new mf thread
- Thread starter Byaxident
- Start date
Byaxident
Diabloii.Net Member
Hmm ya of course failed uniques are great. But getting a bunch of 2x durability martels,bhs,exes isnt fun. Id rather have 190ed rare ones.
it looks like 222 mf is too much for meph in classic, i guess ill just experiment once i get my hands on items similar to the ones i lost in my xfer.
PS: I have the same problem with my mf barb in cs; he has 250 mf and doesnt get as much rares from cs superuniques (-diablo) as my NL 300+ mf does. But he has too much mf for diablo, getting a lot of uniques and failed set items. sigh
it looks like 222 mf is too much for meph in classic, i guess ill just experiment once i get my hands on items similar to the ones i lost in my xfer.
PS: I have the same problem with my mf barb in cs; he has 250 mf and doesnt get as much rares from cs superuniques (-diablo) as my NL 300+ mf does. But he has too much mf for diablo, getting a lot of uniques and failed set items. sigh
MF has nothing to do with stat generation, so there's no reason not to get more MF, short of becoming an ineffective character, in my opinion.
MF also has nothing to do with the base item generated, only the enemy you kill does. Don't be discouraged because of what seems to be a string of rare and unique Broad Swords with more MF. Eventually one of those rare/unique drops will be on a Battle Hammer. There is nothing you can do to effect the base item dropped short of enemy selection. It's not too much MF because you're finding crappy uniques, it's simply the boss dropping the wrong type of unique item.
MF also has nothing to do with the base item generated, only the enemy you kill does. Don't be discouraged because of what seems to be a string of rare and unique Broad Swords with more MF. Eventually one of those rare/unique drops will be on a Battle Hammer. There is nothing you can do to effect the base item dropped short of enemy selection. It's not too much MF because you're finding crappy uniques, it's simply the boss dropping the wrong type of unique item.
Byaxident
Diabloii.Net Member
Hmm It looks like you misunderstood what I was saying. :scratch:
failed set items: the game decides that the item that drops is going to be a set item, it checks but sees no such item exists (as you know theres no such thing as a set battle hammer or martel) so it drops a magic item with twice the normal durability instead.
failed unique: similar thing; ex: the game decides that a martel is going to drop, it then decides that it is going to be unique, since unique martels dont exist in classic, it drops a RARE martel with 3 times the normal durability.
With 222 mf:
Uniques:+117% Sets:+153% Rare:+162% Magic:+222%
In about 50 meph runs with that char, 3 martels dropped: 2 failed set (magic with 100 dur) and 1 failed unique ( rare with 150 dur, of course I dont mind about those). So apparently in this situation 222 is too much, in classic anyway.
EDIT:
a typical meph drop for my sorc, result of too much mf.
Can you guess what the durability was on that death mask? Youre right, 60, 3x the normal dur. :sad2:
failed set items: the game decides that the item that drops is going to be a set item, it checks but sees no such item exists (as you know theres no such thing as a set battle hammer or martel) so it drops a magic item with twice the normal durability instead.
failed unique: similar thing; ex: the game decides that a martel is going to drop, it then decides that it is going to be unique, since unique martels dont exist in classic, it drops a RARE martel with 3 times the normal durability.
With 222 mf:
Uniques:+117% Sets:+153% Rare:+162% Magic:+222%
In about 50 meph runs with that char, 3 martels dropped: 2 failed set (magic with 100 dur) and 1 failed unique ( rare with 150 dur, of course I dont mind about those). So apparently in this situation 222 is too much, in classic anyway.
EDIT:
a typical meph drop for my sorc, result of too much mf.
Can you guess what the durability was on that death mask? Youre right, 60, 3x the normal dur. :sad2:
I have done Meph a few times rescently, teleporting to him with shard and magefists and then switching to gull, 40MF chancies, 30Mf nagel, 3Ptopaz mask and 25MF boots for the killing, but I don't think I have had such a, what seems to me like, good drop. he usually gives me 1 unique tops, mebbe 2 rares. And this is in relatively full games. But, this was on NL.
