Collectable Valuation Project

[SuperNiCd]

...

Feeling a bit guilty about proposing scope creep. I might be willing to help out on this piece if the consensus is that it has value.

[Dielzen]

You say that you're not interested in unbound crafting, so then how are you going to price your rares to sell them on AH?

How is Non_Crafter_001 going to know how much he should pay to get the collectibles in order to provide them to a crafter who will combine for free if the parts are supplied?

How much is Crafter_234 going to charge for his unbound shards that he crafts to make completed items (or shards) to sell to people or on the AH?

Personally, I'm disappointed whenever I pull epic level rares, as I've no plan to take my crafting skill high enough to make unbound stuff, but there's people out there that will, and it should be taken into account.

You can "short-cut" it easily enough, as the only difference is including the rare pulls into the matrix, possibly by including a separate table just for those rare drops only used in unbound crafting.

[SuperNiCd]

You say that you're not interested in unbound crafting, so then how are you going to price your rares to sell them on AH?

How is Non_Crafter_001 going to know how much he should pay to get the collectibles in order to provide them to a crafter who will combine for free if the parts are supplied?

How much is Crafter_234 going to charge for his unbound shards that he crafts to make completed items (or shards) to sell to people or on the AH?

Personally, I'm disappointed whenever I pull epic level rares, as I've no plan to take my crafting skill high enough to make unbound stuff, but there's people out there that will, and it should be taken into account.

You can "short-cut" it easily enough, as the only difference is including the rare pulls into the matrix, possibly by including a separate table just for those rare drops only used in unbound crafting.

Compelling points, these. Personally I've been holding my rares in case I do have a need to make something unbound at some point. But being as they are rare, as long as some percentage of crafters are interested in unbound crafting, some are likely really valuable and could be converted directly or indirectly to something more useful to me.

I kind of wonder what percentage of crafters do unbound crafting, for whatever reason, beyond just making unbound ML shards post U32.

[Xiadais]

On the subject of collectibles builds, I think an 18mnk/2rgr would be optimal for runs which don't need trapping skills or mandatory DDoor. It has 60% movement speed without Haste, fewer sprint boosts (meh), instant recalling (especially useful in Precious Cargo, Baudry, Breaking the Ranks and the like), and either double wings or DDoor (if you use Shadowdancer instead of EA?). Thoughts? Anything I'm missing?

[EllisDee37]

This won't help with speed runs for collectables but I was just running thru Epic Wheloon and noticed a lot of backpacks, lost count at around 12ish. Only 1 was behind a locked door. There were also 2 cabinets, 1 in the hideout and another behind a locked door. Depending on speed/alert and familiarity this may be worth looking into.

Thanks much for the reminder, I will definitely be checking the wheloon backpack bonanza for both heroic and epic Any runs. It's in my notes but I was about to totally space it.

On the subject of collectibles builds, I think an 18mnk/2rgr would be optimal for runs which don't need trapping skills or mandatory DDoor. It has 60% movement speed without Haste, fewer sprint boosts (meh), instant recalling (especially useful in Precious Cargo, Baudry, Breaking the Ranks and the like), and either double wings or DDoor (if you use Shadowdancer instead of EA?). Thoughts? Anything I'm missing?

Some of the best collectable farms require Open Lock, I believe they fixed the fast recall for monks (a guildie tried and failed at it just a few days ago) and DDoor frequently speeds things up.

However, there are still a bunch of collectable farms where it would do great. Way better than a random average character, for sure.

And if you can get fast recall to work then it would be the undisputed champ at ornate charm farming, which alone could make it worth it.

[EllisDee37]

Actually I do have an idea if you'll bear with me through the TL;DR that's going to follow.

I have two ideas that are sort of related to what you're talking about.

First, my next TR life I plan to go from 1 to 30, with no other activity on other alts while I do it. My typical run is all quests one and done on elite / reaper 1. So I'll make a list of all collectables I have before I start, and then a month later when I hit 20 I can see how many of each collectable I pulled during heroics. My epic leveling usually involves a healthy dose of epic dailies, which account for at least 5 million of my 8.25 million epic xp each life. (I don't much like epics.)

