Fencer Archive

* Fencer Special Attack Information *

After several weeks of work cross-verifying all the data, this is complete. All of the following data is extremely accurate (less than 0.5% error factor in worse case) as of the time of this post.
I want to preface this post by saying this wasn't as straightforward of a project as one might think. (As I had thought.) I had to keep starting over for a variety of reasons from server problems; patches; bugs; misconception about the differences between floaty numbers, chat numbers, and real damage; and approaches to verifying the numbers. I'll describe the process below for those that are interested. Those that only want the numbers and don't care where they came from, then just skip down to the results beginning in the following post.

At first I began the testing using a gaderiffi baton, and trying to find the numbers using average numbers. I thought that having no AP, and a higher base damage would allow for more accuracy in the data. Well, that backfired. During the calculations I realized that the greater the spread of base damage (min to max) on the weapon, the greater the potential error factor in the final numbers. Here is why:

For one, averages give good ballpark figures, but the error factor is not truly known, and is potentially quite high. Therefore, in order to make sure I had good data I had to be sure I knew the minimum possible damage a special could produce, as well as the maximum possible damage.
Next, the wide range of results a gaderiffi can produce was a bad choice. If a weapon had a damage range of 1-3, then there is three possible base damages: 1, 2, and 3. That means the weapon above (1-3) with a special multiplier of x2 has 5 possible outcomes: 2, 3, 4, 5, and 6. Which told me something else... the multipliers are applied directly to the base damages BEFORE the random number generator determines the number in the spread. If the multipliers were applied after the determination, then the results could only be 2, 4, and 6. Yet the full range of numbers between 2 and 6 come up... so the multiplier has to be applied first. This was an interesting tidbit that made a big difference in how I would have to collect the data. For the above 1-3 weapon, the likelihood of seeing the lowest number is 1 in 5 (only 20%). To bring the theoretical probability of seeing the lowest (or the highest) possible number at least once up to 100%, I must get at least 5 samples. Of course, that doesn't assure it, but it does gives me a good ballpark number of samples to shoot for.
If that simple 1-3 weapon used a x4 multiplier the new possibilities become 3, 4, 5, 6, 7, 8, 9,10,11, and 12... a range of 10 numbers. I would need 10 samples to get a reasonable chance of seeing the far end numbers, which give the most accurate numbers. Not too bad so far. But now increase the base spread to that of my gaffi, 183-309 (damage sliced) and you get a base range of 127 numbers. Apply a x4 multiplier to that spread and you end up with a range of 505 possible damage numbers. I have to get 505 valid samples to have a 100% (theoretical) probability of seeing the highest and lowest numbers. Multiply that times 16 moves and... well, you can see how fast that adds up.
So I changed my weapon to my brawler noob weapon (yep... still had it in the bank, I'm a packrat). The base damage spread was only 14-28, for a base spread of only 15 numbers. This allowed me to gather all the necessary data within a more plausible amount of valid samples. Instead of hundreds of individual valid tests to have just 1 reasonable shot at seeing the numbers I needed, I could get that shot in less than 100. Me, I didn't trust that 100%, so I calculated how many samples I would need to have a 100% probability of seeing the numbers I wanted and quintupled it. I feel very confident in the results.
There was still a fly in the ointment. The multipliers I was getting were all very odd. Not the simple 1, 1.5, 2, 2.5, etc., we were expecting. The multipliers were calculating out several decimal places, and very unexpected numbers. Ones like 2.78, 3.7, 5.6, and 6.53. I thought it might be due to rounding issues, so I replaced these values with 2.5, 3.5, 5.5, and 6.5... rounding as expected... but the results didn't add up. I was seeing numbers the rounded amounts wouldn't allow. It took a few days to figure out the problem. Actually, I had already been aware of the issue, but I hadn't taken it into consideration as being the complication. It was already apparent that even before a special is performed, you can do more damage than what the listed base damage would allow for. I had always guessitmated this differential to be double what was listed on the weapon... a straight x2 multiplier. Nope. It turns out that Fencers have a damage adjustment on ALL damage (even before the multipliers for specials) of x1.86. Your guess is as good as mine on why this was the number that has been settled on. I guess the extra 0.14 to make this a straight x2 would have made us overpowered. (/rolleyes) When I applied this to all my wacky numbers they all locked in hard to the nice solid numbers I had expected. Mystery solved.

