Notes about Final Fantasy XI

created Sunday, 2005 February 6
updated Tuesday, 2005 March 15

1 What is this?

These are my notes about the online game called Final Fantasy XI. It is a graphical MUD produced by Square Enix. There are client programs for Playstation 2 & for Microsloth Winders.

2 Party Tactics

The most common battle tactic for a party if six is ``camp-&-pull''. The party finds a location & labels it the camp. Five of the party members stay there the entire time. One party member is designated the puller. When there is no battle & everyone is healed enough for the next battle, the puller runs away from camp, finds a beastie, whacks it over the head to get its attention, & runs back to camp where the six members of the party ambush the poor beastie.

2.1 Roles during battle

The important roles during a battle are:

tank

He is built to take damage. His job is to make the monster hate him so it will hit him & not the other party members. A Provoke ability is useful to this role. The best jobs for a tank role are (starting with the best ofthe best) Paladin, Warrior, & Ninja.

Having spent much of my time as a party's healer & observing different tanks in action, I highly recommend against using a Tarutaru as a tank. I don't mean to be funny, & I like Tarus in general, but they are terrible tanks.

Also, I recommend against a backup tank. Have just one tank. Let him take all the hate because your healer's duty is easier if she just has to worry about one character (the tank). Two tanks make it that much more difficult for the healer, & an ineffective tank means that every party member will take damager, which turns your healer's job into a frenzied nightmare.

The only reason for a backup tank, in my opinion, is if your single tank cannot take his share of damage in the fight even with the healer's help (like if you don't have a healer). In that case, get two tanks, but instead of designating one of them a backup tank, designate both of them co-tanks. They should try to split the hate evenly so that each taks $\frac{1}{2}$ of the damage during the fight. Hopefully each can survive the fight that way.

I like to refer to tanks as ``hate hogs''.

healer

This party member's primary responsibility is to use Cure spells to prevent the hit points of other party members from falling too low. She also casts buff spells on the party & enfeebling spells on some mobs.

A White Mage is the primary job for this role. A backup healer works well, & a Red Mage or a character with White Mage as her subjob can fill that role while filling others if that player can pay attention.

If the tank can heal himself, he should do that so that the primary healer can help the other party members. This is one of the reasons Palladins absolutely rock as tanks.

damage dealer

Does a lot of damage. Good damage dealers include Dark Knights, Black Mages (fondly referred to as nukers), Samurai, & Rangers. I guess Theives can be damage delaers with their Sneak & Trick attack abilities. I don't have any experience as a damage dealer, so I don't know which of those jobs is preferrable to the others.

others

Party slots 4, 5, & 6 should be filled with these roles, in this order of preference: second (or backup) healer, damage dealer, & damage dealer. There is no need for a second tank unless your single tank simply cannot obsorb all the damage he needs to during a single fight. See my notes about tanks for more details.

In my opinion, an optimal party would consist of:

1 tank (Palladin/Warrior is first choice, then Warrior)
2 healers (a White Mage being the primary healer)
3 damage dealers, ideally of different jobs. For example, Ranger, nuker, Samurai might be preferable to Ranger, Ranger, Ranger.

Clearly, I'm biased towards parties that hit hard & finish the fight quickly. I suppose a more defensive party would be possible, in which case you might ignore my opinions about two tanks generally being a suboptimal idea.

I'm interested in the idea of a tank-less party, too. Like, could a party use Bind, Silence, Slow, & other enfeebling magic to keep the beastie at a distance or make its melee ineffective while they hit it really hard & really fast? Could a party that was heavy with Black Mages use Bind & geography to keep the monster at a distance, with no danger from melee, while they nuke it to death? I'd like to try.

2.2 To Camp or Not to Camp?

Camping is silly, but for a large party, it makes sense. Here are the advantages to camping:

chaos

You'd think that six people isn't much, but it's amazingly difficult to organize six people. Camping minimizes the chaos by keeping five of those people in a single location the entire time.

Minimizing the disorganization of six people is the main benefit of camping.

rest

Mages can rest between battles. If you didn't camp, if the entire party ran from beastie to beastie, you would have to make special stops for the mages to rest.

Interesting question: So what if you had a six-person party that did not use mages, or maybe that used them only when the party's buffs expire. Yeah, yeah, it seems impossible for a party to work without a healer, at least, but let's jsut pretend you somehow do make a party that works without any mages. Since that party has no mages that need to rest, should that party camp?

A disadvantage with camping is that you are not fighting, therefor not earning experience points, while the puller is seeking the next target. A talented puller can minimize this disadvantage by leaving a fight just before it ends so that she is finding the next target just as the fight does end. Such a puller must be careful not to pull a new target if the rest of the party needs to rest, though.

