If you own a Core Rulebook and a Bestiary, what Pathfinder book should you buy next? A campaign setting book or an adventure module would be good answers, but if you're looking for more character options, the best answer would be the Advanced Player's Guide. This was Paizo's first big player-oriented hardcover to be released after the Core Rulebook, and it's safe to say they knocked it out of the park. This book has stood the test of time and still contains fantastic options for the game even though it was released several years ago. If you're playing PFS on a budget, for example, and you have to be choosy with what books or PDFs you buy, start with the Advanced Player's Guide. You'll find enough options in there to keep you busy for years.
What follows is a chapter-by-chapter review. Do keep in mind that this book pre-dates the publication of classes like the magus, vigilante, kineticist, etc., so you won't find options directly designed for them. In addition, because it's part of the RPG line, it does not contain Golarion-specific flavour (though everything in here is compatible with the setting). As a whole, I would classify the art as in the lower-middle spectrum of what Paizo can do, with a lot of reused mediocre stuff from earlier books. The layout as a whole, however, is quite nice.
Chapter 1 (Races): After an Introduction that's really just an expanded table of contents, Chapter 1 expands the options available for Core races (those found in the Core Rulebook). For each race, a sentence or two describes how each of the Core classes and the so-called Base classes (those found later in this book) are represented within the culture. I found this section was fairly generic and tried too hard to make it sound like each class was common in each race, so there wasn't anything that seemed special. Next up are alternative racial traits for the Core races. These are important in that they allow a player to swap out one of the special features of a race (like an elf's automatic familiarity with elven weapons, or a gnome's resistance to illusion) for a different special feature. In other words, it's a good way to customize your PC just a little more and ensure that not all dwarfs are skilled at stonework, for example. Last, this chapter presents new favoured class options for each of the Core races: instead of the normal rule that a new level in a favoured class provides 1 hit point or 1 skill point, these new options allow a particular race to get something different. For example, a gnome with the favoured class of bard could get an extra round of bardic performance each day, or a half-orc with the favoured class of fighter could get an additional +2 to stabilization rolls when dying. Note that each race only has new favoured class options for handful of classes (not all of them). Unlike the alternate racial traits, I wasn't particularly impressed with the flavour or thought given to the new favoured class options: many of them didn't seem to have any particular tie to the race. Half-orcs, for example, can increase their bomb damage if their favoured class is alchemist, while human paladins can start to get energy resistance--there's nothing in the write-up of these races that make these bonuses seem natural or logical. From an optimization perspective, these new favoured class options are quite useful--I just wish they were better from a storytelling perspective.
Chapter 2 (Classes): One of the most important things that the Advanced Player's Guide brings to Pathfinder is the introduction of six new "Base" classes: the Alchemist, Cavalier, Inquisitor, Oracle, Summoner, and Witch. I don't have a lot of space to review each one, so I'll try to be concise.
The Alchemist fills a real niche in the game, is quite versatile, and would be really fun to play. They get special abilities to rapidly make alchemical items (of course), but also can manufacture bombs, cast magic spells (in the form of drinkable "elixirs"), and temporarily "hulk out" by drinking a "mutagen." As a GM, my only concern is the fact that the bombs resolve against Touch AC, so in games I've run the alchemist PC hardly ever misses and does substantial amounts of damage as an area effect. I also think that perhaps the mutagen feature should have been reserved for a specific "Dr. Jekyll and Mr. Hyde" archetype, as I don't thik it fits well as part of the basic assumption of alchemists.
The Cavalier could probably have been better represented as a Fighter archetype. Cavaliers are mounted knights who swear an oath to follow the precepts of a particular order. Different orders provide different bonuses, Cavalier's mounts are hardier than normal, and the class provides PCs and their allies with some limited use of teamwork feats (discussed below). As written, the class is fairly bland, and I don't think it fills a hole in what could be covered well by other classes. You also see Cavaliers relatively rarely in gameplay because, frankly, they're just inferior to other builds (and I should know, because I've played one for a couple of years now!).
The Inquisitor is one of those classes I'm a bit torn about. The idea is that they're specialists in rooting out corruption and heresy within their faith, which is thematically really cool: but I don't see how that fits naturally with the activities of the vast majority of adventuring parties in the game. The class is conceptually unique and has a lot of cool and useful abilities, some of which seem to fit from a flavour perspective (like Bane) but others that just seem kind of random (like Monster Lore and Cunning Initiative).
The Oracle is another interesting class that I'm unsure about conceptually. Mechanically, they're spontaneous divine spellcasters who don't worship deities per se but instead strive to unravel a particular "mystery." As they advance in level, they get "revelations" which are special powers. Some of the revelations are really cool, and the mysteries are very flavourful. I like the class better after reading it carefully, though I'm still not sure about the name of the class (since divination isn't the focus) nor about the vague relationship they have to deities. They are a divine spellcasting class that is much simpler to play than clerics (though less effective), and thus potentially a good choice for new players.
