Difference between revisions of "User:Twentypence"

From MHWiki
(New page: == Mouse Difficulties Theory ==)
 
m (admin note, link deleted to clear listing on wanted pages after deleting a page no longer needed.)
 
(23 intermediate revisions by 3 users not shown)
Line 1: Line 1:
−
== Mouse Difficulties Theory ==
+
= Introduction =
 +
 
 +
Since I started playing MouseHunt I’ve been interested in how it all works, trying to figure out what would be the best setup to use in a certain area. It’s not an easy game to figure out, and it’s obvious that the developers have put a lot of thought into it to provide a great variety to the game.
 +
 
 +
Below I have included my theories on the mechanics behind MouseHunt, I’ve been assisted in my efforts by many people, including the NVMRC, who started me off with a research grant and Pooflinger's MouseHunt Tools, who set up the survey from the Log Summariser which has provided me information on over 250,000 hunts so far.
 +
 
 +
In an effort to give something back to the community in general, and to help other players I’ve done various analytical work and made it available via the MH Analytics section of Pooflinger's MouseHunt Tools, the culmination of all this work so far has been the Catch Rates Estimator, which is built upon the theories given below.
 +
 
 +
If you're interested in my work and want to get in touch, or if you just fancy having a look to see who I am, here is my {{MHP|781784137|Hunter's Profile}}.
 +
 
 +
=== Quick Links ===
 +
 
 +
[[User:Twentypence/Mouse Populations]]
 +
 
 +
[[User:Twentypence/Survey Results]]
 +
 
 +
= Mechanics of MouseHunt =
 +
 
 +
Much of this is theory which I have tested against the results of the survey, in some places the developers have confirmed or hinted at certain aspects of this, but in most they haven't, so none of this should be taken as fact.
 +
 
 +
== Encountering a Mouse ==
 +
 
 +
Once a hunt is started a mouse is encountered.
 +
 
 +
The breed of mouse depends on two factors, firstly the location the hunter is in, and secondly the type of cheese being used.
 +
 
 +
Each Location has a "pool" of mice that can be encountered in that location and make up the population of that location.
 +
 
 +
I believe that during special events there is a second pool of prize and special event mice that is added to most areas (although some areas do not have them, such as the [[Pinnacle Chamber]]), this second pool means that one area will not run out of prize mice faster than others, there is no "best" place to hunt for the prize mice, and that all hunters have an equal chance of encountering one, as long as they are not in one of the areas that does not have them.
 +
 
 +
In some locations, certain types of cheese can affect the distribution of the mouse population.
 +
 
 +
For example, in the [[Acolyte Realm]], a hunter will encounter a [[Golem Mouse]] with similar regularity with [[Radioactive Blue]] and [[Runic]], but when using [[Ancient]] they are encountered less often. This is because [[Ancient]] changes the distribution of the mice in the area, each mouse has a 50% chance of being replaced with a [[Gate Guardian Mouse]], except for the [[Wight Mouse]], [[Lich Mouse]] and [[Acolyte Mouse]], which are all automatically replaced with a [[Gate Guardian Mouse]].
 +
 
 +
Here is a breakdown of the various populations of mice encountered in areas, with some examples of how the populations are changed due to different cheeses: [[User:Twentypence/Mouse Populations]]
 +
 
 +
== Attracting a Mouse ==
 +
 
 +
Once a mouse has been encountered, there is a chance to attract the mouse to the trap, or fail to attract the mouse.
 +
 
 +
First the type of cheese being used is compared to the mouse encountered, if the mouse is not attracted to that type of cheese, then the cheese automatically fails to attract. This is most obvious in the case of the [[Mousoleum]], where cheese such as [[Brie]] automatically fails to attract any Shadow Mice.
 +
 
 +
If the mouse is attracted to that type of cheese, then the attraction rate of the cheese and the attraction bonus of trap are taken into account.
 +
 
 +
Each cheese has a hidden attraction rate in the game (expressed below as a percentage). This can vary by area, so certain types of cheese will perform better in one area than another.
 +
 
 +
A random number is generated (for example, between 0 and 1), if the attraction rate of the cheese is greater than the number, then the mouse is attracted to the trap, if it is less, then the hunter gets a second chance to attract the mouse based upon the attraction bonus of their trap.
 +
 
 +
This works in the same way as for the cheese, another random number is generated and compared to the attraction bonus of the trap, if it is greater then then mouse is attracted to the trap, if it is less, the hunter fails to attract a mouse.
 +
 
 +
The total attraction of a trap and cheese setup can be calculated like this:
 +
 
