I know I'm not the only player who thinks villager trading is imbalanced. I find it silly that compasses are worth more than diamonds, and overpowered that an hour at my sugarcane farm (yes, it's that large) can net me several doublechests of diamond items. In fact, the level of imbalance, I am told, is why some Bukkit server administrators are considering disabling trading on their servers altogether until there is a rebalance (or a plugin to allow them to rebalance it themselves).
Now, I've seen many suggestions for fixing the balance issues. Some have been very simple, re-assigning about two handfuls of offers based on the poster's notion of the items' worth. Others have suggested the removal of offers, or even the addition of more offers. Others still have suggested traders should imitate the real world, complete with economic ideas of scarcity, flexible supply and demand, and an analysis of the player market on a server.
My idea is neither of these: 1.3 is right around the corner, and any significant changes to trading mechanics would take a bit of time to implement, and even more time to test. And a rebalance of trading based on my own personal values of items will also fail to help. Rather, my idea is the following:
I propose that no offers need to be added or removed, and almost no trading mechanics need to change. Rather, all offers' prices will be modified to match the numbers in the following chart:
(Green is an offer to sell something to villagers, and blue is an offer to buy from them. Purple are a priest's enchantment offer)
The only mechanic of trading which needs to be changed to support this suggestion is that offers requiring more than 64 emeralds would use both input slots, and that offers requiring more than 128 emeralds would use emerald blocks in one of the offer slots. It would actually be nice if villagers automatically equated emerald blocks to 9 emeralds each, but this isn't necessary for the suggestion to work.
Now, how is this any different from me giving my own personal values for items? Because I spent a good 25 hours (across 3 days) putting that together: the prices are the result of a multi-step process which considers every reason an item should have value, and I've asked for numerous second-opinions on it. You can see much of the actual process in the chart below:
Now, what's all this? First, I came up with the value in the "Value Based on Generation Rarity or Ingredients" column. Then, I came up with adjustments to these values, shown in the following columns. The "Adjusted for Usefulness" column is bold because you should look at this when comparing the actual values of items according to this chart - it's after all adjustments are applied except for those specific to villager trading.
Grey text in most areas indicates that I did not factor in any adjustments for that cell (e.g. the difficulty in killing iron golems or zombies for iron ingots is not factored in at all: I acknowledge it's a possibility, but find it too negligible to raise the value in the "Adjusted for Ease of Harvesting" column).
In the "Value Based on Generation Rarity or Ingredients" column, grey italic text means the value in the cell is kind of shot to hell, as comparing the amount of mobs in the world to the amount of ores is like comparing apples and oranges. The value is fully compensated by the cells following it; this just means "rarity" is not an effective way to judge the item's value. Italic text which is not grey in this column indicates that the rarity value is dubious, because I couldn't find hard figures for how much of the item is naturally generated.
The contents of this spoiler explain each column of the chart; yes it is that long. You don't have to read the contents of the spoiler unless you're really into this. In fact, I'm amazed if you've made it this far without skimming.
Although, if you're confused about what a particular column is, consulting the spoiler before asking is advised.
Value Based on Generation Rarity or Ingredients: For items generated in the world, this is their rarity relative to the rarity of gold ore (so diamond ore is 2.4x as rare, etc.). For items which are crafted or smelted, this equal to the "Adjusted for Usefulness" value of their ingredients (the cost of smelting is not included) - literally counting these items as the sum of their parts. For mob drops, as I said, the rarity of mobs on generation is kinda meaningless, but the values for the items are based on how much of the item is dropped and how many mobs drop the item (string also acknowledges Abandoned Mineshafts at this point). Finally, if the item can be earned from a chest (dungeons, mineshafts, and villages. Strongholds are ignored), this is also accounted for, but the result is usually negligible.
Adjusted for Ease of Finding: This acknowledges ore buried deep in the earth is much harder to find than stuff growing on the surface. It also ups the value of biome-specific resources, and compensates for the inaccuracy of mob "rarity" per chunk. Once again, if an item can be found in a dungeon, mineshaft, or village chest, it gets a (negligible) reduction in value.
Adjusted for Capability to be Farmed: Probably the biggest problem in the current implementation of trading is its potential for exploitation - suddenly a sugarcane farm is a diamond pickaxe farm. Items are drastically reduced in value if they can be farmed (this includes hostile and passive mobs). The reduction in value is a factor (whereas most adjustments are just added or subtracted from the item's value). This factor considers how long it takes for a single unit (e.g. tree, 3-tall sugarcane, etc.) to grow, whether bonemeal can be used to hasten the process, and how much a single unit of the farm yields - all on top of an inherent 0.4x multiplier just for the fact that the item can be farmed. The process for mob farms is slightly different, acknowledging that animal farms need wheat, and that hostile mob spawners are ridiculously fast. If an item is farmable due to being a rare drop, this will also slightly factor in - however, auto-cooking farms are considered too negligible and thus do not further depreciate cooked items (its price is already devalued because the raw form is farmable). I have considered gold nugget (pigmen) farms negligible for the purposes of this chart.
