Some primary drop statistics

[Jarulf]

I have updated my program that calculate drop statistics for 1.10. I am still not sure it works correctly though or if there are bugs. I am also not fully sure I have implemented it all correctly, I might have overlooked something. So if anyone can spot errors or strange results, please tell me. Also, if anything else should added or changed, tell, some things can be changed very easilly. For example, the item types list for monsters can behave any of the types from itemtypes.txt.

Anyway, the data exist as text files (and as zip file). For specific monsters, I have mostly calculated only for act bosses and some specific other super unique monsters. It is harder to do it for random monsters with the new system to update for level. The program handles it all correctly, but WHAT map should be assigned to a monster? I would need a generic system, not having specific tables set up for each one.

For the countess mondrop file, the values are NOT completely correct, it doesn't factor in the max 6 items. My program doesn't correctly handle it when the pick number is negative. it do handle it when the pick value is positive though.

The magprob files should handle all possible entries for different unqiue/set/rare/magic values in treassureclassex.txt (it disregards andariles minimal value difference and uses nightmare values for the countess.

The files:

http://www.bolina.hsb.se/hsidor/pedro/magprob.zip
http://www.bolina.hsb.se/hsidor/pedro/mondrop.zip

Individual files (for duriel it is the sub base TC):

http://www.bolina.hsb.se/hsidor/pedro/magprob-norm.txt
http://www.bolina.hsb.se/hsidor/pedro/magp...gprob-champ.txt
http://www.bolina.hsb.se/hsidor/pedro/magprob-boss.txt
http://www.bolina.hsb.se/hsidor/pedro/magprob-actq.txt
http://www.bolina.hsb.se/hsidor/pedro/magp...gprob-super.txt
http://www.bolina.hsb.se/hsidor/pedro/magp...prob-superx.txt
http://www.bolina.hsb.se/hsidor/pedro/magp...rob-council.txt
http://www.bolina.hsb.se/hsidor/pedro/magp...ob-countess.txt
http://www.bolina.hsb.se/hsidor/pedro/magp...ob-hephasto.txt
http://www.bolina.hsb.se/hsidor/pedro/magp...ob-rad-nihl.txt
http://www.bolina.hsb.se/hsidor/pedro/magp...ob-summoner.txt

http://www.bolina.hsb.se/hsidor/pedro/mond...op-andariel.txt
http://www.bolina.hsb.se/hsidor/pedro/mond...p-andarielq.txt
http://www.bolina.hsb.se/hsidor/pedro/mondrop-baal.txt
http://www.bolina.hsb.se/hsidor/pedro/mond...ndrop-baalq.txt
http://www.bolina.hsb.se/hsidor/pedro/mond...drop-diablo.txt
http://www.bolina.hsb.se/hsidor/pedro/mond...rop-diabloq.txt
http://www.bolina.hsb.se/hsidor/pedro/mond...drop-duriel.txt
http://www.bolina.hsb.se/hsidor/pedro/mond...rop-durielq.txt
http://www.bolina.hsb.se/hsidor/pedro/mond...op-mephisto.txt
http://www.bolina.hsb.se/hsidor/pedro/mond...p-mephistoq.txt
http://www.bolina.hsb.se/hsidor/pedro/mond...p-nhilathak.txt
http://www.bolina.hsb.se/hsidor/pedro/mond...-pindleskin.txt
http://www.bolina.hsb.se/hsidor/pedro/mond...thecountess.txt

I might add additional files to the zip.

Comments?

[HunkyDory]

Pedro,

Can you make one for Eldritch please?


Regards, Hunky

[Jarulf]

Hmm, he has no special TC, drops from base TC: Act 5 Super A.

I uploaded the file:

http://www.bolina.hsb.se/hsidor/pedro/mond...herectifier.txt

Please note that I have not verified the file, so can't really tell if the data is correct or not.

[HunkyDory]

Thanks Pedro,

Your work is appreciated.


Regards, Hunky

[whereagles]

This is for each individual drop, right? Individual as in a normal monster gets 1 pick, a champ 2 picks, act boss 6 picks, etc.

[Jarulf]

This is for each individual drop, right? Individual as in a normal monster gets 1 pick, a champ 2 picks, act boss 6 picks, etc.

Yes, it is "per kill" so to speak, not an "per item" basis. In many cases it is not even possible to calculate on a per item (even disregarding the no drop) since it would vary with different items (at least for negative pick values).

[whereagles]

Actually, I meant it the other way around, the charts applying to "per pick".

So, the charts are "per kill", what does it mean, for instance, the council file at 0% mf:

ilvl-qlvl unique set rare magic superior normal low
-20 0.652% 2.672% 4.018% 92.659% 0.000% 0.000% 0.000%

Does it mean that you have a 0.652% chance that he drops at least one unique item?