 +
cheese attraction rate + ((1 - cheese attraction rate) * trap attraction bonus)
 +
 
 +
So for some cheeses, having a large attraction bonus from your trap is unnecessary, whereas cheeses with low attraction rates, such as [[Cheddar]], benefit greatly from it.
 +
 
 +
=== Stale Cheese ===
 +
 
 +
If the hunter fails to attract a mouse, there is a chance the cheese will go stale, this is dependent solely on the [[Cheese Effect]] of the trap, from observation these are the percentage chances of cheese going stale after a fail to attract.
 +
 
 +
{| cellpadding="5" border="1" style="text-align:left; border-collapse:collapse; margin:0 20px 0 20px;"
 +
! Cheese Effect !! Estimated Stale Rate !! Observed Rate !! Sample Size
 +
|-
 +
| Ultimately Stale (1) || 100.00% || 97.35% || 1,359
 +
|-
 +
| Insanely Stale (1) || 100.00% || 97.00% || 934
 +
|-
 +
| Extremely Stale (1) || 100.00% || 94.03% || 201
 +
|-
 +
| Very Stale || 50.00% || 49.33% || 1,490
 +
|-
 +
| Stale || 33.33% || 31.94% || 9,560
 +
|-
 +
| No Effect || 25.00% || 25.30% || 10,220
 +
|-
 +
| Fresh || 20.00% || 20.24% || 2,461
 +
|-
 +
| Very Fresh || 16.67% || 16.95% || 2,555
 +
|-
 +
| Extremely Fresh || 14.29% || 15.25% || 7,851
 +
|-
 +
| Insanely Fresh || 12.50% || 13.45% || 2,870
 +
|-
 +
| Ultimately Fresh || 11.11% || 11.03% || 6,401
 +
|-
 +
| Über Fresh (2) || 10.00% || 7.97% || 615
 +
|-
 +
| Über Fresh+ (2) || 9.09% || 9.30% || 6,088
 +
|}
 +
(Last Updated 29 June 2009)
 +
 
 +
(1) Originally these Cheese Effect ratings went stale less often, however, since the introduction of [[Stale Cheese]] the stale rates seem to have been amended, some of the results to the survey are before this change.
 +
 
 +
(2) It would appear that the maximum Cheese Effect rating that can be achieved is Über Fresh, however there are a few combinations that should be fresher than Über Fresh, which have been referred to here as Über Fresh+.
 +
 
 +
==Catching a Mouse==
 +
 
 +
Once a mouse has been attracted to a trap then the mouse is either caught or escapes from the trap.
 +
 
 +
This takes place in two stages, first the power of the trap is compared to the power of the Mouse and a random number is generated, should this stage fail to catch the mouse, a second comparison involving luck takes place.
 +
 
 +
If either of the two stages results in a success, then the hunter catches the mouse and the mouse may drop some loot.
 +
 
 +
If both of the stages fail, then the hunter receives a message saying they failed to catch the mouse, and the mouse may steal something from the Hunter.
 +
 
 +
===Effectiveness of Power Types===
 +
 
 +
The power type of the trap is compared to the power type of the mouse, and a modifier is applied to the power of the trap. After comparing the various Mouse Powers (see below) for different trap types I believe that the effectiveness modifier is as follows:
 +
 
 +
{| cellpadding="5" border="1" style="text-align:center; border-collapse:collapse; margin:0 20px 0 20px;"
 +
|-
 +
| rowspan="2" valign="bottom" align="left" | Mouse
 +
| colspan="5" | Trap
 +
|-
 +
| Physical || Shadow || Tactical || Arcane || Hydro
 +
|-
 +
| style="text-align:left;" | Physical
 +
| 100% || 100% || 100% || 100% || 100%
 +
|-
 +
| style="text-align:left;" | Shadow
 +
| 25% || 200% || 25% || 100% || 0%
 +
|-
 +
| style="text-align:left;" | Tactical
 +
| 100% || 0% || 175% || 0% || 100%
 +
|-
 +
| style="text-align:left;" | Forgotten
 +
| 0% || 100% || 0% || 200% || 0%
 +
|-
 +
| style="text-align:left;" | Arcane
 +
| 0% || 0% || 0% || 0% || 0%
 +
|-
 +
| style="text-align:left;" | Hydro
 +
| 100% || 0% || 75% || 0% || 200%
 +
|}
 +
 
