Rather than extend the previous thread even more... it occurred to me that I might start another thread... so....
________________

Here's a first stab at making some (public) suggestions.

I assume that we wish to find the best problems the quickest (even if we choose to post only one of the best, and pair a not-so-best with it, on a given day).

It stands to reason, that the best problems are not in the "front 10"... when there are many more than 10 in the total queue. (...working on the assumption that we didn't, for example, happen to submit them in decreasing order of goodness).

To find those best problems quickest, we need some mechanism to go through more problems quicker.

(By "find" I mean that scholars/journeymen are able to peruse it and rate it.)

Now, one obvious mechanism (which I am *not* advocating) is to allow them to see the whole queue, rather than just the "top 10". I think this might be somewhat unwieldy... and I will defer others' judgement, since I am not yet a journeyman.

Another mechanism, that might be more wieldy, would be to increase the range from 10 to 20 problem simultaneously.... and further create at least one new queue.... as follows:

I assume (someone correct me here as needed), that if a problem gets three Thumbs Down (TD), then it gets "deleted"... it is deemed not worthy of Flooble and is taken out of the potentially-posted-problems list.

I suggest that there be at least three choices of voting... TU, TD, and TH "thumb-horizontal, flat palm?"... the third choice indicating that it is worthy of being posted on Flooble, but there may be some more interesting/relevant problems that might take precedence.

This would carry a weight of zero (not impacting the TU/TD weight), but if a problem gets 3-5 of these, it can be moved to this new queue (an 'alternate' list)... This queue may contain a whole bunch of problems, ALL of which have been deemed flooble-worthy, and could at any time be posted.

This new alternate queue, though, would generally grow, as there would likely be more problems placed on this queue than are taken off (much as the current queue is now). At least, it would grow if we continue to have the pace of submission that we've seen as of late.

And then, of course, when posting, problems could be pushed from the "top 10-20", or from the front of the alternate queue.
_________________

You may ask... if we do this, haven't we simply moved from one queue to another? No! ...at least not quite. Because what we've done is... reviewed far more of the flooble problems. Therefore, the better ones get to the front faster, while still granting a reward to those who submit earlier. (e.g., all of Ravi's 100+ submissions from February, will be reviewed before anyone's from September)

And lastly, you may think this would be a lot of work on Levik's part.... but I have faith in Levik's abilities... and I'm sure it wouldn't take much time. ;-)

Criticisms? Compliments? Concerns? Amendments?

--- SK

Gamer, I'm not sure I understand what you're saying. Would you please be more detailed?

This "alternate queue" sounds suspiciously like the backlog idea in the other thread. I'm glad that this idea came up again, though it's really more for journeymen and scholars to discuss.

All these thumbs pointing in every direction, would any of them cancel each other out? Do they currently cancel each other out? Could gamer explain what he's talking about?

Yes Tristan, this is why I proposed 3-5 TH before it gets sent to the alternate queue. No "auto-sending" your own problems.

As for the thumbs cancelling each other out.... as I addressed already... 1 TU cancels 1 TD... +1, -1. But you still need a minimum TH before you can be deemed at least flooble-worthy.

So, to be absolutely clear: one would not be able to place one's own problems in the alternate queue.

I am saying my idea of saying if a problem has a note by someone else (like a journeyman or scholar or such) it would have been in the queue at some time. Then the problems wouldn't jump around.

I am talking about a quick fix, if Levik doesn't want to spend all his time with an alternative queue.

In my opinion, the queue works good enough for me, but I don't know how new people might feel.

I think the queue has been cutting down due to the less amount of problems being submitted. Although I am unsure why it is, I don't think there were as many duplicate/lesser good problems getting TD now than there were when the problems were flooding.

I know that my recent logic problems haven't got much exposure, showing that even less discussion stirring problems can slow down the discussion on a good problem.

I think DJ has a good point which I overlooked as well, the fact that we don't want all our "good" problems (come to think of it, what makes up a "good" problem anyway... they are ususally controversial) in one scoop.

By exposure, I mean problems that have been sitting around the voting queue for days or weeks before making it onto the site. That usually happens (getting into the 'weeks' range is rare) when a problem is being debated or modified; the way the queue is running now, problems are generally posted shortly after they are approved. That's a good thing. There's no reason to have an extra queue of approved problems waiting around to be pushed, when we're not short on problems to push as it is. I also fail to see what wondrous good 'identifying' the best problems sooner is going to do.

The rate of posting isn't increasing, so having all these ridiculous queues is, first, not going to change as much as you seem to think, and second, just going to mean that the voters are looking at more problems sooner. That's what I mean by [over-]exposure, and that's bad, not good.

Getting discussion on the main page is good, which wouldn't happen (as I noted in my previous overlooked comment) if we posted an extrememly challenging puzzle every day, several days in a row.

