Industry and Trading Matrix
Posted: 2012-06-06 Filed under: industry, market, nullsec | Tags: c-level, excel, stats 9 CommentsNumbers > Words
If you can’t tell, I’m a very quantifiable person. Using numbers to guide decisions is a method of problem solving that I have cultivated as my professional career has grown.
I remember this one particular insiadent that really stood out where the numbers really proved valuable.
A few years ago when a group of difficult C-level people were up for a laptop refresh, I sent them a simple survey with a list of considerations for their new machines. Each person had rate weight, speed, storage capacity, ruggedness, maximum resolution, monitor size, etc on a scale of 1-3 (3 being the highest priority).
I then took these results and found the general need of the group. The data told me what machine to buy. I wan’t trying to tailor a machine to their needs, their needs dictated the selection. No fights were had when the machines arrived because they all knew that I had properly weighed everyone’s opinions.
To the Spreadsheets!
Since I have performed a number of industry and trading activities in Eve, I felt that the same type of analysis would offer some insight to the greater community.
I made a list of negative and positive factors and applied a value to each of them. The total value has a little bit of logic built into it to help produce a more meaningful result.
Total =(-(Sum all the bad things)+(Sum all the good things))
The English equivalent of this would be something along the line of, “Considering all the bad and good things, what is the best thing I should be doing to make money in Eve?”
(Yes the scale is 1-3, but the sheer amplitude of risk and profitability of Patch Speculation made me weight it with a 4. The external factor of CCP changing a loot table or rebalancing something on the fly is an enormous risk when speculating)
Note that these weights are based on my experiences and therefore you would most likely apply different values to each activity. Just like with the difficult C-level people, getting a larger survey group would help normalize the data and CYA.
k162space 
Curious, why the risk factor of 2 on mining in high-sec? I’d have though that would be a one, but could see the barrier to entry being a 2 in truth because you can’t exactly consider a frigate as the entry mining ship. Null I’d actually put to a one entirely, as null space is safer than high-sec since you can preemptively kill gankers. As a miner though, I could be a bit biased on those, but I’d had always thought them similar to an industrialists.
I gave it a 2 due to all the wonderful gankings that have been going on.
Ah, the poor non-miners that are mining. Mining while being aligned really isn’t that hard, but heck. I’m sure you probably see the same thing in your realm of industry, where people sell cheaper than materials because they mine for free. Too bad there isn’t a gank campaign for them.
I could take argument with a few rankings, but pretty solid overall. Its a little discouraging that only two categories got positive values though (T3 manufacturing, null trading). Nice post.
Don’t think of the negative as inherently ‘bad’. The negative numbers is just how it worked out. I could have added +5 to all of them so they are nice pretty positive numbers, but I left them alone. Just think of -2 as ‘better’ than -3.
This is what is termed a decision matrix in the classes I took oh so long ago. However, I have to tell you that my instructor would have failed me for this matrix. Let me explain before you jump up and down on me. The first thing I would have been graded down on is the equation -(sum of bad)+(sum of good.) This has opened you up to the biggest flaw in your matrix. It allows the negative to always outweigh the positive. It gives a false representation that everything is more negative than it actually it. To wit, you have 3 columns for negatives and two for positives. Because the equation treats these as equal sub-sets, the negative gets 1 extra “vote” in the matrix. In fact, three 1s in each negative completely negates a single max in a positive column. Next, turn negatives into positives. What I mean is, rather than downgrade a risk or introduce negatives into your matrix, upgrade the advantages. If one column or row is more important than another in your estimation, weight it. Assign it a 1.5x or 2x multiplier. Do this BEFORE assigning any values. If velocity of money is more important than profitability assign it double the weight. If manufacturing is better than any other activity in your view, give all manufacturing rows a 2x weight. Always deal with multipliers, not deductions. Next comes the rating business. Don’t rate them 1 to 3 individually. rank order them. Take each column and assign it a value from 1 to how ever many rows you have. Do not consider any relationship to any other column. Keep them separate. Treat them independently. Do this for every column. Once every row in every column is rank ordered, apply the modifiers for each column or row. Then, do a simple sum of each row. The higher the resultant number, the better the method is for making ISK. That is at least how my instructor would have wanted it. YMMV.
Thanks for the detailed comment. I have never taken a formal stat class as my degree is in Telecommunications. If I have time I’ll re-work the matrix.
Mining in a wh ought to be 3 on the risk level, or the barrier to entry much higher, if your going to have scouts and what not whilst you mine.
would also possibly require a pos for best time spent.
I couldn’t comment on building t3 but t2 is some what of a decline, However its module specific,
Its still possible with some research to find a spot to make decent profit, but at a lower turnover of items.
t1 items typically are a terrible idea, however with the change in loot tables maybe some new people can make small but meaningful profits on some of the based modules required for t2 production.
I don’t see how you came up with a -2 for Competition for Ratting in Wormhole space. When you find a good system, you generally have it to yourself (until that -3 for Risk shows up). Still not enough to put it into positive territory, though, as the risk and barriers to entry are really high.
Also, Mabrick is really right about your bad matrix setup. In both design and hardcore engineering, we follow the method he just described.