Adjusted for Ease of Harvesting: This one counts mining, mob killing, and actual crop harvesting. For farms, replanting is considered here, but usually harvesting is easily automatable or otherwise instantaneous, and thus further depreciates the resource's value. For mining, pickaxe requirements are considered, and for mob killing, the mob's health and attack strength (in Normal mode) are considered.
Adjusted for Crafting/ Smelting: What it says on the tin. If an item is annoying to craft, it will be annoying to craft en masse; this very slightly increases its value. Smelting is more significant, accounting for the cost of 1/8 a lump of coal/charcoal, and the time required for smelting. The time it takes to sign a book-and-quill is also acknowledged here.
Adjusted for Usefulness: This is, in fact, the final modifier worth looking at if you wanted fair trades. Raw materials with many purposes get a large boost from this; their crafted results are thereafter reduced in value. Since an ore is worth the resource it drops, its usefulness is factored in there. Many things modify the usefulness of an item: for example, I considered a lump of coal and a block of glowstone as equally useful; this part is thus rather tricky and open to opinion. The usefulness of tools is based not only on their literal functions, but on their durability - armor also acknowledges its worth in terms of armor points. Food considers the hunger restored, and, to a lesser extent, saturation of the item (these values are applied to the raw form, because it is a precursor to the cooked form). If an item is farmable, even its usefulness modifier is scaled by the farming factor. Bottles o' Enchanting are interpreted as worth their average xp, which is itself considered from a number of factors which I used to determine the worth of an enchantment. Chainmail armor is considered equal in value to iron armor (before its usefulness adjustment is applied), with chainmail's durability and armor values.
Adjusted for Retail: Villagers can be stingy sometimes. Items sold to them are sold for less than their worth, and items bought from them are sometimes bought for more than their worth. This applies especially to the rich priests, but is also applied to diamond items to prevent trading from becoming overpowered. In the case of chainmail, its value is increased because the blacksmith is the only one who knows how to make it. In the case of saddles, however, the value is actually reduced because I presume Jeb intended for trading to be a more reasonable way to obtain the item (also, realistically, making hundreds of tiny chain-links is far more difficult than crafting a saddle . See? the butcher is still being stingier than the blacksmith!).
NPC Trader's Price: The value to look at for what all this would actually change. Villagers would have these prices rather than their current ones. If something costs more than 64 emeralds, the second offer slot may be used. If something costs more than 128 emeralds, emerald blocks will be used in one offer slot. Ideally, villagers would be able to calculate that an emerald block is worth 9 emeralds, and thus accept the two items interchangeably, but this is not necessary to fix trading. Note that the random ranges used here are based on the quantities involved, and capped at 5 in each direction, to not have excessive randomness in offers.
Fair Trade: Iron Ingots Per Stack: Let's say you're a human, not a squidificate. This is the value of a stack of an item, in iron ingots, based on the value in the "Adjusted for Usefulness" column - the retail adjustment is not counted here. If two players wanted to make a trade between a stack of the item and iron ingots, and wanted the trade to be fair (and ingore whatever items each player actually owns), then they might consider the number here to be a reasonable amount of iron. Bear in mind that I said "stack", not 64: ender pearls stack to 16, and many items do not stack (and thus, the price in iron is for one of that item).
Fair Trade: Iron Ingots Per Stack (Rounded): This one is just for convenience; iron is rounded to the closest value and stacked (each stack of iron is 64).
Oh, right, and the golden-colored cell next to Gold Ore? It's the only cell in this chart which contains a value dependent on no other cell. That is, if you make a change to it, all other prices will change accordingly. The values in the last two columns will not change, because the value of iron changes as well. This cell allows for the value of emeralds to be arbitrarily re-assigned with ease.
Now, none of this is any good if I'm the only one suggesting these values (well, some friends and I). If anybody reading this has any suggestion for a change in values, please speak up. Please don't just say something like "diamond should sell for more". Actually give a specific number. Furthermore, give a reason! The whole point of this chart is that every bit of an item's value has some justification behind it. If you don't get what I mean, check the second spoiler.
Finally, my chart is downloadable here, so anybody can play with it (I recommend using OpenOffice Calc, as that's what I made it in). Yes, it's as dynamic as I say it is: Halve the value in the gold-colored cell, and all values are halved (e.g. emerald doubles in value). Make wood less valuable, and suddenly coal becomes less valuable, and thus all items made via smelting and all items using wood become less valuable.
The downloadable chart also contains a column to help generate the code needed for this change, along with a set of cells beginning at S43. Whether you're Mojang trying to implement it, or a mod/plugin developer, I think it can help reduce tedium (although you probably would have done that on your own ).
There are also an extra two spreadsheets in that file: the abridged version of the chart (which I posted at the top of this topic), and a sheet for alternative values for controversial offers:
As one user suggested, here's what happens when emeralds are raised in value a bit:
EDIT: If anybody wants to test out how these changes would pan out, and they own a Bukkit server, I've actually made a plugin implementing this: http://dev.bukkit.or...ance-villagers/
By default, it also disables the removal of offers, as some people have said that makes trading too underpowered and non-renewable. However, the plugin is fully configurable, and you can customize all offers - the chart above can even export the needed parts of the configuration file, so you can go straight from modifying my chart to testing the results ingame!
Hopefully it's not against the rules for me to post this plugin here? I think making the changes testable would help people better decide whether they'd want them, and I still want them in vanilla.
