CharOp is the abbreviation commonly used to refer to the Character Optimisation board on Wizards of the Coast's official forums. It is where players of WotC's games, usually Dungeons & Dragons of variable edition, unite in order to try and break the fuck out of the system any way they can. CharOp is responsible for finding most of the exploitable loopholes and badly-thought-through features in D&D, creating character builds such as Pun-Pun, designing infinite feedback loops that make people infinitely good at something, and otherwise stacking bonuses in unexpected ways to break the game utterly. The only purpose of discourse is to make the best character possible; having a given concept actually make sense is mostly irrelevant, so a charop player would gladly use a half-orc wizard or halfing fighter even though it doesn't make sense in the system, if there were any loophole that would allow for it to be effective.
Before we can go any further in discussing CharOp, we need to narrow things down, and provide some motivation.
Optimization in 4e: What is needed?
Before you can optimize, you must first understand the game. Explaining Pun-Pun, for example, necessitates discussing snippets of rules from eight or more books, and this is certainly an optimized character for Dungeons and Dragons, where characters must deal with game-bending effects and situations too wild to list here. Thus, the thought processes for optimization depend on the game. This is a factor with what one sees in 4e, where very little can affect characters--you cannot even take away their hand weapons in this game, and even the 'rust monster' of 4e doesn't really destroy magic items (it only does so, literally, if the players want it to do so).
Primarily, the only thing you can do to a character is hit point damage, so an optimal character might be concerned about having hit points. Unfortunately, the RAW ('Rules as Written', the only thing an optimizer can use to make judgements) for monster damage is so pathetically low, the healing so jaw-droppingly high, that this isn't much of a consideration.
Similarly, characters primarily, almost exclusively, defeat monsters by dealing hit point damage. The other two methods, exploiting flaws in saves--broken and recently fixed by WoTC--and abuse of the Intimidate rules--broken and likely fixed at some point by WoTC, are too easily exploited to necessitate any lengthy discussion.
The primary way characters do damage is by hitting. And this leads to our fundamental philosophy of optimization in 4e: +1 to hit is everything. 4e is an extraordinarly narrow game, there's really nothing else that's relevant. All of a character's powers are keyed off scoring hits. If you can't hit, your powers are worthless (again, there is an exception, in the form of the 'pacifist cleric', but let's focus on the other possible characters, some 99.9% of the possiblities).
How bad is it? WoTC created a feat, "Weapon Expertise", that grants, just that, a +1 to hit at low levels (more at high levels, but that's besides the point). This feat, this simple +1 to hit, is so dominatingly powerful that it's considered a 'feat tax', as all characters, even non-optimized ones, MUST take this feat. Many campaigns simply house rule "all characters get weapon expertise for free", because, in fact, all characters must take it at some point, the sooner the better.
Someone ignorant of the system might think +1 matters more at low levels than at high, but, 4e uses a treadmill system. A first level character might have +7 to hit, and will attack monsters with a defense of 18 (i.e., they'll have a 50% chance of hitting). A 20th level character might have a +24 to hit, and this might sound better, but the game is designed to keep characters on a treadmill at all times. A 20th level character will fight monsters with a defense of 35, and so still have a 50% chance of hit.
In other words, "+1 to hit" is just as valuable at first level, as it is at any other level. +1 to hit is EVERYTHING 4e has to offer when it comes to optimization.
Another example: the most powerful Epic destiny is considered Demigod. Why? Because it grants a +1 to hit, in the way of a flat bonus to all attributes. No other epic destiny offers anything as game-breaking as this +1.
Yet another example: a +5 magic weapon is valued, according to WotC, at 225,000 GP. If bonuses were worth less at high levels, a +6 weapon shouldn't cost much more. Of course, a +6 weapon (i.e., granting an additional +1 to hit) is valued at 1,125,000 GP, a 900,000 GP increase, obviously worth MORE than any other plus, even worth more than all other plusses put together.
Even WotC acknowledges: +1 to hit is everything in 4e.
Why must you optimize in 4e?
Now that we've established that "+1 to hit" is everything, now we need the motivation for why optimization is important, and critical in 4e:
4e has been designed with optimization assumed, nearly all character abilities are based around combat, and combat is balanced around optimal characters. A typical fight with optimized characters, once you get into paragon, can take 2 hours, more if it's even remotely challenging. Using sub-optimal characters double that time, sometimes more, which is why complaints of 'long fights' are so common on the message boards. Such grindy battles are the results of players falling into stupid traps, such as building half-orc wizards or halfling fighters, or picking +2 proficiency weapons instead of +3, or even picking a cool "+1 Lightning Burst" weapon instead of the dull, but critical to the game, "+2 weapon".
