T9 harvesting: Complaint

Discussion in 'Tradeskill Discussion' started by ARCHIVED-Summona, Sep 10, 2010.

  1. ARCHIVED-Summona Guest

    QuestingCrafter wrote:
    All of these items together doesnt get you over 600 in anything. Only the +50 to fishing item from SH gets you over 600 in one harveting skill. AND if you read my first post, The way the mechanics are set, It doesnt help pretty much, AT ALL.

    Plz read before posting, Ty
  2. ARCHIVED-QuestingCrafter Guest

    Traxor@Lucan DLere wrote:
    1. My first post didn't quote your post -- my responses were to the topic in total.
    2. If you pull out your calculator, my post offers an easy way to exceed 600 skill in Mining.
    3. Your first post indicates nothing about a skill cap (hard or soft).
    ... I'm so pleased YOU could read before posting. =P
    So the ACT plugin to parse harvesting over the past three weeks (T9 molten & infused) : 2119 collects ... 23 rares. Which is right around 1 rare per 92 collects (harvests), or about 1 per full 31 nodes. It *seemed* less frequent, but maybe that's because I need the kaborite much more than the diamonds or brellium. ;)
  3. ARCHIVED-Summona Guest

    QuestingCrafter wrote:
    So reading my post this time now maybe you will read this. I am saying that the +skill items in current game mechanics hardly increase your chance to get a rare. So why is there gear. The gear was implemented to make players work hard to get these rare items to help increase their chances at getting rares. If not then bot programs win by default because most people who bot are naturaly lazy. They do not want to harvest. Their bot harvests all night while they are sleeping.

    Secondly do not post saying you have 50 rares in your bank and you got them all by yourself. No one is picking at you so there are no reasons to get upset. I harvest one time. i get a rare on the second node. I harvest again i can harvest 900 common items without getting a rare. Even if i only get one rare thats a 0.001_% chance of getting a rare. SoE hates telling us game mechanics because they dont like people questioning and finding fault in coding. but i dont think those kind of odds were intended.
  4. ARCHIVED-LivelyHound Guest

    Traxor@Lucan DLere wrote:
    (Somewhere a quote got messed up I think, so if I put a quote on someone by accident I apologise)
    There is gear because it helps increase you harvest rate for raws not for rares as I understand it. The side benefit of added rares, to whatever degree is just that a side benefit. If you tradeskill a lot and require lots of materials these items save you time. If all you harvest for is rares for spells then these items are not designed for you but will over the course of an expansion help out. The items in question are not rare, except for the DWB +20 and the earring +5. They may take time to get but that does not mean rare and the reason for the time invested is to discourage bots from getting them as you happily pointed out.
    Incidentally quester just abbrieviated to ~600. He never said over 600. He did point out that with brell blessing you hit over 600. I have all the items that I know of and end up at fish 593, mining 598, the other 588, which is a moutful to type and it's easier to write ~600 for all. With Brell's I'm at 623 mining if i remember correctly. All of which makes my rare rate 1% rare pulls. I.e. 30 nodes, as I have 15% double harvest aa, and I get a rare. Pretty much without fail if I do a harvest run that is what I get. The trick therefore is to one, use ample harvest and two, have a very high density fast repop harvest loop. Think about it if the rare rate was 2% I would get something like 14 rares an hour... How much of an increase would you like to see?
    "I harvest one time. i get a rare on the second node. I harvest again i can harvest 900 common items without getting a rare. Even if i only get one rare thats a 0.001_% chance of getting a rare. "
    This quote is great you go that 2nd node you get a rare, woot rare rate: 17%. You then happily ignore that fact and go on to say you get a rare rate of 0.001% because the next time it takes you 900 common items. Saying 900 common items is again misleading as it sounds like 900 pulls, i.e. mouse clicks, but it isnt. It's more like 400 pulls, which is 133 nodes, or a rare rate of just 0.75% of pulls (400 because you get 1,3 & 5 raw pulls, and overtime this averages I find to about 2.3 items per pull. 133 nodes is 400/3, assuming zero double harvests. If you have the double harvest then that node value is too high and needs to be reduced by 15% and you up at around 105 nodes. I.e. you have a rare rate more inline with the 1% I have found and that quester has found! Grats lucky you! Because the 2nd node rare pull bonus is pure bonus and lucky RNG!)
    Now the question that needs to be asked and hasn't as far as I know been answered regarding rares, is:
    Without using any of the bonuses what is the rate? 0.5%, 0.75%?
    I would go and do those harvest runs but ... I have the gear and want to maximise my playtime not work the answer out.
    You wrote: I am saying that the +skill items in current game mechanics hardly increase your chance to get a rare.
    I would ask you : how do you know? Have you got the answer to the question I posed 3 lines above?
    Now I will re-ask the question: How much of an increase would you like to see?
    At present you get with everything about 1% of pulls giving rares, 7 per hour
    Up that by 0.1% and you end up with 7 rares per hour maybe 8 on a lucky day, and no real difference.
    Up that by 0.5% and you get 1.5% of pulls, giving 10 rares an hour and most people wont notice.
    Up that by 1% more chance and 2% of pulls, 14 per hour, people would start to notice the bonus.
    Up that by 2% giving 3% of pulls, or 21 per hour! Everyone would have the items and noone would ever sell a rare again.
    The other option is that they give that amount of bonus but reduce the basic harvesting a rare chance, which would lead to having all the gear giving the current rare rate 1%, 7 per hour, but the base rate being so small that you never get a rare without the gear. I cannot see them doing that and thus we have the current state of affairs where rare chance is not affected much but the raws are effected greatly by gear. This is what I think they were initially aiming for and this is what we have. Pretty well done I think. YMMV.
    For me personally if they upped the chance given from higher skill it would be insane, I would love it for the first week because I could get all my alts their rares, after that , I would sit there and wonder what the hell the point in having the items was as every expansion I could have all my experts for all my alts within a week. In other words they would be like the FSSD xp jewelery set. Useless except for that one week I lvl my tradeskilling. At least in their current form I always feel that I'm getiting use out of them as they MAY increase my chances and I still need to harvest.
  5. ARCHIVED-LivelyHound Guest

