The Secret To Item Farming By Jaq

For a long time I have kept my methods of item farming secret ? sharing them only with a select few. Now that I am retired, I?ve decided to open up my best secrets to the public. In this installment, I will reveal the item farming secret which I have used to garner considerable wealth.

There are four basic principles in item farming, and they must be used in conjunction to maximize the chance of success. In addition, there are situational concepts which can enhance your rate of success. Altogether, they can be used to not only get large numbers of desirable items such as teleports and Michelangelo?s scripts, but also specific items such as advanced teleporters.

First Principle: Who

There is no effective way to farm items from npc?s with their notoriously low drop rate. To be able to farm large numbers of items, you?ll need to hit players. The probability of gaining an item is strongly dependent on what items a person has in their inventory. Because of this, people who have large numbers of items in their inventory are significantly more likely to yield an item after being successfully attacked. In other words: those who pay for items or get lucky on the wheel.

How To Apply It:

There are several ways to tell if a person has a large item inventory and what they have in their inventory. First, they will tend to have a lot of level 10 buildings and teleport around a lot. Second, you may see them roll amulets. Whenever a player rolls a high value item off of an amulet, their name and what they won will appear in world chat. This makes it very easy to keep track of who has what. Third, they may bleed items off in attacks/defenses where normally they shouldn?t. For example, I?ve seen a person gain a Michelangelo?s script from defending against 1,000 warriors. The attacker almost certainly had a very large item inventory.

(You may find it more effective to keep track of who has what as an alliance. By pooling your information, you can keep track of juicy targets more easily.)

Second Principle: When

Knowing when to attack is every bit as important as knowing whom to attack. For example, if you see someone win an item on the wheel that you want, you?ll need to attack them before they can use it in order to get it. It?s generally easier to farm items at the beginning of a server than the end ? because most people haven?t used their starting items yet. Lastly, you?ll almost always want to attack when your opponent is offline ? to greatly increase the chance of success.

How To Apply it:

Be as aggressive as you are capable of earlier on. If you are able to kill opponents effectively enough, you should develop quite a store of items by the second month. On one server, I was able to acquire 47 teleports and 13 Michelangelo?s scripts by the end of the first month. Keep a tally of who has what, and focus on players with large inventory stores when/if you go to war with them/their alliance. It?s much easier to check a long list than to make a short one and attempt to hunt them down.

Third Principle: What

Equally important to what a target has in their inventory, is what they have in the town you?re attacking. The larger the defending army killed, the greater the chance of getting an item. The algorithm for the probability of getting an item:

I = Total cent values of items elligible to be rolled. (I will list them later.)
L = The total loss value inflicted in the battle on the victim.
C = Chance in terms of percentage of rolling to get an item.

(L x I)/400 = C

If the value of C is over 100, an additional formula is applied:

(L x I)/2000 = C

In essence, it is possible to get a second item on a successful attack.

In each case the probability of getting an item is capped at 90%. So even if C has a value of 145, the chance of rolling for an item is still only 90%.

I should also note that when you roll to get an item, you are still not guaranteed an item. The specific rules about this and the algorithm for determining which item is given will be discussed later in this post.

L Values Per 1,000 Units:

Archer: .405
Ballista: 3
Cataphract: 2.625
Catapult: 8.7
Cavalry: .855
Pikeman: .27
Rams: 4.125
Transports: .8925
Scouts: .2085
Swords: .405
Warriors: .0915
Workers: .0675

How To Apply It:

Let?s do an example to demonstrate how this formula might work in-game: Player A attacks Player B with 100,000 cavalry and successfully kills the 20,000 archers defending his town. Player B has a total of 500 cents worth of items. So, to calculate the chance that player A rolls for an item, we?d use the following formula:

L = .405 x 20
L = 8.1

I = 500

(L x I)/400 = C
(8.1 x 500)/400 = C
4050/400 = C
10.125 = C

So the chance of Player A rolling for an item is 10.125%.

4th:
There remains one more obstacle in farming items. To get the sort of victory conditions needed to have a high percentage chance of getting an item, you?d need to kill 100,000 archers or equivalent with each successful attack. This can be difficult to do in a single attack. It is much easier to kill those sorts of numbers on defense ? but unfortunately, you cannot control how and when you are attacked. To truly farm items, you must attack. In addition, it must be an attack which can be repeated over and over to be considered farming. If the losses are high each time, it won?t be cost-effective enough to farm. So, we need an attack with the following qualities:

* Kills large numbers of enemy troops with minimal losses.


* Takes no or few losses in doing so.


* Must be usable at will ? cannot be a defense.


Fortunately, such an attack does exist.

How To Apply It:

That attack is called the Jaq Attack. You can read my guide on it by clicking HERE. I generally send between 5-6 out in a run ? prioritizing towns with 100k archers or equivalent and above. This is most effective in tight clusters ? where large numbers of targets are close.