If you have a party of three or less or the party consists of a bunch of people who are used to working together & communicate well, camping is not your best choice. The party could instead run through an area & mow down everything in its path.

2.3 Who Pulls?

Most people with whom I've discussed the camp-&-pull tactics insist that the tank cannot pull, but I've met three tanks who wanted to pull, & they were three of the five best tanks I've worked with. What's more, two of them were the tanks in the two best parties I've worked in.

The reason a tank can pull is that the character which starts the fight gets an extra little bit of hate. Hate is what tanks are there to hog, so why not let the tank get off to a good start with that first bit of hate? If you separate tank from puller, then the tank must provoke the beastie off the puller when the puller returns.

Another reason to let the tank pull is that the battle usually surrounds the tank & the beastie. So the tank places the battle. If the puller is the tank, he has full control over where to stop running, therefor where the beastie stops. Otherwise, the puller & the tank must execute a sort of hand-off maneuver, whether or not its conscious & official, to place the beastie & the fight.

Finally, some people have told me that the tank shouldn't pull because the puller gets damage, & we don't want the tank getting that damage. But in a properly functioning party, the healer spends her time curing the tank, so why not let the tank get that first dose of damage? It's not like he's not built for it (see my notes about the tank).

So I think the tank should be the puller, & I have yet to hear a good reason otherwise.

3 Capitalism vs Socialism

It was obvious to me, & probably to any American¹, that the economy of Vana'diel is an example of a perfect capitalist economy:

The auction houses allow buyers & sellers to execute their transactions efficiently. As a bonus, it provides anonymity.
Beasties usually drop loot when you defeat them, providing an income to anyone who can beat up beasties.
Craft skills combine objects of lesser value to create objects of greater value, providing a greater income to anyone who has the patience to practice the craft skills.
By selling the higher-valued objects at the auction house, the crafter enables the beastie beaters (because the objects are often better weapons & armor).
There is an endless supply of beasties to beat. In a way, civilization is at war with the beasties, an enemy that will respawn forever, never giving up. This drives the economy.
Yet civilization is in no danger of losing the war. Even though the hordes of beasties never end, they will never invade the cities & destroy (or even change) the way of life of the citizens.

It's a perfect capitalist economy.

After noticing these things, I told a friend that it was curious that a game such as Final Fantasy online could not make use of an socialist economy. No player would want to participate in a socialist economy. Curious that, & dissappointing because I'm socialist in real life.

I was wrong that no one would want to play in a socialist virtual world. The economy of Vana'diel is also a perfect socialist economy.

Your character has all the food she needs to live. (Eating to live is not a part of the game, though eating to enhance temporarily your abilities is.)
Your character's home city provides her an apartment (Mog House) free of charge.
Your character is always clothed. If you remove all articles of clothing, your character is still drawn with clothing.
Basic health care is universal. The main form of medical problem in the game is loss of hit points. To regain them, anyone just moves away from combat & performs the ``/heal'' command. The character's hit points will be restored completely & free of charge.²
Even death is universally cured.

So all services necessary for life (food, clothing, shelter, & health care) are provided free of charge to all characters. It is a perfect socialist system.

4 Efficient Strategy for Leveling

Some people will search for minutes to find exactly the right beastie to beat-up. Search time is time you are not fighting which is time you are not moving closer to a reward of experience points.

The question which concerns me is ``Should the puller search indefinitely for just the right beastie, or should she eventually settle for the next beastie she sees?'' How long can you search for just the right beastie before you would be better off taking the next beastie you see? Let's do the math.

Figure 1 shows some assumptions.

**Figure 1:** Assumptions for my calculations of XP-earning efficiency
$\begin{figure}\begin{enumerate} \item The party contains six characters. \item A... ... of calculations, but that has to wait for later.) \end{enumerate} \end{figure}$

The whole operation of battling beastie after beastie can be reduced to the process in Figure 2.

**Figure 2:** The battle-after-battle process
$\begin{figure}\begin{enumerate} \item Puller goes in search of next victim. \ite... ...awarded $X$\ experience points. \item Go to step 1. \end{enumerate} \end{figure}$

The rate of gaining experience points is $\frac{X}{P + B}$ . Obviously, players want to maximize this rate.

Table 1: Rates of earning XP for various beastie difficulties, assuming battle time is always 60 seconds

pull time	20	25	36	50	100	200
0	0.33	0.42	0.60	0.83	1.67	3.33
15	0.27	0.33	0.48	0.67	1.33	2.67
30	0.22	0.28	0.40	0.56	1.11	2.22
60	0.17	0.21	0.30	0.42	0.83	1.67
120	0.11	0.14	0.20	0.28	0.56	1.11
300	0.06	0.07	0.10	0.14	0.28	0.56

Table 1 ³ doesn't tell us much because it assumes that a battle is always 60 seconds. In fact, battles will be shorter for easier foes & longer for tougher foes.