Last night, I finally finished, and hopefully the results are something that will prove useful to the community I care about so much. So, without further delay, I give you the Fencer Special Attack Data. (Delay information was not possible to accurately derive at this time due to the speed cap. HAM information to follow soon... pending some important potential changes.)

~Lyahpar~

La Femme Fencer

"Thrust if you love Fencers!"

Message Edited by Raphayl on 11-28-2003 11:32 AM

Attack: Standard Melee Attack
Macro Name: /attack
Chat Name: “swings at”
Damage Multiplier: Listed weapon damage x 1.86 (Standard Fencer Multiplier, and is applied to ALL Fencer attacks in addition to the listed multiplier.)
Info: This is the damage done when not performing any specials. It is the damage that will appear in the combat chat box, and does not include any reductions due to the target’s defenses. It will randomly target any HAM pool, but appears to have a slight preference towards the Health Pool (about 40-50% of the attacks hit Health, 20-30% Action, and 20-30% Mind.)

Attack: BlindHit1
Macro Name: /melee1hBlindHit1
Chat Name: “Blinding Stab”
Damage Multiplier: x 2.0
Target: Single
Effect: Blind (White “eye” icon on the target.)
Info: Blind status significantly reduces the target’s accuracy, melee defense, and ranged defense. Duration is a 5 second minimum, but is random and varies based upon the target’s Defense Vs. Blind.

Attack: BlindHit2
Macro Name: /melee1hBlindHit2
Chat Name: “Blinding Slash”
Damage Multiplier: x 2.5
Target: All in range
Effect: Blind (White “eye” icon on the target.)
Info: This is the same effect as bestowed in BlindHit1, and is no more potent. The only difference is that this is a multiple target area attack. All targets in weapon range will be attacked, however a valid target must be chosen for the attack to work at all.

Attack: BodyHit1
Macro Name: /melee1hBodyHit1
Chat Name: “Sai Mai”
Damage Multiplier: x 1.5
Target: Single
Effect: Targets Health Pool
Info: This is an efficiency attack. All of the damage is applied directly to the Health Pool every time, rather than randomly across the 3 HAM bars.

Attack: BodyHit2
Macro Name: /melee1hBodyHit2
Chat Name: “Sai Sun”
Damage Multiplier: x 2.5
Target: Single
Effect: Targets Health Pool
Info: Same as BodyHit1.

Attack: BodyHit3
Macro Name: /melee1hBodyHit3
Chat Name: “Sai Tok”
Damage Multiplier: x 3.5
Target: Single
Effect: Targets Health Pool
Info: Same as BodyHit1.

Attack: DizzyHit1
Macro Name: /melee1hDizzyHit1
Chat Name: “Skull Stab”
Damage Multiplier: x 2.0
Target: Single
Effect: Dizzy (White “whirlwind” icon on target.)
Info: By itself this effect does nothing. However, if the target attempts an upward posture change while Dizzy, they can potentially suffer a “Fall Down”. The effects of a “Fall Down” are identical to those of a Knockdown, except that it doesn’t count against the delay timer as a Knockdown. Duration of Dizzy is a 5 second minimum, but is random and varies based upon the target’s Defense Vs. Dizzy.
Special Info: This is designed to be combined with the Novice Brawler “Lunge1” attack to force a posture down (which usually inspires the target to attempt to “get back up”), or a regular Knockdown.

Attack: DizzyHit2
Macro Name: /melee1hDizzyHit2
Chat Name: “Skull Slash”
Damage Multiplier: x 3.5
Target: All in range
Effect: Dizzy (White “whirlwind” icon on target.)
Info: This is the same effect as bestowed in DizzyHit1, and is no more potent. The only difference is that this is a multiple target area attack. All targets in weapon range will be attacked, however a valid target must be chosen for the attack to work at all.