Second, I've been kind of thinking of talking to cru about setting up a wiki page for data that my crafting builder can pull data from to get updated values for collectables. It would ideally be an unformatted, raw data page that can only be edited by me and wiki admins. The idea being to keep the amount of data pulled as minimal as possible, and protected so that people can't go in and change values just before offering a trade.

Not sure if either idea will materialize, but those are things I've been thinking about.

[EllisDee37]

You say that you're not interested in unbound crafting, so then how are you going to price your rares to sell them on AH?

I was actually thinking about that very thing last night after logging off.

My thinking is that rares should be valued as the average of uncommons in their school+tier, or possibly both uncommons and commons. The idea is that their value is heavily reduced by the lack of unbound crafters, but they still have actual value that's related to how many there are available. And that number would be determined by how many of the non-rare collectables of the same school+tier people have or farm.

[EllisDee37]

Did a daily epic run tonight to test:

Tobias EH => VON3 EN => Spies EN => Mirror EN

Total time to log in, run the four quests, decon all the loot and rebuff (to be ready to go the next day) was 60:32, and I generated 699 essences. EH Tobias slows me down a solid 5 minutes compared to EN, but it's probably worth it for the Ornate Charms.

Collectables gathered during those four runs:

1 Ancient Text
2 Brass Censers
1 Crypt Moth
1 Mystical Formula
4 Mortars and Pestles
2 Ornate Charms
2 Pouches of Bones Fragments
1 Slime Mold
1 Smoldering Ember
1 Vial of Dragon's Blood Ink
2 Zygomycota Fungi
---
18 pulls

9 Arcane
3 Cultural
2 Lore
4 Natural


Dispensers from memory:

Tobias (Tier 6): Backpack, backpack, moss, backpack
VON3 (Tier 5): Backpack, altar, fungus, rubble, cabinet, fungus, alchemy, backpack
Spies (Tier 5): Alchemy
Mirror (Tier 6): Cabinet, cabinet, cabinet, cabinet, cabinet


I'm thinking a 5% penalty per pull for epic dailies, pro-rated for Any dispensers.

Tier 5 dispensers:

3 Any
0 Cultural + (3 * 0.5) = 1.5 * 5% = 7.5% penalty
4 Arcane + (3 * 0.22) = 4.66 * 5% = 23.3% penalty
0 Lore + (3 * 0.2) = 0.6 * 5% = 3% penalty
2 Natural + (3 * 0.08) = 2.24 * 5% = 11.2% penalty

Tier 6 dispensers:

3 Any
0 Cultural + (3 * 0.5) = 1.5 * 5% = 7.5% penalty
5 Arcane + (3 * 0.22) = 5.66 * 5% = 28.3% penalty
0 Lore + (3 * 0.2) = 0.6 * 5% = 3% penalty
1 Natural + (3 * 0.08) = 1.24 * 5% = 6.2% penalty

[SuperNiCd]

I've been kind of thinking of talking to cru about setting up a wiki page for data that my crafting builder can pull data from to get updated values for collectables. It would ideally be an unformatted, raw data page that can only be edited by me and wiki admins. The idea being to keep the amount of data pulled as minimal as possible, and protected so that people can't go in and change values just before offering a trade.

To me the ideal data sources would be (in order of preference):

1. Read from an AH web service provided by SSG (but don't think it exists)
2. All players who use this enter AH data points that are shared (but not sure how to solve the problem of cheating)
3. One trustworthy person's data that is updated periodically (I'd trust yours)
4. The user's own hand-entered data points (not shared, their data only affects their valuation projections)

3 & 4 are kind of a toss up with some pros/cons for each - could possibly be combined.

I'm thinking a 5% penalty per pull for epic dailies, pro-rated for Any dispensers.

I'm on board with penalizing collectibles pulled from epic dailies, conceptually - agree with the rationale. But here's what I'm grappling with a bit. Why specifically penalize those only when there are other collectible level/schools that are also potentially deserving of penalty?