 +
However, there are some exceptions to these general rules:
 +
 
 +
{| cellpadding="5" border="1" style="text-align:center; border-collapse:collapse; margin:0 20px 0 20px;"
 +
|-
 +
| rowspan="2" valign="bottom" align="left" | Mouse
 +
| colspan="5" | Trap
 +
|-
 +
| Physical || Shadow || Tactical || Arcane || Hydro
 +
|-
 +
| style="text-align:left;" | Ghost
 +
| 0% || 200% || 25% || 100% || 0%
 +
|-
 +
| style="text-align:left;" | Assassin
 +
| 0% || 0% || 175% || 0% || 100%
 +
|-
 +
| style="text-align:left;" | Student of the Cheese ….
 +
| 50% || 0% || 175% || 0% || 100%
 +
|-
 +
| style="text-align:left;" | Master of the Cheese ….
 +
| 25% || 0% || 175% || 0% || 100%
 +
|-
 +
| style="text-align:left;" | Master of the Dojo
 +
| 0% || 0% || 175% || 0% || 100%
 +
|-
 +
|}
 +
 
 +
===Hunter’s Title===
 +
 
 +
The Hunter’s Title has a small effect on some of the attributes of a trap, I believe it affects the power of the trap and the luck of the trap, I have not seen anything to indicate that it affects the attraction bonus of the trap.
 +
 
 +
It has so far proven very difficult to determine exactly how much of an effect the Hunter’s Title has, from observation I believe it to be less than 3% per Hunter’s Title, so I have estimated it as 1%.
 +
 
 +
===Calculations for catch rates===
 +
 
 +
I believe the catch rates of a trap against a mouse can be calculated as follows:
 +
 
 +
Catch (Power) = (Trap Power * Effectiveness) / (Mouse Power + (Trap Power * Effectiveness))
 +
 
 +
Catch (Luck) = (Trap Luck * Trap Luck) / (Mouse Power + (Trap Luck * Trap Luck))
 +
 
 +
Catch (Total) = Catch (Power) + ((1- Catch (Power)) * Catch (Luck))
 +
 
 +
===Mouse Power===
 +
 
 +
The Mice page contains a value for the difficulty of catching each mouse, this difficulty changes as the hunter changes their trap setup.
 +
 
 +
If we take the value at which the mouse becomes challenging as the base for that mouse’s power then all the difficulties fall into the following range:
 +
 
 +
{| width="300" cellpadding="5" border="1" style="text-align:left; border-collapse:collapse; margin:0 20px 0 20px;"
 +
! Difficulty Rating !! Catch (Power)
 +
|-
 +
| Overawing || 0% to 1/3%
 +
|-
 +
| Near Impossible || 1/3% to 8%
 +
|-
 +
| Overpowering || 8% to 35%
 +
|-
 +
| Very Difficult || 35% to 50%
 +
|-
 +
| Challenging || 50% to 65%
 +
|-
 +
| Moderate || 65% to 90%
 +
|-
 +
| Easy || 90% to 95%
 +
|-
 +
| Effortless || 95% to 100%
 +
|}
 +
 
 +
Please note, in February 2009 all the difficulties were changed, with a drop of 15%, I have seen no conclusive evidence that this made the mice any easier to catch, so I am assuming that the change was made to factor in a rough average estimate for Luck and Hunter’s Title.
 +
 
 +
===Loot===
 +
 
 +
Should the mouse be caught, there is a possibility it may drop loot. Only certain mice in certain areas will drop loot. I am currently looking into how the chance of a loot drop is calculated.
 +
 
 +
===Stealing===
 +
 
 +
Should a hunter fail to catch a mouse, there is a chance that the mouse may steal additional supplies from the hunter.
 +
 
 +
There are three options for the type of supply that can be stolen, additional pieces of cheese, points, or gold, this seems to be randomly determined between the three, except in one of the following situations:
 +
 
 +
If the hunter does not have sufficient gold, then cheese of equal or greater value to the amount of gold that would have been stolen is taken.
 +
 
 +
If the hunter is using either SUPER|Brie+ or a crafted type of cheese, then additional pieces of cheese will not be stolen. An exception to this is Moon cheese, which may have additional pieces stolen, this is presumably a hold over from the time when it was available for sale, rather than as a crafting item.
 +
 
 +
The quantity stolen seems to be an amount up to the maximum reward given for the mouse, although after brief observation it tends towards 1/3 of the maximum award. An exception to this is the Vampire Mouse, which is known to steal more gold than it is worth.
 +
 
 +
==Outcome of the Hunt==
 +
 
 +
Finally the results of all the above stages are written to the journal of the Hunter.

Latest revision as of 18:41, 30 April 2021

Introduction

Since I started playing MouseHunt I’ve been interested in how it all works, trying to figure out what would be the best setup to use in a certain area. It’s not an easy game to figure out, and it’s obvious that the developers have put a lot of thought into it to provide a great variety to the game.

