Guild Wars 2

I did an analysis of the expected value of the contents of some of the Heavy bags, and it turns out they vary widely from bag to bag (for non-mathematicians, I tried to see if you are better off buying the bag and opening it rather than just buying whatever you hope to find in there).

Some of the bags are being sold for amounts that closely match their expected value, but some do not, and the ones that don’t are waaaaaaaaaaay off. For example, the Heavy Ritual Bag has an expected value of only ~5c but sells for ~6s. Any guesses as to why this is? I assume it’s linked to the actual supply of these bags, but since the real value of bags is in what you get when they open, availability is a poor metric.

what are you using for ‘expected value’? (where did you get your drop rates?)
with bags, I’ve found the wiki to be a good indicator of the types of items, but not so good when you ask how many should you expect if you open 10,000 bags. sometimes they might say “these drop x% of the time”, but forget the “when they drop, they drop 3-5 of them”. or they’ll be incomplete (only mention one or two of the cores but be able to drop all of them, for example)
heavy ritual bag has expected value of 5c? I think you mean 5s? even if you do, I agree with you that this bag is overpriced. that is just based off the wiki’s drop rates, and I haven’t done my own research on this particular bag yet.
however, a lot of Tier 5-6 bags actually do give a profit (even when accounting for TP tax). I won’t mention specifics, because then those bags would stop being so profitable keep looking.

I’ve got a spreadsheet with detailed analysis of drop rates from bags (I log every single bag I open) and while a bag can shift back and forth between buy and open for best profit, none of them are “way off”. If you see a bag that appears to be selling for more than you think it is worth, chances are it drops something at a very low rate that is very valuable.

In addition to what Mystic said (5s instead 5c) average value of a lootbag usually more or less reflects the average value of the average loot table.
LEts assume Heavy Ritual Bag costs 5s on buy order and 6s sell listing (which means a price spread of nearly 15%).

If the prices for the mats on the loot table go up, the bags listed between 6 and 7s will get bought out and more bids will be placed and we will land at a price of 6s buy order 7s sell listing, which is the new equilibrium.
Now the average prices of the mats on the loot table go down. Lootbags usually have a latency of adjusting to those lower prices because most bags dont get specifically farmed and most players always do the same thing with lootbags, they either open them or sell them. The average player doesnt know, if its more profitable to open them atm or to sell them, so he is not concerned about markets trends, which only makes those players that have a good idea about droprates of lootbags and their loot tables, the market makers.

As mystic already mentioned, lots of bags are profitable to get on buy order, open and then sell but it takes alot of research to know which ones and at which prices becasue markets are shifting all the time.

During the invasions last year, i could get thousands of light bags of booty within an hour on buy order and would make an average profit of more than 1 silver per bag.

Mystic.5934:

what are you using for ‘expected value’? (where did you get your drop rates?)
with bags, I’ve found the wiki to be a good indicator of the types of items, but not so good when you ask how many should you expect if you open 10,000 bags. sometimes they might say “these drop x% of the time”, but forget the “when they drop, they drop 3-5 of them”. or they’ll be incomplete (only mention one or two of the cores but be able to drop all of them, for example)

I’ve been using the wiki, as I’ve been foolish enough not to log down what happens when I open bags myself (so thanks for the tip, Wanze). It’s a problem for sure on items that could spawn multiple instances. So for example, the silk scraps from the Heavy Ritual Bags are listed at 30%, giving no option but to interpret that as 30% for 1 scrap. Since silk scraps are the most common item to find by a wide margin, it will matter a lot if it is really 10% for one scrap 10% for 2 scraps and 10% for 3 scraps. The discussion page also talks about getting butter and chocolate out of these bags, which are not listed as some of the drops, so this is a source of error. I also notice that people on the wiki will post data even when they haven’t opened at least 100 bags- and to catch items appearing less than 1% of the time you need to open at least 1000. It would be great to start an effort to make this data more reliable, but I can imagine people would be happier if they had the real numbers kept to themselves.