Table 2 is like Table 1 except that it uses longer battle times for tougher foes. The battle times are shown in parantheses after the experience points; the times are in seconds.

Table 2: Rates, in

, of earning XP for various beastie difficulties, using different battle times for each beastie difficulty

pull time	20 (20)	25 (25)	36 (30)	50 (45)	100 (60)	200 (120)
0	1.00	1.00	1.20	1.11	1.67	1.67
15	0.57	0.62	0.80	0.83	1.33	1.48
30	0.40	0.45	0.60	0.67	1.11	1.33
60	0.25	0.29	0.40	0.48	0.83	1.11
120	0.14	0.17	0.24	0.30	0.56	0.83
300	0.06	0.08	0.11	0.14	0.28	0.48

Notice that for any rate in one column, you can find the same or a higher rate in a column to the left at a shorter poll time.

For example, if every 2 minutes you can find a 200 XP beastie, you'll earn XP at a rate of $\frac{0.83}{second}$ , but if you settle for 100-XP beasties & can find one per minute, you'll earn XP at the same rate.

What does this mean? It means that if you've been looking for a 200-XP beastie for one minute, you might as well take the next 200-XP or 100-XP beastie you see.

In fact, the table skews the rates in the favor of the high-XP monsters because it does not take into account death. When your character is KOed, you lose XP. That destroyes the progress you made from your most recent battles, thereby lowering the rate at which you earned XP. The chance of death at a tougher monster is much greater than with a weaker monster.

Also, the table does not take into account the ``take the next 200-XP or 100-XP'' beastie factor. If you have been searching for a minute & decide to take the next 100-XP or 200-XP monster you see, then sometimes the next monster will be worth 200-XP, which will give you a higher rate than if the next monster was always 100-XP.

It would be interesting to see a similar table which takes death & the ``take the next ...'' factor into account. I would also like to use more accurate battle time estimates. (The battle time estimates I'e shown are my gutt fealings from the last week of parties I've seen.)

So much for tables of numbers. Now let's solve for symbolic answers.

Let's make some new variables. They are in Table 3.

Table 3: Variables for comparing experience point rates between ``just the right beastie'' & ``the next beastie you see''

variable	meaning
$R_{t}$	XP rate with tough beasties
$P_{t}$	time to the next tougher beastie
$B_{t}$	time to battle tougher beastie
$X_{t}$	XP awarded to each character for tough beastie
$R_{e}$	XP rate with easy beasties
$P_{e}$	time to next easy beastie
$B_{e}$	time to battle easy besties
$X_{e}$	XP awarded to each character for easy beastie

In Table 3, ``tough'' & ``easy'' do not mean that the monsters checks as Tough or Easy. It just means that Monster A is tougher than Monster B; Monster B is easier than Monster A. You must spend more time in battle to defeat the tougher monster, but you'll earn more experience.

We want to find the pull-times for the tougher & easier monsters such that it's worth your time to take the easier monster even though you'd rather take the tougher monster. In other words, we want to know $P_{e}$ &/or $P_{t}$ such that $R_{e} \geq R_{t}$ . Since we already defined the variables in Table 3, the answer is just some algebra away.

$\begin{displaymath} R_{e} \geq R_{t} \end{displaymath}$

(1)

$\begin{displaymath} \frac{X_{e}}{P_{e} + B_{e}} \geq \frac{X_{t}}{P_{t} + B_{t}} \end{displaymath}$

(2)

$\begin{displaymath} X_{e} \cdot (P_{t} + B_{t}) \geq X_{t} \cdot (P_{e} + B_{e}) \end{displaymath}$

(3)

$\begin{displaymath} P_{t} \geq \frac{X_{t} \cdot (P_{e} + B_{e})}{X_{e}} - B_{t} \end{displaymath}$

(4)

Equation 4 tells us exactly when to stop looking for just the tougher monster & to accept easier monsters, too.⁴, but it doesn't mean much to the human eye unless we plugin some numbers.

But I guess I'll do that later.

5 Brief Observations & Heuristics

A Warrior's Provoke ability is a White Mage's best friend.

Silence is a mage's worst enemy.

A. Other File Formats

This document is available in multi-file HTML format at http://lisp-p.org/relia/.
This document is available in Pointless Document Format (PDF) at http://lisp-p.org/relia/relia.pdf.

Bibliography

Gene Michael Stover 2008-04-20