The Art of Optimization

Check here for guides on how to optimize your builds for maximum fun and efficiency!
Post Reply
User avatar
Fialova
Magical Liopleurodon
Posts: 4745
Registered for: 11 years 2 months
Location: The Great North
Contact:

The Art of Optimization

Post by Fialova »

This was originally written by Phryxian_lawnmower of the WOTC forums, I am merely archiving it here due to that forum's deletion.

It seems a lot of discussion on these board arises from differing viewpoints regarding the nature of optimization. This thread has been erected to provide a place to discuss the nature of optimization and the rules governing the game and thus optimization. When consensus has been achieved over these issues it could pose as a background for discussions, character evaluations and advice. Below I willl formulate my view on optimization as a whole within these boards.
Goals of Optimization
[SHOW]
The goal of Optimization is to make the best character possible. What makes the best character possible however, will often be defined by your group rules. We all know Pun-Pun and no one would bring that to the table seriously. Optimization encompasses role-playing your character in the best way possible and creating a killer backstory, but that is not within the scope of these boards so it is of no issue to us. These boards are concerned with making the best ‘character sheet possible’. Optimization, here, tries to make the best choices regarding the various attributes of your character and is especially interested in the numerical attributes of said character. These attributes can be tested best, in simulations and can thus be judged best regarding their advantages. If one would want to understand what it means to make the best ‘character sheet possible’, one would need a criterion to determine what makes a character good. This could not just be called min-maxing, because one would still need to know what makes certain numbers better than other. So it is within the nature of the game that we need to find the goals of optimization.

There are three factors in the nature of the game: Attributes of your character, attributes of the world around you and the rules governing the relations between all these attributes. Generally speaking, we could say the PHB and the DMG provide, through the intermediation of your DM with the second and third factor. And the Player defines, within the rules governing the relations between all these attributes (PHB and source material), his character and thereby the first factor. Other players’ characters are another factor or could be part of the attributes of the world, but let’s not delve in too deep just yet. Given these three factors, what would define the best character? Probably the character that has been assigned those attributes, that, within the rules governing the relations between all attributes, attains the most success in achieving in-game goals.

Common in-game goals are surviving, winning battles (defeating monsters, helping team-mates, etc…), withstanding traps and hazards, overcoming skill challenges (whether it’s the 4e skill challenge or just challenges that need to be overcome by using skills), puzzles and such. General Optimization is concerned with attaining all the in-game goals, but optimization on these boards is especially concerned with distributing attributes (stats, defenses, attack bonuses, powers, feats, etc…) in such a way as to attain the most in-game goals. Therefore optimization (here) can be defined as distributing attributes in such a way as to make a character that can attain (most of) it’s in game goals. This neither confines optimization to just eking out the most numerical benefits, nor expands it to optimizing roleplay or puzzle-solving. Difficult areas of optimization are tactics, powers such as the druids wild-shape (PHB2) and Rending Fear of Khirad (dragon 366), mobility and status affects. These are difficult areas, because their effects on attaining in-game goals are less obvious than those of attack bonuses, and perhaps more privy to group play style.
Theoretical vs Practical Optimization
[SHOW]
Practical optimization tries to achieve in-game goals in an actual in-game setting. Theoretical optimization is only interested in achieving goals defined by the game. This is why practical optimization is concerned with brokenness, while theoretical optimization merely scoffs at brokenness, even tries to achieve it. To put it otherwise, practical optimization tries to build an optimal and balanced character, theoretical optimization tries to build an optimal character. This is one of the most crucial distinctions in optimization. Prime examples of theoretical optimization are Pun-Pun and Horizon Walker/Warmaster builds. Prime examples of practical optimization are the handbooks and the Lordduskblade builds.
Broad vs Narrow Optimization
[SHOW]
The most general kind of optimization optimizes every aspect of a character. Certain parameters can be set however, regarding which aspects of the character need to be optimized. A build request could specify the class or certain feats for a character for instance. Optimization here has to build the best character sheet possible within the parameters of the request. These parameters themselves (such as race/class choice) could themselves be suboptimal, but it is the nature of this puzzle, that one has to operate within the confines of these parameters. Optimization here has a much narrower scope of operations.

