Page 1 of 6 12345 ... LastLast
Results 1 to 15 of 86

Thread: [BGS] Trading for Influence

  1. #1

    [BGS] Trading for Influence

    This thread details, how trade affects a market controlling minor factions influence. I’m very interested, whether my test results meet others experiences in the game.

    If you are looking for general information on how the BGS works, this thread should be your first stop: https://forums.frontier.co.uk/showth...Background-Sim

    If you are not interested about the mechanics of minor factions influence, this thread isn’t for you. Generally selling commodities with profit gains reputation with the market controlling minor faction for your commander (more profit = more reputation), and brings a market controlling faction into boom state.

    TL,DR:

    1. Influence for a minor faction is gained by selling commodities on a market with profit (Buying has no observable effect).

    2. Influence gain by profit is capped around 700 cr/t profit.

    3. Selling smaller batches of different profitable commodities per load gains more influence in total, than selling 1 highly profitable commodity - even if the total profit is (much) lower.

    4. There is a cap on how much influence can be gained by trade.

    5. Demand has no effect on influence gains.

    6. Boom doubles the influence effect of trade, but does not raise the maximum of influence, which can be gained by trade.

    7. Boom has no effect on the influence loss by smuggling.

    General Test Design

    I took my Trading Python “Almatheia” over several weeks to the same Low-Traffic Low-Population System, carefully selling selected commodities, collecting influence changes, data on profit, demand etc., to find out about the basic mechanics of trading for influence of a minor faction.

    The system in question had a population 150 K, 1 Outpost with Commodity Market, 1 Surface Port with Commodity Market and Black Market, both >3000 LS from entry.

    Some definitions upfront

    “Raw Influence Gain”

    I got the idea for it while working on this thread: https://forums.frontier.co.uk/showth...ting-after-2-3

    “Raw Influence Gain” is the number gained, before influence is recalculated to 100% in system. Basically this allows to compare the influence gain of an action, no matter whether the faction in question is at 50% or 70% influence, as influence is gained slower, the more influence a faction has.

    Applying coatsilvers excellent theory of influence-bucket sizes https://forums.frontier.co.uk/showth...=1#post3778282 one can calculate the Raw Influence Gain with

    New_Influence = ((Old_Influence + Raw_influence_Gain)/(100+Raw_Influence_Gain)*100)

    , if only 1 faction has influence actions.

    As you can see Old and New Influence in system map, and I have discarded any data point which showed other players activities in system, retested most numbers twice, and all actions in system by me only targeted 1 faction (The system and markets controlling one), i could easily calculate the Raw Influence Gain with that.

    “Cap”

    Whenever “Cap” is mentioned, have in mind that it is most probably a “Soft Cap”, not a “Hard Cap” - the cap is the point where i simply could not see any more changes.

    This is effected by both the test design (me being alone, one trade run each day), and the rounding of influence values in system map (and probably even on the servers). It might be, that with a dozen or hundreds of players, or with thousands of tons traded you’d get above that cap. The basic mechanics of gaining influence should be the same nevertheless.

    Caveats

    Rounding and margin of error

    The system map does rounding influence values. Maybe even the server does rounding. An influence of 60,05% as well as an influence of 60,14 will show up as 60,1%. While this has no practical effect, it can lead to a different calculated “Raw influence Gain” of up to 0,2% in my testing numbers.

    The problem gets more severe, if the action has a low Raw Influence Gain (example: selling 10T only), and the higher the factions influence is generally (at 70% influence a 1% raw influence gain only leads to 0,3% Influence gain… while at 50% the same raw influence gain leads to 0,5%).

    So, you shouldn’t take my numbers to calculate/min/max trade effects, especially not where the gain is small. For example the cap on gains by profit might be at 660, 700, or 750 profit/t… or anything in between, slightly higher or lower.

    Also this didn’t allow me to test for example trading a 4 tons trade, - as i need at least a +0,1% influence change as a feedback.

    Small Population System

    While working such a small system allowed me to test the basic effects by myself and without interference, there are caveats coming with it: Caps are probably much lower there.

    So, while for example i can see a rapid growth in raw influence gain up to a total tonnage per run of around 180T, and saturation by tonnage per transaction hitting hard beyond 200-250 T, this might be very different in a large population system - for example a rapid growth up to 20 000 tons, therefore outlasting any ship ingame, while my trading python was already hitting the caps in my testing system.


    Profit

    The influence gain increases rapidly with profit, till it caps out around 700 cr/t.



    Demand has no effect.



    Commentary:
    We can speculate how the curve behaves at very low Profit.

