LK /p3-4 Superchest Drops (The Search for Lo Is Over) ifra
- Thread starter Phosheez
- Start date
skiffcz
Active member
Re: LK /p3-4 Superchest Drops (The Search for Lo Is Over)
Not exactly qualified, but Ill try.
There are two possible ways of LK future (IMO). We know that seed truncation stays (you found the KO match), so:
1, The patterns depends on current rune rarity changes and with trunc seed, HRs still be "plenty" but with recalculated patterns. That can mean anything from 0 to 5 or such.
2, The patterns are somehow hardcoded or hardcode related or just their code is somewhere in dark corner which wont be touched by patch (not unlikely) and theyll stay the same.
Now, I dont know if KO has changed rarity? Didnt they raised runes Mal and higher or such (Im lazy to search for it). On the other hand, changed runes in other TCs can indirectly change rarity of all other things, including other runes.
Thats educated eh? TLDR - I try to sound smart, but I have no idea what this means. The only relatively sure thing is that truncated seed for LK chests stays. Thats good news in any case.
EDIT: pwnt by Pijus on the Mal subject
Not exactly qualified, but Ill try.
There are two possible ways of LK future (IMO). We know that seed truncation stays (you found the KO match), so:
1, The patterns depends on current rune rarity changes and with trunc seed, HRs still be "plenty" but with recalculated patterns. That can mean anything from 0 to 5 or such.
2, The patterns are somehow hardcoded or hardcode related or just their code is somewhere in dark corner which wont be touched by patch (not unlikely) and theyll stay the same.
Now, I dont know if KO has changed rarity? Didnt they raised runes Mal and higher or such (Im lazy to search for it). On the other hand, changed runes in other TCs can indirectly change rarity of all other things, including other runes.
Thats educated eh? TLDR - I try to sound smart, but I have no idea what this means. The only relatively sure thing is that truncated seed for LK chests stays. Thats good news in any case.
EDIT: pwnt by Pijus on the Mal subject
Re: LK /p3-4 Superchest Drops (The Search for Lo Is Over)
It's true the rarity of Ko runes haven't changed. My feeling, though, was that if one thing changed, everything would change. Of course, not understanding how the code works at all, that could easily be wrong. I guess the jury is still out.
It's true the rarity of Ko runes haven't changed. My feeling, though, was that if one thing changed, everything would change. Of course, not understanding how the code works at all, that could easily be wrong. I guess the jury is still out.
skiffcz
Active member
Re: LK /p3-4 Superchest Drops (The Search for Lo Is Over)
Mine too. The change will be small though and the problem with truncatec seeds is that the information could have been easily rounded or left out ... OHM and like with 3 or 2 times greater probability on the other hand ....
In any case, Im pretty guessing here, dont take me seriously
Mine too. The change will be small though and the problem with truncatec seeds is that the information could have been easily rounded or left out ... OHM and like with 3 or 2 times greater probability on the other hand ....
In any case, Im pretty guessing here, dont take me seriously
Pyrohemia1
New member
Re: LK /p3-4 Superchest Drops (The Search for Lo Is Over)
The rune drop process is as follows:
Start with the area's / monster's rune tier. For LK this is Runes 15.
Add up the total probabilities for that tier: Sur: 3, Ber: 2, drop to Runes 14: 11766 (for 1.13) or 27011 (for 1.12a).
Roll a random number between 1 (indexing could actually start at 0, I don't know) and the total, 11771 or 27016, and perform the corresponding action.
If 1,2, or 3 were rolled then drop Sur. If 4, or 5 were rolled drop Ber, otherwise go to Runes 14 and perform the same procedure.
The truncated seed for LK means that there are 65536 unique series of random numbers that are used for item drops. (I believe that there is a second random number generator with an unaffected seed for choosing the item affixes which is why the item stats are different even on the same drop pattern.)
When taking a given series of random numbers the only drop changes with regards to LK are the high rune probabilities. The only way that the drop patterns will change is when something different from 1.12a happens. This is when a high rune does actually drop, because otherwise the drop program just goes to a lower tier like normal. The lower rune tiers are unchanged so the same random numbers being applied to the same probabilities will result in the same actions.
Let us say that for the tier Runes 15, the random number generator produces 68826836. (A 32-bit number because the random number generator is 32-bit even though the seed number was truncated to 16-bits by the LK drop code.)