Optimization in the broadest sense leads to threads about party tactics, calculation theory and theoretical optimization. As noted above theoretical optimization is less constricted than narrow kinds of optimization, such as optimizing build requests and providing solid feat choices. In other cases theoretical optimization can actually be more narrow that practical optimization, because it sacrifices survivability for extra damage, for instance.
(Group) Rules vs Common Sense
[SHOW]
One of the things I hear a lot on these boards is “That can’t work, just use common sense and you’ll see it makes no sense.” It’s true. Most things a character can do just don’t make any sense if you apply common sense. But that’s only because our common sense is constructed on the basis of our common experience. Surely in our experience most of us don’t permit spellcasting, and a lvl 2 character withouth a check penalty, good strength (18) athletics trained and a natural 20 check on his jump check clears 6 squares, that’s 30 feet. Congratulations, you’ve just established a world record! Now, this is just the start, it gets even more epic. Regarded by our normal standards the in-game world and characters are quite extraordinary. It is entirely within the power of the DM or group to determine that some capabilities and effects are too fantastic for their game, but it is another thing to say that something can’t work, because it wouldn’t by common sense rules.

There are ofcourse base assumptions players and DM’s will make about their game world that are challenged by the actual rules. Some optimization shows us where the rules conflict with our base assumptions. The DM or players could decide that their base assumptions are more important, but this can differ from group to group. It is important to remember that the laws of your game world can differ wildly from the laws of our natural world. The normal game assumptions even do on a very base level. Magic is commonplace, and characters are considered to be extraordinary individuals with great power, that only increases as they level. As seen from the game world, as implied from the books, it might not be strange for your character to have certain capabilities. It is therefore unwarranted to dismiss certain ideas on the basis of the ‘common-sense’ argument. It is a more healthy attitude to dismiss certain ideas, because they conflict with the balance of the game or because they conflict with the game setting. Or even better, the rules.

Capabilities that conflict with common sense, cannot however be dismissed because of this alone. One has to first insist that the in-game world has to follow our common-sense rules and this isn’t necessary. It’s even necessary that the in-game world doesn’t follow our common-sense rules, but it’s up to the DM (and players) to determine to what level this happens. For some interesting reading, there was a thread/post on the 3.5 boards regarding roleplaying character development (gaining powers, feats). I believe it was related to the book of nine swords, but I can’t seem to find it however. If someone could post a link I would be very grateful.
Exclusion, Inclusion and implied Inclusion
[SHOW]
One of the main arguments against some ideas reads “Just cause the rules don’t say you can’t, doesn’t mean you can.”. This is a solid argument at it’s core. Sometimes however, it does not apply. The rules of 4th edition work in mysterious ways, just kidding. They basically work like this: in the PHB and other books the rules of play are provided. There are general rules (e.g. combat rules) and specific rules (feats and powers). The general rules provide with a raw network of how things work and the specific rules give you the exceptions to these rules. One way to look at it is this: the general rules specify which things you can’t do, by specifying which things you can. More precisely, they specify how players and DM’s should handle certain situations (if they want to follow the official rules in these situations). Because it is specified how you should handle certain situations, certain ways of handling these situations are excluded. The Specific rules give you alternative ways of handling these situations, they include these alternatives.

To make this more concrete: there are general guidelines for Opportunity Attacks, these state that it has to be a melee basic attack, which things provoke, how many you can take, etcetera. Now feats such as Heavy Blade Opportunity and Polearm Gamble provide exceptions for the general rules of OA’s. They include alternatives. Another interesting rule interaction is between elemental weapons and weapon-implement users. The general rules say weapons are used for weapon attacks and implements for implement attacks, but swordmages and sorcerers can use weapons as implements. They include certain weapons in the implement rules.