Mystic.5934:

However, a lot of Tier 5-6 bags actually do give a profit (even when accounting for TP tax). I won’t mention specifics, because then those bags would stop being so profitable keep looking.

No worries there… I’ve found plenty of places to make a profit, and it’s no surprise that the supply of these bags is usually low (at the price you need for profit). Darn those people with their spreadsheets!

mtpelion.4562:

I’ve got a spreadsheet with detailed analysis of drop rates from bags (I log every single bag I open) and while a bag can shift back and forth between buy and open for best profit, none of them are “way off”. If you see a bag that appears to be selling for more than you think it is worth, chances are it drops something at a very low rate that is very valuable.

Sure, but if you’re buying bags hoping to get that valuable item and the price you paid is too far off the expected value of the bag, then you’re gambling.

Wanze.8410:

If the prices for the mats on the loot table go up, the bags listed between 6 and 7s will get bought out and more bids will be placed and we will land at a price of 6s buy order 7s sell listing, which is the new equilibrium.
Now the average prices of the mats on the loot table go down. Loot bags usually have a latency of adjusting to those lower prices because most bags don’t get specifically farmed and most players always do the same thing with loot bags, they either open them or sell them. The average player doesn’t know, if its more profitable to open them atm or to sell them, so he is not concerned about markets trends, which only makes those players that have a good idea about drop rates of loot bags and their loot tables, the market makers.

I suspect that the prices of a lot of bags are fluctuating based on general changes in materials prices, even though these general trends should not be affecting each bag type equally based on what you are likely to get from them, since as you say, the average player doesn’t care. Example: the price of material X goes up 10%, so a player raises the price of their bag 10% even though that particular material has a negligible effect on the expected value of the bag.

Eggman.1405:

mtpelion.4562:

I’ve got a spreadsheet with detailed analysis of drop rates from bags (I log every single bag I open) and while a bag can shift back and forth between buy and open for best profit, none of them are “way off”. If you see a bag that appears to be selling for more than you think it is worth, chances are it drops something at a very low rate that is very valuable.

Sure, but if you’re buying bags hoping to get that valuable item and the price you paid is too far off the expected value of the bag, then you’re gambling.

If your data set is large enough, you can tell whether buying the bags will pay off given enough volume of bags.

Gambling? Yes, but with good data sets it can be a very educated gamble.

mtpelion.4562:

Eggman.1405:

mtpelion.4562:

I’ve got a spreadsheet with detailed analysis of drop rates from bags (I log every single bag I open) and while a bag can shift back and forth between buy and open for best profit, none of them are “way off”. If you see a bag that appears to be selling for more than you think it is worth, chances are it drops something at a very low rate that is very valuable.

Sure, but if you’re buying bags hoping to get that valuable item and the price you paid is too far off the expected value of the bag, then you’re gambling.

If your data set is large enough, you can tell whether buying the bags will pay off given enough volume of bags.

Gambling? Yes, but with good data sets it can be a very educated gamble.

I probably wasn’t that clear, but basically I think the market prices are being largely driven by gamblers. The people with the spreadsheets aren’t gambling.

http://www.nytimes.com/1992/02/25/us/group-invests-5-million-to-hedge-bets-in-lottery.html
TLDR: a group out of Australia bought 5 million of the 7 million possible combinations in a lottery (for $5 million) and won $27 million.
most definitely: large sample size on gambling something turns it into work instead of a gamble.
As Wanze said: [almost] all the bags are fairly close in price to what they give. there are a few exceptions, like if silver doubloons double in price, the bags that drop them will take longer to adjust. or if a bag has limited-time drop rates such as watchwork boxes.
I think enough people know what to expect out of a bag to be able to buy/sell it at a reasonable price. 5s vs. 6s I would consider within reason, btw. without reason would be more like 3s vs. 6s