iDB Community Forum

Full Version: Defense Testing 101.
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
This is a quick guide to how to test a weapon's defense, intended for someone who is fairly new to the battle dome. In the interest of simplicity and clarity, I'm going to omit a few things. (If you are already an old pro at the BD, or if you want to learn more you might want to see Phoenix's more in-depth article about the intricacies of the defense system.)

The first thing and most important thing you need to know is that THE NUMBER OF DEFENSE ICONS DISPLAYED CAN BE VERY DIFFERENT FROM THE ACTUAL NUMBER OF ICONS DEFENDED!! I know that's counter-intuitive, but we'll get into some of the reasons for it in a bit.

Right now I'm going to start off with the absolute minimum amount of information we here at IDB need from you in order to determine defense.
  • I. A screenshot of the defensive item being used against an attack mix that includes all 7 icon types.
  • II. A screenshot of the defensive item blocking icons of each individual type of icon it blocks. Be sure the attacker uses "constant" weapons which deal the same amount of icons each time. To take screenshots in Windows, press Alt + PrintScreen to copy the image, then open up Paint (usually found in the Accessories menu), paste your image into a new file. Save that file as a .gif or .jpg. (other types can work too, but if you're new to this go with one of those). Then upload the file to a website that can host it and allows image linking (ImageShack and PhotoBucket are popular free image hosting sites that will work). Then post the link here on the forums using the IMG tags -- that should look something like this:{img}http://myhost/mypicture.jpg{/img}, except you'll need to use the square brackets [ ] where I used the curly braces { }. Important reminder: for screenshots of other people's pets and/or weapons, always get permission before posting.
  • III. The Strength and Defense boosts of both pets. We also need species, but that should usually be obvious from the screenshots
  • IV. Information on whether the displayed icons and numbers were the same over multiple tests of the weapon.

That will usually be enough information for us here at IDB to tell how much defense an item offers, but if you're still reading, you probably want to know how to figure it out yourself.

To do that, you need to know the basic formula for damage in the battledome:

Damage_Done = {
( Attacker's_Strength_Boost_Mutliplier * Attacker's_Number_of_Offensive_Icons * Attacker's_Stance_Multiplier * Defender's_Stance_Multiplier )
-
( Defender's_Defense_Boost_Multiplier * Defender's_Number_of_Defensive_Icons * Defender's_Stance_Multiplier)
}
*
Defender's_Species_Modifiers.

That seems complicated (even though it omits several small details), but it is really important and there are ways to simplify it. The best way to simplify is to make sure that both pets use the "Cautious Attack" option, as that has a Stance Multiplier of 1.0. It's also helpful to use pets with no species resistances or vulnerabilities when that's possible, and to test defense with a pet that has the same defense boost as the attacking pet's Strength boost. (If you are unsure about what boosts and boost multipliers are, please see the beatmetoapulp petpage.)


If you can do all of those, the equation simplifies down to:

Damage_Done= Boost_Multiplier ( Number_of_offensive_icons * Boost_Multiplier - Number_of_defensive_icons )

However, since we are testing here, we're looking for the number of defensive icons, so you'll want to rearrange it like this

Number_of_Defensive_Icons = Number_of_Offensive_Icons - ( Damage_Done / Boost Multiplier )

Thus if you are testing a new item's dark defense by having a 55 Str boost ( * 2.0) gelert use Scroll of the Dark Star ( 6 dark icons ) to cautious attack ( * 1.0 ) a 55 defense boost ( * 2 0 ) kyrii who is using the new item (unknown icons) and cautious attack( *1.0) and you get a result of 6 damage, you can conclude that the item deals -- no wait, you stop and figure it out first.


Please note this this is only going to be accurate for damage values greater than zero. Please also note that this is assuming a mono-icon attack.

Next let's assume that for various reasons you can't do all of the recommended simplifications. Maybe you're testing a species weapon for a species with a resistance or vulnerability or maybe you can't find an opponent with a Str boost that matches your Def boost, or maybe the opponent you have needs to use berserk to deal enough damage to get anything through, or maybe all three.

There are more factors, but the method is exactly the same. Go back to the longer formula, solve it in terms of Defender's_Number_of_Defensive_Icons


and then plug in the numbers you do have.

For a full example, let's assume that the same 55 boost (* 2.0) gelert from before is using a Fiery Sun Scroll (6 fire) and Berserk Attack ( *1.5) against a 200 def ( *4.5) Scorchio (30% species resistance against fire, or 0.7 * damage taken) who is using an untested defense weapon and Berserk Attack ( * 1.5) and the damage result is 5 points.

Plug in the numbers and try that one out. Just watch out, it's tricky.


Got all of that?

