Original Post
Real-time market approximated price updating and other item related things
I was thinking that the market could be updated in such a way where players can see how much an item would approximately be worth and how often that particular item is selling based on the statistics of people buying. Meaning that the market would display how much an item is "really" worth currently based on how often it is put into the market, what price it is put at in the market, and how often it is bought from the market as well as the torishop. Making it a bit easier on the players who sell their items often to find a good price to buy and sell at in order to make their profit, also making it easier for the consumer to know how much they could be buying this item for. For example: an orc force is 150tc in the torishop but since it is currently out of stock, it is being sold in the market in a minimum of 395tc. So this market tab system takes in consideration all the orc forces in the market, the lack of orc forces in the torishop, how many people are actually buying the orc forces and creates an approximate price range from minimum to maximum amount that this item "should" be sold/bought at.
I realize the Items forum is the prime place for marketing players to do their business, but as far as I can tell there is also heavy usage of the market itself in order to get better deals that they might not make if they had posted a thread and got their price haggled down. It also may be beneficial to be able to add rarity of the item in circulation to the mix of this approximated price formula. So that even though there isn't a huge demand in void forces because most people don't have the qi, it is still worth a large amount do to there not being as many void forces owned.
So, a summary if this is tl;dr - I'd like to see a system integrated into the current market system where a player can tell the approximate 'street' prices of an item based on: how many people own it (rarity), how many people buy it straight out of the market, how many people sell it, what price is it being bought at more often, etc.
The one downfall I see of this currently is that it is a lot of effort to make this efficient, probably a lot of coding, or maybe even so much effort that it's not even possible with the amount of work force that toribash has with the ability to do this. Anyway, just a thought, feel free to shoot holes in my idea as you see fit, or point out any flaws. Also point out any grammatical mistakes D8
I realize the Items forum is the prime place for marketing players to do their business, but as far as I can tell there is also heavy usage of the market itself in order to get better deals that they might not make if they had posted a thread and got their price haggled down. It also may be beneficial to be able to add rarity of the item in circulation to the mix of this approximated price formula. So that even though there isn't a huge demand in void forces because most people don't have the qi, it is still worth a large amount do to there not being as many void forces owned.
So, a summary if this is tl;dr - I'd like to see a system integrated into the current market system where a player can tell the approximate 'street' prices of an item based on: how many people own it (rarity), how many people buy it straight out of the market, how many people sell it, what price is it being bought at more often, etc.
The one downfall I see of this currently is that it is a lot of effort to make this efficient, probably a lot of coding, or maybe even so much effort that it's not even possible with the amount of work force that toribash has with the ability to do this. Anyway, just a thought, feel free to shoot holes in my idea as you see fit, or point out any flaws. Also point out any grammatical mistakes D8
Originally Posted by Thrandir
I was going to write such a application (ruby based one) but I failed as I couldn't log at toribash.com properly (I am programming newb, mechanize didn't work as I cannot into session saving using cookies).
API provided by Fish is absolutely enough to make app gathering data and showing them using graphs, which would let to predict proper price for any item put into market. It just needs anybody with proper python / ruby / anything skills to write it. My friend said it would need less than 100lines in ruby (downloading data, saving them properly, generating graphs). I wanted to do it, but failed at logging in which is needed to download torimarket data in json format.
by the way, I am pretty sure the only factor such an app should use to predict proper price is lowest market price. And to get average price, it would just take differences in minimum price on given period of time.
if talking about buying- it is up to devs to create such a thing, as regular user cannot determine whether item was taken from market or bought. However, I am sure that average minimum price would be accurate enough- and it could be done by any user. main con is the fact that it won't take "popularity" into account, which may make rare/ not really popular items price predictions struggle over time ( their price vary harder).
oh, and- yeah, over time prediction IS important. It would advise some users to wait with their buy as for example item is way higher than it used to be. however, it shouldn't be long-time analysis- it could show price change over a week.
Only going to point out so parts I don't quite agree on.