    I’d argue, that it is a sigmoid https://en.wikipedia.org/wiki/Sigmoid_function with point zero at small loss (so at a very small loss, no influence is gained) for three reasons: a) the BGS has a lot of sigmoids b) people buying the wrong commodity and selling to the same market not nuking the factions influence and b) for a bug recently patched with BM trade, where small profits in illegal goods lead to influence gains.

    But, as 73 cr/t profit was the least profitable trade i could find for my testing system, and you’ll always find something with more profit, this is largely an academic question.

    I personally think that, since state effects on prices have been upped (and security level plays a role, too), profit effect caps out a bit too early - i have had a hard time finding commodities that don’t cap out the profit effect. On the other hand i suspect this cap being lowered to mitigate the effect of zero-purchase-commodities (mission reward). This makes rare trading for influence superfluous, and reduces the effect of p&p&p mining. I do hope FDEV tweaks the system to bring rare trading and mining back to where it was.


    Tonnage

    The influence gain by tonnage follows a sigmoid curve:



    Commentary:

    This design is neat, as it mitigates the effect of random trade on the one hand, allows small ships to have any effect while random trade goes on, ups the effect of dedicated trading activities, while it does not unbalance the trade game fully towards huge ships. So: either bring a dedicated trader with huge cargo capacity, or trade smaller batches in a single class 5 cargo rack on your mission runner!

    Different Commodities (“Transactions”)

    Selling 250 T of the same commodity is less effective, than selling 250 T of different commodities.



    This also applies to Smuggling (negative influence, raw influence loss).



    Commentary:

    Once upon a time, there was an exploit called the “1-t-trading-bug”. Basically, by selling 1T of a commodity repeatedly, you could gain massive influence - basically, the BGS being weighted on transactions, not CR value, allowed to artificially up the trade transactions. This was patched out twice. We don’t know how the fix looked like, but have suspected, that the BGS now adds up all trade transactions per CMRD per instance/per docking. Anyway: the fix works.

    Selling different commodities still count as a different transaction, though. Which means, selling different commodities per load is more effective, than selling only one type of commodity.

    How much more effective? Generally i could gain 2,6 raw influence by selling the most profitable single commodity (“Single commodity cap”). I could gain 3,9 raw influence by selling 4 different commodities of 1/4th the profit. As that was the cap by trade anyways, it is at least 50% more effective to sell different commodities. Caveat is the small population of my system.

    I generally like the concept, as it gives a role to many lower profit commodities, and it also mitigates the effect of the usual trader maximising profits (therefore only trading 1 commodity repeatedly). It explains why strip mining is such a heavy hitter for influence gain (all commodities profitable, most of them beyond the profit cap, and a lot of different commodities).

    I got a new idea how to test the effect without hitting the maximum influence cap, and will post an update here.


    Boom Trading

    During Boom, raw influence gain by trade is doubled. While it is even doubled for single commodity trade, the maximum influence gain by trade does not change.



    Boom does not effect the influence loss by smuggling:



    Commentary:

    If not opposed in a system, and if you already hit the max influence gain cap by trade usually, you can simply trade less during boom for the same effect. that is a bit nonsensical to me.


    Trading vs. Smuggling

    I’m currently planning more tests on this, but generally it looks as if smuggling is as effective to reduce influence, as is trading to gain influence. So selling 250 illegal goods with a profit of ~1500 cr/t reduces as much influence, as selling 250 legal goods with a profit of ~1500 cr/t gain influence.

    I’ll post any update in this thread - one of my next tests will be, to trade and smuggle at the same time.


    General Commentary:

    I was very surprised, how effective trading is, and how fast caps are reached at the same time.

    My conclusion is, i will continue to use trade as a tool, and even use it a bit more - but i will do less trade runs in number, as i probably hit the cap much earlier than i suspected before this test.

    I’m now always bringing a bunch of different commodities each run.

    I have also switched my cutter for a trading python, to have better sources for more different commodity types, instead of huge cargo capacity.

    Basically: less repetition, more effect, more variety.
    ___

    Ceci n'est pas une signature.

  2. #2
    "Deploy the PowerPoint!"


    Edit: that may have come across as flippant or disparaging, which wasn't my intent. Looks like you've put a heck of a lot of work into that. Thank you.

    Orbiting inside the Roche Limit of sanity

  3. #3
    uh -
    BRB
    ...
    okay so generally I only "trade to order" and let the boffins at HQ pass my lazy butt a route like a noob. There are other ways to push influence for a medium ships guy like me.

    BUT - I am surprised by your findings. When I have traded for influence on the grounds that I'm going that way anyways why travel with an empty cargo hold - I trade to demand. Not profit.

    So I'm surprised that you don't find demand to be the key.

    I don't like to argue with PowerPoints but my superstition is that a variety of high demand goods sold in numerous smaller quantities with any amount of non-negative profit was the way to conquest.

