Jineapple
Squire

posted 040117 06:58 AM CT (US)
How Monks really work v.2
Introduction
As some of you might know, there's an article in the University that talks about Monk mechanics. It's a good article, however not all questions are answered, and it seems that some values still refer to an earlier state of balance than 1.0c AoC. With this article I hope to expand and correct that information.
Basic Facts
As explained in the original article, conversions happen in fixed intervals of about 1.2 seconds (the exact length varies) which we will call Monk seconds. There is a minimum and a maximum number of Monk seconds during which all conversions take place For regular units, Monks take between 4 and 10 Monk seconds (or about 512 seconds) to convert a unit. Buildings take 1525 Monk seconds (roughly 1728 seconds). No conversions take place before the minimum time and units are guaranteed to be converted after the maximum time After each Monk second, the chance of a conversion taking place is about 28%. The seconds until the minimum time count as a single conversion attempt.
Conversion resistance
There are a number of effects that change conversion rates. These can influence the minimum time a monk needs to convert a unit, the maximum time needed, or it can lower the chance of conversion after each conversion interval. This last effect is called "conversion resistance" in the data files and the basic formula is that if a unit has x conversion resistance, the chance of conversion during each monk second is roughly 1/x of the original 28%. These bonuses affect both buildings and units.
Teuton Team Bonus: This bonus increases the minimum conversion time by 1 monk second, the maximum conversion time by 2 monk seconds, and conversion resistance by 2 (resulting in ~13% chance of conversion in each interval) Faith: This bonus increases the minimum conversion time by 2 monk seconds, the maximum conversion time by 4 monk seconds, and conversion resistance by 3 (resulting in ~9% chance of conversion in each interval) Eagle Scouts/Scout Cavalry: These units have a basic conversion resistance of 8, resulting in a significantly lowered chance of ~4% during each interval. However, they will still be converted during the regular 410 Monk seconds. Building: These units have a basic conversion resistance of 3, resulting in a chance of ~9% during each interval. They have a separate minimum/maximum time of 1525 Monk seconds (~1728s) Inquisition (Spanish unique technology introduced with Forgotten Empires): This lowers the minimum and maximum conversion times for enemy units by 1 Monk second and for buildings by 6 Monk seconds. It also slightly increases the chance of conversion against buildings by 5%, though the conversion resistance of buildings lowers this to effectively about 2%. Stacking of multiple effects: These effects can be stacked. The minimum/maximum times are simply added (So Faith+Teutons increases the minimum time by 3, maximum time by 6 Monk seconds). Conversion resistance is also added. This means the Faith+Teutons will give all units 5 (3+2) conversion resistance, resulting in a 56% chance of conversion, which is 1/5 of the original chance.
Stacking of conversion attempts
One very interesting and important fact to know is that the threshold for minimum conversion time is only reset once a monk does another action besides conversion. This has a few interesting effects:
If a monk is converting a unit that runs outside of his conversion range, the minimum time is not reset. Only running completely out of sight of the monk so he stops the chase will do that. This means that if a monk attempts to convert a unit, it runs outside his range, but then doubles back and gets into range again, the minimum time is now shorter or even 0 if the monk was converting for at least 4 intervals before. Therefore, if you run from a Monk to escape conversion, make sure he stops chasing the unit before going back The same effect also applies when switching conversion targets. If Monk A and Monk B convert Unit C together for 4 monk secs and Monk B then switches to converting unit D, unit D can be immediately converted instead of needing a minimum time. This is also the case when switching from units to buildings or vice versa. This is why the monk micro tactic of tasking all monks on one unit, deselecting one monk, targeting the next unit etc. works so well  part of the minimum time is eliminated. However, unless you have theocracy, you should click the ground or stop about every 45 seconds to avoid all monks losing faith on a succesful conversion.
Multiple Monks Multiple monks are all calculated individually. Three monks converting a unit means it will be converted as soon as one monk converts it, independently of the others. This also concerns conversion resistance  If multiple monks would convert a unit at the same time, there is an individual 1/x chance for each conversion to go through. So for two monks, against a unit with two conversion resistance the chance for a conversion per monk sec would be (1/2*0.28)[chances of first conversion working]+(1(1/2*0.28))*1/2*0.28[chances of first one not working, but second one working] = 26.04%.
Quirks When the conversion resistance goes above 10 (e.g. Scouts with Faith researched), there is a curious effect. Until the maximum conversion time, everything behaves as previously described, however conversion is no longer guaranteed at the maximum time. Instead, it's about a 10/x chance of your units getting converted, with x being the conversion resistance, so some units may take slightly longer than the maximum conversion time.
Testing methodology and pure numbers
I have a very similar setup to Neilkaz in his original tests. I have two hundred monks with their respective targets. I will start the game, pause and assign the targets (using an Autohotkey script). I then queue a villager in a TC for more exact timing (1 villager takes 25s, so each % is worth 0.25s). Then I let the scenario run with pausing to get the exact times of conversions (again using AHK for more precise pausing). If you want to see the results of my tests, I've uploaded my spreadsheet with them here, but I haven't had time to make it look nice.
Thanks for reading and let me know if there's anything else you want to know!
