Dismiss Notice
This Section is READ ONLY - All Posts Are Archived

Big Thread of Mining Suggestions

Discussion in 'Release 27 Feedback Forum' started by Noric, Mar 11, 2016.

Thread Status:
Not open for further replies.
  1. Noric

    Noric Avatar

    Messages:
    1,328
    Likes Received:
    1,822
    Trophy Points:
    113
    I've spent a lot of time mining between R23 or so and now including having 2 mining related GMs both before and after the latest wipe. Currently, I am sitting 78 mining proficiency, 115 mining swiftness and 100 mining meticulous. I also spend a lot of time crunching numbers and optimizing stuff and thus have found a number of design issues which I also have some suggestions for. I would note that although some of these concepts apply to all gathering, mining can vary in a few ways and this means some of the these ideas may not apply across the board.

    Problems:
    1. Proficiency ineffectiveness
    2. Experience Penalty for Meticulous
    3. Swiftness scaling too backloaded
    4. Wide disparity of skilling order viability


    Proficiency Ineffectiveness:

    Success chance has 2 primary factors. The first is mining proficiency and the second is producer
    level (as displayed by V). I do not know what the equation that uses these is, but it has some embedded issues. With 10 proficiency and no producer levels I start out at about 12% chance to mine a tier 5 node (average metal node found in a mine). With 10 proficiency and 70+ producer level I have something like a 50% chance to mine a tier 5 node. With 78 proficiency and 70+ producer level I have a 69% chance of mining a tier 5 node. 68 levels of proficiency only increased my success chance by less than 19%( which is about half the impact that my producer levels had). Skilling above 78 has little added bonus (I had GM before the wipe).

    One of two things is happening. The first option is that producer level just plain means more than
    proficiency. The second is that some aggregate value based on the two is used, and that the value has diminishing returns. I think the latter is more likely, but the end result is essentially the same as the former. Producer level is based on total producer experience gained so anyone who is training mining skills high will naturally get producer levels high. This means that except for training proficiency at very low producer levels, producer level will always have a larger impact.

    Fix: The fix is pretty easy. Change the success chance formula to value producer level less and proficiency more. If diminishing returns are involved, they might need to be reduced to balance out success chance to around where it is now.


    Meticulous Experience Penalty:

    This issue is pretty obvious. Producer experience is gained through mining success and spent through mining attempts. Training meticulous slows both of these processes and therefore slows skill gain.

    Fix: The obvious fix is to make meticulous success give experience. Now the slowdown only occurs after a normal success, so granting full xp for each meticulous seems a bit weird. I would say for each meticulous success adding .1 to .25 times the normal node value. This would mean 5x meticulous success would grant 1.4 (1 + 4*.1) to 2 (1 + 4 *.25) times the experience of a normal success.


    Swiftness Scaling too Backloaded:

    The current issue with the scaling on this skill is that it does an additive reduction rather than a multiplicative one. By this I mean that instead of reducing the current mining time by some amount of the current time, it is based on the the original time. This means that every point in swiftness reduces the mining time of a tier 5 node by about half a second. So leveling from 0 to 10 takes a player from 60ish to 55ish seconds for a tier 5 node but leveling from 100 to 110 goes from 12ish seconds to 7ish seconds. This means that training swift gets almost all of its advantages from training the skill really high which limits skilling order choices and penalizes those not willing to grind swift extremely high. The other issue with the current design is that the skill is effectively hard capped at 125 which is a lot lower than most skills.

    Fix: The fix is to shift to a multiplicative reduction. I actually worked out a formula that keeps the reduction at GM the same while doing this shift.

    T2 is the new time, T1 is the base node time and S is the Swiftness skill level:

    T2 = T1 / (1.016225^S)

    This formula would make each level reduce the current mining time by the same amount and provide a smoother benefit curve.


    Wide Disparity of Skilling Order Viability:

    As things stand right now, the order in which you effectively train skills is obvious. For the biggest benefit one should train Swift exclusively until it is as high as you possibly want it. Then you train proficiency as much as you want to and finally train meticulous. This is because training mining skills is more or less dependent on the amount of producer xp you have and the time it takes you for a gathering attempt. Although, proficiency does potentially increase the amount of producer xp you get (more successes means more producer xp), that increase is probably going to be smaller than the number of success you will get from mining faster (leveling proficiency to 50 might gain you a ~10% chance to mine nodes, but compared to removing 40% of the mining time while still gaining some increased success chance from producer levels it usually will lose). Additionally, swiftness getting stronger further up means that rushing towards extremely high swiftness will make training the other skills a lot easier. Finally, meticulously is dead-weight from a training perspective so training it last is ideal.

    Note: The survey skill are all universally so unimportant that I would probably advocate removing them entirely and adding a minor glimmer by default.

