XP mechanics and XP rate of different areas

Spreadsheet (read post below for description)

In this thread I will describe some fundamental aspects of XP in this game which I believe are essential for competitive racers, as well as people who wish to excel at long term leveling and improve their decision making. It occurred to me while playing the game that I lack some fundamental knowledge regarding the XP that I'm getting and was making guesses on the fly instead of educated decisions. This set me out on some study which I will now present here.

First before going further, make sure you read the mechanics thread, it is prerequired knowledge to this one. That thread is fantastic and I would love if other companies did as good of a job as GGG did in explaining the fundamental rules governing their game. However there are some things they didn't mention, possibly deliberately but just as possibly from omission.

The very first thing that factors into a mob experience is the mob level. I have a very good estimation of the formula the game uses, either I'm spot on or it's off ever so slightly.

Experience = 1.02*Mob_Level^2.2

On top of this there is an experience multiplier for the various mob types. Here is the list of the mob types and their multipliers:

multiplier type
1 bandit, melee
1 bandit, suicide
1.2 bandit, archer
1.5 bear
1 crab
1 goatmen
1.5 golem
1.2 merveil's daughter
1.5 monkey big
0.65 monkey small
1.4 necromancer
1 pea shooter crab
1 pirate, melee
1.2 pirate, ranged
1 rhoa, scavanger
1.4 rhoa, other
1.2 samurai
1.5 samura, no helm
1 skeleton, melee, white
1.2 skeleton, other
1.25 snake
1 spider
0.8 spider, small
0.55 squid
1 rock throwing humans
1 totem
1.4 vaal construct
1 vaal fallen
1.2 water elemental
1 wolf
0.9 zombie
1 zombie, alira's dearest

(As a sidenote to GGG, I think some of these could do with a makeover to better reflect mob difficulty. For example the goatmen have same multiplier as melee white skeletons, which is counter intuitive and just means goatmen will forever be avoided over skeletons if the players have a choice in the matter)

Rare mobs give x5.5 times the amount of a normal mob, and have +2 levels.
Magic mobs give x3.5 times the amount of a normal mob, and have +1 level.
The base XP value uses the base level of the mob.

Currently monsters gain +50% base HP and +65% base XP for each additional member in the area (hopefully GGG will change the XP mechanics in party soon). This XP gets shared by all nearby partied players according to the relative player's shares.

Player_shares = (Player_Level+10)^2.71

Finally XP penalty is applied in case of high level difference between player level and mob level.

Percent_XP_Given = ((PlayerLevel +5)^1.5) / ((PlayerLevel+5+(EffectiveDifference^2.5))^1.5)

EffectiveDifference = IF(ABS(Player_Level-Monster_Level)<Difference_Allowed, 0, ABS(Player_Level-Monster_Level)-Difference_Allowed)

Difference_Allowed = 3 + ROUNDDOWN(Player_Level/16)

Now that the technicalities are out of the way...
I went through each area of the game compiling a list of the mob types in that area. I then proceeded to assign each area with two values, 'XP modifier' and 'density modifier'.

XP modifier is an average of the combined XP modifiers of the mobs in that region, weighed by the expected amount of mobs from each category. For some regions I approximated it without testing. For example the Dark Forest area has Melee bandits (1 modifier), spiders (1 modifier), Archer bandits (1.2 modifier) and snakes (1.25 modifier) I approximated the average modifier as 1.11. For other areas no testing was needed, for example the weaver's warren has only spiders in it which all share the modifier of 1. However for the more interesting areas I did testing counting the different amount of mobs I encountered from each type until I was satisfied I had a decent sampling size. For example in a full Ledge clear I encountered 119 mobs with a modifier of 1 (melee skeletons, rock throwers, totems) and 29 mobs with a modifier of 1.2 (archer skeletons), giving an average XP modifier of 1.04 for the area.