Nice! most of these are good, but some tweaks are needed, like chainmail should be made cheaper. It's worth less than iron, actually. Just because you used to be unable to get it legitimately doesn't make it worth more. Also, the costs for bookshelves need to be fixed. I think it should be 1-2 per emerald.
Nice! most of these are good, but some tweaks are needed, like chainmail should be made cheaper. It's worth less than iron, actually. Just because you used to be unable to get it legitimately doesn't make it worth more. Also, the costs for bookshelves need to be fixed. I think it should be 1-2 per emerald.
Alright, I removed the retail adjustment for bookshelves, which leaves them at 1.3664, or 1-2 emeralds. I wouldn't make them 1-2 per emerald, as this places it at the same value as 1.5 books. They cost 3 books, and 6 wood as well; that cost can't be left out.
Chainmail is a bit trickier. Chainmail helmets should not be worth less than iron helmets: they have the same durability and armor points, making them literally the same. Other pieces of chainmail are just half an armor bar less than their iron equivalents, and have the same durability as iron. You'll see the "Adjusted for Usefulness" column acknowledges this.
However, villagers do not need to price equally with the actual values of items: that's the whole point of the "Adjusted for Retail" column. If you were the only one who could make chainmail armor, why value it less than iron armor? I suppose it could be equal, as it presumably has a lower material cost and is harder to make. But it shouldn't be less. Any ideas?
Alright, I removed the retail adjustment for bookshelves, which leaves them at 1.3664, or 1-2 emeralds. I wouldn't make them 1-2 per emerald, as this places it at the same value as 1.5 books. They cost 3 books, and 6 wood as well; that cost can't be left out.
Chainmail is a bit trickier. Chainmail helmets should not be worth less than iron helmets: they have the same durability and armor points, making them literally the same. Other pieces of chainmail are just half an armor bar less than their iron equivalents, and have the same durability as iron. You'll see the "Adjusted for Usefulness" column acknowledges this.
However, villagers do not need to price equally with the actual values of items: that's the whole point of the "Adjusted for Retail" column. If you were the only one who could make chainmail armor, why value it less than iron armor? I suppose it could be equal, as it presumably has a lower material cost and is harder to make. But it shouldn't be less. Any ideas?
Oh, I forgot about retail. It was still too expensive compared to books, being adjusted for retail twice in you calculations. Also, I checked the wiki, and chain is more enchantable than iron. Iron and chain should have the same cost.
Oh, I forgot about retail. It was still too expensive compared to books, being adjusted for retail twice in you calculations. Also, I checked the wiki, and chain is more enchantable than iron. Iron and chain should have the same cost.
Actually, the way my chart works, the retail adjustment can't ever be applied twice: the value of books that booksheves are based on is from the "Adjusted for Usefulness" column. I suppose I'll also make that column bold to reduce confusion; the values from that column are what determine crafted items' values.
The way the retail adjustment worked in this case, I halved the value of a book being sold to a villager, and added one emerald to the value of a bookshelf bought from a villager. I suppose the latter adjustment wasn't really needed, as it would be good if trading were a decent way to get shelves. But the adjustment to books has no bearing on shelves, as they used the value of books from before retail adjustment.
At any rate, bookshelves shouldn't be lowered in value because a player can simply buy some shelves at 2 per emerald, break the shelves into three times as many books, and whether the book-buying offer is 4 or 5 per emerald, they would make a profit and make the villager look like a fool
And making chainmail and iron equal would make sense I suppose; the reason I felt a higher price is suitable is because it's more of a novelty. The odds of a chainmail offer showing up are incredibly rare anyhow; by the time one does usually, you have more than enough emerald to afford it. A friend of mine agrees with me on that note; maybe we should have several versions for alternate pricing on controversial items, and let Mojang make the call on whichever they prefer? You aren't the only one to think chainmail is overpriced.
The only thing I find wrong with your post is with chainmail.It's only now obtainable in vannila and I doubt anybody would even use it because of its price and weakness.With 1.3 a full set of chainmail will be a sign of wealth and talent with trading.In my opinion I think chainmail should cost a near exorbanent price somewhere between iron and diamond.
Chainmail isn't meant to be worn for show. I would never buy any chainmail but a helmet if it was more expensive than iron.
Anyone can play around with it and try correcting things they feel need correcting. I'll merge changes I support into the main post; hopefully nothing too extreme is in need of changing.
I would say that the costs for Enchanting should be significantly higher. I don't have an actual number in mind, but when you consider that you are, in essence, renting the usage of an Enchanting Table, an item which takes several Diamonds and Obsidian (a material only obtainable with a Diamond Pick), which doesn't even factor in making bookshelves to set the level. Three emeralds seems not even close to the value of what you would need to do it yourself, even factoring in that the table is a one-time cost. I'm also surprised by the "usefulness" adjustment on Enchants, which lists them as less useful than a Compass or Clock. This seems unreasonable to me. A compass is of some use, and more use in making a map (I am quite surprised that you give it lower usefulness than a Clock) but ultimately one can navigate without it, particularly if they have a system of signs and/or established roads set up, and a clock will tell you the time, but so can looking at the sky. Comparatively, enchanting any piece of equipment is going to make it more effective in some way, possibly (maybe even likely) granting capabilities that can not be approximated by other means. That surely warrants more than just pocket change as a cost.