The plot thickens ...
The plot thickens ...
norvi11erogers
Diabloii.Net Member
q: have you finished the quest, or do you join say cs games to kill him?Byaxident said:Hmm It looks like you misunderstood what I was saying. :scratch:
failed set items: the game decides that the item that drops is going to be a set item, it checks but sees no such item exists (as you know theres no such thing as a set battle hammer or martel) so it drops a magic item with twice the normal durability instead.
failed unique: similar thing; ex: the game decides that a martel is going to drop, it then decides that it is going to be unique, since unique martels dont exist in classic, it drops a RARE martel with 3 times the normal durability.
With 222 mf:
Uniques:+117% Sets:+153% Rare:+162% Magic:+222%
In about 50 meph runs with that char, 3 martels dropped: 2 failed set (magic with 100 dur) and 1 failed unique ( rare with 150 dur, of course I dont mind about those). So apparently in this situation 222 is too much, in classic anyway.
EDIT:
a typical meph drop for my sorc, result of too much mf.
Can you guess what the durability was on that death mask? Youre right, 60, 3x the normal dur. :sad2:
I was going to work out an exact relation of meph dropping stuff, but it seems to be dependent on the item type level. I may still do so for say a bhammer, but i'm at work right now...
I understood what you meant perfectly, Byaxident. Diablo was dropping a bunch of items that were initially rolling to be set or unique, whether they could or couldn't be. This is exactly what you want. The more he drops uniques, the better the chance the exceptional weapons he drops try to roll unique. How is that too much MF? The one thing I've never understood is the idea of "too much MF." How can more rare and unique drops combined be bad?
The only reason I can think of to ever limit your MF is on virgin boss kills when wanting rare jewelry.
The only reason I can think of to ever limit your MF is on virgin boss kills when wanting rare jewelry.
norvi11erogers
Diabloii.Net Member
He's saying that the diminishing returns kill unique drops faster than they do set.
Assume one only wants to find exceptional rares, maybe even certain exceptional rares. Knowing that once an item type is decided upon, items attempt to drop in the order unique, set, rare, magic, one could imagine a scenario wherein adding some amount of magic find barely adds to the chance to find a unique and adds slightly more to the chance for a set. If magic find is so high that, if a unique or set fails to drop, an a rare almost always drops, then one can easily imagine a scenario where the number of exceptional rares goes down with the addition of magic find.
There are other scenarios that might make this true as well, this is just one I thought of off the top of my head.
If this is true in practice, there might be a "sweet spot" for finding, say, rare battle hammers. Sweet spots for other items might be at other locations. In any case, the actual results would require working through the numbers, which I may be curious enough to do...
Assume one only wants to find exceptional rares, maybe even certain exceptional rares. Knowing that once an item type is decided upon, items attempt to drop in the order unique, set, rare, magic, one could imagine a scenario wherein adding some amount of magic find barely adds to the chance to find a unique and adds slightly more to the chance for a set. If magic find is so high that, if a unique or set fails to drop, an a rare almost always drops, then one can easily imagine a scenario where the number of exceptional rares goes down with the addition of magic find.
There are other scenarios that might make this true as well, this is just one I thought of off the top of my head.
If this is true in practice, there might be a "sweet spot" for finding, say, rare battle hammers. Sweet spots for other items might be at other locations. In any case, the actual results would require working through the numbers, which I may be curious enough to do...
My understanding of how MF works is as follows. I'll use theoretical base numbers, since I don't know the real base chances for item types to be dropped.
Let's say hell Diablo, by default, has a 200 out of 10000 chance to drop a unique, 500 out of 10000 for a set item, 500 out of 10000 chance for a rare, and the rest (8800) out of 10000 for a magical item. The server rolls a few random-10000's. 1-8800 is magical, 8801-9300 is rare, 9301-9800 is set, and 9801-10000 is unique.
You get 222 MF for Uniques:+117% Sets:+153% Rare:+162% Magic:+222%.
