Latest Diablo 3 News
DiabloWiki Updates
Support the site! Become a Diablo: IncGamers PAL - Remove ads and more!

1.09 superchests?

Discussion in 'Single Player Forum' started by pharphis, Jul 6, 2016.

  1. pharphis

    pharphis IncGamers Site Pal

    Joined:
    Aug 1, 2011
    Messages:
    13,594
    Likes Received:
    1,546
    Trophy Points:
    283
    For whatever version of 1.09 allows up to /p64... Does anyone know if /p9-10 functions differently than /p7-8? What about /p30? /p64?

    I would love to know if it's plausible to get a zod in 1.09 through A5 superchests and think that this would be the way to more or less guarantee a pattern exists, if plausible. Cmon code junkies.

    Also do 1.09 patches differ for superchest drops between one another?

    I think this is one of the last conquests for D2 vets here: finding a reproducible means of getting an older version zod rune since iirc they function differently.

    Of course, this isn't very plausible outside of simulating runs with bots or whatever but it would be incredible to know!

    Of course I'm also interested in 1.07 and 1.08 (even more due to crafting junk from chests and charms) but 1.09 could in theory allow a lot of possible drops.
     
  2. RadTang

    RadTang IncGamers Member

    Joined:
    Mar 8, 2016
    Messages:
    489
    Likes Received:
    64
    Trophy Points:
    107
    P64.. I would check out 1.09 mostly for this reason ! Summonmancers are terrible as far as I have read / heard. No synergies - a lot of possibilities, much more than in 1.10+. Thankfully, it's easy to play older versions - and one does not even have to reinstall the game. This (forbidden here) addon allows to play as early as 1.07 LOD. No Classic.
     
  3. Brak

    Brak IncGamers Member

    Joined:
    Aug 19, 2006
    Messages:
    1,948
    Likes Received:
    674
    Trophy Points:
    233
    I thought drop mechanics dont change after p78, just exp and monster stats

    could be wrong
     
  4. pharphis

    pharphis IncGamers Site Pal

    Joined:
    Aug 1, 2011
    Messages:
    13,594
    Likes Received:
    1,546
    Trophy Points:
    283
    I'm not sure if patterns are tied to no drop chance, or if the no drop chance is already 0 at p8, but I wasn't sure where to find accurate info on no drop chance if that's relevant anyway
     
  5. Brak

    Brak IncGamers Member

    Joined:
    Aug 19, 2006
    Messages:
    1,948
    Likes Received:
    674
    Trophy Points:
    233
    jjscud probably already figured it out. Maybe forum search him and 1.09
     
    Pb_pal likes this.
  6. pharphis

    pharphis IncGamers Site Pal

    Joined:
    Aug 1, 2011
    Messages:
    13,594
    Likes Received:
    1,546
    Trophy Points:
    283
    From the 1.13 beta superchest thread (where he covers his simulations) I saw mention that he helped with 1.12 and 1.13 and something else (edit: see quote below). I don't think 1.09 was one of them. He didn't comment on that post directly so I don't see much reason to doubt the validity considering he was active in the thread.

    Here is the quote:

     
  7. helvete

    helvete IncGamers Member

    Joined:
    Jul 17, 2004
    Messages:
    6,805
    Likes Received:
    907
    Trophy Points:
    534
    I've only played around 60 minutes of 1.09b, but from what I can tell, the nodrop chance from monsters continues to decrease as players settings go past 7/8 and even 19/20. This to me is an indication that superchests would be at least slightly different, since they would get closer and closer to always having 4 items per drop cycle as players setting increases.

    I really don't think we have the resources here to get, what? 300k runs in for each of the 32 possible players settings for each of the 3 acts which might produce some good results. It's 1.09 though, so no cubing. Might be worth to check with some boring but efficient jjscud type method.
     