I would say that the costs for Enchanting should be significantly higher. I don't have an actual number in mind, but when you consider that you are, in essence, renting the usage of an Enchanting Table, an item which takes several Diamonds and Obsidian (a material only obtainable with a Diamond Pick), which doesn't even factor in making bookshelves to set the level. Three emeralds seems not even close to the value of what you would need to do it yourself, even factoring in that the table is a one-time cost. I'm also surprised by the "usefulness" adjustment on Enchants, which lists them as less useful than a Compass or Clock. This seems unreasonable to me. A compass is of some use, and more use in making a map (I am quite surprised that you give it lower usefulness than a Clock) but ultimately one can navigate without it, particularly if they have a system of signs and/or established roads set up, and a clock will tell you the time, but so can looking at the sky. Comparatively, enchanting any piece of equipment is going to make it more effective in some way, possibly (maybe even likely) granting capabilities that can not be approximated by other means. That surely warrants more than just pocket change as a cost.
Compass has a usefulness modifier of -.45, while clock has -.04, and enchantment has -1. The usefulness modifiers of compass and clock are based on the loss in usefulness compared to their source ingredients: you can't get your iron back out of your compass and use it for something else. I never said a clock's more useful than a compass; the reason its loss is less than a compass is because gold's own usefulness was very small to begin with. In fact, clock loses 100% of its gold-use, while compass only loses a fraction of its iron-use because it can still be used for maps and such.
Unfortunately, at any rate, "usefulness" is probably where the most opinion ends up in this chart. But negative usefulness is based on the loss of a potent material, as the material's own value (including its usefulness) is used in the calculation of a crafted item's base value. I suppose I should have not made crafted items the sum of their materials from "Adjusted for Usefulness", but perhaps one column to the left instead. Then there would be no need for negative usefulness, and the resulting confusion.
As far as enchantments, the level a priest provides is random, between 5 and 19, inclusive (thus 85-341 xp). I assumed the average of the required xp here, which puts the level at 9. The value of a level 9 enchanment pales in comparison to a level 30, which is why I felt the service was not all that useful. Of course, I neglected to consider the free enchanting table and bookshelves at work.
My basis for having a negative usefulness at all was that an unenchanted item has much more potential (for better enchantments), but then, I suppose the fact that you see the enchantment you're paying for removes all doubt, which might make up for that. I'll remove the negative usefulness modifer for enchantment offers, placing them at +1 emerald of usefulness with your other considerations. That makes them 5-6 Emeralds instead of 3; they need not be more expensive or the player may as well have built themselves an enchanting room already (actually, at these low levels, bookshelves may not be necessary). For diamond items, however, the value of en enchantment is reduced to +.5 usefulness, or 4-5 Emeralds: they're less enchantable, and lose much more of their potential when not enchanted at the maximum level, plus you'd really have an enchantment table by then.
Hmm...that seems reasonable, and your explanation makes sense. Given that it's just low-level enchantments, I suppose that five or six emeralds sounds about right. I'm not sure if I would make it cheaper for diamond, myself, though. The resulting tool will still be more effective than another tool with the same enchantment. I suppose that it does make sense, since it's a bit of a waste, but on the other hand you know what you're getting so it makes up in reliability what it lacks in potency. If the priest is selling an enchantment you want, it's more effective to buy than do it yourself.
Hmm...that seems reasonable, and your explanation makes sense. Given that it's just low-level enchantments, I suppose that five or six emeralds sounds about right. I'm not sure if I would make it cheaper for diamond, myself, though. The resulting tool will still be more effective than another tool with the same enchantment. I suppose that it does make sense, since it's a bit of a waste, but on the other hand you know what you're getting so it makes up in reliability what it lacks in potency. If the priest is selling an enchantment you want, it's more effective to buy than do it yourself.
Yeah, I guess I'll make diamond and iron enchants equal again; I left my alternate idea in the alternate pricing chart at the bottom of the page. Also, I made a mistake in my last comment: in 1.3, that much xp is actually level 12, not 9
I think this is a good idea because it's silly that villager trades really can either make you rich or turn you bankrupt. I would say that this is an easier way to think of the colors:
I've added a poll. People have been telling me that these prices still leave trading very overpowered. People have also been saying that these prices make trading very underpowered.
I can't reconcile that. I can't reconcile that any better than I can reconcile the opinions of one person who has told me I undercharge for diamonds and overcharge for items made of diamonds, when I've already valued diamond items as less than the sum of their parts.
I don't think anybody can reconcile that. So I've made a poll, where people can say whether this does too much or not enough. It can't sensibly do too much and not enough simultaneously. Ideally, the votes in the poll would have an equal amount of people at both extremes, and a decent amount of people in the middle. If there are too many people at one extreme and not enough at the other, that says my suggested prices have failed to fix the issue. If the votes balance out, that says this chart's done as good as it can at reworking the prices for everyone.
If you don't feel like voting because you disagree with some part of the chart, that's fine. You could even wait until I reply to you about your concerns and (possibly) modify the chart, before you cast your vote, if you like.