[Jarulf]

Ehh, I think it depend on what file you refer to. The "magprob" ones are for a specific item (that CAN be magic) and the chance the game would decide to pick a specific quality. It doesn't mean you actually DO get the quality, for example, if there is no set item of the tpye, (or too high level), it won't drop but it was still randomly decided intially to be a set.

The "mondrop" files are "per kill" though. Giving the over all ammount of each quality from a kill of the monster, factoring in all possible rules the game use, actually calculating on a per otem basis the possibility of each quality and adding it up. Failed uniques are specified under rare for example, all jewlery that would have been superior, normal or low quality will be moved to magic since jewlery has to be at least magic and so on.

[whereagles]

I'm afraid I didn't understand that too well B)

Suppose I have a drop from a champion monster. Its default drop is (I think) 2 potions, 2 magic items. Let's say those 2 magic items are a short sword and a long sword. Each sword will have the probability shown on magdrop.txt (which depends on ilvl, qlvl, mf) of being unique/set/rare. Right?

What about the mondrop files? Baal drops 6 items (I think), so what does mondrop say about that drop?

[Jarulf]

Suppose I have a drop from a champion monster. Its default drop is (I think) 2 potions, 2 magic items. Let's say those 2 magic items are a
>short sword and a long sword. Each sword will have the probability shown on magdrop.txt (which depends on ilvl, qlvl, mf) of being
>unique/set/rare. Right?


Yes, that file will have tables where you can check the probabilities of those sword to be of a specific quality. Note that as I said, there can be slight actual changes, lets suppose the unique versions of the sword has such a high level so that a specific monster can't drop it, then of course the final chance for a unique sword is 0%. However, the chance that the game tried to make it unique will be as stated in that file (in that case, it will be a failed rare with higher durability by the way).

Also, lets assume it says the chance for a unique is 1%. That does not mean 1% of all items will be unique, it mean that 1% of those that could be unqiue (diregarding potions, keys and so on) will be unique.


>What about the mondrop files? Baal drops 6 items (I think), so what does mondrop say about that drop?

The first table will list the ammount of items that will be of a specific type. For example, out of ALL items droped by Baal, almost 10% will be class specific ones. A bit more than 7% will be gold and so on (that table is actually on a "per item" basis.

Next comes a table similary to the one in magprob. But here it WILL factor in that some items (for example potions) can't be magical in any form (they will end up under "other". As you see, the sum will always be 100%. These values will typically be lower than those in magprob of course. For example the act ,ag drop file lists values for unique in the 5-6% range (with no MF) while the actual ammount will be arround 4-5% as seen in mondrop. The difference is mainly in the 20% "other" that is not accounted for in magprob, plus limited things such as failed uqniues.

The final tables for which runes, uniques and set that can drop should be self explaining.

[whereagles]

Allright! I see it all now. Obrigadíssimo! (I take it understand this one, no? B) )

So let me make a corollary:

At 400% mf, champions have around 2,9% chance to drop a unique. Since they have 2 picks, they'll drop at least 1 unique around 5,7% of the time.

Compare with Mephisto. At 400 mf, Meph has around 5% to drop a unique. On his 6 picks, this means around 26,5% chance to drop at least one unique.

So, the two "item farming" possibilities (kill Meph over and over, or look for champs/bosses across the map), seem to be quite similar in effectiveness. Neither is particularly "better" than the other: the time it takes to kill Meph is more or less the time needed to find and kill 7-8 champs/bosses.

[Nastie_Bowie]

This thread is up as a news item.

Nice work, Jarulf.

NB :rolleyes:

[IKKE]

What exactly is the differnce between Mondrop and Magdrop?


and whats the diff between theese ?
http://www.bolina.hsb.se/hsidor/pedro/mo...dariel.txt
http://www.bolina.hsb.se/hsidor/pedro/mo...arielq.txt

plz help ? :D

[Jarulf]

What exactly is the differnce between Mondrop and Magdrop?


and whats the diff between theese ?
http://www.bolina.hsb.se/hsidor/pedro/mo...dariel.txt
http://www.bolina.hsb.se/hsidor/pedro/mo...arielq.txt

plz help ? :D

Magprob simply gives the chance for the game to pick a spcific quality for an item. It is sort of generic tables, not really specifying the item or the monster (there is a level difference column for it though). It is based on the different TC types that exists.

The mondrop are really for specific monsters and will calculate for each specific item the monster can drop factoring in everything like for example a ring can't be superior in quality (even if the game at first might pick it with the chance calculated as in magprob). It also holds tables detailing what distribution of item types it can drop and what unique, set and runes.

The difference between the two are that the one with a q in the end is for the quest drop.

[MrBunnyBan]

The links don't seem to work at the moment. Has the programme been taken offline?