  8. pharphis

    pharphis IncGamers Site Pal

    Joined:
    Aug 1, 2011
    Messages:
    13,594
    Likes Received:
    1,546
    Trophy Points:
    283
    I did some googling yesterday and saw that no drop chance is like 1% at p8, so I imagine the formula drops that lower for higher p settings (barring other weird caps or operators for rounding) but I'm not sure if that is the reason itself that patterns vary.

    I think simulation with bots is the only feasible means with a community this small (and only 1 chest at a time for A5, the only act I'd say is worth checking for this specific goal) though I'm willing to contribute in 1.07 (I've done ~2k runs in the past with nothing beyond mid runes which is unsurprising). If only jjscud were around
     
  9. helvete

    helvete IncGamers Member

    Joined:
    Jul 17, 2004
    Messages:
    6,805
    Likes Received:
    907
    Trophy Points:
    534
    Well, I tested it roughly.

    Scenario 1: 1.09b players8 normal, first area
    20 monsters killed, 16 of them dropped items. Had a streak of 3 monsters not dropping anything.

    Scenario 2: 1.09b players20 normal, first/second areas
    50 monsters killed, 49 of them dropped items. Had to wait until monster #39 to see a nodrop.

    I realize this is very rough, but it at least FEELS like there is a noticable change in nodrop chances past players8.
     
  10. pharphis

    pharphis IncGamers Site Pal

    Joined:
    Aug 1, 2011
    Messages:
    13,594
    Likes Received:
    1,546
    Trophy Points:
    283
    Ah, found the formula here:
    http://diablo2.diablowiki.net/Item_Generation_Tutorial

    For monsters:
    NewNoDrop=int( ProbSum/(1/((NoDrop/(NoDrop+ProbSum))^N)-1) ), where N=int (1+AdditionalPlayers/2+ClosePartiedPlayers/2), AdditionalPlayers — the number of all additional players in the game, where the monster is killed, ClosePartiedPlayers — the number of players in your party near you (not far than two screens away), ProbSum=Prob1+Prob2+Prob3+...+Prob10, int — operation of truncating fractional part of a number.

    In SP the Players X command sets the number of NON-partied players, e.g. /players8 command sets N=4.

    For example: a character kills a non-boss Devilkin in The Pit level 1 in Hell in /players8 mode (or in a full game on Battle.net). The Devilkin has Mlvl=85. In monstats.txt file Devilkin is figured under "fallen3" name. Non-unique fallen3 drop items from "Act 1 (H) H2H B" TC in Hell difficulty. In TreasureClassEx.txt file "Act 1 (H) H2H B" TC has group=7 and level=66. Because Mlvl=85 the TC is upgraded to "Act 5 (H) H2H C" TC, which has level=85 (less or equal to mlvl). So Devilkins in The Pit drop items from "Act 5 (H) H2H C" TC.

    It has 1 pick from following possibilities:

    • NoDrop (100);
    • gld (21);
    • Act 5 (H) Equip C (16);
    • Act 1 (H) Junk (21);
    • Act 5 (H) Good (2).
    The numbers in (parenthesis) are probabilities.

    Lets calculate NewNoDrop:

    • ProbSum=21+16+21+2=60.
    • NoDrop=100.
    • N=4.
    • NewNoDrop=int( 60/(1/((100/(100+60))^4)-1) )=11.
    So The Pit Devilkin has fallowing chances:

    • 11/71 for dropping nothing
    • 21/71 for dropping gold
    • 16/71 for dropping from Act 5 (H) Equip C
    • 21/71 for dropping from Act 1 (H) Junk
    • 2/71 for dropping from Act 5 (H) Good

    So someone could plug the numbers in to see at what p setting this stops changing no drop chance. That said, how does this work for chests, and is the no drop chance what actually affects the patterns available?
     