Bolded part - the current lowest market price isn't the only time it has been bought for that price
(Last paragraph) - giving the users something that predicts how an item might do over time would kill the item forum completely, which isn't something I want to do. Even more threads will go un-viewed, specifically on items that the market tells them will be a lower price later. Making it near impossible to sell items that would currently be a high price. As an example: the market tells the users that a Supernova Force will be worth 8k 2 months from now, but is currently averagely 13k. Anyone wanting to sell their Supernova Force would get no buyers simply because everyone knows that the price will be 8k later on and wants to get the better deal. The point isn't to have the exact price of an item and destroy any element of marketing. The idea is to make marketing a little more 'accessible' to everyone. Rather than the big market-head users controlling item prices, this will tell everyone the range of prices on an item based on what people have bought, so they themselves can decide whether they are getting a good deal or not.
Anyway, I'm back with some results from the data Fish gave me. This is based on ALL purchases of Supernova Force from the beginning of the market (May 1st, 2011). I have been given the prices and amount of times this item has been bought for that price. There are currently 249 Supernova Forces in Toribash. The data tells me that the highest probable price a Supernova Force can be sold 'on the street' is 18k, the lowest price is 10.2k, and the average price that this item has been sold at is 12.9k. All of these amounts are rounded up to make them easier to read and to keep the aspect of marketing alive, since basically everyone will know not to buy much lower than 10.2k or much higher than 18k, but can still make considerable profit. This has also factored in how many people have bought their Supernova Force for the market price.
Originally Posted by Thrandir
Of course, but it should be just an estimation. Wouldn't it be enough to get average price You should be looking for? It's not perfect for sure, but in my opinion- accurate enough.
Oh, You are wrong! It wouldn't kill items in any way. Market has one major con- it takes time to sell items there. That is why prices are higher at market. Fast toricredits = items, higher price = market. Market is basically fast buy while items are fast sell. And also trends often change, so it wouldn't really let anyone predict how high item will be tomorrow/ 2 months later. Market constantly changes, predicting its flow is impossible. Some items may go high for short time, some go high once and for long period (vide: platinum force at constant 30k price at market till march -> since then: stable 38k+ price). But such an graph would let people know "hey, if You are not in hurry- wait. This item used to be way lower this week, maybe it will fall again?".
If talking about "threads staying unviewed"- come on, most of people buying items are marketers who constantly check prices at market and items section. They do those predictions anyway, even without graphs. I don't think it would affect items section hard, but it could let many newcomers sell their items for proper price (as marketer couldn't simply tell them- trust me, this item was lower so please sell it for lower price).
Getting average prices in this case is absolutely inaccurate as those are 2+ years. Analyzing that long period of time- especially considering stock price fixes which happened- makes no sense in my opinion. It would be misleading, as people would think "my god this item is 2k under average right now, I will buy it from market and make profit!" while for last 2-3 months price was constantly under average.
However data You mentioned would be extremely useful if used properly (shorter periods of time, taking into accounts only "regular" prices- without weird "2k for platinum force / 100k for orc force" ones, talking into account only under-stock prices if talking about items which are in stock etc). Most of those things can be done using ToriShop API from Fish by the way (however, it is impossible to gather data about buying with it, which is crucial).
I think it would be cool to give people the full range so there is still some form of the users deciding on what they will sell and buy for. Average price is irrelevant. Yes, the users still get the same amount of deciding but it's just the same as looking at the average price on the market currently and selling your item for much lower than that.
I just checked the past few months as you suggested in order to be more accurate and the average looks just about the same. As a matter of fact, the probable high amount and probable low amount are about the same too. though the numbers are simple closer to the mean. Even from 2 years ago the data is still accurate, I could try it with an item that has vastly changed in price from a long time ago. Or better yet, only use data from the start of all the torishop prices and market prices being changed.
Originally Posted by mwah
Okay, this seems like the same thing I guess? If it is the same thing how is this easier than what I've done already or more accurate? Would you intend to use a different graph for each item?
yes indeed
Originally Posted by mwah
I think that would be a little confusing to figure out for the majority of users.
if majority of users can't be arsed to learn to read japan candles graphs, then it's their own fault.
if you spent 5 minutes to get grip of concept, you may get some use out of it, afterall this is the type of graph used pretty much everywhere.