Now, 4rries might claim that any race/class combination is viable, but this is pure rubbish. If you pick something that doesn't have the right attribute bonus (that is to say, a +2 to your attack attribute), then you're permanently gimping your character in a way that can never be fixed. Not taking Weapon Expertise at first level, for example, can be fixed by taking it at 2nd level, and similarly other mistakes, like not picking a high proficiency weapon, can be fixed as the character gains levels. But if you're the wrong race/class combo, it's all over, that character can only be a drag on the rest of the party. So '+1 to hit' needs to be focused on immediately at character generation, and never abandoned as a principle of character building.
So, next time you're caught in a 4 hour long battle in 4e, and the fight is grinding away endlessly, realize this is because someone forgot to grab a +1 to hit. Perhaps there's a halfling fighter or half-orc wizard in the party?
The Poor Human: A Case Study
Summary: Though +1 means everything in 4E, the humans having +1 to all their defenses and thereby nullifying a +1 on every attack against them is completely useless and makes the race awful.
Unlike D&D, humans have no real place in 4e, due to poor design. Let's look at humans from a CharOp point of view, to see why this must be the case.
Humanity's main bonus in 4e is their floating +2 attribute, unlike all other races that have a +2 to two fixed attributes. In theory this might seem good since the bonus could go anywhere, but realize the value of this drains away as soon as your character chooses a class (i.e., in a few seconds). As soon as you pick 'fighter', for example, your +2 has to go to strength, so immediately your human is behind every single race that has a +2 to strength and a +2 to some other attribute. Since this happens regardless of what class you pick, the 'floating' benefit is completely nonexistent in practice. Not getting that secondary +2 to an attribute means humans are behind on all secondary attacks and effects, an insurmountable penalty just for being human. Do the other advantages make up for this somehow? Let's take a look.
Humanity's next supposed advantage is a 'free' feat. This is negligible at best. All non-human races receive a built-in feat, such as "Elven Accuracy" or halfling's "Second Chance". Humans get no built-in feat. So this 'bonus' simply makes up for the lack of built-in feat. Granted, this is a floating feat, and if WoTC printed broken feats, this could be an advantage. Unfortunately, there is only one broken feat, Weapon Expertise; as all characters get a feat at first level, the human advantage amounts to nothing. If there were TWO broken feats, then humans would be ahead...until second level, when all characters could pick those two broken feats. Similarly, no matter how many broken feats WotC theoretically creates, humans must fall behind. Even if one considers racial feats, half-elves have access to all human feats, so, once again, humans have no way to get ahead there, either. Recent splatbooks have given more and more powerful racial feats to non-humans, putting them very far behind even if they could catch up.
Humanity's next supposed advantage is a bonus skill. Unfortunately, all non-human races get a +2 bonus to two skills. 4e is designed via the 'treadmill' discussed earlier, so that it's much better to be good at a few things than it is to be mediocre at several things. Thus, humans are mediocre at three skills, while non-humans are good at two skills...this is not an advantage, and is usually a drawback.
Additionally, humans get a bonus at-will power. This, too, is a fairly minimal advantage. This isn't simply because at-will powers are among the weakest powers of the game, or because this 'extra' at-will will be a power that's available to all other races anyway, but also because a character can generally only take one standard action a round. Much like the human 'floating' ability modifier, this 'bonus' ends as soon as the player makes a decision for what to do. The game is designed so that all classes can have a fine number of options with 2 at-wills (the default), humans getting a third at-will, while sometimes, occasionally, is helpful, in very special situations, usually means nothing, and there's no way to 'optimize' around using the weakest powers in the game, powers that any other race could use just as easily.
Humanity's final supposed advantage is a small bonus to NADS, the non-armor defenses. At least 60% of all attacks (arguably more) are against armor, so this might seem like a small bonus against 40% of attacks, but it's not nearly this good. Non-humans get an additional +2 to an attribute, which raises one of the NADS. Poor design means all characters will be vulnerable in one NAD, and the +1 bonus won't be of much help here (being hit on a 2 or better, either way). So, arguably, the humans get a bonus that might be a factor 5% of 13% of the time (i.e., 00.65% of the time), soaring all the way up to 1.3% of the time for an extremely defensive build. This is nonetheless an advantage, but pales in comparison to the many ways humans are inferior.
