failed . . .

*wait! more than 50% of my videos ARE failing according to a quick analysis of my discard/published piles. what the hell am i doing here?

The SiftGods should give the collectives a week or two to work at the 10 level to see if the success rate is skyrocketing (I suspect that it is increasing). If an average of more than 70% of the videos submitted are skyrocketing to the front, then bump the number up to 11 and wait a week and see what happens.

I don't think the issue should be framed as "I can't keep up with the front page any more in this Brave New Sift!" so we need to increase the votes needed. As PM wrote, there is no way humanly possible to 'keep up' anymore.

I do think the issue should be framed as "Are the Sift's holes too big? And are too many Sifts getting through?"

I agree there are *more* videos out there, but are there really that many more "good" videos? I've seen a lot float to the top just because more sifters are queue voting.

I would say the metric is how many queue votes per week total of all sifters. And then peg the escape limit to that. Also pluck out high flyers sooner (five in 24 hrs) and dump the bad videos quicker. If we have twice as many eyes on the queue as we did six months ago, we can change the 24 hr and 48 hr numbers lower. So IT SHOULD BE 12 VOTES! See, we NOW have more hands and sifters in the queue, so it can happen quicker (no need for 4 days). So the queue size shouldn't be growing, unless the ratio of submitters to queue sifters is changing.

Dag: "Lets roll back the queue escape to 10."

Done.

Dang, so many videos from this morning. DOH!! 5 pages!

Comment hidden because you are ignoring dag. (show it anyway)

Ant in your profile, change your posts per page to 50. There, now only one page.

i changed my option to 50 the very first day it was offered ....

Comment hidden because you are ignoring dag. (show it anyway)

I'd just like to comment that there are now 90 comments on this thread.

Dag is wrong. It's 91.

Re: anon queue

Actually, I try - esp. with save calls - to focus on un-starred or lower ranked members, and try to follow a similar approach with vids (but I usually only zero in on ones that look interesting rather than on how sifted them - I try to ignore the ID thing, now that I think about it, except when saving...)

Secondly, i think DD's idea about the queue containing anonymous post information a really really cool idea. I've mentioned before that diamond and 100 gold star posters generally get their stuff unfairly voted up by default, so why not level the playing field?

The main problem to solve: how do we prevent queue rot?

A few ideas:

1. Insert three or four randomly chosen videos on every queue page. Videos that user hasn't voted on (and by implication, hopefully hasn't seen)

2. Change the static 10 vote escape to a log based on time. For example, let a video escape if it's been in the queue for T time and has V votes:

time < 01 hr: 5 votes time < 12 hrs: 7 votes time < 24 hrs: 8 votes time > 01 day: 10 votes

Alternatively, we could use the same scheme, but instead of keying by time, use the queue page.