This changes the chances to 434 out of 10000 unique, 1265 out of 10000 for set, 1310 out of 1000 for rare, and the rest (6991) out of 10000 for magic. The server rolls a few random-10000's. 1-6991 is magical, 6992-8301 for rare, 8302-9566 for set, 9567-10000 for unique.
Let's say a hell Flayer, by default, has 40 out of 10000 chance to drop a unique, 100 out of 10000 for a set, 100 out of 10000 for a rare, and 1000 out of 10000 for a magical item. The server rolls a few random-10000's. 1-8760 is a white item, 8761-9760 is magical, 9761-9860 is rare, 9861-9960 is set, 9961-10000 is unique.
You get 222 MF for Uniques:+117% Sets:+153% Rare:+162% Magic:+222%.
This changes the chances to 87 out of 10000 unique, 253 out of 10000 for set, 262 out of 1000 for rare, and 3220 out of 10000 for a magical item. The server rolls a few random-10000's. 1-6178 is a white item, 6179-9398 is magical, 9399-9660 is rare, 9661-9913 is set, 9914-10000 is unique.
If something like Arrows is selected to drop before rolling the random-10000, it either skips it or gets defaulted to white, anyway. If something that can't be unique drops, that random range turns into a rare with triple durability. If something that can't be set drops, that random range turns into a magical item with double durability.
The idea of "sweet spots" sounds like superstition and only noticing certain streaks to me, especially are there are no hard numbers to back it up. The same deal as all the superstition around the random raid drops and streaks in World of Warcraft. If my assumption on how MF works it wrong, please correct me so that I can adjust my gear accordingly.
Let's say hell Diablo, by default, has a 200 out of 10000 chance to drop a unique, 500 out of 10000 for a set item, 500 out of 10000 chance for a rare, and the rest (8800) out of 10000 for a magical item. The server rolls a few random-10000's. 1-8800 is magical, 8801-9300 is rare, 9301-9800 is set, and 9801-10000 is unique.
You get 222 MF for Uniques:+117% Sets:+153% Rare:+162% Magic:+222%.
This changes the chances to 434 out of 10000 unique, 1265 out of 10000 for set, 1310 out of 1000 for rare, and the rest (6991) out of 10000 for magic. The server rolls a few random-10000's. 1-6991 is magical, 6992-8301 for rare, 8302-9566 for set, 9567-10000 for unique.
Let's say a hell Flayer, by default, has 40 out of 10000 chance to drop a unique, 100 out of 10000 for a set, 100 out of 10000 for a rare, and 1000 out of 10000 for a magical item. The server rolls a few random-10000's. 1-8760 is a white item, 8761-9760 is magical, 9761-9860 is rare, 9861-9960 is set, 9961-10000 is unique.
You get 222 MF for Uniques:+117% Sets:+153% Rare:+162% Magic:+222%.
This changes the chances to 87 out of 10000 unique, 253 out of 10000 for set, 262 out of 1000 for rare, and 3220 out of 10000 for a magical item. The server rolls a few random-10000's. 1-6178 is a white item, 6179-9398 is magical, 9399-9660 is rare, 9661-9913 is set, 9914-10000 is unique.
If something like Arrows is selected to drop before rolling the random-10000, it either skips it or gets defaulted to white, anyway. If something that can't be unique drops, that random range turns into a rare with triple durability. If something that can't be set drops, that random range turns into a magical item with double durability.
The idea of "sweet spots" sounds like superstition and only noticing certain streaks to me, especially are there are no hard numbers to back it up. The same deal as all the superstition around the random raid drops and streaks in World of Warcraft. If my assumption on how MF works it wrong, please correct me so that I can adjust my gear accordingly.
Byaxident
Diabloii.Net Member
Obviously, highest mf possible is good if you plan to kill a lot of normal monsters and champions which have a low chance of giving rares (very low in the case of normal monsters considering their no drop % is high). Im talking about mephisto however. He usually gives my sorc a couple rares even if she has no mf on her.
Many times ive seen sorcs in cs asking to get the last hit on D, saying they have 400+ mf - wearing tancs and the whole shabang - and then their blizz gets the last hit and poof, a bunch of blues(and perhaps some uniques). Im not looking for uniques.
