|
|
|
|
|
|
|
#1
Building Groups That Make Sense |
||||||||||
|
||||||||||
04/16/08, 5:12 AM
|
#2 (permalink) |
|
Glass Joe
|
Rationality and preferences
Being interested in rationality and preference relations (and wow) I think this is an interesting idea. There are, however, some problems involved in such a project at least as you have stated it. And since you people are elitist jerks, then this is probably the only place I can vent my concerns.
The idea seems to be that we are supposed to derive a (group) decision from individual preferences or needs. We could state that a class' needs and desires are the indifference sets describing the alternatives of which the members would be indifferent between. And that all members of the need set are preferred to all members of the desire set. Example: if xIy describes indifference between alternative x and y, and xPy describes strict preference between x and y then we can describe a healers's preferences as following. Healer: Restoration Shaman I Paladin, Shadow Priest I Drums of Restoration, Restoration Shaman P Shadow Priest, Paladin P Shadow Priest. What we are interested in, once we have identified needs and desires sets, is to aggregate those into a raid-leaders decision. In order for any aggregation rule to be rational we might say that it should satsify some conditions, say: Pareto condition: If all prefer x to y, then so should the group. For example, if everyone prefers that the resto shamans and holy pallas are in the same grp, then so should the grp. Non dictatorship: It should not be the case that the grp prefers x to y whenever individual i prefers x to y. That is, i should not be a dictator. Unbounded domain: The rule should be able to generate a rational group decision for all possible rational individual preference profiles. It should always be able to give us an answer. (Rational preference orderings require that they are complete and transitive). Irrelevance of independent alternatives: The group decision should not be affected by independant alternatives. The raid leader might for example suddenly prefer to have Illidan in his group, but since this is an independent alternative it should not affect the group decision. Ok, you probably see where I'm going with this: Arrow's impossibility theorem which states that it is impossible to create a group decision rule that satisfies the four conditions. Now, what relevance does this have to wow and assembling raid groups? Well, in wow we are not really interested in people's preferences. Or rather, people's preferences ought to reflect the needs of the raid group. So, what is at issue is to construct a raid group that maximizes the chance of success in a given setting. And to do this we don't need to go via desires or needs. A more promising way to approach the problem is to specify and solve a optimization problem where healing, dps, and survivability are what we want to maximize. Thus, we might be able to deduce the optimal (rational?) raid composition. This is, however, nothing that I'm interested in doing, but I hope that there are some engineers out there who will head the call... |
|
|
|
05/13/08, 11:57 AM
|
#3 (permalink) |
|
Glass Joe
|
Confused
...
Healing: Needs All: Enchant Weapon: Spellsurge Druid, Priest: Atiesh, Greatstaff of the Guardian I dont understand how Spellsurge is a necessity. Even if it IS a necessity, who would you put it on? Neither do I understand why Atiesh, Greatstaff of the Guardian, a level 60 legendary with far from healing quality stats, is a necessity. The spell crit rating? I understand the game has changed but TBC has been out for a while and I would beg to differ on the both of those stats. MAYBE those would be "wants" but there is no way those are needs. In TBC there are plenty of other procs that are more beneficial then spellsurge (i.e. set bonus' [namely druid] and/or Superior Mana Oil) |
|
|
|
|
05/14/08, 2:10 AM
|
#5 (permalink) |
|
Captain N
|
Have you considered putting 'Improved Expose Amor' into the encounter-specific section? On bosses being tanked by a Druid or Paladin, it provides more armor reduction than Sunder without the threat penalty imposed.
|
|
|
|
|
|
|
05/20/08, 3:33 PM
|
#6 (permalink) |
|
Glass Joe
|
Melee DPS and Tank desire
Might want to add the Balance Druid ability Improve Faerie Fire to the tank/melee desires list in the case that they are not hit capped.
|
|
|
|
|
07/13/08, 8:41 PM
|
#8 (permalink) |
|
Glass Joe
|
question
Healing: Needs
... Paladin: Concentration Aura Restoration Shaman: Mana Tide Totem Holy Priest: Circle of Healing I don't understand the common point between paladin & shaman (give buff for their group) and the circle of healing (holy priest). It's a heal for each group of the raid and not only the healer's group. Can you explain ? |
|
|
|
|
07/14/08, 6:28 AM
|
#9 (permalink) |
|
Si Tibi Narraremus Te Interficere Debemus
|
I can think of one of two things.