    Artemiz@The Bazaar wrote:
    Have you done this rinse repeat in order to find that 0.04% increase? Just curious as to how you got your figure and was that for SF harvesting tables, or for the slightly better chance ROK/TSO tables? I ask as this is the only ever figure I have seen posted regarding the difference between with gear on and without gear on at cap.
  6. ARCHIVED-QuestingCrafter Guest

    Traxor@Lucan DLere wrote:
    So, let's look at the bottom of each Traxor post in the thread ...
    Traxor@Lucan DLere wrote:
    Traxor@Lucan DLere wrote:
    Traxor@Lucan DLere wrote:
    Traxor@Lucan DLere wrote:
    Traxor@Lucan DLere wrote:
    Traxor@Lucan DLere wrote:
    Traxor@Lucan DLere wrote:
    Traxor@Lucan DLere wrote:
    ... you don't seem to have the same point "every time" at the bottom of your posts.
    You might recall that you believed:
    Traxor@Lucan DLere wrote:
    ... I showed you how you COULD get your Mining skill higher. You did not say "I don't think I can get it any higher via non-temporary gear" -- so I informed you how you could get it higher. Sheesh, try to help some ESL fellow out ...
    Traxor@Lucan DLere wrote:
    Fantastic -- for once you clearly state your position. Huzzah! I happen to agree that your chance at getting a rare increases a very tiny amount (we could probably argue how much exactly -- but we agree it's tiny). But in a game based around numbers ... "every little bit helps" as they say. If you're going to harvest thousands of times per week (per month, whatever) ... an extra rare here and there is beneficial.
    If I could demonstrate that the skill increase from item "X" yielded me one extra rare in 1,000 harvests, then we could do a cost-benefit analysis, relating time to acquire +skill items, to increase in harvests, and its relative value in time/coin. For some people, getting an extra rare (to use, or sell for 15p) every week, in exchange for a few hours of work one time, is worthwhile.
    Sadly, no one appears to have definitive data they can share, as to the increased rare harvest rate for upper-end harvesting.
    Traxor@Lucan DLere wrote:
    Never did either -- I can only guess you were 'quoting' someone else. Or maybe you didn't bother to read. Again. ;)
  7. ARCHIVED-Summona Guest

    Im putting together harvest data. Currently it looks like around a 0.008 rare chance.
  8. ARCHIVED-S1lence2 Guest