Sub-Optimization: A Case Study of Dwarves
Summary: If you roll a party of Dwarf Fighters, fighting an encounter one level below you will take 12 hours.
It's often argued by those ignorant of the game that you can pick a race that doesn't necessarily have the appropriate +2 bonus to a primary attack attribute, but let's take an example to see what happens in this situation.
Consider someone who falls into a few very common traps (by 'trap' it is meant a flaw in the game design by which a player can be tricked into making a bad decision), and doesn't take Weapon Expertise because it's "just a +1". Keep in mind, there was a time when Weapon Expertise wasn't even an option, and it's quite possible WoTC will add more feats in the future--the end result is still the same, however.
Trap 1: The player chooses Dwarf fighter. This is a common error to make, since the PHB on page 36 says dwarves make good fighters. This starts him with a 16 strength.
Trap 2: The player picks a +2 proficiency weapon, like a warhammer. This, too, is an easy trap to fall into, as dwarves get bonuses with warhammers, via feats (we'll assume he picks such feats, as well, which up his damage, but not, alas, the all important to-hit).
Trap 3: Instead of picking the highest bonus possible for his magic weapon, the player takes a lower bonus, in exchange for a whiz-bang power. In this case, the cool +2 Lifedrinker weapon (level 10), as opposed to a +3 magic weapon (a level 11 weapon, but possible he could have chosen it).
By level 10, this dwarf fighter will have a 4 (str) + 2 (prof) + 5 (level) + 1 (fighter bonus) + 2 (magic) = + 15 to hit. He'll be dealing around an expected 15 points of damage a hit, assuming optimal (for his level) magic items.
Now, let's assume we have a whole party of such characters, and pit them against a pair of level 11 vampire lords. This is a "level -1" encounter, which should be pretty easy. To keep combat simple, each vampire will simply sit in a corner and only perform basic defense.
Vampires have an AC of 29, so the dwarves will only hit 35% of the time, let's just call that 6 points a round, rounding generously. Three dwarves attack one vampire, two attack the other, since they're in corners. Let's focus on the vampire with three dwarves attacking it.
The vampire starts with 186 hit points, but has an action point and can 'second wind' to gain another 46, so that's really 232 hit points. It also regenerates 10 a round, so the three dwarves put together will effectively deal 8 points of damage a round to it. This means 39 rounds of combat, assuming at-wills. Encounter powers will reduce this somewhat, dealing nearly double damage, so let's call that 36 rounds until the first vampire drops, neglecting the vampire's Dominating Gaze and Blood Drain, either of which could vastly extend combat time. We won't use Dailies, since, after all, Dailies are for 'tough' encounters, and this is a simple one.
A round generally takes 15 minutes, so that's a 9 hour combat, right there...and that's just the first vampire. The second vampire will die around 3 hours later.
So, 12 hours for a 'level -1' fight, with characters that are 'sub-optimal'. Optimal characters would likely end the fight in 2 hours or so.
Yes, you can make a dwarven fighter, but you'll be crippling the party by doing so, and should reasonably expect dirty looks from the other games around the table if you use it, because you're literally screwing everyone by using such a character.
Because character optimization in 4e is limited, the game design lends itself to a fascinating and entirely new form of optimization: optimizing the entire party.
One example is the "four orb wizard" party. Because orb wizards get spells that can completely debilitate a monster, or even groups of monsters, for two rounds or so, a group of 4 such wizards can completely shut down a typical encounter for many rounds. While the party could be entirely composed of wizards, commonly such parties have one very high damage dealer (eg, two weapon melee ranger), who can 'coup de grace' for extreme damage; by the mid-teens, such a character can deal four hundred or more points of damage to a helpless monster, enough to slay even major dragons in a few rounds.
Another example party is the 'Radiant Whore' party. This includes 4 paladins, and a warlord. The paladins can deal a large amount of radiant damage, and can inflict high levels of 'radiant vulnerability' on a chosen monster. The warlord can multiclass and take a path that allows it to exploit this vulnerability, as well as several powerful abilities that work best with melee characters. The end result is a party, with no 'striker' characters, that nonetheless deal 'striker' level damage consistently, even with their basic attacks, with all the benefits of being a party full of defenders and a leader.
Optimization in 3.5
1) Play a wizard, cleric, druid, or artificer
2) Know the multitude of options the multitude of splats offers well enough to find broken combinations
3) Make sure your DM is semi-retarded
4) You're done!