Also, if we did that, the queue wouldn't keep up; frankly, we don't generally see five 'exceptional' puzzles submitted a week, so we'd eventually be cutting off our own legs.

So, I propose that your assumption is misplaced. There is no reason to even find the 'best' problems more quickly, as they won't be posted any more quickly; they're assorted more or less randomly as is, so that the problems are distributed roughly evenly with regard to 'goodness.'

By the way, I just worked white suspenders and really liked it

Thanks
Fatboy

Do we have to go back to the reason we don't have more than 2 problems per day, AGAIN? I thought we had covered that already.

(Actually what I was saying above was there are LESS TDs than there were before... It's not as important to the discussion.)

I really don't see any problem with the current queue other than problems jumping in and out of it due to queue weight. I don't think there is as many problems as when this thread was started.

The issue of changing the posting rate has been discussed and, I thought, closed not too long ago. We could go into it again, but I think any pertinent questions were brought up and addressed in the previous thread.
I stated that we're not going to increase the posting rate simple because, well, we're not, at least not any time soon.

I also don't think, as I stated before, that putting all the best problems on the site as soon as we can is necessarily a 'better' decision. Sure, we'd have good problems for a while, but none of them would get the viewer attention it probably warrants (especially so if the posting rate were to increase as well), and eventually we'd end up with a long run of 'mediocre' problems. Given SK's idea, we'd put all the 'great' problems on the site first, and when those ran out, we'd go back to the alternate queue and push the 'okay' problems. I think it's plain enough that this isn't the most favorable situation, unless you alternate posting problems from the 'good' queue and the 'alternate' queue -- which in essence is what happens now, but with limits as to the number of problems that are being examined at any given time.

There are two underlying ideas you are overlooking.
The first is longevity; there's no reason to assume that the ratio of good to okay problems is increasing when they are being subitted, so there's no reason to increase that ratio in the problems that are being posted now.
The second idea is the primary purpose of this site; if the point were to give people as many puzzles as quickly as possible, we would just go through the entire queue, post the problems with the solutions as soon as we could, and let people just browse through the puzzles. There are a lot of sites that do that, which is fine. The way this site is set up, however, with the comment system, holding back the solutions, even the aesthetics of highlighting a few porblems at a time on the main page, is geared to show good puzzles, but also to foster the sense of a web community, putting out ideas even if you don't have a full solution to the problem, that sort of thing.

I think that, while the current set up is not flawless, it is very close to ideal, in that:
Problems are able to be sorted, filtered, or edited as necessary before they are posted.
We see a good mix of problems, by category, difficulty, length (of problem and solution), discussion, etc, posted to the site throughout the week (I don't know what you want to change as far as this goes, but if we start pulling problems from farther than farther back in the queue, the imbalance will eventually catch up to us).
The way levik has determined the posting should be done (weekday, weeknight, and weekend problems) is very good to allow front page time and exposure to the problems that most warrant it.
Problems move through the active voting queue relatively quickly, so that they are generally not 'overexposed' to the scholars and journeymen who review them, before the problems are pushed live to the main site (this is the biggest thing that your suggestion would probably affect, detrimentally, and I don't really see the benefits as far as anything else is concerned in the long run).

****

To Tristan: I'm not sure what you are asking. If we look at more problems sooner, as SK is suggesting, without upping the rate of problem posting, then it would have to follow that we are seeing problems for longer before they are pushed onto the site, and more of them at that. Now, of course we have to see the problems in order to approve/disapprove them, request any necessary changes, etc. Ideally, we want them posted, once they're approved, fast enough so that they're still relatively new to us when they hit the main site, and so that we don't have a slew of voters with pre-typed solutions before anyone else even sees the problem.
Personally, I won't attempt to solve a problem that's in voting; there are some problems for which the solution comes immediately upon reading the problem, but that's often an idication that it's too easy or trivial for the site. Other people do figure out the solutions in voting, or at least insofar as to emsure that one exists and is satisfying (we have come across, in the past, a few problems that could not be solved or were indeterminate, or just weren't as good as the problem itself sounded), which is good and sometimes needed.
Anyway, that's what's meant by preexposure to the problems in queue, and why we try to minimize it.

As DJ said, the goal isn't to distribute "great" problems more appropriately; it's to distibute all the problems at an appropriate time. While occassionally there is a situation where none of the problems ready to be pushed are (for example) "weekend" problems, and a "morning" problem has to be pushed on Saturday, this really does not happen all that often. If it does start happening more frequently, the other, forgotten part of your suggestion, increasing the "top ten" to the "top twenty" (or even only "top fifteen") would probably be sufficient to cure it.

Seems like if we combine our problems together it would make the queue smaller. If the bulk of the information is the same, the problems should be clustered together. Also, if it isn't pulling them apart is likely better (as I have observed)