For example, right now the game would appear to be slanted toward heroic runs while people rack up RXP and racial PLs. There may be some heroic quests that are being spammed more than epic dailies at this time. There are surely other examples, and I would think they change over time based on game changes and other factors.

I don't much like epics.

Straying off topic I know, but I am curious... Why so?

[EllisDee37]

For example, right now the game would appear to be slanted toward heroic runs while people rack up RXP and racial PLs. There may be some heroic quests that are being spammed more than epic dailies at this time. There are surely other examples, and I would think they change over time based on game changes and other factors.

Highly doubtful. Reaper XP is best gained with first-time bonuses during heroic leveling, which means not repeating quests. Essentially one-and-done elite (reaper) streaking from 1 to 20.

[SuperNiCd]

I am pretty curious about how these numbers stack up against what's being offered on the AH, so I hacked together a proof of concept to inject this data in. It's pretty crude at this point. No error handling, doubt it will handle more than one data point per collectible. But it's interesting to see how the numbers compare. Feel free to mess with it and if you break it (highly likely lol) just refresh the page to reset. Just added the collectibles that have a tentative baseline from this project at this point.

You can change the going market rates for essences up at the top (I'd set these first). Then add rows of (AS)AH market data using the button, and things should update. Orange colored data are based on the market data with the baselines in black.

This screenshot depicts real data points from Sarlona today. Used the lowest price point I could find in the ASAH. Doesn't mean there were necessarily mass quantities available at that price but most were stacks of 5,10, or 15. What do you make of it?

[Xiadais]

Some of the best collectable farms require Open Lock, I believe they fixed the fast recall for monks (a guildie tried and failed at it just a few days ago) and DDoor frequently speeds things up.

However, there are still a bunch of collectable farms where it would do great. Way better than a random average character, for sure.

And if you can get fast recall to work then it would be the undisputed champ at ornate charm farming, which alone could make it worth it.

I use the instant recall every single time I recall from anything... But I was unsure if I'd been on DDO in the last few days, so I just checked. Still works fine.

[FuryFlash]

What I'm thinking is to assign a value between 1 and 10 to each category, then add up the totals for every shard a collectable is used for to get their demand value.

Universal: 10 points
Universal effects are desirable for every possible build.

Common: 8 points
Common effects are used by many builds, possibly (probably?) a majority of them.

Uncommon: 6 points
Uncommon effects are used by a fair number of builds, though clearly a minority.

Niche: 4 points
Niche effects are very desirable, but only for a small handful of build types.

Obsolete: 2 points
Obsolete effects are better off being replaced by using augments. Still, you might have a reason to craft these if you need it on a lower ML than the augment allows, or maybe the augment is too expensive to acquire.

Rarely Used: 1 point
There is probably someone, somewhere, who will want to use any one of these effects, but for the most part they have little value.

Awesome work so far, EllisDee - I've always been a fan of your projects and use your character planner all the time. Thanks for all the contributions you make to the community.

As for the demand numbers, values between 1 and 10 may be pretty, but since these numbers are an important factor in the final calculation these might need more specific tuning.

For one thing, to reach more reasonable numbers as to what we're seeing right now on the market in game, more categories will probably need to be added. It's just too hard to separate these into such a small number of categories and expect realistic results; looking at the niche category, I don't think "Shield Spikes" and "Spell Penetration" are anywhere near the same supply, yet they are in the same category and therefore are using the same number in the calculation.

Perhaps something more like the following?

Rather than specifically named categories like common or niche, just a simple 1 to 10 scale where 10 is the most common. This should not be the number used in the final calculation!

Rating	10	9	8	7	6	5	4	3	2	1
Examples	Constitution	Strength	Wizardry	Spell Penetration	Combustion	Natural Armor	Bluff	Incite	Bashing	Elf Bane

Filled it with some examples which are just off the top of my head and would preferably be refined over time. Agree or disagree, the point is that the effects in the lowest categories are no where near the top categories, whereas many fairly popular shards are only slightly different values, leaving the middle categories to determine slight differences. I would expect the very lowest values would be very full and the middle and top categories would be relatively evenly spread - of course, it's all relative so the actual magnitude of the values doesn't really matter.

