Expeditions : a study from the French community

  • Hi,

    Let me share this with you : https://board.fr.ogame.gamefor…tID=11997610#post11997610 (scroll up)

    I can't really said whether it is applicable to your servers or not, well, I assume it is.

    You'll get a fairly good translation of it with Google, except as for technical terms (see the lexicon below). I hope that you'll be able to grasp the meaning of the documents attached as well, especially the spreadsheet. If not, go and learn French! :biggrin: Or ask me, i'll be happy to help.

    As a start :

    • métal : metal ;
    • cristal : crystal ;
    • deut : deuterium ;
    • pt : small cargo ;
    • gt : large cargo ;
    • clé : light fighter ;
    • clo : heavy fighter ;
    • cro : cruiser ;
    • éclaireur : pathfinder ;
    • vb : battleship ;
    • traq : battlecruiser ;
    • dd : destroyer ;
    • faucheur : reaper ;
    • edlm : death star.

    Here's a recent update of the figures :


    You'll get to know why there are so many lines in the table by reading the full document. In short, it is a very basic attempt to express the fact that some players aren't aware of solar systems depletion.

    Lexicon :

    • rss : get ressources ;
    • vsx : bring back ships ;
    • rien : get nothing ;
    • retard : come back late ;
    • avance : come back early ;
    • AM : darkmatter ;
    • item : get an item ;
    • mrchd : get a merchant ;
    • pirates : fight pirates ;
    • aliens : fight aliens ;
    • TN : encounter a black hole.

    Well, I think that's pretty much it. Have a good read! :beer:

    DeepMiningInc You've written somewhere : "there are a total of four levels of depletion : no depletion, minimal depletion, medium depletion and maximal depletion." Where did you get that from?

    The post was edited 1 time, last by Grandpa ().


    The first part displays general information about expeditions. Based on unreliable statistics, it will have to be revised.

    The second part tackles the question: is it mathematically viable to send military ships in v7 expeditions? It introduces a spreadsheet that I make available to you as an attachment.

    All of the following is only valid for explorers.

    PART I


    On expedition, one can: collect resources, bring back ships, come home empty-handed (usually on time, sometimes early, sometimes late), get an item, get darkmatter, get a merchant, fight pirates, fight aliens, lose one's fleet in a black hole.

    The probabilities of these events have evolved over the versions and are not yet well known for v7. I gathered above 70000 expeditions from various sources, knowing in advance that the results would be biased for numerous reasons. Anyway, here are the results:


    It seemed to me (I'm possibly be wrong) that the "Nothing" percentage was a little high, so I decided to vary it and see what would then happen to the probabilities :


    "Bullshit, man, that's not how you make stats!". I'm aware of that and we will have to start from scratch to get reliable results, the good news being, a friend of mine has crafted his own website. He retains a high level of details with very high reliability. I'll be using his results for further investigation.


    When one get resources, it is a small find (89%), a large find (10%) or a very large find (1%).

    One find metal, crystal or deut respectively in 50%, 33.33% and 16.67% of cases.

    Use the benneb simulator to find the “metal ceiling”, namely the maximum amount of metal that can be brought back: http://mviv.hebergratuit.net/expe71_prod.htm?i=1. From there :

    • a small metal find equiprobably corresponds to 5%, 6%, ..., or 25% of the metal ceiling;
    • a large find, equiprobably, to 26%, ..., 50% of the ceiling (maybe starting at 25%);
    • a very large find, equiprobably, to 51%, ..., 100% of the ceiling (maybe starting at 50%).

    The same principles apply for crystal and deut:

    • crystal ceiling = metal ceiling / 2;
    • deut ceiling = metal ceiling / 3.

    C) SHIPS

    The 89/10/1 ratio is applied in the same way as for the "resources" event.

    We have ships ceiling = crystal ceiling, the latter corresponding to the max amount of structure points that can be brought back. 1 metal = 1 crystal, a cruiser accounts for 20000 + 7000 + 0 = 27000 points.

    As for resources:

    • a small find equiprobably corresponds to 5%, 6%, ..., or 25% of the ceiling;
    • a large find, equiprobably, to 26%, ..., 50% of the ceiling (maybe starting at 25%);
    • a very large find, equiprobably, to 51%, ..., 100% of the ceiling (maybe starting at 50%).


    The pirate (resp. aliens) weapons technologies are 3 levels lower (resp. higher) than ours. There are 6 fight configurations:

    1. meet weak pirates: 27%, 28%,… 33% of our fleet + 5 light fighters;
    2. meet strong pirates: 45%,…, 55% of our fleet + 3 cruisers;
    3. meet very strong pirates: 72%, ..., 88% of our fleet + 2 battleships;
    4. meet weak aliens: 34%,…, 46% of our fleet + 5 heavy fighters;
    5. meet strong aliens: 54%, ..., 66% of our fleet + 3 battlecruisers;
    6. meet very strong aliens: 81%, ..., 99% of our fleet + 2 destroyers.

