Thoughts on tradeskill platinum macro exploits

[furnok]

A note that I sent to Sony:

The problem with the platinum exploits on macroing tradeskills, as I see it, is this.

The Playing field
* Every recipe is known to exist
* Every component has a vendor buyback and sell price
* Recipes can be manufactured through several layers of subcombines

The Challenge
* Identify all of the vendor sell prices for base components (Those that cannot be manufactured)
* Identify all of the vendor sell prices for intermediate combines
* Identify the manufactured price for intermediate combines
* Identify all of the vendor buyback prices for intermediate combines
* Identify all of the vendor buyback prices for final combines

The Analysis
* Find the intermediate combines where the buyback price is greater than the manufactured cost
* Find the final combines where the buyback price is greated than the manufactured cost using the smaller of the vendor sale price or the manufactured cost for items that are intermediate combines

What the macroers to do after a patch
* Look for price changes
* Update the database
* Rerun the analysis
* Macro it a bunch

The Fix
* Return the price of the base components (medium quality ore, poison vials, etc.) back to their previous levels.

What the player community will end up doing
* Get more frustrated at folks like Yantis who exploit the game for personal advantage
* Make a public database that demonstrates the loop holes and run the QA for you

What I see happened
Some tradeskill developer decided to lower the price on a singular item so that some combines would be cheaper. As a result, he destroyed the entire economy because he did not follow the chain.

For example, sealed poison vials used to cost 6 pp each. Farming the zombie skin and making the sealed poison vial yourself cost you 3-4 pp each (I forget the exact number). Therefore, you could farm the skins, make the vials yourself, and sell them in the Bazaar for 5 pp each to rogues who did not want to work up pottery, that was a good thing.

Then they changed the price of sealed poison vials to 1 gp. Why? Now, it costs more to make the vial yourself than buy it from the vendor.

Same thing happened with the ore.

The solution is to put the prices of the raw components back to where they were. Put the medium quality ore back to 2 pp each. Then the entire plat macro problem is solved.

What I want to know:
Do you have a QA tool that can do the analysis described above to find and eliminate these exploits? Since you already have the database, you don't even have to do the research step.

Recipe
Component A - Vendor Sell Price / Vendor Buy Price
Component B - Vendor Sell Price / Vendor Buy Price
Product C - Cost = Component A + 2 * Component B
Product D - Cost1 = 2 * Product C Cost + Component C
Product D - Cost2 = 2 * Componet A + 4 * component B + Component C

If Product D Buyback > Product D Cost1, then pricing error.
If Product D Buyback > Product D Cost2, then pricing error.

The current round of platinum exploits is being worked from the second equation - making the expensive componets from the cheaper base components and then selling it back.

Vendors sell farmed components. So this equation works just fine for every recipe. You can consider the price of a NO DROP item to be zero for the purposes of this evaluation because you cannot macro recipes that use NO DROP items, because you cannot obtain an unlimited supply of the item.

The fact of the matter is that every item in the game has a vendor buyback price and a vendor sell price. Anyone can do a catalog of prices from the vendor, get the buyback prices, build the database, and RUN THE ANALYSIS. Sony, with the base database to start with, doesn't even have to do the research.

Think of what Allahkhazam has done. For a premium member, every recipe can be broken down to its most basic components. The cheapest manufactured price can easily be calculated with such a breakdown.

[Saleani]

It looks like the "correction" to the last pricing discrepancy was to drastically reduce the merchant buy back price for the items made with those products. The proper correction would have been to put the items back to the original sale price. In the old days a pricing screw up like this would have merely resulted in the vendor being permakilled until the next patch, and the price restored to the original (and correct) value.

Now, with the coldain velium temper issue and the small brick of MQ ore issue, it looks like the corrections seem to be to reduce the sell-back price of the items made. And as the macroers decide to make further combines, the next recipes in the chains will be messed with by having their sell-back price fooled with. Woohoo, the buy-back price of coldain tempers used to be 4pp, now it is 1 copper. And you still have to make them. That will fix the folks who got a few free stacks.

The end result of this sort of price fixing will be to make the assumption that all tradeskillers will be macroing or always be making combines from the cheapest source. As well as the assumption that you have to do all of the subcombines to prevent bleeding plat. How many of you make fine plate from blocks of MQ ore that you buy for 20pp? Due to the pricing screw up, you can still make those blocks for 1gold. Should the buy-back prices of dozens of fine plate and LDON plate armors get reduced because the assumption is that you made the folded sheets from blocks you made or bought? Will there be some mistake, so that large 3-resin blue legs don't get changed, but the rest will? How many possible combinations are there? 13 slots x 3 sizes x 7 dyes x 3 resins for just fine steel. And 13 slots x 3 sizes x 5 LDoN tempers. That is 1014 recipes that have to have the pricing changed and checked. It does not count the weapons or shields you can make from MQ ore. Could you afford to even make the stuff for skilling up if you lost 30pp-70pp per success instead of only losing 15pp? Would you bother making them at all? How many arrow recipes have to be investigated if the assumption is that you will make your shafts for 2gp per stack rather than buy them for 25pp per stack? 3 heads x 6 fletchings x 3 nocks = 54 recipes, that is not too bad. Would we ever see a thread again where folks ask where they can farm pelts for making leather padding again? I ended up burning through 26kpp to get fletching from 202 to 220 (for the aid gimel quest), will future fletchers hemorage that kind of plat just getting to 200?

I am working on getting smithing to 188 and fletching to 202 (well, I will stop at 200) on all my chars before the next patch. I shudder to think what the next "correction" will be. The pattern is that we are getting the solutions to the WRONG problems. Making a quick fix to satisfy an irate boss has almost always caused trouble down the road. Because you did not have the time to fix the problem correctly. One of the things I keep in mind when writing software is always: if you there isn't time to do it correctly now, when will you have the time to do it correctly?

I sent some emails to Sony offering to make a VB app that could do the data mining that you were referring to. I don't need to have all the data to make the application, and it is simple enough that any code monkey can modify it to point to the databases in actual use. It would let folks find these things right away, instead of watching the economy get trashed. So far all of my emails have been ignored. Since I am moving in just over a week, I don't expect to see any answers (or be playing) for 2+ more weeks.