Now, most importantly, the actual "value" of each category. This can't be some arbitrary number because it's used in the final calculation and is one of the few values making it very impactful on the result. Getting the right values here would again require refining and discussion.

The problem with your values of 1 to 10 for the calculation is that "False Life" isn't crafted 10 times more than "Elf Bane". It's crafted at least 100 times more. We can't show that kind of variation with whole numbers between 1 and 10. In fact, the suggested table I made above might not even have enough columns for ranking demand, because of the variation that some shards have. Still, using my table above, let's say we give the 10 category a value of 10. We can't give the 1 category a value of 1, but depending on what is determined based on actual in-game popularity maybe something like 0.1 would be appropriate for the least crafted shards. This is an internal value, so I wouldn't be afraid to use decimals. We still have the pretty 1 to 10 categories or whatever is decided for number of categories, which is irrelevant to the calculation.

For example, with 10 columns, you might have the following actual values:

Rating	10	9	8	7	6	5	4	3	2	1
Actual Value	10	8	6	5	4	3	2	1	0.5	0.1

This is all just numbers I came up with off the top of my head. The main thing is that the difference between the highest and lowest should be very high.

Hopefully this helps - I might just be spouting non-sense.
I look forward to following your progress on this project!

[FuryFlash]

This screenshot depicts real data points from Sarlona today. Used the lowest price point I could find in the ASAH. Doesn't mean there were necessarily mass quantities available at that price but most were stacks of 5,10, or 15. What do you make of it?

I think this just highlights the fact that it would be very helpful to have a calculated value for collectables so that people are less likely to post at prices that don't match the actual rarity or value of the item. If the work in this thread was a finished product that was easily accessible to all AH and ASAH users, I'd suspect we'd see a serious change in prices within a week. There will always be people who don't care and post at any random price though.

[EllisDee37]

Finished Wheloon Prison, which was a major undertaking. So much so that I created its own thread.

Heroic is the best Tier 4 Cultural farm, and epic gives Tier 5s if you need something that elementals don't drop.

[EllisDee37]

Added an estimate for Terminal Delirium for Tier 6 Lore based on my paladin running it a couple times just now. The time of 2:20 involves skipping the locked room entirely, so it misses out on 3 more arcanes, but that's appropriate for a lore farm, I think.

My collectable farmer could probably do it faster, so once the store issues get squared away I'll buy a guest pass (only 70 TP) to capture video. I say "probably" because the collectable farmer dps is very low compared to a capped character, and there is a fight at the start, but we'll see.

As for the demand numbers, values between 1 and 10 may be pretty, but since these numbers are an important factor in the final calculation these might need more specific tuning.
[...]

Excellent thoughts, thanks much for the feedback.

One thing to keep in mind is that I think you're putting more emphasis on the effect of the lower end of the demand scale than is warranted. Consider Elf Bane using the current formula, which only costs Cryptmoss Worm Larva and a soul gem:

Cryptmoss Worm Larva = 40 * (1 + 8 + 1 + 6 + 1 + 1 + 4 + 8 + 6) * (700 / 3600) = 280 essences

If we set elf bane worth 0 instead of 1 in the demand table, that number changes to:

Cryptmoss Worm Larva = 40 * (1 + 8 + 6 + 1 + 1 + 4 + 8 + 6) * (700 / 3600) = 272 essences

I mean, it's different, sure, but not a ton. Worrying about whether the exact demand value for unused bane shards should be 1 or 0.1 or 0.01 or whatever would be missing the forest for the trees, I think. The 1-10 scale for demand was deliberately chosen to be not super granular because demand values are so subjective. I wanted to make it as generalized as possible. A player who only runs S&B builds might think your example on Shield Spikes is off the mark, for example.

Let's see how the numbers shake out as it is now, and then adjust as appropriate. It'll take me a bit to add the data to my crafting planner so I can generate the final table, but once that's done we can revisit the demand scale.

[EllisDee37]

For Cultural Tier 1 to 3, I'm kind of assuming I'll just manually set their value to something super low. There is no farm for them, really, because pretty much no matter how you play the game you'll likely end up with tons of all of them.

