This post presents some research I did to figure out the effects of the Bonus Item system, to provide information for myself and other people wanting to figure out what Monster Power is most efficient for them to farm for them sweet brimstones.
The main goal was to figure out how much faster does MPs 7-9 need to be farmed to better for legendary farming than similar run on MP 10.
TL;DR: Jump to section 4. Table 2.
TABLE OF CONTENTS:
- Data collection
- Analysis& Results
- Updates&Future plans
There was some very interesting discussion on US monk forums about which MP is the best for Crypt of the Ancients farming. While many good points were raised, the discussion resulted in a ‘who knows’-result due to the fact that no-one knew exactly how the Bonus Item system introduced with the Monster Power system affects loots. When doing similar farming run it is common and quite accurate to compare Life Time Kills per hour and the effects of the changes in Magic Find, when changing the Monster Power, are easily calculated. None of this helps much when the average amount of gear items varies between monster power and no-one really seems to know how.
The discussion topic for those interested:
The only information originally given about the Bonus Items was this quote from the original Monster Power post by Blizzard:
It was later confirmed by blue post that the Bonus Items are always gear drops. This is extremely important in understanding the benefits of the Bonus Item system. Basically this means that 100% Bonus Items not only mean another piece of gear drops for every piece of gear dropped, but a piece of gear also drops for every time a monster drops gold, potions, tomes etc. Generally it is the amount of gear drops that people interested in farming efficiency want to know.
Here is where the problem comes in: No-one seemed to know drop chances or the relative drop chances for gear items and other drops. This means we have no way to mathematically determine which MP is best even though we can figure out the kills per hour and the effects of the changes in Magic Find.
So I set out to calculate kills and gear drops from white mobs on Monster Powers 0 and 10. Since Life Time Kills in profile update only when you leave a game, I needed to figure out a run where I don’t kill anything extra. I chose to do this in the Crypt of the Ancients in act 1.
I only calculated loots and kills from runs where the elite spawned in lower area and therefore I could leave it be. This way I only kill the same 2 types of monsters, with a very strong emphasis on the Returned. On every run I let the Returned Summoners spawn all the returned they would, so that the amount of Returned-kills in relation to the amount of Returned Summoner-kills was equal on both Monster Powers. By choosing the The Broken Blade-quest I was able to take the portal straight to CotA every time so I faced no other monsters that would ruin my Life Time Kills calculations.
The gear drops were calculated trough looting them and then calculating them in my inventory, since it is practically impossible to count the items on the ground, especially on MP 10.
I decided to collect data of 1000 white mob kills on both MPs. While this is certainly not a big enough sample to figure out accurate drop chances gear- and other drops, this does give a good idea of the effects of Bonus Items on high MPs.
It is possible that there is difference in the relative gold and gear drops chances on different monsters. If this is the case, the data collected or the results concluded can’t be generalized to farming other areas. Personally I don’t believe there to be any significant differences between other ‘normal’ white mobs as to their relative chance to drop gold, gems, tomes etc. or gear.
Here is the data I collected in a very simple format:
Monster Power : 10
White mobs killed: 2062
Gear drops: 974
Gear / Mob: 0,472
Monster Power: 0
White mobs killed: 2496
Gear drops: 237
Gear / Mon: 0,095
As mentioned, the sample size is obviously not big enough to come up with accurate drop percentages. But using these numbers we can get quite a good general idea as to the differences between Monster Powers.
With the data we can calculate a lot of interesting things as to what’s the difference between a mob killed on MP x compared to similar mob killed on MP y (or 10 000 of each).
On average one needs to kill 4,97 times more mobs on MP 0 than on MP 10 to get the same amount of gear drops.
79,9% of gear drops on MP 10 are bonus items.
59,8% of gear drops on MP 10 are bonus items that ‘proc’ from drops other than gear.
The bonus item chances are announced by Blizzard in the 2nd link of the introduction.
These are as follows:
TABLE 1: Bonus Item chance (%)
Since we now know approximately what percentage of the MP 10 gear drops are caused by the Bonus Item system, we can make a simple comparison of Monster Powers: On average, how many mobs on MP x do we have to kill the get the same amount of gear drops as killing 1000 mobs on MP 0. With this table we can easily compare other MPs as well.
Adding the effects of Magic Find complicates the calculations to some degree. The relative benefit of the increased MF on higher MP is affected by ones MF from other sources. In the table below the comparison is done with the assumption of 300 MF from Paragon Levels and gear but no Nephalem Valor stacks.
The second table presents the following data for all MPs that should help you to calculate the MP with your own MF.
- The amount of kills that on average yields the same amount of gear drops as 1000 kills on MP 0
- The average amount of gear drops per 1000 white mob kills. (use this to calculate the relations for your given MF)
- A comparison of MPs relative to MP 10 with 300 base MF and no NV stacks.
TABLE 2: Results
I hope this was of some help to those of you who are interested in optimizing your farming efficiency. I plan to work on the presentation format in the coming days and any feedback on the issue is welcomed. Also please let me know of any possible math failures that I may have done.
6. Updates & Future Plans
MP 0 Sample size increased from 1008 to 2496 kills. All calculations were adjusted accordingly, although the differences were very small so some rounded numbers did not even change. The old table is available here: https://dl.dropbox.c...opstaulukko.jpg
MP 10 Sample size roughly doubled. All calculations and the Results table were adjusted accordingly. Again the changes to percentages were minor at best.
Keep increasing the sample size.
Test the difference in amount of gear drops to Keep Depths Level 2 monsters.
Write a guide how these numbers can be used for efficiency testing.