Also, I'm removing the retail modifier on mined items; it's already hard enough to get emeralds without farming as it is.
It's hard to answer the second question on the poll. Half of the vanilla trades are overpowered, the other half underpowered.
That's a good point. Should I make both questions multi-choice? (as in, a checkbox for every option?)
Or would it be better if I added an "It's imbalanced in both ways: some offers are overpowered and some are underpowered." to both questions? I'm thinking of doing the latter; fortunately this forum lets people change their votes.
In my opinion, the vanilla system overpowers trading overall. While some offers may be undervalued, it's easy to abuse an overpowered offer to create a surplus of emeralds, and than use that surplus to buy up offers that would otherwise be unreasonable. So, "good" offers will most likely benefit the player more than "bad" ones will take away. Overall, that skews the balance toward the player.
It's just my opinion, but I would say that it's better to err on the side of making the balance generally skewed toward trading being a poor exchange, since if people don't use trade, the overall difficulty of the game is unaffected, while making deals favour the player runs the risk of making things generally easier than they are now. Getting an item without having to find the materials...that is, a guaranteed "right now" item, is already more convenient than getting what you need to make it, and will always allow plentiful resources to be exchanged for scarce ones, even with poor exchange. While limited availability of offers does some to eliminate this, it isn't to hard to force new villagers to spawn so that you can get a fresh set. Which means that even if offers are poor, there will always be some value to trading over DIY, and it's just a matter of how frequently people are inclined to take advantage of the convenience.
For the most part I like this rebalanceing of costs. Things really are just too far out of whack at the moment when I can buy diamond pants for 13 emeralds and a diamond sword for 11 emeralds, among other things.
At the same time I don't think that it should be purely mathematical, there is still the issue of availability of the items that you want from the villagers that exist because you can't just buy anything whenever you want. I think that points could be made that prices should be made more or less than their calculated rate due to convenience or availability. For example it makes since to trade with a villager since they will have items that would be time consuming to hunt for, but if the price is more than the effort required to hunt for an item, it becomes pointless to trade with them. Maybe a random variable included in the calculation (ingame) so that good/bad deals will exist say 50% +/-. I was thinking 20% initially, but thought that a wider range might make better deals stand out.
For the most part I like this rebalanceing of costs. Things really are just too far out of whack at the moment when I can buy diamond pants for 13 emeralds and a diamond sword for 11 emeralds, among other things.
At the same time I don't think that it should be purely mathematical, there is still the issue of availability of the items that you want from the villagers that exist because you can't just buy anything whenever you want. I think that points could be made that prices should be made more or less than their calculated rate due to convenience or availability. For example it makes since to trade with a villager since they will have items that would be time consuming to hunt for, but if the price is more than the effort required to hunt for an item, it becomes pointless to trade with them. Maybe a random variable included in the calculation (ingame) so that good/bad deals will exist say 50% +/-. I was thinking 20% initially, but thought that a wider range might make better deals stand out.
I think it should be 30%, to prevent overpowered trades. Also, I think trades should be assigned a random stock, instead of have a chance of removal, and they run out even if the trade GUI is still open. Blacksmith sells would tend to have low stock. One reason for this is so they can be a set amount in custom maps.
EDIT: I'll post this in small suggestions.
Now, I've seen many suggestions for fixing the balance issues. Some have been very simple, re-assigning about two handfuls of offers based on the poster's notion of the items' worth. Others have suggested the removal of offers, or even the addition of more offers. Others still have suggested traders should imitate the real world, complete with economic ideas of scarcity, flexible supply and demand, and an analysis of the player market on a server.
My idea is neither of these: 1.3 is right around the corner, and any significant changes to trading mechanics would take a bit of time to implement, and even more time to test. And a rebalance of trading based on my own personal values of items will also fail to help. Rather, my idea is the following:
I propose that no offers need to be added or removed, and almost no trading mechanics need to change. Rather, all offers' prices will be modified to match the numbers in the following chart:
(Green is an offer to sell something to villagers, and blue is an offer to buy from them. Purple are a priest's enchantment offer)
The only mechanic of trading which needs to be changed to support this suggestion is that offers requiring more than 64 emeralds would use both input slots, and that offers requiring more than 128 emeralds would use emerald blocks in one of the offer slots. It would actually be nice if villagers automatically equated emerald blocks to 9 emeralds each, but this isn't necessary for the suggestion to work.
Now, how is this any different from me giving my own personal values for items? Because I spent a good 25 hours (across 3 days) putting that together: the prices are the result of a multi-step process which considers every reason an item should have value, and I've asked for numerous second-opinions on it. You can see much of the actual process in the chart below:
Alternate upload: http://i50.tinypic.com/2dtpw6s.png
Now, what's all this? First, I came up with the value in the "Value Based on Generation Rarity or Ingredients" column. Then, I came up with adjustments to these values, shown in the following columns. The "Adjusted for Usefulness" column is bold because you should look at this when comparing the actual values of items according to this chart - it's after all adjustments are applied except for those specific to villager trading.
Grey text in most areas indicates that I did not factor in any adjustments for that cell (e.g. the difficulty in killing iron golems or zombies for iron ingots is not factored in at all: I acknowledge it's a possibility, but find it too negligible to raise the value in the "Adjusted for Ease of Harvesting" column).