  11. Brak

    Brak IncGamers Member

    Joined:
    Aug 19, 2006
    Messages:
    1,948
    Likes Received:
    674
    Trophy Points:
    233
    Important thing to remember is that no drop really has nothing to do with Superchest high rune patterns pre-1.13. It's just which truncation won the high rune lottery. From the sounds of it, there should be 32 unique 65k sets. So about 2.1 million patterns. If I'm correct that pre 1.13 Zods are 1 in 3,000,000, there's a decent chance of 1 Zod pattern buried in there somewhere. But by the time you find it, you could farm numerous ones in 1.13 I'd imagine.
     
  12. pharphis

    pharphis IncGamers Site Pal

    Joined:
    Aug 1, 2011
    Messages:
    13,594
    Likes Received:
    1,546
    Trophy Points:
    283
    1.09 zods are different, though. From what I've heard they remove durability on an item altogether (not atma bugging) so that you can make any socketed ethereal item indestructible even after removing the zod. Mostly just a pvp thing, but interesting nonetheless.

    That said, I suppose the socketing of the item would have to take place before 1.10 as well...?
     
  13. Brak

    Brak IncGamers Member

    Joined:
    Aug 19, 2006
    Messages:
    1,948
    Likes Received:
    674
    Trophy Points:
    233
    So you found a new loophole to not-ATMA bug items
     
  14. helvete

    helvete IncGamers Member

    Joined:
    Jul 17, 2004
    Messages:
    6,805
    Likes Received:
    907
    Trophy Points:
    534
    I *believe* that the zod was changed only realm-side in 1.10 (as the tp+hel recipe was introduced with 1.10) so you'd still be able to cube a zod and remove the durability completely as long as you do it before 1.11, I don't remember there being sub-versions of 1.10 except for the betas.

    IIRC there were items in 1.10 ladder which had been zodbugged, most noteably a high end HoZ upped, zodded, then hel'd and ber'd.
     
  15. pharphis

    pharphis IncGamers Site Pal

    Joined:
    Aug 1, 2011
    Messages:
    13,594
    Likes Received:
    1,546
    Trophy Points:
    283
    If only eth jmod were easy to find... :p

    It would be great to get confirmation on when/how zod was changed for single player, but I suppose it could be tested simply enough once one is found
     
  16. helvete

    helvete IncGamers Member

    Joined:
    Jul 17, 2004
    Messages:
    6,805
    Likes Received:
    907
    Trophy Points:
    534
    It can be tested by other methods, but I don't want to do that myself.
     
  17. Brak

    Brak IncGamers Member

    Joined:
    Aug 19, 2006
    Messages:
    1,948
    Likes Received:
    674
    Trophy Points:
    233
    Don't remember, does the 3PGem recipe maintain ethereal status?
     
  18. helvete

    helvete IncGamers Member

    Joined:
    Jul 17, 2004
    Messages:
    6,805
    Likes Received:
    907
    Trophy Points:
    534
  19. Treeharl

    Treeharl IncGamers Member

    Joined:
    May 22, 2016
    Messages:
    488
    Likes Received:
    221
    Trophy Points:
    120
    In this case there are three scenarios possible (I am totally unsure what scenario is true)

    1. (Most likely) players 9-64 function the same way as players 7/8 and as such all patterns are the same.
    2. (Less likely) players 9-64 have 65536 patterns but players 9 and players 48 result in the same 65536 patterns it is different from 7/8.
    3. (Less likely) players 9/10 have different patterns as players 7/8 as players 63/64. If this scenario is true there are over 2 million of patterns in total.

    Also the expected value of a Zod in a superchest on 1.09 is 0.051 which tralslates in to the fact that on a given playersetting there is a 2-3% chance of a Zod pattern on act 5(not 5% as you would think!)
     
  20. pharphis

    pharphis IncGamers Site Pal

    Joined:
    Aug 1, 2011
    Messages:
    13,594
    Likes Received:
    1,546
    Trophy Points:
    283
    What reasoning did you use to come up with scenario 2?
     

Share This Page