Attack: HealthHit1
Macro Name: /melee1hHealthHit1
Chat Name: “Shiim”
Damage Multiplier: x 1.5
Target: Single
Effect: Bleeding and Wounding (Red “droplets” icon on target.), Target’s Health Pool
Info: This is an efficiency attack. This attack produces two consequences. The first, but least significant, is the high probability of infliction of Wounds in the target’s Health Pool. The primary effect is that of Bleeding, which drains the Health Pool over time. The damage listed in chat is the amount of the Bleed established, but the attack itself does NO immediate damage. 15 seconds after the attack, the indicated damage is automatically deducted from the target’s Health pool, and again every 15 seconds after. This will continue whether you are fighting, incapacitated, or dead. The Bleeding will stop on its own after 10 cycles. First Aid and Meditation can remedy Bleeding sooner.
Special Info: HealthHit1 stacks with HealthHit2, HealthShot1, and HealthShot2 with each on their own timers. Multiple hits with HealthHit1 are NOT cumulative, and do NOT reset the timer unless the damage is higher than the current HealthHit1.

Attack: HealthHit2
Macro Name: /melee1hHealthHit2
Chat Name: “Shiim Shiak”
Damage Multiplier: x 3.0
Target: Single
Effect: Bleeding and Wounding (Red “droplets” icon on target.), Target’s Health Pool
Info: Same as HealthHit1.

Attack: Hit1
Macro Name: /melee1hHit1
Chat Name: “Cho Mai”
Damage Multiplier: x 2.5
Target: Single
Effect: None
Info: This is the standard special attack, designed to simply boost damage output.

Attack: Hit2
Macro Name: /melee1hHit2
Chat Name: “Cho Sun”
Damage Multiplier: x 3.5
Target: Single
Effect: None
Info: Same as Hit1.

Attack: Hit3
Macro Name: /melee1hHit3
Chat Name: “Cho Mok”
Damage Multiplier: x 5.0
Target: Single
Effect: Special, see below.
Info: Same as Hit1, except that there is the added possibility (around 30-40%) of inflicting Blind on the target.

Attack: ScatterHit1
Macro Name: /melee1hScatterHit1
Chat Name: “Scatter Stab”
Damage Multiplier: x 3.0
Target: Single
Effect: Targets Multiple Simultaneous HAM Pools
Info: This attack spreads its damage across 2 - 3 HAM Pools. The targeted HAM Pools are randomly chosen, as is the damage distribution.

Attack: ScatterHit2
Macro Name: /melee1hScatterHit2
Chat Name: “Scatter Shiak”
Damage Multiplier: x 4.0
Target: Single
Effect: Targets Multiple Simultaneous HAM Pools
Info: Same as ScatterHit1.

Attack: SpinAttack1
Macro Name: /melee1hSpinAttack1
Chat Name: “Slash Spin”
Damage Multiplier: x 2.0
Target: All in range
Effect: None
Info: This is the basic Area Attack. All targets in weapon range will be attacked, however a valid target must be chosen for the attack to work at all.

Attack: SpinAttack2
Macro Name: /melee1hSpinAttack2
Chat Name: “Blind Spin”
Damage Multiplier: x 3.0
Target: All in range
Effect: Blind (White “eye” icon on the target.)
Info: Same as SpinAttack1, except that this also inflicts Blind, just as in BlindHit1.

***Additional Attack Note: All of the above Fencer Special Attacks impart an Accuracy bonus of approximately 20-30%.

~Lyahpar~

La Femme Fencer

"Thrust if you love Fencers!"

Awesome job! Thank you !

Thanks Raph! -- This is essential stuff to those of us who really want to maximize your er... profuctivity..... as it were....

Seena

---

M0G4N wrote:

Are you certain that the damage multiplier on the DizzyHit2 is greater than that for the SpinHit2 (3.5 > 3.0) ? I understand that you have said that the data collected has a 0.5% error margin, but that doesn't really include the error of typoes

---

---

KaliMirchi wrote:

Thanks Raph! -- This is essential stuff to those of us who really want to maximize your er... profuctivity..... as it were....

Seena

---