    QuestingCrafter wrote:
    Lol this is obviously the T9 whiffleball server kit, The T9 Nagafen kit is basically a full toughness set of pvp gear supplemented with capped run/nount speed and as many available evac items as possible. Possible even a 2boxed Conji in immunity for quick Coh evacs:)
    And if sighted: You should always run out ahead of every botter and take 2 out of 3 pulls off of every node in the zone and then sit back and rofl while they scratch their heads :)
  9. ARCHIVED-Summona Guest

    Type# Collects% Collects# Items% ItemsAvg Items/Collect# Rares% Rares
    All4039100.00%9166100.00%2.2742100.00%
    amber174643.23%398343.45%2.2800.00%
    bamboo shoot30.07%50.05%1.6700.00%
    brellium ore20.05%20.02%1.0024.76%
    caynar nut20.05%80.09%4.0000.00%
    ethereal material521.29%520.57%1.0000.00%
    giant octopus meat10.02%30.03%3.0000.00%
    kaborite cluster190.47%190.21%1.001945.24%
    kejekan palmfruit10.02%30.03%3.0000.00%
    lapsang tea leaf10.02%10.01%1.0000.00%
    Marr cherry20.05%80.09%4.0000.00%
    mottled pelt210.52%600.65%2.8600.00%
    quicksilver cluster166241.15%377941.23%2.2700.00%
    spotted pelt10.02%10.01%1.0012.38%
    sumac lumber400.99%780.85%1.9500.00%
    titanium ore46611.54%114412.48%2.4500.00%
    ulteran diamond200.50%200.22%1.002047.62%
  10. ARCHIVED-Summona Guest

    This is with 598 mining and bountiful harvest on. 0.5% rare chance. Proving +harvest skill does nothing
  11. ARCHIVED-kdmorse Guest

    All data is good, but you must break it out by node type to be meaningfull.
    The different node types are known to have drasticly different rare rates. So without knowing how many pulls were at each node type specifically, an actual rare per node rate cannot be calculated. An 'overall' rare rate is of course meaningless.
    (We can sortof infer it from the results, amber vs quicksilver, etc... - and I will try to sort through that. But some things, like ethereal material could have come from any node, so there's no way of knowing what you pulled).
    -Ken
  12. ARCHIVED-kdmorse Guest

    Traxor@Lucan DLere wrote:
    While the data you have provided is excellent and is greatly appreciated, I *strongly* disagree with both conclusions here.
    You have mixed the two mining nodes roughly 50/50. Your results make sense - as spellcasting rares are expected to be about a 1% rate, and brellium is expected to be about a 0.1% rate. Combine the two, and you end up with a 0.5% average.
    EDIT: /math fail. The correct math is actually much more interresting....
    But as I said, an overall rare rate is meaningless since it varies dramaticly based on the type of node you are harvesting. Comparing overall rates between two different harvesting gears on an inconsistant mix of node types makes any comparison between the two severly tainted.
    Regardless, only 2000 pulls on a given node is statistically insufficient to demonstrate a conclusion on the effectivness of harvest skill. You are still well within an expected standard deviation caused by random sampling. If the difference is as small as we suspect it is - we're going to need approximatly 12,000 pulls, of a given node, at each harvesting level to demonstrate a difference at any level of certainty. (Maybe as low as 8,000 if the difference is larger)
    Damnit - someone come do my day job - I enjoy this sort of thing, and need more time to spend on it.....
    -Ken
  13. ARCHIVED-QuestingCrafter Guest

    I'd almost wish to bot -- just to do 10,000 collects of a specific node type (e.g.: molten) at 450 Mining, then another 10,000 of that same specific node type at 598 Mining (or any fixed number over, say, 490). Those data would be interesting to see, from a set like that, could we "prove" something.
  14. ARCHIVED-Summona Guest

    take out the extra's and re do the %'s urself. Or present data to argue my claim.
  15. ARCHIVED-Vieray Guest

    Ynnek@Kithicor wrote:
    All you say is true...not sure why you continue to respond to this. The internet is a great place to find people who think they are right no matter how much the data stacks up against them. Experiment design and data interpertation can be very difficult and are usually to hard for the average eq player.