In 1.12a:
68826836 modulo 27016 = 17 084
This is not 1,2,3,4, or 5, thus the program drops to Runes 14.
In 1.13:
68826836 modulo 11771 = 1 799
This is also not 1,2,3,4, or 5, thus the program also drops to Runes 14.
In both cases Runes 14 gets the same next random number from the random number generator, and is likely to also act in exactly the same way in both versions.
Given the high chance of dropping to a lower tier it is likely that the drop program still falls below the revised drop probabilities and then acts in the same way as 1.12a.
Summary: I expect that for any lower rune patterns that do exist in 1.13 they will be identical to 1.12a. Not all of the patterns from 1.12a may exist in 1.13 as some of them might instead drop a high rune which would alter the pattern. Also, the high rune patterns from 1.12a are very likely to no longer exist in 1.13 because of the change in probabilities.
(It is possible for the patterns to remain unchanged, however, because 5 is a 32-bit number too and if the drop program received the number 5 on tier Runes 15 it will always drop Ber whether in 1.12a or 1.13. The 16-bit seed makes this highly unlikely to actually occur.)
The rune drop process is as follows:
Start with the area's / monster's rune tier. For LK this is Runes 15.
Add up the total probabilities for that tier: Sur: 3, Ber: 2, drop to Runes 14: 11766 (for 1.13) or 27011 (for 1.12a).
Roll a random number between 1 (indexing could actually start at 0, I don't know) and the total, 11771 or 27016, and perform the corresponding action.
If 1,2, or 3 were rolled then drop Sur. If 4, or 5 were rolled drop Ber, otherwise go to Runes 14 and perform the same procedure.
The truncated seed for LK means that there are 65536 unique series of random numbers that are used for item drops. (I believe that there is a second random number generator with an unaffected seed for choosing the item affixes which is why the item stats are different even on the same drop pattern.)
When taking a given series of random numbers the only drop changes with regards to LK are the high rune probabilities. The only way that the drop patterns will change is when something different from 1.12a happens. This is when a high rune does actually drop, because otherwise the drop program just goes to a lower tier like normal. The lower rune tiers are unchanged so the same random numbers being applied to the same probabilities will result in the same actions.
Let us say that for the tier Runes 15, the random number generator produces 68826836. (A 32-bit number because the random number generator is 32-bit even though the seed number was truncated to 16-bits by the LK drop code.)
In 1.12a:
68826836 modulo 27016 = 17 084
This is not 1,2,3,4, or 5, thus the program drops to Runes 14.
In 1.13:
68826836 modulo 11771 = 1 799
This is also not 1,2,3,4, or 5, thus the program also drops to Runes 14.
In both cases Runes 14 gets the same next random number from the random number generator, and is likely to also act in exactly the same way in both versions.
Given the high chance of dropping to a lower tier it is likely that the drop program still falls below the revised drop probabilities and then acts in the same way as 1.12a.
Summary: I expect that for any lower rune patterns that do exist in 1.13 they will be identical to 1.12a. Not all of the patterns from 1.12a may exist in 1.13 as some of them might instead drop a high rune which would alter the pattern. Also, the high rune patterns from 1.12a are very likely to no longer exist in 1.13 because of the change in probabilities.
(It is possible for the patterns to remain unchanged, however, because 5 is a 32-bit number too and if the drop program received the number 5 on tier Runes 15 it will always drop Ber whether in 1.12a or 1.13. The 16-bit seed makes this highly unlikely to actually occur.)
jjscud
Active member
Re: LK /p3-4 Superchest Drops (The Search for Lo Is Over)
I agree. Once the roll has gotten past the changed runes, then the patters should be identical, except that a few more will likely result in HRs and those will be missed. Though its likely that most of those will come from lower runes (because there are more) and we'll never notice.
Now on to more concrete evidence.
I have about 10 Fal - Um matches that are identical to existing drops.
I have also found that there is a NEW Vex drop at p7/8 which corroborates all this.
In fact, look for a thread in a few more minutes.
I agree. Once the roll has gotten past the changed runes, then the patters should be identical, except that a few more will likely result in HRs and those will be missed. Though its likely that most of those will come from lower runes (because there are more) and we'll never notice.
Now on to more concrete evidence.
I have about 10 Fal - Um matches that are identical to existing drops.
I have also found that there is a NEW Vex drop at p7/8 which corroborates all this.
In fact, look for a thread in a few more minutes.
