What do you folks know about the cap so far? I mean I can still go over 400 MF, but I'm not sure if it's "working" as over 400%.
From personal observation, the +300% MF cap applies to the sum of:
(1) Player gear MF
(2) Paragon levels
(3) MF Shrine
It does not include NV stacks or follower bonus. The MF displayed in character sheet does cap, based on the restrictions listed above. I think the highest possible MF is +406% (300% + 75% NV stacks + 31% templar follower).
Not too sure about the follower, but anyway, how exactly do you get your templar to 31%? Giving a templar a 300 spear (25 MF), shield (20 MF), neck (40 MF) and 2x rings (36 MF) gives a grand total of 121 MF on the follower; 20% is applied to you so that's 24 MF... Am I missing something?
*New* Sun Keeper 1h mace: +45% MF
*New* Litany of the Undaunted i62 set ring: +18% MF
*New* The Wailing Host i62 set ring: +18% MF
Two ring set bonus: +15% MF
Amulet: +40% MF
Shield: +20% MF
Total: +156% MF ---divided by 5 and truncated---> +31% MF
This is good info! Thank you! Sad to see that WD are no longer the kings of MF :-)
I know there are a bunch of changes incoming but I already have gathered this data. It is 200 Elite Kills (100 with 5NV) in A2 with 400MF. The results are very close to the ones from A1. I posted both spreadsheets here: This is where I will keep all my data points from now.
Do Legendary and Set pieces adhere to their respective item levels when determining what Acts drop them? For example all set pieces are 61s. There isnt a whole lot of difference from the rate 61s drop in Act 1 compared to Act 3 but the difficulty jump is massive. Does that mean Act 1 is the best place to farm Set pieces? Or are legendaries and set pieces like recipes and only the higher acts drops the best ones? I have put in 300 hours farming Act 1 with 350 MF with 5 stacks and have yet to get a decent legendary or set piece. I know they are supposed to be rare but when I hear people say they get them all they time in Act 3 using only NV stacks that cant possibly be true.
I have gotten quite a few set pieces and legendaries probably around 20 since launch and not a single one has sold for over 500k and 17 of them have either been vendored or salvaged. So out of 20 only 3 actually made it to AH and didnt even sell for 1 million combined. Is that because they were all from Act 1 and therefore could only be crappy rolls? For instance I always seem to get the same worthless garbage ones every single time. I always get crappy Blackthorn legs with minimum stat rolls for example and they never sell for more than 150k. For legendaries I got mostly Skorns, so again 100% worthless along with those 57 fist weapons, again total garbage. For the good stuff like the rings and necklaces or Natalya's Mark that sell for millions are they only from Acts 3 and 4?
We are not 100% certain how the setpieces work. They share the same integer-identifier as legendary items and is therefore considered to be grouped with them. But in practice I'd say that you've just had bad luck. I found skull grasp from A1 and that little bugger sold for well over 50euro for me.
At this point, however, you're not off to hit the money-making monkey by aiming at legendary farming: Only a small % of the items will be sellable and the fact that they drop so rarely (I usually find 1 per 5 hours straight with 300 MF in A1) just makes it not worth the grind. Hopefully this will change when new legendary items get implemented.
That being said: None of us found a set item except for Timza (and he even found it with no NV). Shurafa also found one while getting the 5xNV stacks. Either we are very unlucky as a group, but set items do seem more rare than legendary items.
I have actually found 3 set pieces while doing my research in A1 (300ish elite kills). 2 of which came off of trash mobs and both of which sold on the AH for $20-$40 (IK Belt with good stats and high armor was the one for $40 it was the only one in its class on the RMAH). I also found a 1200DPS 1 hander and a sick 800dps LOH Life Steal Critical hit 60+ both of which sold for $75-100 so there is really good stuff to farm out there even if its not a set piece. Hopefully this will change in the patch. (Hopefully they dont mess with MF too much or we will have to start our research over).
Set pieces are soooo rare though we dont have anywhere enough data to extrapolate much useful info on their drop rates as far as where the best place to find them is.
I take that back great work Murskautuminen! I wonder how many kills we need to verify that Legendary/Set drop rate. It seems a bit high. If it is correct though once 1.04 comes out MF gear will be super important if you want to farm rares/set pieces.
Yes warrior´s rest act 1. The order of drops (you will learn to recognize stuff really fast after some views) is:
gear drop 1 guaranteed (mostly blue, some yellow)
gear drop 2 guaranteed (mostly white, some blues, mostly 1 affix)
tome / no drop
gear drop 3 / no drop (mostly white, some blues, about half 1 affix)
gem / no drop
potion / no drop
gear drop 4 guaranteed (same as 1)
I´m currently working on another batch of 71 kills, will post updated results in the evening.