Good! Big Grin

Those are the basics, and knowing them will let you accurately test the defense for 90% of neopian weapons. However, there are a wide variety of possibilities in the battledome and you may run into an oddball weapon in the other 10%. There's too much material for me to discuss all possibilities in detail, but if you find that your test results aren't making sense, here are some other things to watch out for:

  • The weapon may have Variable Defense - this just means that the weapon can defend different amounts when used. This is why it is important to test weapons multiple times to determine the average. An example of a weapon that can defend varying amounts is Carved Sansetsukon
  • The weapon may defend Fractional Icons which may round up or down differently depending on the boosts and stances of the pets using them. The best way to test for this is to have pets with different boosts and stances test the same weapon. An example of a weapon which defends fractional icons is Shiny Shoal Shell Shield
  • The weapon may provide Full Defense. This means that the weapon blocks all damage from icons of one or more types, regardless of the defense boost of the pet using it. The best way to test for this is to have a pet with a high strength boost attack a pet with a low defense boost. An example of a full defense weapon is Ring of the Lost
  • The weapon may provide Percentage Defense This is very similar to full defense in that it does not matter what the defense boost of the defender is, and the best way to test for it is again to have a pet with a high strength boost attack a pet with a low defense boost. The classic example of percentage defense is downsize!, although Thyoras Tear and Thick Smoke Bomb are similar.
  • The weapon may provide Reflection in place of, or in addition to normal defense icons. Phoenix has an article explaining reflection in detail here, but the short version is: we need screenshots with all relevant info for both pets, and who clicks "go" first matters very much for reflection. When testing a weapon for reflection it is helpful to be in contact with your opponent via IM so you can co-ordinate the turn order more precisely. An example of a reflective weapon is Double Dryer
  • The weapon might do an amount of Self-Damage. This means that when used, the weapon deals damage to the pet using it, usually in addition to more worthwhile functions. The most well known weapon that self-damages is Winged Scarab
  • The weapon may display Fake Icons. That is to say it might display blocked icons that are not actually blocked. Many defensive species abilities are like this, and something similar happens with the light reflection of Jhudoras Wand
  • The weapon may actually have display errors. For example, Accelerator Helmet displays defense icons on the wrong side of the battle log and the fire defense of Rod of Dark Nova is not displayed at all.
  • The weapon might have defense that varies against combinations of icons. This is why it is important to try all seven icon types against a weapon. For example, Patched Magic Hat defends extra physical icons when the wielder takes air damage.
  • The weapon might require Intelligence or Movement Prerequisites to work, or to behave differently for different values of such stats. An example of a weapon with stat prerequisites is Tunic of Warding
  • The weapon may function differently than any current weapon. Neopets does change constantly and they love to throw us curveballs. You're on your own figuring out how to test for this, and obviously I can't provide any examples.
Thanks for the guide. Smile
That's a nice guide.

It's also a good idea if this forum supported spoilers within spoilers.
I wondered why there wasn't a comprehensive guide to defense testing before. Great article, thanks a lot!
word of rabid, we alabate you rabid
GJ, ohh almigthy rabid by giving us this guuide...

rabid_schnauzer Wrote:
[*]III. The Strength and Defense boosts of both pets. We also need species, but that should usually be obvious from the screenshots


... unless of course your pet is painted invisible, hence we REALLY DO need to know the species Toungue

Great one, but you forgot to mention the fixed icon defence like Ice Shield Smile
thumbs up Big Grin

neomail Wrote:
I went ahead and read your article about testing defence. My question is: how do you account for the randomness factor (reported as 10% or 15% in different sources) while testing? if it is done by repeating the test, how many times you have to repeat before you are comfortable that you have averaged out the randomness?


There is no global randomness factor anymore. Constant weapons deal constant icons with no variation.

I'm unsure if there at one time was one and the workings of the BD were changed, or if the randomness factor was reported due to an incomplete understanding of factors such as species resistances/vulnerabilities or petpet assistance (disabled long ago) at the time those guides were written.

For individual weapons with variable results, yes you do have to run multiple tests. Usually 50 trials is enough to be reasonably sure of the average, but bear in mind that some weapons, for example Cardboard Baby Cabbage Cannon, do have effects that occur very rarely and far more trials may be needed to investigate those.

rabid_schnauzer Wrote:

neomail Wrote:
I went ahead and read your article about testing defence. My question is: how do you account for the randomness factor (reported as 10% or 15% in different sources) while testing? if it is done by repeating the test, how many times you have to repeat before you are comfortable that you have averaged out the randomness?


There is no global randomness factor anymore. Constant weapons deal constant icons with no variation.

I'm unsure if there at one time was one and the workings of the BD were changed, or if the randomness factor was reported due to an incomplete understanding of factors such as species resistances/vulnerabilities or petpet assistance (disabled long ago) at the time those guides were written.

For individual weapons with variable results, yes you do have to run multiple tests. Usually 50 trials is enough to be reasonably sure of the average, but bear in mind that some weapons, for example Cardboard Baby Cabbage Cannon, do have effects that occur very rarely and far more trials may be needed to investigate those.


Statistically, you need at least 10 of every possible outcome before you have a good, robust sample.

Amazing guide, thanks for the help I'll retest the Anit Gravity Boots
Quite useful for a n00b like me XD Hope no1 gives me a negative point for posting XD
Reference URL's