    Assistant Cultural Attaché for G 14-6
    Alliance Elite Diplomatic Corps - Data Entry Clerk Second Class
    Alliance Office of Statistics - Kitchen Staff Supervisor

    -[----->+<]>+.---.++++++++.+[-->+<]>+++.++[->+++<]>.+++++++++.+++.[-->+++++<]>+++.---[->++++<]>.------------.---.--[--->+<]>-.[->++<]>+.[--->+<]>+..---.--------.+++++++++++++.-----------.++.--[--->+<]>.

  4. #4
    Sterling work ! Glad you didnt post it on Discord, my eyes would have melted.
    "The Monster" - "The Canonn Material and PPOI Sheet" - "Formidine Rift"
    ... --- --..-- / .. -. / - .... . / ..-. .- -.-. . / --- ..-. / --- ...- . .-. .-- .... . .-.. -- .. -. --. / --- -.. -.. ... --..-- / .. .----. -- / .-.. . ..-. - / .-- .. - .... / --- -. .-.. -.-- / --- -. . / --- .--. - .. --- -. ---... / .. .----. -- / --. --- .. -. --. / - --- / .... .- ...- . / - --- / ... -.-. .. . -. -.-. . / - .... . / ... .... .. - / --- ..- - / --- ..-. / - .... .. ...

  5. #5
    Excellent. Thanks for your work on this.
    Bookmarked.

  6. #6
    Originally Posted by DNA-Decay View Post (Source)
    I don't like to argue with PowerPoints but my superstition is that a variety of high demand goods sold in numerous smaller quantities with any amount of non-negative profit was the way to conquest.
    I dont think you ARE arguing, you seem to basically agree (your high demand will normally give you geomon's profit) but geomon has provided some numbers to nuance what you are already doing, if you want.
    "The Monster" - "The Canonn Material and PPOI Sheet" - "Formidine Rift"
    ... --- --..-- / .. -. / - .... . / ..-. .- -.-. . / --- ..-. / --- ...- . .-. .-- .... . .-.. -- .. -. --. / --- -.. -.. ... --..-- / .. .----. -- / .-.. . ..-. - / .-- .. - .... / --- -. .-.. -.-- / --- -. . / --- .--. - .. --- -. ---... / .. .----. -- / --. --- .. -. --. / - --- / .... .- ...- . / - --- / ... -.-. .. . -. -.-. . / - .... . / ... .... .. - / --- ..- - / --- ..-. / - .... .. ...

  7. #7
    Originally Posted by DNA-Decay View Post (Source)
    superstition is that a variety of high demand goods sold in numerous smaller quantities with any amount of non-negative profit was the way to conquest.
    i would have assumed the same, but i could see no effect of demand at all in my test. i'd argue that your way to conquest was a variety of profitable commodities in numerous smaller batches, with demand having an effect only in so far, as it makes a profit more likely in most cases.

    see above - i tested with a no demand and a high demand commodity of same profit - both trades had the same effect.
    ___

    Ceci n'est pas une signature.

  8. #8
    Extremely useful info once again Goemon. Did you find that there was a cap on the influence amount that could be lost by smuggling in a single tick?

  9. #9
    Originally Posted by DNA-Decay View Post (Source)
    uh -
    BRB
    ...
    okay so generally I only "trade to order" and let the boffins at HQ pass my lazy butt a route like a noob. There are other ways to push influence for a medium ships guy like me.

    BUT - I am surprised by your findings. When I have traded for influence on the grounds that I'm going that way anyways why travel with an empty cargo hold - I trade to demand. Not profit.

    So I'm surprised that you don't find demand to be the key.

    I don't like to argue with PowerPoints but my superstition is that a variety of high demand goods sold in numerous smaller quantities with any amount of non-negative profit was the way to conquest.
    I understand demand as a modifier on how quickly a price rises again in the market on sustained trading in that good; i.e. how many market ticks it takes for the price to recover. Profit sustainability is most heavily influenced by system population. High demand with a tiny population will always lose profitability quicker, but will also recover quicker.

    夏草や
    兵どもが
    夢の跡
    Summer grass
    All that remains
    of young warriors' dreams

    i7-6700K @4.4GHz -- Zotac GTX 1070 Amp! -- 16GB DDR4-3000 -- Asrock z170 Extreme IV -- Iiyama 32" 1080p -- OSVR HDK2 -- X52 Pro -- 3DConnexion Space Explorer

  10. #10
    Interesting...

    So smuggling will always beat trade because smuggling has no cap ?

