This subject has been touched on in other threads but we've never sat down and really thought about this.

When the time comes for a swap to begin the actual selection process, where you go through and say "I'll take this" for all the stuff you want: how many per day should you be able to call dibs on?

It can't be a set number that applies to all swaps, because some swaps will have a small number of items, and others will have thousands of items. So it needs to be something that scales well.

One idea I had was to have the system count how many entries are available, and then let you choose X% of them per day. So if there are 500 items, perhaps you can call dibs for 5 per day. Or if there's 10,000 items, maybe you get to call dibs for 50 items.

What do ya'll think? Any ideas?

I like 5 per day .

Is it going to be equal dibs for all participants? A percentage makes sense, otherwise it could take way too long with large amounts of seed available.

5 per day for this swap seems about perfect, I think.

I'm actually looking for feedback for how to program this for the future. We're going to be hosting many swaps out of this feature and they will all be different. I need to program it once to work for them all.

So I can't say "5 dibs per person per day" because that won't work for all swaps.

I like the percentage.

Can it also factor in how many people are in on the swap? Say two different swaps each had 500 packets. But one swap only has 10 people participating but the other has 50 participating. Amount of dibs per day should reflect that?? Not sure that is possible. Just throwing it out there.

Thinking more I guess my idea maybe is not possible since people might join late OR seeds might get added as the swap goes on

That's an interesting idea.

I think there's an algorithm here that I can make that makes sense. Something like:

Average amount of offers per member / Number of Participants / Number of days of trading still ahead of us.

So as you get nearer to the end you can do more dibs since by then everyone has gotten the things they can't live without.

Once the trading begins, nobody can join the swap.

I like 1% of items to trade

I was thinking earlier a number that might work, but as already mentioned, it could take too long in some swaps. What you came up with sounds good to me, something that would work for all swaps.

A percentage that is adjusted upward throughout the swap should work.

What happens if Sam Seeder requests 40 packets incrementally over the allotted swap period, but Susie Swapper et al. only selects 15 of Sam's packets? Will there be a fail-safe once the swap is over, or nearly so, to make the "unearned" seeds available to all of the participants? I'm okay with Sam getting all of his requests but I can see how that could be a problem in this set up.
And how would the seeds Sam forfeits be determined? Automatically based on which seeds he selected first which would likely be his favorites? Or do we want to talk about that yet?

dave said:That's an interesting idea.

I think there's an algorithm here that I can make that makes sense. Something like:

Average amount of offers per member / Number of Participants / Number of days of trading still ahead of us.

So as you get nearer to the end you can do more dibs since by then everyone has gotten the things they can't live without.

That should work very well, especially since there will be no late participant entries.

What about late seed entries, will that be possible?

How would this work if we go with the idea that you can't ask for more packets than you offer? Will that be programmed in or will the moderator have to keep watch to make sure no one joins only offering 10 packs of seed but asking for 50?

I think the consensus was that if you're participating, you can request as many packets as you think is reasonable. I'm not planning on programming a limit based on how many packets you're offering.

dave said:I think the consensus was that if you're participating, you can request as many packets as you think is reasonable. I'm not planning on programming a limit based on how many packets you're offering.

That sounds good to me.

The problem I see is setting a flat number per day is the members could be giving away more packs of seeds per day than they can hope to get in return. For example: 19 members each select five packs of seed from me the first day. That's 95 packs of seeds requested from me, but I can only pick 5. If the swap runs for two weeks,the most I would be able to pick is 70 packs total, less than I would have give away on the first day.
Here is a thought if it's not to difficult to program. Start everyone with 5 packs per day. If a member receives more than 5 request from there list, then their available dibs would increase by that amount up to a set maximum, maybe 10 or 15.
So if I have 5 packs or less requested from my list then I get 5 dibs. If I receive 8 request for my seeds then my dibs would increase to 8. It would increase every time someone picked from my list until the daily maximum was reached.

>> So as you get nearer to the end you can do more dibs

I agree strongly. In many swaps, the most-desired things will be picked in the first few days. In a 30-day swap, I might set a limit like "5" for the first week, "10" for the second week, and "no limit" after that.

I would let the host pick the daily limit. Maybe suggest something based on your formulas.

However, just because one swap has 100 offerings and another has 10,000 offerings, doesn't necessarily mean there are more "desirable" or "competed-for" offers in the big swap.