This is really GREAT work. Great find. Thank you! Looking forward to the results...
Sup guys, you went into really nice detail with this MF study. I've been doing a little of my own just for fun. When I started to gear a MF whirlwind barb I noticed that the average quality of the items that dropped from the first 5 champs tended to be better than the guaranteed rares that dropped. I made a little spreadsheet to keep track of the items that I've collected. I would normally be able to farm 100+ items in each category per day but I have been really busy the last 3 weeks. Things are starting to settle back down and i suspect to be able to get 1000 data points in both categories in the next week.
Some things about my data:
It does not include blue/white/legendary/set items as a percentage of the drop. So for example, when you see 11.2% for non-guaranteed 6 affix rare, that means, with an average of 234 MF the rares I found had about an 11.2% chance of being 6 affix. That does not mean that 11.2% of the items dropped were 6 affix, that wouldn't even make sense. I'm sure there are a few more assumptions and parameters but I will edit that later. In fact as I type this I want to add that I did not include any drops when I had fortune shrine. I averaged my MF by taking my base MF and averaging that with 75 MF from 5 NV. I know that is not super scientific but it allows me to collect more non-guaranteed rares.
Let me know what you guys think.
This is extremely interesting! I would love to see more data on this. Your data suggests that Affixes are affected by MF for non NV rolls and 5NV rare rolls drop without taking MF into account.
If this were in fact found to be true. It would drastically change peoples farming habits...
We definitely need to do more testing but from what I have found If you are just looking purely at the number of rares per hour you are MUCH MUCH better off playing with less or even ZERO MF and efficiently clearing A3 than you are doing A1 or A2 with MF. NV makes this a no brainier as the drop rate for iLVL 63 items is x4 what is is in A1. Its impossible to make up this difference with MF (not even close if you have to sacrifice any dps and/or survive ability). Given the premium price of MF gear in the AH. You are clearly better off spending your money on gear with extra resists and crit so that you can farm A3 more efficiently. This will net more items overall. However this changes once you reach the upper limits of gear (only a handful of players have reached this plateau). Once you reach your DPS and resist limits MF becomes viable again as its a method of indirectly increasing your farming efficiency. This is because you are not sacrificing any dps or survive ability as MF is now your 5th ot 6th stat rather than your 1st 2nd or 3rd.
If you are hunting for legendaries MF is the way to go. Just as difficulty affects the iLVL drop rate MF directly affects the legendary drop rates. Also if it is found that MF affects Affix distribution this will be a big win for MF stacking assuming that a strong correlation is found.
I think the above is worth mentioning with a disclaimer that it will take some time to gather enough data to support these claims.
PS. Is anyone gathering info from other Acts? It is possible that the drop rates are different. I have started farming A2 with 400MF. I have noticed some interesting trends lets see if they hold up over time.
I dont have a screenshot of my two legendary drop but not sure why it matters; someone could very easily drop two unid legendaries on ground if they wanted to lie and back it up with a SS.
kiting away the last elite would be doable sometimes but there are times when they all kinda die together as meteors rain down on them when frozen. If I am not sampling all elites would that still be valid? certain affixes might be easier to kite & kill then others.
Not sampling all elites is just fine... Sometimes I make a mistake and kill off an extra mob along with the elites so I am not sure which one dropped which items. For these these encounters I just skip writing down the data.
[Update 2012/7/22]: Added a new data point by contribution from Shurafa (MF=400, NV5, 200 elites); much appreciated! Added a chart in 3.5 showing rarity distributions as a function of MF (NV rare not included). Corrected few data errors and updated datafile
OBS! NV RARES NOT INCLUDED IN FOR THE CHART
What a great update to the thread on 0722. The OP is officially epic! Thank you again.
I know you dont have tons of time right now but I would love to see a graph like this one you just posted comparing all the data points. However this time with NV included as that is how most people typically farm.
All kills were in Act1 Inferno.
For the 383MF Data point an Enchantress was used.
For the 400MF Data point a Templar was used (his shield adds 4 more MF).
Affix Counts were done by identifying the item and then visually counting the number of affix on the item.
No Fortune Shrines were used.
No gear swapping
I found 2 legendaries and 1 set piece (Immortal Kings Belt, this was the first set piece I ever found) off of random mobs during the farming process. In addition I found a good number of blacksmith plans.
I found it interesting that the drop rate percentage between 20 kills and 300 kills averaged out to less than one percent for white/magic/rare/set and legendary drops. Is anyone else finding this trend? I found the numbers to be very stable during the grind. (Statistically how many runs do we need to do to have our data be accurate to within 1% with a 99% confidence level for rare drops only.)
Looking forward to getting all of this data organized. Cant wait to see the trends that arise! Big Props to Cyeron for spearheading this en devour!