  11. #11
    Thanks a lot for sharing your insights.
    Especially the cap mechanics in regards to number of different commodities is interesting.
    Now we can greatly speculate whether this is a
    - trading cap
    - commander cap
    - general cap

    That is a great test and great resultpresentation! I am impressed.
    Results are similar to our impressions and longstanding theory that "It's the profit that matters."


    EDIT: The good thing about this results is that you can easily do effective influence management by flying a nice trading route within your systems. I like the consequences of that.

  12. #12
    Originally Posted by call_me_trent View Post (Source)
    Extremely useful info once again Goemon. Did you find that there was a cap on the influence amount that could be lost by smuggling in a single tick?
    i have no indication of such a cap, even if profit cap and tonnage softcap probably apply. i'm running more tests on smuggling currently, but are limited to the illegal goods in testing system... found out that a bunch of often illegal rares wasn't illegal in system, so for exampel a stress test of profit and types of commodities similar to legal goods wasn't possible. anyway, the influemce loss by smuggling 4 illegal commodities exceeded the max influence gain by trade, that's why i think there is no similar cap in place.

    i will post an update as soon as i have more data on smuggling.

    Originally Posted by BL1P View Post (Source)
    Interesting...

    So smuggling will always beat trade because smuggling has no cap ?
    could be, but also couldn't be the case, depending on whether the cap on influence gains of trade is applied before or after the negative influence of smuggling is calculated in.

    i assume, that the first is the case, like:

    trade actions of +10 in system, cap of +4 = faction gains +4 (cap)

    trade action of +10, smuggling action of -5, 10-5 = 5, factions gains +4 (cap)

    trade action of +10, smuggling action of -12, faction looses -2 (+10-12= -2)

    i'm currently testing exactly this, and will post an update.

    Originally Posted by _Flin_ View Post (Source)
    Thanks a lot for sharing your insights.
    Especially the cap mechanics in regards to number of different commodities is interesting.
    Now we can greatly speculate whether this is a
    - trading cap
    - commander cap
    - general cap

    That is a great test and great resultpresentation! I am impressed.
    Results are similar to our impressions and longstanding theory that "It's the profit that matters."


    EDIT: The good thing about this results is that you can easily do effective influence management by flying a nice trading route within your systems. I like the consequences of that.
    as we do know, that you can gain more influence than 3,9% for a faction in a 150k population system by positive influence actions, i conclude there is a cap on trade itself.
    (with trade counting positive, and smuggling negatively). i'd even jump to the conclusions that similar caps are in place for other action-influence buckets. basically that the general cap is applied by applying caps to those buckets. i plan to test, what counts towards that "trade-influence-bucket" by, for exampel, running a trade with max. gain + a trade mission - problem here is my testing system only gives out non-superpower donation missions ...
    ___

    Ceci n'est pas une signature.

  13. #13
    So the big gain is selling 2 goods rather than 1, more than 2 grants only a minimum advantage.

  14. #14
    Originally Posted by Sentenza View Post (Source)
    So the big gain is selling 2 goods rather than 1, more than 2 grants only a minimum advantage.
    ... in this small population system.

    this could look very different in large population systems, for three reason:


    1. the influence gain by tonnage could be much slower. let us assume in a large population system the rapid growth of tonnage sits at 1800 t, not on ~180 t as in my testing system - in that case you would be better off selling 20 different commodities of 20T from your trade conda each run, than bringing in 2 commodities of 200t per run, as you'll anyway sit on the low level of influence gain from tonnage.

    i can't really test that alone, but i have long decided that i do these tests by myself only, as i want to know when i pressed the wrong button (or forget to sell anything like two nights ago, with the surprising outcome that zero trade actions have zero influence effect )

    2. caps of different commodity types/diminuishing returns of it could apply at much larger numbers in large population systems. i hit the max. gain already with 3 commodities - that number could be 30 or 300 in a large populatiom system. our chat on discord gave me an idea how i can test that without hitting the max. influence cap, and i will post an update as soon as i have numbers.

    3. as any action is generally less effective in a large population system, it could well be that you'll make a bigger dent when racking up actions.

    basically my assumption is, that the influence bucket growth exponentially, by calculating with 100, 1000, 10000, 100 000 influence points in the bucket depending on population size.

    if we assume, that there is a similar cap in place in large population systems, you'd need an awfull lot of actions to make any effect - racking up commodity types could be the only way, as you can have only 700t of cargo in one run, but potentially a hundred or more of profitable commodities.

    i'm most interested to hear what works for people working large population systems!
    ___

    Ceci n'est pas une signature.

  15. #15
    Hum, so this means I'll hit the cap in a single run if I fill up my Cutter with different commodities?

    Good stuff as always Goemon, +rep


    -CMDR Sali Vader // Join SINC today! // Discord

Page 1 of 6 12345 ... LastLast