Below I have included my theories on the mechanics behind MouseHunt, I’ve been assisted in my efforts by many people, including the NVMRC, who started me off with a research grant and Pooflinger's MouseHunt Tools, who set up the survey from the Log Summariser which has provided me information on over 250,000 hunts so far.

In an effort to give something back to the community in general, and to help other players I’ve done various analytical work and made it available via the MH Analytics section of Pooflinger's MouseHunt Tools, the culmination of all this work so far has been the Catch Rates Estimator, which is built upon the theories given below.

If you're interested in my work and want to get in touch, or if you just fancy having a look to see who I am, here is my Hunter's Profile.

Quick Links

User:Twentypence/Mouse Populations

User:Twentypence/Survey Results

Mechanics of MouseHunt

Much of this is theory which I have tested against the results of the survey, in some places the developers have confirmed or hinted at certain aspects of this, but in most they haven't, so none of this should be taken as fact.

Encountering a Mouse

Once a hunt is started a mouse is encountered.

The breed of mouse depends on two factors, firstly the location the hunter is in, and secondly the type of cheese being used.

Each Location has a "pool" of mice that can be encountered in that location and make up the population of that location.

I believe that during special events there is a second pool of prize and special event mice that is added to most areas (although some areas do not have them, such as the Pinnacle Chamber), this second pool means that one area will not run out of prize mice faster than others, there is no "best" place to hunt for the prize mice, and that all hunters have an equal chance of encountering one, as long as they are not in one of the areas that does not have them.

In some locations, certain types of cheese can affect the distribution of the mouse population.

For example, in the Acolyte Realm, a hunter will encounter a Golem Mouse with similar regularity with Radioactive Blue and Runic, but when using Ancient they are encountered less often. This is because Ancient changes the distribution of the mice in the area, each mouse has a 50% chance of being replaced with a Gate Guardian Mouse, except for the Wight Mouse, Lich Mouse and Acolyte Mouse, which are all automatically replaced with a Gate Guardian Mouse.

Here is a breakdown of the various populations of mice encountered in areas, with some examples of how the populations are changed due to different cheeses: User:Twentypence/Mouse Populations

Attracting a Mouse

Once a mouse has been encountered, there is a chance to attract the mouse to the trap, or fail to attract the mouse.

First the type of cheese being used is compared to the mouse encountered, if the mouse is not attracted to that type of cheese, then the cheese automatically fails to attract. This is most obvious in the case of the Mousoleum, where cheese such as Brie automatically fails to attract any Shadow Mice.

If the mouse is attracted to that type of cheese, then the attraction rate of the cheese and the attraction bonus of trap are taken into account.

Each cheese has a hidden attraction rate in the game (expressed below as a percentage). This can vary by area, so certain types of cheese will perform better in one area than another.

A random number is generated (for example, between 0 and 1), if the attraction rate of the cheese is greater than the number, then the mouse is attracted to the trap, if it is less, then the hunter gets a second chance to attract the mouse based upon the attraction bonus of their trap.

This works in the same way as for the cheese, another random number is generated and compared to the attraction bonus of the trap, if it is greater then then mouse is attracted to the trap, if it is less, the hunter fails to attract a mouse.

The total attraction of a trap and cheese setup can be calculated like this:

cheese attraction rate + ((1 - cheese attraction rate) * trap attraction bonus)

So for some cheeses, having a large attraction bonus from your trap is unnecessary, whereas cheeses with low attraction rates, such as Cheddar, benefit greatly from it.

Stale Cheese

If the hunter fails to attract a mouse, there is a chance the cheese will go stale, this is dependent solely on the Cheese Effect of the trap, from observation these are the percentage chances of cheese going stale after a fail to attract.

Cheese Effect Estimated Stale Rate Observed Rate Sample Size
Ultimately Stale (1) 100.00% 97.35% 1,359
Insanely Stale (1) 100.00% 97.00% 934
Extremely Stale (1) 100.00% 94.03% 201
Very Stale 50.00% 49.33% 1,490
Stale 33.33% 31.94% 9,560
No Effect 25.00% 25.30% 10,220
Fresh 20.00% 20.24% 2,461
Very Fresh 16.67% 16.95% 2,555
Extremely Fresh 14.29% 15.25% 7,851
Insanely Fresh 12.50% 13.45% 2,870
Ultimately Fresh 11.11% 11.03% 6,401
Über Fresh (2) 10.00% 7.97% 615
Über Fresh+ (2) 9.09% 9.30% 6,088

(Last Updated 29 June 2009)

(1) Originally these Cheese Effect ratings went stale less often, however, since the introduction of Stale Cheese the stale rates seem to have been amended, some of the results to the survey are before this change.