    Fix: Well the first 3 solutions all go to fix this somewhat, but there are a couple things that still could use addressing. Adding experience with meticulous is a net increase in experience, which means that the base amount should probably be reduced somewhat. Additionally, I'm a bit skeptical that even after retuning success chance that proficiency will be fully desirable and in that event adding a slight experience multiplier to it makes sense to me (very slight... and obviously rebalancing base xp if needed)

    These things combined should allow players to viably choose different skill training orders (including training them all at once or switching back and forth between them).


    Dev shoutouts(not really sure who makes decisions on this subject) :
    @Chris @DarkStarr @Bzus @oneandonly @Chaox


    If anyone has any feedback or questions, i would be happy to explain/debate more on the subject.
     
    Last edited: Mar 26, 2016
  2. Daxxe Diggler

    Daxxe Diggler Avatar

    Messages:
    2,692
    Likes Received:
    5,711
    Trophy Points:
    153
    Gender:
    Male
    Location:
    Virtue Oasis - Hidden Vale
    Fantastic post Noric! Being a fellow GM miner (101 proficiency, 109 swift gathering, 43 meticulous) I have seen similar results with regards to the effects of training the particular mining skills. I haven't done the research you have to produce numbers like you did, but in general, I would say I've noticed the same types of disparities as you.

    Swift gathering definitely feels like it gets better the higher you take it... more so than the increased benefits you see from increasing the other skills that high.

    Survey is indeed meaningless and frankly its a waste of 10 skill points just to get to meticulous. If you can't see a mining node without the minor sparklies that survey provides, then you need to get your eyes checked. :eek:

    I also agree that meticulous is sort of a deterrent to gaining producer xp and also increasing your skills faster because, as you said, the meticulous chances don't give you bonus xp and take extra time that could be spent on starting another node instead to get more xp faster. Besides that, meticulous chances seem to only provide you with extra base ores/granites and don't give extras of rares like tin and nickel (at least not that I've seen yet). If you need extra copper/iron/silver/gold/granite... it's worth having. But if you are looking for extra rares to make the better armors and weapons, meticulous is actually a hindrance for collecting those.

    Again, great write up on the mining skills and I concur with pretty much everything. The only things I can't say are 100% correct are the data points you provided for percentages at levels, etc. because I didn't take notes to compare and verify them. But they do seem to be accurate to me based on my experiences in mining.
     
    Aldo and Beaumaris like this.
  3. agra

    agra Avatar

    Messages:
    1,501
    Likes Received:
    3,489
    Trophy Points:
    113
    Regarding meticulous, I snagged this screenshot during R26. True, it may have changed in R27, but here it is anyway:
    [​IMG]
    That was a Meticulous x 2, with Iron, Nickel and 2 Granite from the first, and Iron, Tungsten, and 2 Granite from the second.
     
  4. Noric

    Noric Avatar

    Messages:
    1,328
    Likes Received:
    1,822
    Trophy Points:
    113
    The cap is on duplicates of the SAME rare ore. I've done hundreds of meticulous groupings including several very large sprees (6-9x) and i don't ever see 2x tin, 2x zinc etc. I bug reported this in r26 but this thread is more on the feedback side of things so I didn't include anything on that.

    Additionally, I honestly generally find myself more capped by common materials than rares considering it takes 4-8 common ore to make a single rare ingot.

    The numbers are approximations. However, the mathematical properties of the swift differences (12 to 7 second and such) are consistent whether or not the original mining time is exactly 60 seconds or +/- 10 seconds of that value. Additionally there are other factors such as wind-up time that I am not addressing (refer to Umuri's bug report thread for more on swift gathering : https://www.shroudoftheavatar.com/forum/index.php?threads/extreme-swift-gathering-bugs.47272/ ) . The numbers for producer level versus profic are estimations - though i have had variable proficiency versus producer level for 2 different wipe iterations so they are not blind estimations.
     
    Lord Baldrith and Sara Dreygon like this.
  5. Daxxe Diggler

    Daxxe Diggler Avatar

    Messages:
    2,692
    Likes Received:
    5,711
    Trophy Points:
    153
    Gender:
    Male
    Location:
    Virtue Oasis - Hidden Vale
    Yep, I know you can get tin and zinc or nickel and tungsten in the same loot with meticulous. But I was referring to not being able to get more than 1 of each type.

    I've never seen 2x tin or 2x nickel from one node. But I have seen up to 10x copper or granite from say a 5x meticulous run. Why can't we get extras of the rares (that we can actually use in a recipe)?
     