The chances of getting a specific quality(rare,set,unique) in relation to a certain amount of mf is indeed dependant on the item's type (its qlvl to be more exact). A certain amount of mf might give the best chance of getting rare daggers, but a very poor chance of getting rare battle hammers; just an example obviously =o. Thats why Creon's hypothesis': "Let's say hell Diablo, by default, has a 200 out of 10000 chance to drop a unique, 500 out of 10000 for a set item, 500 out of 10000 chance for a rare, and the rest (8800) out of 10000 for a magical item." is flawed.
I guess for that reason its pretty pointless to try and find a sweetspot for ALL items at the same time but instead focus on items with very similar qlvls such as martels and execs for example. It also depends on the situation ie highest mf might be good for flayer runs or cs runs(except D) but really bad if all you ever kill is mephisto - unless of course you want uniques and set items, then it would be great.
Im sorry im quoting myself here: With 222 mf:
Uniques:+117% Sets:+153% Rare:+162% Magic:+222%
In about 50 meph runs with that char, 3 martels dropped: 2 failed set (magic with 100 dur) and 1 failed unique ( rare with 150 dur, of course I dont mind about those).
Logically, one might think that with slightly lower mf the 2 failed set martels couldve been plain rares and the failed unique couldve been a failed set.
I wouldve gotten 2 rares and 1 failed set with lets say 210 mf vs 2 failed set and 1 failed unique (rare with 3x dur). It seems pretty obvious to me that the first situation is better, especially in the long run. So this could very well be a situation where lower mf is better.
Im not 100% sure my mf is too high yet since from a statistical point of view, I havent killed him often enough(with the same amount of mf). Im pretty certain my claim is founded however since Ive gotten A LOT more failed sets than rares from him. Ive gotten a lot more failed set gscythes than rare ones eventhough they have the highest qlvl of all the weapons in classic.
Many times ive seen sorcs in cs asking to get the last hit on D, saying they have 400+ mf - wearing tancs and the whole shabang - and then their blizz gets the last hit and poof, a bunch of blues(and perhaps some uniques). Im not looking for uniques.
Shes a level 77 baron. Im doing solo runs only so im trying to find a solution for that specific situation, without + players bonuses.norvi11erogers said:
The chances of getting a specific quality(rare,set,unique) in relation to a certain amount of mf is indeed dependant on the item's type (its qlvl to be more exact). A certain amount of mf might give the best chance of getting rare daggers, but a very poor chance of getting rare battle hammers; just an example obviously =o. Thats why Creon's hypothesis': "Let's say hell Diablo, by default, has a 200 out of 10000 chance to drop a unique, 500 out of 10000 for a set item, 500 out of 10000 chance for a rare, and the rest (8800) out of 10000 for a magical item." is flawed.
I guess for that reason its pretty pointless to try and find a sweetspot for ALL items at the same time but instead focus on items with very similar qlvls such as martels and execs for example. It also depends on the situation ie highest mf might be good for flayer runs or cs runs(except D) but really bad if all you ever kill is mephisto - unless of course you want uniques and set items, then it would be great.
Im sorry im quoting myself here: With 222 mf:
Uniques:+117% Sets:+153% Rare:+162% Magic:+222%
In about 50 meph runs with that char, 3 martels dropped: 2 failed set (magic with 100 dur) and 1 failed unique ( rare with 150 dur, of course I dont mind about those).
Logically, one might think that with slightly lower mf the 2 failed set martels couldve been plain rares and the failed unique couldve been a failed set.
I wouldve gotten 2 rares and 1 failed set with lets say 210 mf vs 2 failed set and 1 failed unique (rare with 3x dur). It seems pretty obvious to me that the first situation is better, especially in the long run. So this could very well be a situation where lower mf is better.