(2) It would appear that the maximum Cheese Effect rating that can be achieved is Über Fresh, however there are a few combinations that should be fresher than Über Fresh, which have been referred to here as Über Fresh+.

Catching a Mouse

Once a mouse has been attracted to a trap then the mouse is either caught or escapes from the trap.

This takes place in two stages, first the power of the trap is compared to the power of the Mouse and a random number is generated, should this stage fail to catch the mouse, a second comparison involving luck takes place.

If either of the two stages results in a success, then the hunter catches the mouse and the mouse may drop some loot.

If both of the stages fail, then the hunter receives a message saying they failed to catch the mouse, and the mouse may steal something from the Hunter.

Effectiveness of Power Types

The power type of the trap is compared to the power type of the mouse, and a modifier is applied to the power of the trap. After comparing the various Mouse Powers (see below) for different trap types I believe that the effectiveness modifier is as follows:

Mouse Trap
Physical Shadow Tactical Arcane Hydro
Physical 100% 100% 100% 100% 100%
Shadow 25% 200% 25% 100% 0%
Tactical 100% 0% 175% 0% 100%
Forgotten 0% 100% 0% 200% 0%
Arcane 0% 0% 0% 0% 0%
Hydro 100% 0% 75% 0% 200%

However, there are some exceptions to these general rules:

Mouse Trap
Physical Shadow Tactical Arcane Hydro
Ghost 0% 200% 25% 100% 0%
Assassin 0% 0% 175% 0% 100%
Student of the Cheese …. 50% 0% 175% 0% 100%
Master of the Cheese …. 25% 0% 175% 0% 100%
Master of the Dojo 0% 0% 175% 0% 100%

Hunter’s Title

The Hunter’s Title has a small effect on some of the attributes of a trap, I believe it affects the power of the trap and the luck of the trap, I have not seen anything to indicate that it affects the attraction bonus of the trap.

It has so far proven very difficult to determine exactly how much of an effect the Hunter’s Title has, from observation I believe it to be less than 3% per Hunter’s Title, so I have estimated it as 1%.

Calculations for catch rates

I believe the catch rates of a trap against a mouse can be calculated as follows:

Catch (Power) = (Trap Power * Effectiveness) / (Mouse Power + (Trap Power * Effectiveness))

Catch (Luck) = (Trap Luck * Trap Luck) / (Mouse Power + (Trap Luck * Trap Luck))

Catch (Total) = Catch (Power) + ((1- Catch (Power)) * Catch (Luck))

Mouse Power

The Mice page contains a value for the difficulty of catching each mouse, this difficulty changes as the hunter changes their trap setup.

If we take the value at which the mouse becomes challenging as the base for that mouse’s power then all the difficulties fall into the following range:

Difficulty Rating Catch (Power)
Overawing 0% to 1/3%
Near Impossible 1/3% to 8%
Overpowering 8% to 35%
Very Difficult 35% to 50%
Challenging 50% to 65%
Moderate 65% to 90%
Easy 90% to 95%
Effortless 95% to 100%

Please note, in February 2009 all the difficulties were changed, with a drop of 15%, I have seen no conclusive evidence that this made the mice any easier to catch, so I am assuming that the change was made to factor in a rough average estimate for Luck and Hunter’s Title.

Loot

Should the mouse be caught, there is a possibility it may drop loot. Only certain mice in certain areas will drop loot. I am currently looking into how the chance of a loot drop is calculated.

Stealing

Should a hunter fail to catch a mouse, there is a chance that the mouse may steal additional supplies from the hunter.

There are three options for the type of supply that can be stolen, additional pieces of cheese, points, or gold, this seems to be randomly determined between the three, except in one of the following situations:

If the hunter does not have sufficient gold, then cheese of equal or greater value to the amount of gold that would have been stolen is taken.

If the hunter is using either SUPER|Brie+ or a crafted type of cheese, then additional pieces of cheese will not be stolen. An exception to this is Moon cheese, which may have additional pieces stolen, this is presumably a hold over from the time when it was available for sale, rather than as a crafting item.

The quantity stolen seems to be an amount up to the maximum reward given for the mouse, although after brief observation it tends towards 1/3 of the maximum award. An exception to this is the Vampire Mouse, which is known to steal more gold than it is worth.

Outcome of the Hunt

Finally the results of all the above stages are written to the journal of the Hunter.

Retrieved from "https://mhwiki.hitgrab.com/wiki/index.php?title=User:Twentypence&oldid=152180"