Help with a class project

Hey all, I'm taking an introduction to statistics class and for a project I've decided to do an observational study of our market. I'm going to track a high demand, medium demand and low demand piece of salvage for about three weeks and then create charts tracking the price fluctuations or lack of them.

My question to you all is there any advice of what would be good pieces to monitor?

I decided on salvage so I won't have to worry about all the level ranges but I'm not 100% sure on what pieces qualify as high, medium and low demand. Any suggestions on what to track would be greatly appreciated.

Thank You,

Revanchist

Part of the problem I think you're going to run into is that it's very hard to infer demand based on the information available to us. For example nevermelting ice is in quite high demand (since it's used in endurance reduction and run speed common IOs) but the supply at the moment is so high that it's largely down to vendor trash pricing.

If you want to do it with common salvage my recommendation is to use the Tier 2 salvage since that tends to be where you get significant price variations. Tier 1s tend to be moderately oversupplied and Tier 3s tend to be heavily oversupplied unless there's an AE exploit doing the rounds.

For Tier 2 commons I would say go with:
High demand: Alchemical Silvers
Medium Demand: Masterwork Weapon or maybe Iron (the price of iron seems to fluctuate a lot).
Low Demand: Commercial Cybernetic or Circuit Board

The other option would be to do rares in which case I'd look at:
High demand: Magical Conspiracy
Medium Demand: Mu Vestment
Low Demand: Any of the Tier 1 rare salvages or possibly Synthetic Intelligence Unit if you want something with low-ish demand but some value.

OP, you are going to find that prices are nowhere near normally distributed!

I'd consider tracking recipes or IOs rather than salvage. The prices are a little more stable, and easier to track over time. When I was niche-watching for Archie, I found I could catch most of the sales of a particular recipe or IO by recording prices every hour.

The problem I could see with tracking recipes is that something that's in low demand might not give enough data points for his project. With salvage, everything has enough volume to make the task of obtaining good data samples a breeze.

Revanchist, have you considered monitoring one common, one uncommon, and one rare salvage? That might give some interesting results.

Try common IOs, level 50. Finding high and low volume items will be easy, middle, maybe less so.

It has been a while since I looked, but when I was watching purple prices I found that sales were slow enough that I could see all of them if I checked the market a couple of times a day. Something like Hecatomb is probably a high demand purple, the confuse one is probably low. I'm not sure what would be medium. And you'd want to pick a specific enhancement, and decide if you want the recipe or the crafted version.

Thanks for the replies so far.

Yeah, I'm on a pretty strict timeline, I'm in an accelerated college program, each class is held twice a week for 5 hours at a time with the term lasting for 5 weeks before I start my next set of classes. At least I'll have my bachelors in only 2.5 years though.

I fully understand the invention system, with my only two lvl 50 heroes being fully tricked out, I'm just not an avid market player and with all my alts I normally have all the salvage I need that's why I'm unsure what salvage is considered "hot" (I keep my salvage categorized by tier and type on certain character across my two accounts on my main server).

Thanks for the heads up about the double xp weekend though, didn't realize it was that time of year again, I guess I'll monitor both salvage and a couple of purples and decide what looks the best when I'm done with data collection.

The teacher doesn't seem too picky, since this is only an into to stats course and she already signed off on the project. I think the main purpose teach us how to collect and compile data, even if the results turn out to be wonky.

I'll be collecting 4 times a day on weekdays and 5 times on weekends so I hope to get some decent readings over the next 3 weeks. Maybe I'll post the results when I'm finished if anyone is interested.

Even Mathematic Proofs have a lot of throughput. There's probably at least a thousand a day bought at Wents. Now there might be TWO thousand a day that people are trying to sell. Which is why there are like 16,000 sitting on the shelf and a large number beyond that which are listed for 1 infl or deleted on sight.

The current lowest listed price is higher than the current highest bid price, or the system would sell things to people.

Interesting Project.

For my part, I think I'd side with the idea of monitoring Common IOs at a
particular level for several reasons.

1> As others have noted, salvage is tiered, and some salvage is used in
many recipes (both common and set) and some is used in few (if any).
So, you have two extra "artificial" variables thrown into the mix - namely
extra demand for multi-use salvage, and variable demographics of
production (due to the salvage tiers and the levels at which salvage drops
vs how many players run in that level range, as well as other means of
production - ie. AE tix, for instance).

2> For Common IOs, I'd think that "demand" is much more in line with AT
distribution, so for instance, most of the power 10 Commons (Acc, Dmg,
etc) will be more in demand across the board since everyone uses them,
and Resist, Confuse, Hold, Taunt and others will have less demand due to
fewer ATs in play which would use them.

3> Only one probable supply path. Pretty much the only way crafted
Common IOs pass through the market is through player effort - they
can't be bought from vendors, they don't drop (whole), and I doubt anyone
would obtain Common IOs with merits (if you even can). So, the supply
of them is more "pure" in the sense that someone actually took the time
to gather components and craft the item. That effort is also pretty
constant accross the Common IO population, so there's also less liklihood of
skewing because some are inherently easier to get than others (unlike our
A-D recipe pools, for instance).

4> As for what level to pick, personally, I'd look at L30-L35's because, they're
cost effective, they're better than even SO's, and they're cheaper than sets,
for most non-marketeers. I'd expect more of the game's population to use
these... Before then, they're probably using DO/SO's, and after that, they
are probably wealthy enough to start switching to sets. Also, with mid-
level set recipes scarce, L30-L35 Common IO's are easier to obtain and
therefore more generally convenient initially.

5> Finally, I think the data collection will be somewhat easier to manage
than tracking salvage, although, you may run into the issue of too little
data for your low demand IOs.

I also agree with the others, that 2XP weekend will probably skew things
significantly. That might be a good time to test the data collection methods
rather than treat it as representative. On the other hand, noting the time
of key events in-game during the project might actually provide some
interesting insights, into the data patterns themselves... Who knows?

GL with the Project in any case,
4

Turned the paper in this morning, anyone want me to post the raw data?

Hmmm, where do I post it for viewing? Google docs doesn't like my formulas and charts

Skydrive.

skydrive.live.com

Hold 25g of data, you can setup public folders.