In fact, very discerning and picky traders might gravitate toward tiny swaps where just a few people offer just a few things - RARE or specialized things. They are the ones that need careful throttling to give everyone a fair chance.

Hence my suggestion that the host may over-ride the suggested daily limit, and specify how fast the limits ramp down. If that gets complex, it could even be just an on-off option: limit requests to N per day for the first D days, then no daily limits.

Also:

It will be very time-consuming to find our "most wanted" 5 things out of 1,000 offerings, hence I wish we could flag or prioritize ALL our "must-haves" in one or two passes, even if we have to go back later and formally request them in priority order, a few at a time, once per day, for the next 10-20 days.

I think that, as things are now, we would have to keep a manual list on paper or in some offline doc format, of each "must have" that we discover each day after reaching that day's daily limit.

Horntoad said:The problem I see is setting a flat number per day is the members could be giving away more packs of seeds per day than they can hope to get in return. For example: 19 members each select five packs of seed from me the first day. That's 95 packs of seeds requested from me, but I can only pick 5. If the swap runs for two weeks,the most I would be able to pick is 70 packs total, less than I would have give away on the first day.
Here is a thought if it's not to difficult to program. Start everyone with 5 packs per day. If a member receives more than 5 request from there list, then their available dibs would increase by that amount up to a set maximum, maybe 10 or 15.
So if I have 5 packs or less requested from my list then I get 5 dibs. If I receive 8 request for my seeds then my dibs would increase to 8. It would increase every time someone picked from my list until the daily maximum was reached.

But, on the other hand, if I have 50 packs of seeds to offer and no one picks any of my seeds for five days in a row, I'm up to 25 packs of seeds that I'm getting... and then the next five days, all 50 packs get requested? Then I'm even. So, my seeds weren't the first pick but they were desirable. I guess it could be fair that I get more dibs since I offered more desireable seeds.

Edit to say I wasn't finished but somehow my message was sent - and now I don't remember what else I was going to say....

Jay, I think the first motivation or plan for "you can't request more than have been requested FROM you" was to apply it at the END of the swap, mainly to filter out Greedy Seeders who intentionally offer little but request lots.

I think you're suggesting using that same policy of "request no more than requested" to increase fairness of access to the most popular seeds offered. Maybe that is a good idea, I would have to think about it.

If so, maybe there would be some place for a two-phase approach ... I'm just spit-balling here.

Say the first D days of some swaps were focused on fair access to the most-desired items.
The host could select whether or how strongly to go in this direction.

Low daily limits. Maybe reward those who get oinked firstest and fastest, since clearly they offered the tasty stuff.

After that first phase, focus the swap on convenience. Increase or remove the daily limit and fall back on total-number-of-requests as the main fairness criterion for the second phase.

Just an idea. The host could select a daily limit for the first D days, the daily limit AFTER Day D, and a value for "D".

I don't think that would be TOO complex for most hosts ... but I do love details.

Maybe Dave could give hosts a "host-startup Wizard". First pici a "flavor" of swap, and Dave would fill in each field automatically with representative values for that KIND of swap.

Then the host could edit those choices, after seeing what the typical format would be.

1. Small swap with tightly focused theme, for Collectors who know exactly what rare and unusual things they want, and just hope any are offered and they get a fair shot at getting that rare thing. They want what they want and not much else. Low daily limits that last through 90% of the swap.

2. Big generic swap with lots of fairly available stuff for lots of people. For example, "vegetables". Fairness still matters a little because things like pkt size and freshness still matter, and there is always something that "goes first" even if nothing is Collector-Level rare.

Medium daily limits that ramp down before the swap is 1/2 obver.

3. Bottom-of-the-barrel swaps, where nothing is too old and no ID is too vague. Cross-pollinated seeds fine.l Re-trades fine. Large daily limits that go away after the first week.

My thought is to have three phases or rounds with slightly diff rules. Maybe each lasts for 5 or 7 days.

Round 1: say there are 400 items with 20 users and round 1 is 5 days. 400/20/5 = a limit of 4 items per user per day. But since I only contributed, say 8 items, I don't think I should be able to pick more than 8 items.

Round/phase 2: maybe we increase the daily limit. Am not sure how to handle users who have now picked the number that they contributed.
Round/phase 3: do we say no limit?

Whatever is done, I see a green or red light shown than alerts me that I'm red "met my allowance" or green "limit 4, or whatever, per day" .

Anyway,this is a tough one.