In the "Value Based on Generation Rarity or Ingredients" column, grey italic text means the value in the cell is kind of shot to hell, as comparing the amount of mobs in the world to the amount of ores is like comparing apples and oranges. The value is fully compensated by the cells following it; this just means "rarity" is not an effective way to judge the item's value. Italic text which is not grey in this column indicates that the rarity value is dubious, because I couldn't find hard figures for how much of the item is naturally generated.
The contents of this spoiler explain each column of the chart; yes it is that long. You don't have to read the contents of the spoiler unless you're really into this. In fact, I'm amazed if you've made it this far without skimming.
Although, if you're confused about what a particular column is, consulting the spoiler before asking is advised.
Value Based on Generation Rarity or Ingredients: For items generated in the world, this is their rarity relative to the rarity of gold ore (so diamond ore is 2.4x as rare, etc.). For items which are crafted or smelted, this equal to the "Adjusted for Usefulness" value of their ingredients (the cost of smelting is not included) - literally counting these items as the sum of their parts. For mob drops, as I said, the rarity of mobs on generation is kinda meaningless, but the values for the items are based on how much of the item is dropped and how many mobs drop the item (string also acknowledges Abandoned Mineshafts at this point). Finally, if the item can be earned from a chest (dungeons, mineshafts, and villages. Strongholds are ignored), this is also accounted for, but the result is usually negligible.
Adjusted for Ease of Finding: This acknowledges ore buried deep in the earth is much harder to find than stuff growing on the surface. It also ups the value of biome-specific resources, and compensates for the inaccuracy of mob "rarity" per chunk. Once again, if an item can be found in a dungeon, mineshaft, or village chest, it gets a (negligible) reduction in value.
Adjusted for Capability to be Farmed: Probably the biggest problem in the current implementation of trading is its potential for exploitation - suddenly a sugarcane farm is a diamond pickaxe farm. Items are drastically reduced in value if they can be farmed (this includes hostile and passive mobs). The reduction in value is a factor (whereas most adjustments are just added or subtracted from the item's value). This factor considers how long it takes for a single unit (e.g. tree, 3-tall sugarcane, etc.) to grow, whether bonemeal can be used to hasten the process, and how much a single unit of the farm yields - all on top of an inherent 0.4x multiplier just for the fact that the item can be farmed. The process for mob farms is slightly different, acknowledging that animal farms need wheat, and that hostile mob spawners are ridiculously fast. If an item is farmable due to being a rare drop, this will also slightly factor in - however, auto-cooking farms are considered too negligible and thus do not further depreciate cooked items (its price is already devalued because the raw form is farmable). I have considered gold nugget (pigmen) farms negligible for the purposes of this chart.
Adjusted for Ease of Harvesting: This one counts mining, mob killing, and actual crop harvesting. For farms, replanting is considered here, but usually harvesting is easily automatable or otherwise instantaneous, and thus further depreciates the resource's value. For mining, pickaxe requirements are considered, and for mob killing, the mob's health and attack strength (in Normal mode) are considered.
Adjusted for Crafting/ Smelting: What it says on the tin. If an item is annoying to craft, it will be annoying to craft en masse; this very slightly increases its value. Smelting is more significant, accounting for the cost of 1/8 a lump of coal/charcoal, and the time required for smelting. The time it takes to sign a book-and-quill is also acknowledged here.
Adjusted for Usefulness: This is, in fact, the final modifier worth looking at if you wanted fair trades. Raw materials with many purposes get a large boost from this; their crafted results are thereafter reduced in value. Since an ore is worth the resource it drops, its usefulness is factored in there. Many things modify the usefulness of an item: for example, I considered a lump of coal and a block of glowstone as equally useful; this part is thus rather tricky and open to opinion. The usefulness of tools is based not only on their literal functions, but on their durability - armor also acknowledges its worth in terms of armor points. Food considers the hunger restored, and, to a lesser extent, saturation of the item (these values are applied to the raw form, because it is a precursor to the cooked form). If an item is farmable, even its usefulness modifier is scaled by the farming factor. Bottles o' Enchanting are interpreted as worth their average xp, which is itself considered from a number of factors which I used to determine the worth of an enchantment. Chainmail armor is considered equal in value to iron armor (before its usefulness adjustment is applied), with chainmail's durability and armor values.
Adjusted for Retail: Villagers can be stingy sometimes. Items sold to them are sold for less than their worth, and items bought from them are sometimes bought for more than their worth. This applies especially to the rich priests, but is also applied to diamond items to prevent trading from becoming overpowered. In the case of chainmail, its value is increased because the blacksmith is the only one who knows how to make it. In the case of saddles, however, the value is actually reduced because I presume Jeb intended for trading to be a more reasonable way to obtain the item (also, realistically, making hundreds of tiny chain-links is far more difficult than crafting a saddle . See? the butcher is still being stingier than the blacksmith!).
NPC Trader's Price: The value to look at for what all this would actually change. Villagers would have these prices rather than their current ones. If something costs more than 64 emeralds, the second offer slot may be used. If something costs more than 128 emeralds, emerald blocks will be used in one offer slot. Ideally, villagers would be able to calculate that an emerald block is worth 9 emeralds, and thus accept the two items interchangeably, but this is not necessary to fix trading. Note that the random ranges used here are based on the quantities involved, and capped at 5 in each direction, to not have excessive randomness in offers.