    The game determines whether you encounter pirates or aliens. Then you fight weak, strong or very strong opponents according to the 89/10/1 ratio we've mentioned for ressources and ships.


    Is it mathematically viable to send military ships in V7 expeditions?

    The problem is not to determine whether the battles against pirates and aliens are profitable: they are. The real problem is to know whether those profits outweigh both the losses caused by black holes and the deut consumption. This is what I studied.

    I will not go into details here. You can dive into the paper attached if you like (it's in French though).


    You can dowload the spreadsheet here: http://www.okok.fr/information…52-40f1-a0ba-2cd8d34dfd08.

    I opted for a "Nothing" percentage of 22%:


    It is almost an arbitrary choice, feel free to choose another percentage.

    1 Civil expeditions, 2a, 2b, 2c: leave these sheets as they are.

    2d) Military expeditions

    Calculates whether a military expedition is profitable or not according to the consumption factor of the universe and a given exchange rate. If you want to adjust these factors permanently, do so from the “Synthesis” sheet.


    On the left, profitability (or not) is expressed in metal, crystal and deuterium; on the right, in deuterium equivalent. "LF" means you only send LF, "0.4 SC + 1 LF" means that you send a combo of SC and LF in this proportion (example: 40,000 SC + 100,000 LF).

    Vary n, the number of ships sent, to get meaningful figures (you will notice that Delta is proportional to n):


    "Compared to a civil expedition, 100000 LF bring back on average a bonus of 950951 units of metal, a bonus of 316983 units of crystal and a malus of 437703 units of deuterium; in other words, a bonus of 249095 units in deuterium equivalent 2/1,5/1. "


    "Compared to a civil expedition, 100000 battlecruisers bring back on average a bonus of 11710406 units of metal, a bonus of 15613875 units of crystal and a malus of 9249921 units of deuterium; in other words, a bonus of 7014532 units in deuterium equivalent 2/1,5/1."

    "I knew it man, LF, they suck." Easy, easy, boy, the comparison table is to be found at the bottom of the sheet:


    The absolute coefficient is valid for a given exchange rate. It also depends on the consumption factor of the universe and the probabilities of the events "Pirates", "Aliens" and "Black hole". It allows to compare the performances of the compos. A composition is all the more profitable as its absolute coefficient is positive and high; it is not profitable when its absolute coefficient is negative.

    It turns out that the most successful military fleets In expeditions are lights fighters (LF), battlecruisers and small cargos (SC) + LF, large cargos (LF) + LF and LC + SC + LF combos.


    As you can see, I have devoted two lines in the table to each and every combo:


    The first ratio is profitable and realistic; the second one gets the best absolute coefficient but remains theoretical to me, because if you want to make profit, you need above all to send a lot of ships. If you send 10,000 LF, you’ll only bring back peanuts. Having 0.4 SC for 1 LF is absolutely not a problem: it is versatile and harmonious. But who would like to have 1.75 SC for 1 LF? 0.5 LC for 1 LF? ... Up to you!

    Let's take a closer look at the strategic options we have considering the example of a SC + LF combo:


    Interpretation: the absolute coefficient of a SC + LF combo, compared to a LF compo, increases as we add more SC, up to the best theoretical ratio: 1.75 SC + 1 LF. After which, it decreases. So you can basically choose any value of X between 0 and 1.75 and get good results.

    The same principles applies for other combos since you would get the same type of curve.

    4) Synthesis

    Gathers the information from "Civil expeditions" and "Military expeditions":


    If you send LC exclusively, look at the LC line; if you send military ships, look at the line that corresponds to the compo of your choice. Vary n, the number of ships sent, adjust the factors to your taste (metal ceiling, consumption ...) : you'll know how much you're supposed to bring back on average.

    Don't forget though, the sheet won't be fully functional before the probabilities are reliable.

    :!:IMPORTANT: make all your permanent modifications from this sheet. If you want to make changes elsewhere, whatever they are, be sure to Ctrl-Z to keep the formulas in place.

    Note: just because volume is the most important factor doesn't mean you should send a maximum amount of ships in one slot. It is a losing strategy. A poker player, as good as he may be, can't afford to go "all in" in every hand: if he loses once, it's over. The same goes for expeditions. You have to fragment your fleet as much as possible to reduce the variance of the profitability.


    If the probabilities are favorable (it's more and more likely they are), the best military compositions are LF, battlecruisers and LC + SC + LF combos. However, the v7 is to be revised, and the results displayed aren't fully reliable yet.

  • dex let me answer here if you agree ?

    LF get better average results in fights and their deut consumption is lower. That's the reason why. HF are not doing well enough in fights and they consume too much deut to be really profitable as a whole.