Im not 100% sure my mf is too high yet since from a statistical point of view, I havent killed him often enough(with the same amount of mf). Im pretty certain my claim is founded however since Ive gotten A LOT more failed sets than rares from him. Ive gotten a lot more failed set gscythes than rare ones eventhough they have the highest qlvl of all the weapons in classic.
norvi11erogers
Diabloii.Net Member
While I agree that people tend to attribute meaning to slumps that are really not statistically significant, I don't think that it is really necessary to refuse the possibility for a "sweet spot," especially when your evidence has as much foundation in numbers as the original poster's.
Also, that is not exactly how item generation works, although it is sort of similar to that. Rather, the game tries to roll unique. If it succeeds, then it will give you a unique or a rare with 3x durability. If on the other hand, the unique roll fails, it will try to give you set. If the set roll succeeds, you will get a set item or a magic item with 2x durability. And so on for rare and magic. Check either hrus' guide to mf or warrior of light's item generation tutorial for more info.
So let's do this for a battle hammer dropped from Meph. Let's assume the values from the text files linked by Hru's mf guide for monster level, etc, are true for classic. Then it's just a blind application of the formula:
We know that
Effective MF=(MF*Factor)/(MF+Factor),
where Factor=250 for unique items, 500 for set items and 600 for rare items.
and the base chance, divisor, and minchance are:
unique: 400, 1, 6400
set: 160, 2, 5600
rare: 100, 2, 3200
quest: unique, set, rare
993,993,1024
normal: u,s,r
983,983,983
I have graphed the chance for a rare, given a bhammer (lvl 48) with meph's level at 87. As can be seen, the chance for quest drop rare given a battle hammer is dropped is monotonically decreasing, whereas that for normal drop is monotonically increasing. I fiddled with numbers, and this result appears to be true (all things also true) for many different levels for meph, even at or below the bhammer's level.
If you want to check these results, i can email the spreadsheet to you. Otherwise, it appears that if you are doing the quest drop, don't put mf on for exceptional rares. Otherwise, mf won't hurt you.
Also, that is not exactly how item generation works, although it is sort of similar to that. Rather, the game tries to roll unique. If it succeeds, then it will give you a unique or a rare with 3x durability. If on the other hand, the unique roll fails, it will try to give you set. If the set roll succeeds, you will get a set item or a magic item with 2x durability. And so on for rare and magic. Check either hrus' guide to mf or warrior of light's item generation tutorial for more info.
So let's do this for a battle hammer dropped from Meph. Let's assume the values from the text files linked by Hru's mf guide for monster level, etc, are true for classic. Then it's just a blind application of the formula:
We know that
Effective MF=(MF*Factor)/(MF+Factor),
where Factor=250 for unique items, 500 for set items and 600 for rare items.
and the base chance, divisor, and minchance are:
unique: 400, 1, 6400
set: 160, 2, 5600
rare: 100, 2, 3200
quest: unique, set, rare
993,993,1024
normal: u,s,r
983,983,983
I have graphed the chance for a rare, given a bhammer (lvl 48) with meph's level at 87. As can be seen, the chance for quest drop rare given a battle hammer is dropped is monotonically decreasing, whereas that for normal drop is monotonically increasing. I fiddled with numbers, and this result appears to be true (all things also true) for many different levels for meph, even at or below the bhammer's level.
If you want to check these results, i can email the spreadsheet to you. Otherwise, it appears that if you are doing the quest drop, don't put mf on for exceptional rares. Otherwise, mf won't hurt you.
norvi11erogers
Diabloii.Net Member
I'm pretty sure you will almost always get more failed sets than failed uniques. Didn't bother to plot that, but the chance for a set is just (1-chance for unique)*chance for set.
norvi11erogers
Diabloii.Net Member
Keep in mind what i was plotting. An item type (short sword, battle hammer, etc) is decided before its quality (unique set rare or magic). What I was plotting is AFTER an item type is decided, in this case battle hammer, what the chance for it to be either a failed unique or a rare.Byaxident said:
In terms of 2 uniques per run... At 222 mf, your chance for a unique per drop is 15%. If you assume 6 drops, your chance for at least 1 (failed or not) unique is 1-.85^6= 62%. For at least 2 uniques, provided i get the combinatorics correct, the chance is
1 - chance for 0 - chance for 1 =
1-.85^6-6*.85^5*.15 = 22%
So, if meph is level 87, and the unique is level 48, you should be seeing at least 2 uniques drop only about a quarter of the time, if my numbers are right. Of course, since you are seeing actual uniques, the level is much lower, could estimate it to be around 30 or so.