Fair Trade: Iron Ingots Per Stack: Let's say you're a human, not a squidificate. This is the value of a stack of an item, in iron ingots, based on the value in the "Adjusted for Usefulness" column - the retail adjustment is not counted here. If two players wanted to make a trade between a stack of the item and iron ingots, and wanted the trade to be fair (and ingore whatever items each player actually owns), then they might consider the number here to be a reasonable amount of iron. Bear in mind that I said "stack", not 64: ender pearls stack to 16, and many items do not stack (and thus, the price in iron is for one of that item).
Fair Trade: Iron Ingots Per Stack (Rounded): This one is just for convenience; iron is rounded to the closest value and stacked (each stack of iron is 64).
Oh, right, and the golden-colored cell next to Gold Ore? It's the only cell in this chart which contains a value dependent on no other cell. That is, if you make a change to it, all other prices will change accordingly. The values in the last two columns will not change, because the value of iron changes as well. This cell allows for the value of emeralds to be arbitrarily re-assigned with ease.
Now, none of this is any good if I'm the only one suggesting these values (well, some friends and I). If anybody reading this has any suggestion for a change in values, please speak up. Please don't just say something like "diamond should sell for more". Actually give a specific number. Furthermore, give a reason! The whole point of this chart is that every bit of an item's value has some justification behind it. If you don't get what I mean, check the second spoiler.
Finally, my chart is downloadable here, so anybody can play with it (I recommend using OpenOffice Calc, as that's what I made it in). Yes, it's as dynamic as I say it is: Halve the value in the gold-colored cell, and all values are halved (e.g. emerald doubles in value). Make wood less valuable, and suddenly coal becomes less valuable, and thus all items made via smelting and all items using wood become less valuable.
The downloadable chart also contains a column to help generate the code needed for this change, along with a set of cells beginning at S43. Whether you're Mojang trying to implement it, or a mod/plugin developer, I think it can help reduce tedium (although you probably would have done that on your own ).
There are also an extra two spreadsheets in that file: the abridged version of the chart (which I posted at the top of this topic), and a sheet for alternative values for controversial offers:
As one user suggested, here's what happens when emeralds are raised in value a bit:
EDIT: If anybody wants to test out how these changes would pan out, and they own a Bukkit server, I've actually made a plugin implementing this:
http://dev.bukkit.or...ance-villagers/
By default, it also disables the removal of offers, as some people have said that makes trading too underpowered and non-renewable. However, the plugin is fully configurable, and you can customize all offers - the chart above can even export the needed parts of the configuration file, so you can go straight from modifying my chart to testing the results ingame!
Hopefully it's not against the rules for me to post this plugin here? I think making the changes testable would help people better decide whether they'd want them, and I still want them in vanilla.
Alright, I removed the retail adjustment for bookshelves, which leaves them at 1.3664, or 1-2 emeralds. I wouldn't make them 1-2 per emerald, as this places it at the same value as 1.5 books. They cost 3 books, and 6 wood as well; that cost can't be left out.
Chainmail is a bit trickier. Chainmail helmets should not be worth less than iron helmets: they have the same durability and armor points, making them literally the same. Other pieces of chainmail are just half an armor bar less than their iron equivalents, and have the same durability as iron. You'll see the "Adjusted for Usefulness" column acknowledges this.
However, villagers do not need to price equally with the actual values of items: that's the whole point of the "Adjusted for Retail" column. If you were the only one who could make chainmail armor, why value it less than iron armor? I suppose it could be equal, as it presumably has a lower material cost and is harder to make. But it shouldn't be less. Any ideas?
Oh, I forgot about retail. It was still too expensive compared to books, being adjusted for retail twice in you calculations. Also, I checked the wiki, and chain is more enchantable than iron. Iron and chain should have the same cost.
Actually, the way my chart works, the retail adjustment can't ever be applied twice: the value of books that booksheves are based on is from the "Adjusted for Usefulness" column. I suppose I'll also make that column bold to reduce confusion; the values from that column are what determine crafted items' values.
The way the retail adjustment worked in this case, I halved the value of a book being sold to a villager, and added one emerald to the value of a bookshelf bought from a villager. I suppose the latter adjustment wasn't really needed, as it would be good if trading were a decent way to get shelves. But the adjustment to books has no bearing on shelves, as they used the value of books from before retail adjustment.
At any rate, bookshelves shouldn't be lowered in value because a player can simply buy some shelves at 2 per emerald, break the shelves into three times as many books, and whether the book-buying offer is 4 or 5 per emerald, they would make a profit and make the villager look like a fool
And making chainmail and iron equal would make sense I suppose; the reason I felt a higher price is suitable is because it's more of a novelty. The odds of a chainmail offer showing up are incredibly rare anyhow; by the time one does usually, you have more than enough emerald to afford it. A friend of mine agrees with me on that note; maybe we should have several versions for alternate pricing on controversial items, and let Mojang make the call on whichever they prefer? You aren't the only one to think chainmail is overpriced.