    Sara Dreygon likes this.
  6. agra

    agra Avatar

    Messages:
    1,501
    Likes Received:
    3,489
    Trophy Points:
    113
    Yeah, that's a good question. To me, I would say it's a bug, but it could be working as intended, and just punitive. It's a strange place to put a punitive mechanic, though, so... yeah, I'm not sure.
     
    Lord Baldrith and Sara Dreygon like this.
  7. Noric

    Noric Avatar

    Messages:
    1,328
    Likes Received:
    1,822
    Trophy Points:
    113
    It was a change that i think happened either before or after r24 or so. Not sure if it is intended or not, but again i don't personally find it to be a high priority. I have tons of rare ores for practical purposes and the value would drop if they got more common - so it wouldn't change much for me.
     
    Sara Dreygon likes this.
  8. Noric

    Noric Avatar

    Messages:
    1,328
    Likes Received:
    1,822
    Trophy Points:
    113
    Just wanted to post a small update. I've been deleveling my mining prof (as my % seems moderately stuck at 69%). My plan is to delevel until i drop to 68%. Along the way i noticed something interesting that i thought was worth mentioning. With my 78 mining proficiency i have a 69% mining chance on tier 5 nodes... but my forestry chance against tier 5 nodes with 10 proficiency is actually 59%. So instead of the 68 levels of proficiency providing 19% increase my previous estimate gave - it is only a 10% increase for me now(76 producer level). This means that GM proficiency is potentially worth less than a 15% increase for someone with high producer levels. As a percent of percent (.15 / .59) that would represent the act of leveling mining proficiency from default to GM would only provide around a 25% increase in the rate of mining success. This is a rather wide contrast to the 500% speed increase that GM swiftness gets from level 0.
     
  9. Daxxe Diggler

    Daxxe Diggler Avatar

    Messages:
    2,692
    Likes Received:
    5,711
    Trophy Points:
    153
    Gender:
    Male
    Location:
    Virtue Oasis - Hidden Vale
    Interesting findings Noric. So basically, the producer level helps more than the individual "Proficiency" skills?

    If this is true, then it might be a good plan to just train proficiency up to 10, pick up the swift gathering, and turn off proficiency while leaving swift on until you get fast enough to be "bearable" to you. Then move on to another gathering type and do the same, then onto another gathering type and do the same, etc., etc.

    If you do that, then your producer level will continue to increase, which will increase your % chance for all gathering skills at the same time. And, all of your producer xp will be focused entirely on the swift skills... making you faster at gathering everything sooner!
     
    Sara Dreygon likes this.
  10. Noric

    Noric Avatar

    Messages:
    1,328
    Likes Received:
    1,822
    Trophy Points:
    113
    Few Things Daxxe:

    1. I'm not sure about the most recent new characters but last wipe i started with 10 proficiency in the gathering skills - so no leveling required if you don't want higher.

    2. I don't know if proficiency directly matters less or if its is just a calculation based on them both that has significant diminishing returns. Either way, there may be a small window of time during which proficiency matters enough to train. Given that the lower levels of skills take a lot less xp, then something like training proficiency to 50 or so after training swift to the same (250% speed increase is still likely hard to top with the difference from 10 to 50 prof).

    3. There isn't really any reason to train more than 1 gathering skillset. I'm not saying it cant be nice to gather the resources quickly, but the fastest way to gain producer experience(and thus producer levels) would either be mining(in somewhere with a lot of tier 5 nodes) with high swift or skinning with high swift. In the case of mining, training both swift and meticulous beyond 100 have notable payoff so it isn't like you need to spend the xp on other trees if you don't want to.
     
    Oninoshiko and Sara Dreygon like this.
  11. Jefe

    Jefe Avatar

    Messages:
    905
    Likes Received:
    1,007
    Trophy Points:
    93
    I think the major question is, how will experience gain in crafting occur in the next release, and beyond that? Per my understanding, we are only at the tip of the iceberg for crafting. I just hope that they find what makes them happy and stick with it.

    Refactoring systems from the ground up after release will just frustrate people.
     
  12. Noric

    Noric Avatar

    Messages:
    1,328
    Likes Received:
    1,822
    Trophy Points:
    113
    While i agree with you, i think that it is largely unrelated to most of the content in this thread.

    The gathering system is not really reliant on the crafting system for how it functions. If crafting becomes a source of producer xp - then we might see across the board gathering xp decrease. However, that doesn't really invalidate any of the changes mentioned here or address the concerns.
     
  13. Jefe

    Jefe Avatar

    Messages:
    905
    Likes Received:
    1,007
    Trophy Points:
    93
    Ok, to be more on point.