Advanced Item Farming Concepts:

Now that we?ve covered the four basic principles in item farming, let?s cover some of the more advanced concepts. Let?s start with the algorithm for getting a specific item:

Getting A Specific Item:

The probability of getting any particular item is dependent upon the proximity of its cent value to the ?O? value. The formula is enormously complicated ? due to the fact that all possible items are included in the determination of the probability. So, instead of attempting to look at it through a broad mathematical model, let?s compare specific examples instead. For example:

Player A attacks Player B and kills 80,000 archers. Player B has 1000 cents worth of items. Player A knows that Player B has both teleports (and nothing else at 30 cents value) and Michelangelo?s scripts. He wants to know what he?s most likely to get between the two. So, using the formulas:

O = Odds value which determines the probability of all items when rolling.

C/2 = O

L = .405 x 80
L = 32.4

I = 1000

(L x I)/400 = C
(32.4 x 1000)/400 = C
32400/400 = C
81 = C
81/2 = O
40.5 = O

(The probability of getting any one item is never above 50%.)

So the probability of Player A getting a teleport scroll or a Michelangelo?s script is about equal ? leaning ever so slightly towards getting a Michelangelo?s script.

The exact slope is unknown due to the limitations of my data set, and inconsequential. If you?re attempting to get a specific item, your focus will be on maximizing the probability of getting that item ? the minute probabilities of disparate items won?t factor into your estimations. Even if you want to know what other items are likely for a given attack, you only need to see which are closest in cent value to know which are most probable.

Depending on what items are in the victim?s inventory, certain rules may apply:

If the item selected in the roll isn?t in the victim?s inventory, the item is not awarded. It will then conduct two more checks, each time checking for the next closest item below of lesser value. If items of equal value exist, one will automatically be selected and awarded without additional checks ? each with an equal value of being selected.

Returning to our previous example, let?s assume that the game selects a war ensign for Player A from successfully attacking Player B. However, Player B doesn?t have any in his inventory, so the game checks to see if Player B has any items of equal value. He has 1 war ensign and 1 truce agreement and no other 30 cent items. In this instance both the truce agreement and war ensign have a 50% chance of being selected.

Here is a list of cent values and corresponding items:

Cents:

3 ? Speaker
5 ? Amulet, Hero Hunting
6 ? National Flag
8 ? Anabasis
10 ? Corselet, War Horn
20 ? Holy Water, Dynamite
30 ? City Teleporter, Epitome of Military Science, Truce Scroll, War Ensign
50 ? Michelangelo?s Script, Civil Code
60 ? Ivory Horn, Ultra Corselet
90 ? Adv. City Teleporter
100 ? On War, Speech Text
120 ? Penicillin
200 ? New ID
500 ? Ritual of Pact

Note: The above items are the only items eligible to be rolled, and the only ones included in the ?I? value.

If you are awarded an item from the ?C? value, but your victim fails the 3 checks, you?ll get an npc roll instead. The game treats the person as an npc and grants an npc item roll based off the level of their townhall. Because npc?s have such low drop rates, this is generally undesirable. However, in rare instances you can be granted an item that you could not normally get off a player, such as a medal or speed up.

Principle Of Multiplicity:

This is an important advanced concept, and one you?ll be applying at multiple levels. Items gained are not removed from the victim?s inventory. As you and/or your alliance gains in items, so does your ?I? value. Gaining items necessarily increases the chance of others getting items. This means your opponents will have a better chance of getting items from you as well. In practice, this means it?s actually a GOOD thing if your enemies are trying to farm items too ? because it?ll raise their ?I? values and make it easier for you to get items from them. Put simply ? the more people that know how to item farm, the easier it will become to item farm.

Feeding Your Farms:

When you raise your ?I? value from successful farming, you?ll eventually start bleeding off items in normal suicide attacks. Each time you do so, you increase the ?I? value of the person you are trying to farm. In practice, this means you can ?give? a victim items that you can then win off of them.

Cost And Profit:

Sometimes it takes items to get items. I like to teleport around frequently with 1 city. It takes me about 2-3 teleports to land in an enemy cluster, and I?m able to get 5-6 teleports from a cluster on average. With a well used advanced city teleporter, I can do even better.

Conclusion:

Heavy item farming may not be practical for most players, especially newer players. However, even the most inexperienced player may be able to acquire a few extra items using these techniques. I realize that this article contains quite a lot of information ? it may take some work to learn and fully apply. Think of it as being like high risk/high reward npc farming. You?re committing a much larger number of troops to potential doom ? but the rate of success is much better (up to 90%). Happy hunting!

Research Notes:

The above conclusions derive from 536 pieces of raw data and 278 pieces of indirect data. Numerous formulas were tried during the testing phase ? the one presented within this article is the only one which accurately reflected all of the data and coherently explained outliers. I do not expect that it is the exact formula used by the Evony game engine, but that it accurately reflects the probabilistic nature in which items are rewarded for actions. It has been tested on servers 14, 31, and 57. It is possible the effectiveness of the above techniques may differ on different servers ? much like medal drops. It is also possible that Evony may change the drop rate at any time ? certain peculiarities in the current formula convince me this has occurred at least once before.