update... At lvl 25, the chance for at least 2 uniques is about 25%.
norvi11erogers
Diabloii.Net Member
Big oops - was using the LOD values for Base chance, div, and minchance. Here's the actual graph for the chance that an exceptional item will be either a failed unique or a rare:
The discontinuity is because the set chance drops below minchance, and thus its value is readjusted to minchance. Were it not for this discontinuity, there would be a "sweet spot" for finding rares.
Because of this discontinuity, however, we come to the same conclusions as before when it comes to finding exceptional rares:
For the non-quest drop, there IS NO SWEET SPOT! MF is always GOOD.
For the quest drop, there IS A SWEET SPOT! MF is always BAD.
(I love when my idiocy cancels to give me the right answer)
One further note: we see that if one only wants to find exceptional rares, one should repeatedly quest drop with 0 mf for the best results.
Edit: Because the base chance for unique for non-exceptional items is 400, the same as the value for LOD, the analysis for getting 2+ uniques (about a 20-25% chance at your mf) is still more or less correct at your level. Keep in mind that this % still comes AFTER the game decides what type of item to give you. Thus, the percentage that you will get real uniques is slightly lower, as any exceptional item that rolls to be a unique will be a rare.
Alles klar Kommissar?
In terms of you seeing more failed set Martels than failed uniques, first keep in mind that your sampling, 3 martels, gives nothing close to statistical certainty. This is part of the problem with testing these formulas - while they may be correct in that they are the way bliz intended, there could be random bugs to ruin things, liek the floor eating drops or fangskin being of the wrong TC, and actually finding enough items for any kind of statistics takes forever.
In any case, the probabilities that a martel (lvl53) that drops from meph (lvl87) at your mf will be:
failed unique: ~24%.
failed set: ~43%
rare: ~25%
magic: ~7%
Which adds to 99% - rounding errors. These numbers compare favorably with your results - it is about twice as likely a failed set will roll than a failed unique. But as I said, your data set is way too small to "prove" or "disprove" this.
The discontinuity is because the set chance drops below minchance, and thus its value is readjusted to minchance. Were it not for this discontinuity, there would be a "sweet spot" for finding rares.
Because of this discontinuity, however, we come to the same conclusions as before when it comes to finding exceptional rares:
For the non-quest drop, there IS NO SWEET SPOT! MF is always GOOD.
For the quest drop, there IS A SWEET SPOT! MF is always BAD.
(I love when my idiocy cancels to give me the right answer)
One further note: we see that if one only wants to find exceptional rares, one should repeatedly quest drop with 0 mf for the best results.
Edit: Because the base chance for unique for non-exceptional items is 400, the same as the value for LOD, the analysis for getting 2+ uniques (about a 20-25% chance at your mf) is still more or less correct at your level. Keep in mind that this % still comes AFTER the game decides what type of item to give you. Thus, the percentage that you will get real uniques is slightly lower, as any exceptional item that rolls to be a unique will be a rare.
Alles klar Kommissar?
In terms of you seeing more failed set Martels than failed uniques, first keep in mind that your sampling, 3 martels, gives nothing close to statistical certainty. This is part of the problem with testing these formulas - while they may be correct in that they are the way bliz intended, there could be random bugs to ruin things, liek the floor eating drops or fangskin being of the wrong TC, and actually finding enough items for any kind of statistics takes forever.
In any case, the probabilities that a martel (lvl53) that drops from meph (lvl87) at your mf will be:
failed unique: ~24%.
failed set: ~43%
rare: ~25%
magic: ~7%
Which adds to 99% - rounding errors. These numbers compare favorably with your results - it is about twice as likely a failed set will roll than a failed unique. But as I said, your data set is way too small to "prove" or "disprove" this.