The Summoner as presented in this book is infamous as the most overpowered class in all of Pathfinder, to the point where most GMs and PFS disallow it. "Unchained" Summoners (as they're usually called in contradistinction to a different type from another book) are, of course, really good at summoning lots of monsters, which is annoying for everyone at the table because it dramatically slows down gameplay. But more problematically, each Summoner gets an "eidolon" which is a bit like a completely customizable and incredibly powerful monstrous animal companion. If you have an Unchained Summoner, you may as well be playing a solo campaign because you probably don't need anyone else in the party to win most encounters. I'm not sure how the Unchained Summoner ever made it through playtesting, but it stands as an example that even great companies like Paizo can make major mistakes.
The Witch is a full (up to 9th level spells) spellcasting class that receives special powers called hexes. Some of the hexes are really flavourful and cool, and the concept of the class as a whole is one I really like. There are two things about the class I'm not a fan of: first, familiars are a major part of the class and as both a player and a GM I find familiars really annoying to deal with (because they rarely contribute positively to a play experience); second, each witch receives bonus spells depending on what "patron" they choose, but the patrons are just abstract concepts (like "Agility" or "Water") and have no substance or flavour to them, and no real potential for story development. I think it was a bland and almost forgettable way of implementing a really cool idea (mysterious forces granting a character power in exchange for . . .?). I should also note that one of the witch's hexes, Slumber, has proven overpowered and problematic at a lot of tables.
So as a whole, I think the Alchemist is a real success, while Witches, Oracles, and Inquisitors are solid additions to the game. The Cavalier is mostly forgotten, while the Summoner is a good example of what not to do in terms of game design.
The Classes chapter then continues by offering each of the Core classes something special, often in the form of "archetypes." If you don't already know, archetypes are packages of abilities that swap out some of the features of a class in exchange for other features, and they've become an important part of most builds for experienced players. Here's a summary of what each Core class gets.
1) Barbarians receive a lot of cool new options for rage powers (though, oddly, a lot of them relate to consuming alcohol) and several archetypes that don't change a lot of class features but that are quite good;
2) Bards get some fantastic and (sometimes quite dramatic) archetypes, at least as written--but admittedly, I don't hear about them being played very often;
3) Clerics receive the introduction of "subdomains", which are, as the name indicates, "branch" domains. A cleric with the Sun domain, for example, could now choose the replacement special power and domain spells of the Light subdomain. It's a way to allow the further customization of clerics since they don't have a lot of class features to trade out for archetypes;
4) Druids get archetypes that are all terrain-based and quite formulaic, along with a handful of "animal shaman" archetypes that have the same essential ability to gain an aspect of a particular animal's powers.
5) Fighters get a lot of archetypes, most of which are poor in terms of flavour ("Archer" or "Two-Handed Fighter") but some that are quite nutritious, as it were, to aiding particular combat styles;
6) Monks get a lot of archetypes, most of which are pretty bland but some, like the Zen Archer, the Monk of the Four Winds' Slow Time ability, and the Monk of the Healing Hand's capstone power are pretty cool;
7) Paladins get archetypes that are okay, but there's some clunky features for the Divine Defender and Sacred Servant. There's also the introduction of the Antipaladin (formally an "Alternate" Class) which I know a lot of people demanded but I'm just not a fan of the concept because I think it devalues the essential goodness of the Paladin idea;
8) Rangers get new archetypes and some new combat styles. I really like the Guide archetype, as the Terrain Bond feature seems much truer to the niche that rangers should fill as wilderness experts. The Infiltrator and Skirmisher archetypes also get some cool stuff;
9) Rogues receive 30 new rogue talents and 12 new advanced rogue talents to choose from, though most are of the "1/day, roll two d20s and take the better" on a specific skill check type. I like the Fast Getaway talent (allowing a rogue to sneak attack and then withdraw), and imagine it would keep a lot of rogues alive. The class also receives several archetypes, but most are pretty thin and forgettable (though the Cutpurse could be used to devastating effect depending on GM discretion);
10) Sorcerers receive 10 new bloodlines, and although I'm not an expert on the class, they look useful and meaningful;
11) Wizards get new elemental schools to specialize in, and some of the special powers look like a lot of fun (like the Air school's Cyclone power or the Water school's Wave power). There's also the introduction of "Focused Arcane Schools" which you can think of as "super specialization" in a particular aspect of a School in order to gain replacement powers.
Whew! A lot of stuff in that chapter. Moving on.