  • i see. but as a miner wouldnt doubling the crystal profits be worth the bigger deut consumption since you dont need deuterium for a bigger attacking fleet?

  • The absolute coefficient (the only figure which allows you to make comparisons) relies on a given exchange rate, it already grants a better value to crystal, compared to metal. You can download the spreadsheet and change the 2/1,5/1 rate I used if you want to.

    But if your goal is to bring back crystal, i'd rather advise you to send a SC+LF combo, or a LC+LF combo, or even battlecruisers.

    Also, if your fleet is really low in volume as a miner, I recommend that you stick to LC as a small military fleet (no matter the compo) isn't very profitable (as a whole), you'd have to use additional slots to collect DF for almost nothing.

  • I go with SCs in the 1x unis I play in, and LCs in the faster flight speed unis.

    I add cargos to the combat fleet until I see my losses increasing a bit, then I add more LFs. I want as few LFs as possible to protect as many cargos as possible and kill as many aliens as possible with as little loss as possible. I think that makes sense.

    No-merchant unis require creative measures to increase crystal income, and this is one I use. Right now Im around 4500 LFs + 6000 SCs in Mensa.

    Brute force, if it's not working, you ain't using enough.

  • Oh yes it makes sense! Actually, mixing cargos and light fighters is pretty much like crafting an optimized battlecruiser in the sense that you increase the defensive qualities of your expedition fleet without any deut losses apart from the consumption. 1,75 SC + 1 LF and 0,5 LC + 1 LF are the absolute best combos according to the calculations.

    However, to my point of view, you don't bring back enough ressources with 6000 SCs + 4500 LFs, regardless how good that combo is:


    See? 43987 units of deut (equivalent deut 2/1,5/1) per expedition on average. You have to be there to collect the DF for peanuts. I don't think that's worth it. Volume is the key, all the more as there's no limit ...

  • My uni's top player has less than 5 million points, so the rewards are substantially lower than older Unis.

    I dont play actively in any "older" (still has merchant) uni at this time.

    Brute force, if it's not working, you ain't using enough.

  • I've been doing 7 waves of10K LF/600 LC/1 DES/1 PATH/1 ESP roughly 6 to 7 times per day in a 1x uni and I have been fairing pretty good...but it is a fairly young uni, with rank 1 (me) <25KK (currently 8.2KK). Adding addiitonal LC and SC can get a bit more riskier, but I might try to plus up my combo with SC and see how that goes. The sims seem to favor it.

  • I send 1m LF and 500k SC in a slot / 8 slots for pirate.

    but the biggest problem is that I can't meet them.....

    I burn more than 500m duet per day to find pirate, but only 1-2 battles in 80 exps.

    I'm very puzzled of that.....:(

  • JorjX Oh I see, then it makes sense, 43987 units of deut per expedition is already a fair amount of resources in such a uni.

    Cathedral Riskier? I mean of course you add a little bit of variance since you engage more ships, but that's ok given the expected return, the latter being exactly proportional to the size of your expedition fleet.

    gufengqu You're just being unlucky, that's part of the fun and you have to accept it. It's not going to last forever. Also, for your friend, even if they consume nothing, DS fight too poorly against themselves to be profitable as a whole.

  • The wiki has been written by players, not developers, and it is getting outdated. I don't know how accurate it was at the time it was written, I can only tell you about my own work. The figures comes from TopRaider, Expé 3000 (a French add-on) and "home stats". I am aware that both TopRaider and Expé 3000 fail to recognize messages from times to times, so the results are possibly biased (and there are other reasons). That being said, the "home stats" sample, which I'm 100% confident about, does not state that the black hole rate is higher than 0,203% (on the contrary).

    Now, of course, a stat and a figure which is contained in the source code of the game are not the same thing. The best you can do is to estimate the actual rate based on the stat and the sample size for a given confidence level, thanks to a confidence interval - which I have done. If my numbers are correct, it is almost certain (99,9%) that the black hole rate is included between 0,156% and 0,263%. So I'd say it is very likely that the black hole rate is not 0,33%, nor 0,3%.

    With a 3/2/1 exchange rate, the absolute coefficient of profitability decreases, yeah, but LC+LF compos are still profitable.

  • Granpa i saw your english version table.

    Could you tell me the EV ROI per expo in pirates/aliens?

    Assume i got absolutte coefficient for 2/1 LF LC: 12.30 or 2/1 LC HF 8.04

    what those numbers means?

    i want the EV Profit per expo with LF LC and LC HF

    Do you want success? just play smart

  • Absolute coefficient is ROI*10000 (a factor that I added for players who don't like numbers to read the table easily).

    Theoretically, you gradually deplete a solar system if you exp too much in it (which means the percentage of "nothing" increases). At least it was true till v6. For v7 I've got to say I don't know (I haven't made any calculation yet), but I'd recommend that you exp in separate systems.