Here's how many I have of each right now:

Common
722 Strings of Prayer Beads
952 Blades of the Dark Six
339 Funerary Tokens (I spent like 500 on inflict pots or harm scrolls for my pale master at some point)
736 Amulets of the Lost Empire
280 Small Planar Crystals

Uncommon
162 Small Wooden Idols
203 Marks of the Keeper
223 Signets of the Devourer
193 Amulets of the Six
89 Planar Spoor


I can do timed runs for farming them, but even with that I wouldn't plug them into the formula just as that. They need a manual override of some kind to lower their value. Any ideas welcome.

[SuperNiCd]

For Cultural Tier 1 to 3, I'm kind of assuming I'll just manually set their value to something super low. There is no farm for them, really, because pretty much no matter how you play the game you'll likely end up with tons of all of them.

I can do timed runs for farming them, but even with that I wouldn't plug them into the formula just as that. They need a manual override of some kind to lower their value. Any ideas welcome.

Agree, I'm not sure it's worth your time to do timed runs for them. My counts are similar to yours. They do have some value but most will have far more than they reasonably need. If someone asked for a few in the trade window, I'd probably assume they were a newer player and would probably be inclined to just go hand some to them.

[SuperNiCd]

I think this just highlights the fact that it would be very helpful to have a calculated value for collectables so that people are less likely to post at prices that don't match the actual rarity or value of the item. If the work in this thread was a finished product that was easily accessible to all AH and ASAH users, I'd suspect we'd see a serious change in prices within a week. There will always be people who don't care and post at any random price though.

I don't think the AH is completely uninformed - sure, there will always be deals and ripoffs in there, whether intentional or due to an uninformed seller. But in general if selling something valuable, if it doesn't sell, you'll probably relist, lowering the price until it does. If it sells really quickly you likely under-priced. So it does give feedback that leads to a true market value.

When I gathered this data yesterday, there were multiple listings for Ornate Charms right around that price point - 9.9 AS/ea. Today there's only one listing, and I think the time remaining on the others was longer than 1 day, so I assume they sold for that. Since a seller agreed to sell it for that, and a buyer agreed to buy it for that, and assuming no one was under duress , that actually is the definition of the market value. Granted, the sample size is pretty small here.

To me, the biggest value of this model will be to help me make informed decisions. If I need a collectible, I can either trade resources for it (plat, AS, collectibles, essences, other) or trade time for it.

Right now the model is telling me that if Ornate Charms are available for 9.9AS/ea, I shouldn't farm for them. It is not worth my time because someone else is willing to do the work at a discount.

Does that ring true, or does the model need adjustment?

Is there value in me collecting more AH data - over a longer time period, across more servers, for more collectible types? Would that be a value add for this project?

[SuperNiCd]

To Aarows_of_Fire's point, I think the AH data might be helpful in adjusting some of the demand coefficients. Right now, they are the most arbitrary part of the model. Just from the initial small sample, seems like there might be a cap somewhere on the value of each collectible - somewhere around 10 AS (~500k plat, 530 essences) for one unit of the most valuable collectible, and people may not be willing to pay more than that no matter how long it takes to earn one. But a larger sample might be more revealing.

Wanted to establish a data driven baseline for the cost of essences. I feel like there's enough data here, and enough consistency between the points that these are pretty good numbers. Granted this is only the supply side of the equation but the consistency makes me think these listings will probably sell at these price points if left on the market long enough for a buyer. Threw a couple of obvious outliers away, but they're marked in the data.

Here's how it shakes out. Poor Wayfinder fell out completely due to it's only data point looking like an outlier lol.

Server	Essence Cost Avg (AS)	Essence Cost Avg (PP)
Argonessen	0.019	798
Cannith	0.015	1,040
G-Land	0.018	1,388
Khyber	0.008	927
Orien	0.035	766
Sarlona	0.017	956
Thelanis	0.018	816
All Worlds	0.019	935

We could then also say that 1 AS = 49,475 PP = 53 Essences.

Or maybe 1 AS = 50k PP = 50 Essences is close enough.


Link to Source Data