Chapter 3 (Feats) contains a *lot* of new feats. The summary table which gives a one-line description of each one fills four pages. Many of the new feats are standalone things, but others can be grouped by type: several give an additional use of class features ("Extra Rage Power", "Extra Rogue Talent", etc.), make it easier to use the new combat maneuvers introduced at the end of the book, create new metamagic options for spellcasting (with "Dazing Spell" responsible for a lot frustration to GMs), etc. A new type of feat, Teamwork Feats, are introduced for the first time in this chapter. The idea with Teamwork Feats is that if two PCs (or allied NPCs) have the same feat, they both get bonuses in particular situations: for example, if two PCs have the "Allied Spellcaster" teamwork feat, they each get a +2 bonus on caster level checks to overcome spell resistance. I do like the concept, but the proven problem is that it's often hard to get other players at the table to have their PCs take the same one that you're taking, and the bonuses provided by the feats aren't so amazing that groups are inclined to carefully coordinate.
Chapter 4 (Equipment) contains about 25 new weapons (including some of those fun, weird polearms D&D veterans will recognize), a handful of new types of armor, a lot of new pieces of adventuring gear, and several new alchemical items. There's not a lot here that's earth-shattering, though some items, such as Weapon Blanch, have become de rigeur for every smart adventurer. It would have been nice if more of the equipment was illustrated, and that better choices were made on what was essential to illustrate: I know what an hourglass looks like, for example, and don't need a picture, but seeing what a "light detector" looks like would have been interesting.
Chapter 5 (Spells) has 57 pages of options for spellcasters of every stripe. Reading through, I noticed a surprising number of cool Paladin spells, a lot of Bard "finale" spells (that are cast and instantly end bardic performance), and a lot of ninth level spells. Some of the spells I really liked include Blaze of Glory, Fire Snake, and Hero's Defiance, and the picture of Cacophonous Call on p. 209 is hilarious. Every spellcaster is bound to find something useful, but there are some problematic ones introduced in this chapter, like the Create Pit line, that GMs need to be aware of.
Chapter 6 (Prestige Classes) introduces eight new options that PCs could, but probably won't, strive for. Pathfinder long had a reputation for not making much of the prestige class concept, and that's only recently begun to change. Really fast verdicts: 1) Battle Herald: Love the concept, but everything is tied off an "Inspiring Command" bonus which just progresses too slowly, making the entire prestige class weak; 2) Holy Vindicator: no design room for the concept, and the abilities don't help; 3) Horizon Walker: the bonuses in some terrains are fantastic and in others completely "meh"; 4) Master Chymist: Classic Jekyll & Hyde alchemist; 5) Master Spy: I liked this more than I thought I would, and could see it used for a lot of NPCs or maybe a PC (in just the right campaign). Gets clever and useful foils to most means of detection, but abilities come on line much later than they should for most adventures; 6) Rage prophet: Not impressive. 7) Stalwart Defender: Good, cool abilities that fit the theme, and a good capstone power.
Chapter 7 (Magic Items) has something of everything: magic weapons, armor, wondrous items, minor and major artifacts, etc. The new metamagic rods are really powerful considering the price, the new staves are pretty boring, and there's a lot of stuff geared specifically for the new classes, which makes sense. If you've dumped Strength and are relying on Muleback Cords, you've got this book to thank. My only regret is that the chapter introduces so many fun cursed magic items, and I hardly ever get an opportunity to use any in a game.
Chapter 8 (New Rules) is an important chapter containing three new concepts: additional combat maneuvers, hero points, and traits. The additional combat maneuevers are Drag, Reposition, Steal, and Dirty Trick. Your experience may differ, but in a few years of regular play I've honestly only ever seen the last type of maneuver used, and only in a single session. The circumstances in which doing one of these maneuvers in place of a regular attack are just too rare, and using the maneuvers effectively requires an investment that's rarely worth it. The second new concept are hero points, which some groups swear by. They are basically an optional mechanic that give each player a limited pool of points to gain a big bonus on a roll, do something really cool, or save their character from certain death. I imagine they do dramatically lower the lethality of the game, which some gamers would really appreciate. Last is traits, which, although officially optional, have become a normal part of character creation for PFS and most home groups. Traits are little bonuses a PC can get (each ostensibly half as powerful as a feat) that tie into and help form that character's backstory. I really like the concept of traits; I just wish more players took the "backstory" part of the trait more seriously instead of just parsing lists for the most advantageous mechanical bonus.
The Advanced Player's Guide is a 335-page book, loaded with tons of options for PCs. The crunch it provides is well-written and sound in terms of design, with a few exceptions pointed out above. It has become an integral part of Pathfinder, and a book that's almost impossible to ignore. Even though it was published seven years ago and thousands of pages of Pathfinder material has come out since, you really can't go wrong adding this book to your collection.