    Agree. Skill should be more of a factor than producer level.
    Linear growth does make swift much more viable than base mining proficiency. This is buffeted by producer level being the larger impact. I don't know if I agree on making uniform speed improvements though. While the number on the skill simply increments by one, what goes into getting that increment increases. I believe amount of usefulness of the skill should increase more as you get higher in the skill. Meaning the higher you climb the mountain, the more worth there is to making the climb.
    Agree that Meticulous should give some xp.
    On swiftness scaling, depends on how meticulous is calculated. If it is precalculated, I.E. generated while hitching to the node then yes this will work. If it is calculated when the bar completes then M would be unknown at T sub 1. I have noticed a frame stutter at completion of the first bar, so I suspect the latter currently.

    My line of thought above is that once base equipment needs are reached, then the primary reason for improving gathering is to increase experience gain rate.
     
    Noric and Daxxe Diggler like this.
  14. Noric

    Noric Avatar

    Messages:
    1,328
    Likes Received:
    1,822
    Trophy Points:
    113
    I understand where you are coming from here - but it is a direct contradiction of how almost any other skill currently works in the game. Most adventurer skills seem to have base and scaling aspect and it takes GM skill before the scaling is even valued as much as the base. This means that the benefit of the skill is a very slow linear while the experience cost is growing. Even in other crafting skills nothing really ever scales better than linear increments (crafting gets a success chance every 4 levels it seems like, without a cap proficiency gets a .5 (something?) gain every level). I personally, could live with a slightly building benefit(less aggressively than it currently is), but the formula i provided is my estimation of how Portalarium would design something inline with other skills in the game.
    Sorry, I do believe you caught a typing error. Did not mean that equation to have anything to do with meticulous.

    I'm honestly not sure what the situation is going to be here. Base equipment needs seem pretty modest on a long term(unless they keep changing stuff) and as you mentioned before, we don't know what the experience situations is going to be like until they start tweaking the crafting system. Until I have more information, I have a limited opinion on it. I definitely don't think meticulous should penalize experience, and i think the skills should not be totally imbalanced in improving meticulous intake, but I'm not sure about the end state of crafting.
     
  15. Jefe

    Jefe Avatar

    Messages:
    905
    Likes Received:
    1,007
    Trophy Points:
    93
    A. It may be a case of what is in a name then? I think then the concept of Grand Master may need some evaluation. I am a grand master of smelting (at 100 skill), yet I have a 43% chance of making at Bronze ingot ATM. IIRC, there was talk about multi-tier such as Master and then Grand Master. The term Grand Master implies that there are few that are your peer. I think they need to think about what a Grand Master implies, otherwise GM seems to be about the level of Journeyperson. If short linear growth is all it will ever be, then that is an argument for removing or pulling the GM marker to the right IMO.

    In Adventurer skills, specifically active combat, they definitely feel/seem unfinished in progression as the same values are often what you get when you reach the next 10-20+ levels. I assume that observation is where there are people stating that it is best to stop at 80 on a skill. Agree wrt innates.

    I think largely to growth concerns do stem from producer level being the larger weight. I am curious to see if that will invert as time goes on.

    B. No worries. Based on my interpretation, I wasn't sure if the plan was to pre-calc meticulous in the time equation and set speed based on total time (base plus meticulous).

    C. 3 x Agree.
     
  16. Noric

    Noric Avatar

    Messages:
    1,328
    Likes Received:
    1,822
    Trophy Points:
    113
    The GM debate is ongoing, but instead of the base being half of the GM value, i really would prefer the base being 1/3 GM value. That way level 80 would be around 85% of GM instead of 90%. Anyway, the adventurer side of that is definitely off topic for the thread.
     
  17. Jefe

    Jefe Avatar

    Messages:
    905
    Likes Received:
    1,007
    Trophy Points:
    93
    Fair enough. Perplexingly, I believe I've said all I need to say.
     
    Noric likes this.
  18. Noric

    Noric Avatar

    Messages:
    1,328
    Likes Received:
    1,822
    Trophy Points:
    113
    More or less how I feel on the subject. Not that i mind discourse with other people on it, just that there's only so much to say. Most of the high level mining people i speak with seem to think the problems I've identified are at least accurate, though some dislike portions of my solutions or think that the end target should be slightly different(whether all skills should increase xp gain somewhat proportionally, whether skills should have linear benefits, etc).

    However, all of our design conversations only have so much importance. While better solutions than are actually implemented might actually occur in feedback threads, we don't get the option to implement them. My major purpose with writing this was to help identify design issues that exist and if possible give Portalarium some starting points to cut down the time needed to implement fixes if they want to(there's obviously more than one solution for every problem, and I don't claim to have the best solutions nor those that are "right" for the game). No idea if the thread will fulfill that purpose, but i suspect the odds are low given that it will be deleted in another 4 days.
     
    4EverLost, Elwyn and Daxxe Diggler like this.
Thread Status:
Not open for further replies.