    Oh and is the brellium pull rate really around 10 times lower then pulling a spellcasting rare? Is there a reason for this? I know you need more rares for experts then you do for armor or weapons but it isn't 10 times as many.
  16. ARCHIVED-LivelyHound Guest

    Traxor@Lucan DLere wrote:
    To me the most important numbers here are the first row: 4039 collects, 42 rares = 1.04% rares much as I stated upthread it would be.
    Followed by : 2.27 that is average items/collects for all, which is identical to my data.
    Followed by amber + quicksilver collects = 3408 compared to titanium ore collects of 466 or a node harvest differential of 3408 : 466 or approx 7.3 soft metal nodes to hard metal nodes, that you then need to multiply brellium rares by to get the rare rate that is comparable to your spell rare rate. That is 7.3 * 0.07 = 0.51% 7.3 * 0.05 = 0.37% compared to your 0.5% ulteran diamond, and 0.47% kaborite. Wow they are virtually identical. (of course this would be simpler if the node break down was given, as i couldnt be bothered to add in the rares to amber + quicksilver + rares : ...)
    Edited to correct very odd formatting errors
    2nd Edit as I noticed a mistake, took the wrong number (1 line up oops) from the chart.
  17. ARCHIVED-Te'ana Guest

    Some folks here are under some misconceptions about how testing for the rate of rares pulled works.
    The probability of getting a rare should not be different from one node type to another node type. So combining the results is valid unless some Dev says the probability for obtaining a rare varies significantly between types of nodes.
    Using 2,000 pulls is sufficient for statistical sampling. In fact, its overkill. Generally speaking, using a sample size of about 1,500 should be sufficient for this type of experiment to be valid. Expanding the sample size will not increase accuracy in a meaningful way since you can never test the entire "universe." Some statisticians would argue that a sample size of 50 is all that is required, but I am sure no one here, including me, would ever trust results obtained from that small a sample :)
  18. ARCHIVED-kdmorse Guest

    Vieray wrote:
    That's actually a good question. The short answer is that I love this sort of thing. Here we have a black box, pulls go in, harvests come out, and we potentially have a dial (harvesting skill) that affects it, but we're not sure by how much, and if it's signifigant. It's facinating, and I wish I could work on deciphering it's mysteries. Unfortunatly, I don't evem have the time to transcribe those results into my own spreadsheet.
    So, all that leaves, is yammering about it on the internet. And I don't mean to come off as overly critical (too late) - I just feel the need to jump on statements like "This proves this", when, really, it doesnt.
    Look at every set of results posted here so far, including the ones just posted above, and they're all in that catagory. My own data matches Traxors, and the data someone else posted earlier.
    2211 pulls at infused deposits, 2 brellium ores - for a rate of 0.09%
    4877 pulls at molten formations, 51 spell rares - for a rate of 1.04%
    All together we've got nearly 10,000 pulls amongst us, and they all show the same general 10:1 discrepency between the two mining node types.
    -Ken
  19. ARCHIVED-kdmorse Guest

    Lateana@The Bazaar wrote:
    I agree that it *should* not. But it is. Maybe it's as simple as someone making a typo in a harvest table somewhere.
    I agree with you if the goal is to determine an approximate harvest rate, to get a sense of the nature of the beast. For that, I'd accept as few as 1,000-1,200. That is more than enough to say that the rare rate for spellcasting rares is about 1%.
    The extra pulls are required however if the goal is to determine the difference between two different harvesting skills, when the difference is believed to be as low as 0.04%. (I personally believe it to be higher, but let's go with that). To demonstrate - or disprove - a difference as small as that, requires many, many more samples.
    For example, I'm up to (only - sadly) about 1000 controlled harvests at known skill rates. My current results are:
    At skill 583 - 1.35%; At skill 450 - 0.78%; 1000 samples at each
    You look at those numbers and say "Wow - skill makes a difference!", but 1000 samples is far to low to jump to that conclusion. If you expand the data with error bars and account for the random nature, all you can really say to a confidence level of 90% is:
    At skill 583 - the actual rate is somewhere between 0.45% and 2.25%
    At skill 450 - the actual rate is somewhere between 0% and 1.13%
    The ranges are so large, it's borderline meaningless, and only 1000 samples at each is cleary insufficient from a statistical point of view.
  20. ARCHIVED-Valdaglerion Guest

    Xalmat wrote:
    The thing I hate about T9 harvesting is the geometry. It's to the point where you have to climb up all the rocks and look in the crevices for nodes. Bunches of strangeness. Perhaps it was done on purpose to thwart botters but it sures makes it a pita for normal players too :(