Originally Posted by mwah
Plus, I'm not sure if that method can factor in rarity of an item that I am basing off the amount of qi that the general public has. Wouldn't this show that price of a Void Force is constantly going down because it's not really being bought often, if it is based on trends?
nope. it shows the trend of how does the X items sells.
let's say we will have a timeframe of 1 hour
in this case you will be able to see for how much X item was sold for in previous 12 hours and see the pricing trends, for example during last 6 hours orc was sold for 10 - 15 tc, looking at the graph you can tell the trend if price for orc will rise or fall in next 6 hours. time frame may be adjusted to fit tb pace, but the general idea is, you can tell how much this or that item may cost aprosimately in the future.
rarity is redundant. if the item is rare - the price will go up drived purely by supply / demand model. this is really basic stuff. no need to overcomplicate anything.
making graphs that have some degree of uncertainity will improve market since it will have some trading element, instead of having an average value which will be used as baseline to haggle the item.
Originally Posted by mwah
I'm open to new methods mathematically that may be easier than the way I've used in the example but I don't know about showing everyone a chart for each item, or even the result of the chart since it cannot judge things by it's rarity. Plus, this chart seems more useful if you are wanting to look over long periods of time which isn't the case. The way I'm proposing would update a price range all the time simply based on the math of for how much people are buying and selling and how often they are on certain prices.
i can't be arsed to explain really basic things, but seriously, this type of graph is used for all exchanges and show dynamics in price daily. in short - these many bars are usually actually representing time frame of 5 / 15 minutes, so it's used for very fast paced trades, like foreign exchange.
meh, get to uni, this is very basic type of crap, that even i remember.
Originally Posted by snake
yes indeed
if majority of users can't be arsed to learn to read japan candles graphs, then it's their own fault.
if you spent 5 minutes to get grip of concept, you may get some use out of it, afterall this is the type of graph used pretty much everywhere.
nope. it shows the trend of how does the X items sells.
let's say we will have a timeframe of 1 hour
in this case you will be able to see for how much X item was sold for in previous 12 hours and see the pricing trends, for example during last 6 hours orc was sold for 10 - 15 tc, looking at the graph you can tell the trend if price for orc will rise or fall in next 6 hours. time frame may be adjusted to fit tb pace, but the general idea is, you can tell how much this or that item may cost aprosimately in the future.
rarity is redundant. if the item is rare - the price will go up drived purely by supply / demand model. this is really basic stuff. no need to overcomplicate anything.
making graphs that have some degree of uncertainity will improve market since it will have some trading element, instead of having an average value which will be used as baseline to haggle the item.
i can't be arsed to explain really basic things, but seriously, this type of graph is used for all exchanges and show dynamics in price daily. in short - these many bars are usually actually representing time frame of 5 / 15 minutes, so it's used for very fast paced trades, like foreign exchange.
meh, get to uni, this is very basic type of crap, that even i remember.
You didn't answer any questions as to why having a chart/graph that most users may not understand is better than simply telling them a price range based on data.
You do realize some items are rare purely based on their qi right? Does that really fit with supply/demand if the majority of users MAY have the demand for it but simply cannot obtain it because they have a lower qi? How does that factor in with this economic model? I understand what it is, because it happens even in toribash obviously, but it doesn't apply to the fact that an item is just unobtainable based on it's qi. So there must be a way to input this so that the item does not become worthless because of it's high qi.
The same way this graph changes quickly, so does my method, just as quickly if not more so. I'm not saying my method is better than one used across the world, but it works and doesn't need an updated graph for every item because not everyone is going to want to view an update graph for EVERY ITEM.
You really only had to explain why this is a better method, currently the two seem similar. Both work, it's just one gives a result in plain text, and the other in a graph.
You don't always have to throw in a casual insult. If you couldn't be arsed to explain why the graph is better, you shouldn't have posted at all.