Either it should have been Prayer of Healing all along (which would fit because it's a group-only heal), or what I think is more of the case, at least one priest who is specced for CoH, because of its sheer utility. The second is how I'm actually reading it, because the 'needs' are what should be expected by 99% of the raid groups out there. |
|
|
|
|
07/18/08, 4:31 AM
|
#10 (permalink) |
|
Glass Joe
|
I guess he meant that you should have the AOE-healing in mind while setting up groups.
Putting people in a group who will never stand reasonably close together makes it impossible to Circleheal them. Prime example beeing Bloodboil where AOE-Healing is a major concern when you build groups you don't plan to heal with Chainheal. |
|
|
|
|
07/19/08, 5:04 AM
|
#11 (permalink) | |
|
Solution complicated; Dispense enlightening graph.
|
You gain a lot of utility from the existing game by paying attention to things of this nature. Small changes like "stand near this symbol" make big differences in healing requirements in a lot of places. |
|
|
||
|
|
|
|
08/24/08, 1:30 PM
|
#12 (permalink) |
|
Glass Joe
|
The most interesting portion of this article is also the hot spot of contention in my guild.
Melee vs. Hunters getting Leader of the Pack (LOTP). It would seem that Hunters and Warriors both have abilities that are keyed to critting, so those two classes are going to get the biggest benefit from the buff. If we were building an optimal raid with only one feral druid, which group gets it?? I have seen the melee automatically get it and the hunters have to pray for melee players not coming so one of us can get pulled into the melee group. Is it more advantageous to have the LOTP buff in a "hunter" group and let the melee be composed of a retadin and enhance shammy? Or are the melee going to be greatly gimped and we would miss the prime optimization of the buff? Or, should the group be stacked with top dps from the 3 classes regardless? Ehance shammy Feral Druid Combat Rogue Fury Warrior BM Hunter I have also heard much talk of Hunters being content (or even preferring) a shadow priest in place of LOTP. This is garbage in my opinion. Use potions, mana oil and get your pally tank to judge wisdom. Learn to switch Aspects until a pot CD is up and keep on the pally to get wisdom on the boss as much as possible (easily done with a ret pally in the group). My mana concerns are rarely enough to warant a shadow priest. My main concerns are always LOTP. |
|
|
|
|
08/25/08, 12:46 PM
|
#13 (permalink) |
|
Piston Honda
|
Your melee group should be
Enh Shaman Ret Paladin DPS Warrior Rogue Rogue There's no room for a feral. The 5% crit is more of a novelty than anything else. Windfury is wasted on the druid, so that's already a dps loss compared to giving WF to your Ret pally. A Ret Paladin allows JoW to be up near 100% thus means the hunters will not need a shadow priest (as a side note, switching to aotv is not viable for maxing dps and is not worth bringing up). It seems that many lesser guilds do not like giving their hunters proper support. Perhaps they are caster heavy, or perhaps they do not understand group composition, I don't know. In a 3 hunter raid, giving them a feral druid and resto shaman is key. They are the only two "group buff" classes that really benefit them. LotP will allow for higher Expose Weakness uptime (from the survival hunter), thus benefiting the melee already. For BM hunters, it means their pets will crit more, giving their group higher uptime on ferocious inspiration, giving the group more dps overall. Plus, with more crits, the hunters get more gftt procs allowing their pets more damage (if they are able to dump all the focus). For a 2 hunter raid, it's still preferred to give them the feral, and perhaps throwing a protection warrior into the group in place of the 3rd hunter. Even totem twisting, it's not a big increase for a hunter to get an Enh shaman over a resto. Yes, the pet gets unleashed rage (more ap, yay!) and the totems are stronger (twisting enhanced agi is not considerably better than straight normal agi totem) but you lose out on mana tide (which helps for BM in mana intensive fights such as mommy shahraz or council). Long story short: Don't ever put your single feral with the enh shaman when you've got melee who can benefit from WF. |
|
|
|
|
|
 |
|
||||||
| Article Tools | |
|
|
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Common Sense / Account Security | Praetorian | Public Discussion | 41 | 12/09/07 7:25 AM |
| Does it make sense to use Blade Flurry on bosses? | Jo | Public Discussion | 15 | 09/26/06 8:41 AM |
| Does it make sense to use Blade Flurry on bosses? | Jo | Public Discussion | 3 | 09/25/06 7:37 PM |