Byaxident
Diabloii.Net Member
lol im kinda lost at the moment to be honest.
Although it sounds great to have the highest chance of finding rare exceptional weapons, putting on 220 mf more (going from roughly 130 to 350 mf) only gives you about 15% more chance of getting a rare (if we assume that your graph is correct). However, having such high mf means that you're going to get a LOT more uniques,failed uniques, sets and failed sets version of items with lower qlvl. Higher mf also means slower runs.
(I just got 3 failed set defenders in my last drop
).
But with that amount of mf Im obviously going to get a lot of unique and set items versions of normal items which isnt that great. For that reason its probably very hard if not impossible to find the best amount of mf which would result in the highest amount of rares, all types of items considered.
:head explodes:
PS: I guess I should just switch to LOD. So much simpler there, just slap on as much mf as possible, 1400+ if possible and you're set. :undecided:
Although it sounds great to have the highest chance of finding rare exceptional weapons, putting on 220 mf more (going from roughly 130 to 350 mf) only gives you about 15% more chance of getting a rare (if we assume that your graph is correct). However, having such high mf means that you're going to get a LOT more uniques,failed uniques, sets and failed sets version of items with lower qlvl. Higher mf also means slower runs.
(I just got 3 failed set defenders in my last drop
).Theres something wrong here, there is no doubt about it IMO. Theres a 67% chance Im going to find a martel of higher quality then rare. Actually its not all that bad when I think about it since that means I have a 49% chance of getting a rare martel (24% 3x dur mart + 25% normal dur rare mart) once the game decides a martel is going to drop.norvi11erogers said:
But with that amount of mf Im obviously going to get a lot of unique and set items versions of normal items which isnt that great. For that reason its probably very hard if not impossible to find the best amount of mf which would result in the highest amount of rares, all types of items considered.
:head explodes:
PS: I guess I should just switch to LOD. So much simpler there, just slap on as much mf as possible, 1400+ if possible and you're set. :undecided:
norvi11erogers
Diabloii.Net Member
I'm 99% positive the qualitative aspects of this graph are correct and about 80% sure that they are quantitatively correct. (Is meph actually lvl 87 in classic?) That doesn't make too much difference, just changes how fast things go up, etc. The other thing i'm not sure about are the mf multipliers. I think theyre 250, 500 and 600, but not completely sure.
Also, doing for all items at once is not possible unless you're willing to give a weighting function. Maybe you like breast plates, marts, execs, and bhammers a lot, and hate splints and, I dunno, leather. With coefficients for every item, you could do some averaging and whatnot to get a good mf. You are right though that the difference between normal and exceptional is pretty huge for rares. Not only are there sets and uniques, but all the numbers for base chances etc change. I would guess that less mf in this case would almost certainly be better, but i really don't want to run through the numbers.
Actually, the way to compare drops is like this: from ~150 to ~350, the percentage of bhammers that drop that will be rare goes from around 45 to around 58. That's an icrease of 58/45 ~= 30%. Then again, it's all about killing speed. If you can do 30% more runs in the same time with 150 mf, then it evens out.Byaxident said:lol im kinda lost at the moment to be honest.
Although it sounds great to have the highest chance of finding rare exceptional weapons, putting on 220 mf more (going from roughly 130 to 350 mf) only gives you about 15% more chance of getting a rare (if we assume that your graph is correct). However, having such high mf means that you're going to get a LOT more uniques,failed uniques, sets and failed sets version of items with lower qlvl. Higher mf also means slower runs.
(I just got 3 failed set defenders in my last drop
Also, doing for all items at once is not possible unless you're willing to give a weighting function. Maybe you like breast plates, marts, execs, and bhammers a lot, and hate splints and, I dunno, leather. With coefficients for every item, you could do some averaging and whatnot to get a good mf. You are right though that the difference between normal and exceptional is pretty huge for rares. Not only are there sets and uniques, but all the numbers for base chances etc change. I would guess that less mf in this case would almost certainly be better, but i really don't want to run through the numbers.