Chainmail isn't meant to be worn for show. I would never buy any chainmail but a helmet if it was more expensive than iron.
http://www.mediafire.com/?5db2gqz4jxbrdfe
Anyone can play around with it and try correcting things they feel need correcting. I'll merge changes I support into the main post; hopefully nothing too extreme is in need of changing.
Compass has a usefulness modifier of -.45, while clock has -.04, and enchantment has -1. The usefulness modifiers of compass and clock are based on the loss in usefulness compared to their source ingredients: you can't get your iron back out of your compass and use it for something else. I never said a clock's more useful than a compass; the reason its loss is less than a compass is because gold's own usefulness was very small to begin with. In fact, clock loses 100% of its gold-use, while compass only loses a fraction of its iron-use because it can still be used for maps and such.
Unfortunately, at any rate, "usefulness" is probably where the most opinion ends up in this chart. But negative usefulness is based on the loss of a potent material, as the material's own value (including its usefulness) is used in the calculation of a crafted item's base value. I suppose I should have not made crafted items the sum of their materials from "Adjusted for Usefulness", but perhaps one column to the left instead. Then there would be no need for negative usefulness, and the resulting confusion.
As far as enchantments, the level a priest provides is random, between 5 and 19, inclusive (thus 85-341 xp). I assumed the average of the required xp here, which puts the level at 9. The value of a level 9 enchanment pales in comparison to a level 30, which is why I felt the service was not all that useful. Of course, I neglected to consider the free enchanting table and bookshelves at work.
My basis for having a negative usefulness at all was that an unenchanted item has much more potential (for better enchantments), but then, I suppose the fact that you see the enchantment you're paying for removes all doubt, which might make up for that. I'll remove the negative usefulness modifer for enchantment offers, placing them at +1 emerald of usefulness with your other considerations. That makes them 5-6 Emeralds instead of 3; they need not be more expensive or the player may as well have built themselves an enchanting room already (actually, at these low levels, bookshelves may not be necessary). For diamond items, however, the value of en enchantment is reduced to +.5 usefulness, or 4-5 Emeralds: they're less enchantable, and lose much more of their potential when not enchanted at the maximum level, plus you'd really have an enchantment table by then.
Yeah, I guess I'll make diamond and iron enchants equal again; I left my alternate idea in the alternate pricing chart at the bottom of the page. Also, I made a mistake in my last comment: in 1.3, that much xp is actually level 12, not 9
Green = >
Blue = >
Purple = >
Well maybe not.
I can't reconcile that. I can't reconcile that any better than I can reconcile the opinions of one person who has told me I undercharge for diamonds and overcharge for items made of diamonds, when I've already valued diamond items as less than the sum of their parts.
I don't think anybody can reconcile that. So I've made a poll, where people can say whether this does too much or not enough. It can't sensibly do too much and not enough simultaneously. Ideally, the votes in the poll would have an equal amount of people at both extremes, and a decent amount of people in the middle. If there are too many people at one extreme and not enough at the other, that says my suggested prices have failed to fix the issue. If the votes balance out, that says this chart's done as good as it can at reworking the prices for everyone.
If you don't feel like voting because you disagree with some part of the chart, that's fine. You could even wait until I reply to you about your concerns and (possibly) modify the chart, before you cast your vote, if you like.
Also, I'm removing the retail modifier on mined items; it's already hard enough to get emeralds without farming as it is.
That's a good point. Should I make both questions multi-choice? (as in, a checkbox for every option?)
Or would it be better if I added an "It's imbalanced in both ways: some offers are overpowered and some are underpowered." to both questions? I'm thinking of doing the latter; fortunately this forum lets people change their votes.
It's just my opinion, but I would say that it's better to err on the side of making the balance generally skewed toward trading being a poor exchange, since if people don't use trade, the overall difficulty of the game is unaffected, while making deals favour the player runs the risk of making things generally easier than they are now. Getting an item without having to find the materials...that is, a guaranteed "right now" item, is already more convenient than getting what you need to make it, and will always allow plentiful resources to be exchanged for scarce ones, even with poor exchange. While limited availability of offers does some to eliminate this, it isn't to hard to force new villagers to spawn so that you can get a fresh set. Which means that even if offers are poor, there will always be some value to trading over DIY, and it's just a matter of how frequently people are inclined to take advantage of the convenience.
At the same time I don't think that it should be purely mathematical, there is still the issue of availability of the items that you want from the villagers that exist because you can't just buy anything whenever you want. I think that points could be made that prices should be made more or less than their calculated rate due to convenience or availability. For example it makes since to trade with a villager since they will have items that would be time consuming to hunt for, but if the price is more than the effort required to hunt for an item, it becomes pointless to trade with them. Maybe a random variable included in the calculation (ingame) so that good/bad deals will exist say 50% +/-. I was thinking 20% initially, but thought that a wider range might make better deals stand out.
https://www.youtube.com/playlist?list=PL0OtPNZX22RvZVeq4-dHa8GYKOc5lojKX
I think it should be 30%, to prevent overpowered trades. Also, I think trades should be assigned a random stock, instead of have a chance of removal, and they run out even if the trade GUI is still open. Blacksmith sells would tend to have low stock. One reason for this is so they can be a set amount in custom maps.
EDIT: I'll post this in small suggestions.