Density modifier is an estimate of how many mobs you're expected to encounter in a given amount of time. Initially I approximated it for all the areas, but I ended up testing it for 17 of the more interesting areas in the game out of the 41 available. The way I tested was to run in the area attempting to cross as many mobs as possible, counting the amount of mobs I encountered. After a while I backtracked and ran through all the piles of corpses and timed how long it takes me to run through it. Do note that both for this and the XP modifier testing I did no more than a single run through an area, so that my sampling may not necessarily represent the average values expected after multiple runs. In particular some areas are maze like, for example the waterfall caves level 1, lower and upper prison, and for these regions I generally recorded a fairly optimal run through. i.e. no backtracking. Thus the values I chose represent mostly an upper bound on the expected values in a typical run of these areas.

I then multiplied the XP modifier with the density modifier and got what I call 'Area Value'. This should be a value that enables players to compare regions among each other for optimal XP gathering speed, with higher values representing better XP yields. There are however a few cadavets:
- As mentioned, the values are either approximated, or at most gathered from a single run through. I mark in green all the values which are tested.
- Mob HP and difficulty is ignored. A bear may have a 1.5 multiplier, but it usually takes longer time to kill. This can not be factored in the spreadsheet and is left to the player to judge the difficulty factor of the area for his character by himself, as well as his killing speed. The longer the player takes to kill mobs in an area, the less the area is worth in terms of XP modifier for the player.

Finally taking all this into account, I was able to make a spreadsheet that given player level, will report the 'player XP rate', which is the Area Value after the monster level of the area and the player XP reduction are taken into account. Note that rare and magic mobs are ignored for this calculation, as their +2 and +1 levels may yield reduced (or increased) XP reduction for players who are under or over leveled.

You can grab the spreadsheet from here

The file has some other things which I'm using, which I left in there for my own comfort, feel free to ignore them or snoop around. The thing that's relevant to this post is the 'Areas' sheet. Enter player level at top left, and difficulty (1 to 4, 1 for normal 4 for merciless), and it will list your 'player XP rate' for each region in the game.

Also relevant is the 'XP Values Cruel' sheet, which contains the list of regions in the game with the mobs you can find there, and their multiplier. Also at the bottom of the 2nd sheet you can find the full list of mob types and their XP multipliers.

Download spreadsheet, goto 'areas' sheet, enter player level and difficulty number (1-4) in top left, and 'player XP rate' will tell you optimal area for max XP.
My Marauder 3 hour race video guide - http://www.pathofexile.com/forum/view-thread/52146
XP rate for different areas - http://www.pathofexile.com/forum/view-thread/53056
Last edited by NotThat on Oct 8, 2012 10:47:11 PM
My Marauder 3 hour race video guide - http://www.pathofexile.com/forum/view-thread/52146
XP rate for different areas - http://www.pathofexile.com/forum/view-thread/53056
Great work, that's a lot of testing required there to figure out all those mob type multis and so forth.
I do agree that some of the more "annoying" to fight against monster times could use an xp boost over your basic slow walking skele types.
Winner of the First Ever Cut-Throat Server Test, and Winner of the First Ever Cut-Throat League Event!
Awesome work! I don't want to estimate how much time you spent creating it.

There is just one thing I would like mention. Just because the XP rate is great in "Waterfall Cave 1" I would not want to spend a lot of time in it. ;D
Very impressive, you are doing a good job for the community.

Keep on going! :D
good read. thanks panda
Build of the Week 14
The first Righteous Fire/Non-Shavronne's/Shavronne's HC
Shameless self-proclaimed theory-crafting extraordinaire and forum crusader
This is invaluable information. Thank you for the detailed compilation.
great info, and great spreadsheet!

I noticed you even had a speedrun chart to measure how well you're going this time around! :)

Quite a scholar and free time you got there :)

I appreciate the effort. Getting the race down to science is the only way to make it fair. Win or lose is now determined only by luck (and the player's skills involved).

I saw you calculation of density involved number of mobs pulled and time spend in an area, this may not be accurate for a different play-style (that's not ground slam marauder)
Unviable build tester.
Fuse mechanics:
95% Crit Build Without Charges [0.10.1c]:

Report Forum Post

Report Account:

Report Type

Additional Info