JC + DE + TSM = Profit

[seafunk]

HB + JC*DE + TSM = 5mio/month turnover

Started logging after a break that lasted about half year and ended with me quitting the gold selling thing. I still love to play around with ideas surroounding botting and the game economy though.

After realizing that i lost gm status in 5 out of 7 lvl25 guilds I had, the month turned out to be unexpectedly rewarding. Might just be TSM bugging out on me, I can't tell, it "feels" like I made gold at a considerably higher rate than ever before. I have caught myself indulging in sudden outbursts of lavish behaviour, usually spend about 100k on gear alone for each toon hitting 85 before restricting them to a single stable profile.

The basic principle I am following is this:
A toon either is a pure gather toon or receives mats from other toons, transforms them to whatever is profitable, sends the products off to the ah toon (or other toons for further processing), logs to the next toon, restarts hb, rinse and repeat.

Most AH players have told me that diversity is key, but despite the diminishing returns problem of investing too much time into a single small capacity market niche focussing on a single market in order to master it is well worth the invested time. Implementing bug free dynamic micromanagement in botting profiles usually takes me many hours, but once it's up and running it'll quadruple your MOARGOLDZ!

I'll attach a screenshot of my TSM statistics. I could've easily edited the TSm files easily to show these numbers, so please don't point to the "obvious", it won't lead anywhere and I'm doing this to share information that has been valuable to me in the past.

For those too lazy to clickThe biggest impact seems to be caused by the frequency of AH scans and exactly how the ore buying thresholds were set. This is imho the riskiest part since you are almost forcing your non-botting competition to report. This is why the AH toon is on a vanilla account and constantly running either levelling or gathering profiles between AH runs or sometimes just idling in a low profile spot close to the AH,

It would be so fucking cool to train a neural network with AH/toon/playtime data etc., analyse for maximum profit, repeat, finetune for an eternity! Just dreaming... Eventhough there is a good neural network with excellent documentation.

 

[Quarian]

Any chance of sharing your settings since they seem mad good ;-)

 

[seafunk]

@Quarian
I love your sig!

The point I was trying to make, albeit a bit clumsily: Every server is different. Try focussing on a single production chain that's not hog-botted, see how long the concept pays out for you until you see
-your products lose in value because you're flooding the market and undercuts add up, the price never has time to recover because of frequent reposting. Best stop here unless you are trying to influence your competition.
-the mats become too expensive because you are buying all of them.

While the former cries for the market to have a couple of days rest, the latter is wonderful if you have a gathering toon for the needed mats because it can be exploited:

After a few days people will be used to the higher mats price they're selling for, say Ele and Obs go from 2g to 4g which is realistic and can be stable ona high pop server due to 1 persons messing around.
Here this one person comes into play with a 2nd toon who farms said ore himself, passes it to the Gather toon who only sells little of the product since the AH/production chain should be in recovery.
This scenario leads to the central psychological loop hole we exploit: The ore farmers will keep posting at the high price after having gotten used to it, wetting their pants over the goldz or euroz they're gonna get. They will sell a lower volume while farming habits don't adapt as quickly (or at all), which results in ore stacking up in their storage. You were pretty much the only one buying at their ridiculous prices. During this time these things have happened:
-Demand has previously been increasing when people were unwilling to buy ore at at your upper cutoff price, a circumstance masked by the 1 person spendings 100s of thousands on ore. ---> The ore price will for these reasons fall slower
-Your product has reached a peak price. With Greater Celestials this can be between 20g and 150g or so, in this case on the high end, 80g is common during this time, over 100g can still be seen once or twice a month. It really depends on who's using the same production chains and how cleverly they behave.

Now, this is your chance, people have been reluctant to buy GCE at these constantly rising prices, demand is high. youve crafted a shitload by farming and buying all the ore off the ah, financing your ore habit with uncut inferno rubies alone which you've caused to rise from 60 to 220g (hehe, cant wait for alch transmute ).
While people buying products often depend on them quite urgently the GCE will remain pretty high along with the "overpriced" ore sold by greedy farmers who probably just adjusted their price thresholds.
People underestimate the ore value. I will try to integrate personalized spreadsheets like TSM does if I can ever find the motivation and empowering mental clarity I need for programming. Lack of familiarity with a market, fear of it's potential instability etc. cause reluctance in most to just invest all their gold into a production chain (or ideally into more than one).

So with the ore price slowly recovering you can sell 100s of GCE's at a price much higher than it would be had you constantly posted them, but still low enough for people to buy them by the dozen without hesitation. I found it makes a huge difference during this time, whether I set TSM to 30 auctions and do my AH cycle more frequently or set more auctions and less do frequent visits. The former will give people the illusion of a rare and therefore greaz deal. Then it's basically rinse repeat and focussing on other chains.

This will only get really interesting with neural networks and random AH experiments, training and execution looping after each other with focus on a single chain or so. It would be need to just search AH data for clusters of correlating elements with consideration of scantime. Sigh if only there was more time and a better teacher than just myself for this...