PS. Thank you! To whomever made the original version of this Google Docs Data Sheet. (I suspect it originally came from Cyeron's Data sheet but someone modified it for GD).
400 MF (Shurafa is on it). Btw, Shurafa, out of curiousity, what class are you playing?
I am one of the very few WD's out there. (Which is why I am able to afford the gear I have). When I started this two weeks ago I had 383MF with only 17k life and Zero resists outside of my Intel. All of this while using a 400DPS weapon without a socket. It was a challenge but extremely fun! This was my build... http://us.battle.net/d3/en/calculator/witch-doctor#ajXkZU!ZVW!abbZaZ Got almost 100 kills with this setup.
At Cyeron's request I upped my MF to a nice round 400. Upgraded to a 500dps socketed weapon with about 100 to resists and 30k life. The new gear, along with a tweak to the old build made things really easy. Farming has become extremely efficient. Here is the updated build. (bad medicine is for when I play with friends) http://us.battle.net/d3/en/calculator/witch-doctor#aSXkZU!cVW!aZbZaZ Switching to Pestilence plugged a lot of holes in my build (finally an efficient AOE spell and a way attack without LOS) this greatly increased the speed of my runs.
What class/build do you guys play?
Also I was able to verify that at 400MF with 5NV stacks I have not seen a single non superior white drop on any mob. Looks like it may have dropped off the loot table completely. (or at the very least its a very very low drop percentage) This is really interesting! I wonder at what point this happens exactly?
I would like to ask to the persons farming with 400% MF if they ever see a normal white drop from elites or if they only see superior items (0,2 * (1+400MF) = 100% superior, 0% white chance from elites.
I will take note of this today as I am going to try and hit 300 kills today and submit my data. I have noted that I only see inferior drops from weapon racks and chests. None from Elites or regular mobs.
I never really thought about the NV in regards to Goblins, but that is good info. The diminishing returns on MFw/NV is disheartening. There was always a point in D2 where is had diminshing returns, but with an unavoidable mechanic like NV it makes it difficult to determine the "sweet spot" for farming.
The NV effect on goblins was also quite a surprise to me.
It is true that the determination of a "sweet spot" has become rather difficult, however, there are two spots as of now that have at least some relation to what slightly resembles a "sweet spot":
At 225 MF you will on average see rares just as commonly as white items (without the guaranteed rare drop from NV).
At 300 MF you will on average see ~50% of the total gear dropped being rare items (with the guaranteed rare drop from NV).
If I were to suggest anyone regarding what mark of MF they should be at I would say: If you're new at 60 / inferno or have low leveled gear get higher dps as the NV guaranteed rare is superior to MF. When you then get money for MF gear try hit 125-150 MF (you will hit 200-225 with NV by then) and you should notice a decent change. From that point on it is just a matter of dps and survival loss versus magic find gain.
Notice that I found quite a decent bunch of legendary items at 300 MF. It may be coincidence, but when Shurafa returns with data ~400 MF we should see if the trend continues.
Edit: fixed a horrible grammar mistake.
I suppose it is worth clarifying, are you wanting only data on affixes dropped from elite packs, or does affix data from any rares work, as long as the MF value is known/constant? For example, if we have 2 lists, one for rares from elites, one for ilvl vs affixes, both at a constant 5NV+ some MF on gear, would you be able to make use of both? I suppose it would depend on if we thought the affix distribution was different for rares from elites vs other sources and if we're not sure. I think what I'll do for my future data collection is to record ilvl, item type, affixes, and source so I'll basically record all of my rare drops instead of just those from elites, but in a way that the any of data can easily be extracted.
The more variables you add to a system the less certainty you will have with the results. Keeping things as consistent as possible is super important.
The variable we are trying to isolate is the affects of MF. Everything else should be as similar as possible.
1. The location that is being farmed (I suggest Act 1 Inferno).
2. The methodology of collecting the data. Using the Identification exploit discussed earlier, or just counting the total visible affix on the item. (I suggest doing both).
Or course once the data has been verified to show a clear pattern we can then expand the sample to include many other data points in different areas with different monster types.
Cyeron is the one who spearheaded this experiment. I suggest we try and follow his lead as much as possible so we can keep things organized.
The point is that the extra rare drop from NV makes MF slightly redundant (we want to maximize the average number of rares per elite pack). Actually, maybe redundant is too poor a word; devaluing is much better.
Having the guaranteed one drop will affect the numbers when you add more MF to your farming when compared to not farming with NV:
This is the conclusion I have come to as well. At this point. Unless it is found that MF also increases your Affix quality then it should be considered a luxury stat. It looks as though you will likely be better off stacking another "useful" stats and completing your runs faster rather than going for higher MF. It would be nice to quantify this at some point once all the base drop rates have been figured out.