These specific inclusions however are not clear enough on how far the inclusion goes. Are the weapon powers also available to someone who uses the weapon as an implement? The rules that make the weapon available as an implement don’t specify it and neither do the general rules exclude the use of weapon powers, when the weapon is used as an implement. So what does this leave us with? Is the use of weapon powers implied or is it restricted by the general rules?
In this case CS says it’s restricted, but the books aren’t clear at all, therefore errata is necessary.
RAW vs RAI
[SHOW]
One other major point of discussion is RAW vs RAI. With Raw we mean the Rules As Written and with RAI the Rules As Intended. The problem with RAW is that most rules are unclear and need some interpretation to make them work. This would leave us with RAIn, Rules As Interpreted, not RAW. The strength of the argument using RAW then becomes as strong as the interpretation of the RAW. RAI on the other hand is not without it’s problems. It is very hard to know what the intent of the writers was, since we don’t have a direct line of communication to the developers. Customer Service is not very reliable either and does not guarantee that we heard from the developers themselves.
(Un)balance
[SHOW]
A game requires a certain balance to be a game. A game needs challenge to be a game, and if there is no question of failing a challenge we can hardly call it challenging in any way. Likewise dnd is a cooperative game, wherein the DM and players cooperate to make a pleasant game experience (ideally). If a player would become so powerful that he eradicates the need for cooperation between the players the game could also seize to be a game (for all the players), because the overpowered character could just do everything by himself or he’d have to hold back.

Just imagine a party with Pun-Pun in it. Either Pun-Pun would deal with all the party’s challenges or he’d have to refrain from using (most of) his abilities. In this case the player has upset the balance between the power of the players and thereby limited the freedom of the players to play as they desire. Playing an overpowered character can force certain play patterns on players. It can thus destroy the game.

Thus there are two ways in which the balance can be destroyed, by eradicating the challenge and by eradicating the need for cooperation. Eradicating the challenge technically is a greater danger to balance, than eradicating the need for cooperation, because a player that does not need to cooperate can still cooperate, but a player that does not need to cooperate would probably have eradicated the challenge too.
Brokenness
[SHOW]
Anything that tips the scales to your favor is optimization, anything that knocks them over or places so much weight on them that there’s no counterweight or only a counterweight that would make the other players contributions insignificant is broken.
If something eradicates the challenge it is obviously broken. If something would require the challenges faced to become too challenging to remain challenging at all to such a degree that the challenges become too challenging to the other player’s it breaks the game. It must be noted that there are multiple degrees of challenge and that the player’s can face. The problems ( such as hitting AC, acquiring stealth, defending from attacks) that players can face can have a varying challenge, and the power of the characters influences the degree of challenge experienced by players. A character with +15 to hit has a better chance to hit AC 20 then a character with +10 to hit, so hitting AC would be less challenging. Something is only broken if it tips the scales in favor of the player too much, but we don’t really have good guidelines for deciding how much is too much.
Fallacies
[SHOW]
Stormwind Fallacy

Good roleplayers do not optimize. This is a fallacy because roleplaying and optimizing can coexist.

They say that (a) if a character is optimized it cannot be roleplayed well. Or inversely that (b) if a character
is roleplayed well, it cannot be optimized.

Those who believe tend to assume (1), for if they would not, there would be a middle ground.

(1) Either a character is roleplayed well and is not optimized or a character is not roleplayed well and is optimized
(2) Character X is not roleplayed well
(3) Character X is optimized

But we don't believe that bad roleplay necessarily breeds good optimization.
Really all one needs to do is bring an optimized character to the table and roleplay him beautifully.
(Good roleplayers don’t bring pun-pun or the apocalypse twins to the table, however.)

Mayfly Fallacy

I’m at the endgame so I can have an overpowered character. Just because you're close to the endgame doesn't mean something has the right to be broken. Whether you’re at the start or the end of the game, having a broken character breaks the game. See

Common Sense Fallacy

This is discussed above. When someone says “You can’t do this or that, because common sense disallows it” he is making the common-sense fallacy. Remember, it can make sense even though it doesn’t make common-sense.
Post Reply