An interesting aspect of 3rd edition Dungeons and Dragons character advancement is that it lends itself particularly well to forward planning. Indeed, I've seen that game system criticised because such forward planning is almost mandatory if you are interested in having a maximally efficient character, want to "pull your weight" in a group that contains other players who optimise their characters, or aspire to have your character take levels in a prestige class or classes. As a multiclass character design from the point of conception onwards, planning and implementing Edwin's character design and advancement has been an interesting learning experience and an intellectually enjoyable exercise. So much so, in fact, that I've decided to document the entire process on this page.
Edwin's original character concept was the dashing, acrobatic, elusive and lightly armed fighter. I'd read on the rec.games.frp.dnd newsgroup that Fighter/Rogue multiclasses could be very effective, so I decided that was the thing to try instead of a single-classed Rogue. Also, I was already playing a single-classed character in another D&D game, Sheyana of Karicksbridge, and I wanted to try out multiclassing. I decided against Rogue 0/Fighter 0 at first character level because I wanted the initial class to be Rogue (for the skill points, of course!) and I didn't like how the special abilities were distributed between first and second character levels.
As you can see from the skills selection, I was focussing on Wis and Dex based skills, plus Bluff. At the time, I expected I'd have to feint much more often than I actually have to in play: using Tumble to flank turned out to be a much more reliable way of getting the sneak attack bonus dice, particularly when I realised the feint manoeuvre only provided a benefit to the first attack in an iterative attack sequence.
At this stage, the plan was to advance evenly in both classes to Rogue 10/Fighter 10, though I hadn't gone much further than planning skill point allocations for the first three levels. And then, at third character level, Edwin died.
Edwin at 3rd level (Rogue 2/Fighter 1)
When he got better from being dead, Edwin was a pixie. This change led to a strong preference for Rogue advancement, as the natural invisibility made every shot with the crossbow a sneak attack. If my second character level hadn't been Fighter 1, I probably would have dropped the multiclass idea at that point and gone for a single-classed Rogue: the first seven Rogue levels are an almost constant source of class abilities and increasing sneak attack dice, and the Fighter level was slowing Edwin's Rogue advancement down. I even made the (in hindsight) mistake of taking Point Blank Shot as my feat at third character level, weakening the character design's focus on being an effective Dex-based melee combatant. I don't remember using that feat at all once Edwin was wished back to being a human at fifth character level.
As I wrote earlier, being an invisible flying pixie was fun at first but quickly became dull. I was also somewhat worried about being reduced to negative hit points, automatically turning invisible (if Edwin wasn't already), and bleeding to death while the other characters fumbled around looking for me and their Cure spells failed to overcome my innate Spell Resistance. The benefits of being a pixie weren't outweighing the disadvantages, and having a character with ECL +4 was skewing gameplay as well. When the opportunity arose, I was happy for Edwin to regain his human form in exchange for owing two powerful spellcasters a service. It set a precedent for the player characters exchanging additional promises of service for other magical assistance, but I see creating future plot hooks for the DM as a courtesy and not a problem, so that didn't bother me at all.
Edwin at 6th level (Rogue 5/Fighter 1)
By this stage, two things had changed about my plans for Edwin's advancement. Firstly, I had decided against an even split between Rogue and Fighter in favour of Rogue 16/Fighter 4. I wanted Weapon Specialisation with Rapier for the extra damage with every melee attack, and I wanted BAB +16 at 20th character level so that Edwin had four iterative attacks. By this time, I'd read reviews of the Epic Handbook on rec.games.frp.dnd that confirmed iterative attacks didn't increase at epic levels, and the GM had confirmed the campaign would eventually go epic (probably up to 30th character level). Having as many iterative attacks as the Ranger/Fighter multiclassed character was an important part of making Edwin seem a better fighter than the Cleric and the Sorcerer/Rogue.
Secondly, it had become clear that there was a great need for someone who could Open Locks, Disable Devices, Search for traps, and perform the other Int-based Rogue skills. The Sorcerer/Rogue multiclassed character was advancing solely in Sorcerer levels, so her player's original intention of covering those skills had fallen along the wayside of her pursuit of spellcasting ability. As the primary Rogue in the group, Edwin was the obvious character to cover those gaps in the party's resources. I'd also decided it was important for the group to have reliable access to Use Magic Device, so I started investing skill points in Decipher Script for the synergy bonus to casting spells from scrolls with Use Magic Device.
Around this time, the GM and I bought the five splatbooks, and one of the prestige classes in Sword and Fist caught my eye.
Edwin at 9th level (Rogue 6/Fighter 2/Ranger 1)
The Duelist prestige class was intended by the designer for Dex-based fighters, and as a result the class abilities and features meshed reasonably well with the sort of things I wanted Edwin to do in play. Taking a level of Duelist looked good, and there wasn't much benefit in having 3 levels of Fighter, which made a a level-drop into Ranger attractive as well. At the end of the day, I was balancing Weapon Specialisation and +1 to all saves against Track, Ambidexterity and Two-Weapon Fighting, Canny Defence, and +2 to Fortitude and Reflex saves. Accordingly, I gave away Fighter 4 for Fighter 2/Ranger 1/Duelist 1.
At first, I planned to get up to Rogue 10 as fast as I could for the Evasion class feature. However, I was spending a lot of my game time flanking and sneak attacking, and not much time making Reflex saves, so I took Rogue 6 (for Uncanny Dodge, can't be flanked), Ranger 1 for the extra (sneak) attack, then Fighter 2 (using the bonus feat to complete the prerequisites for Duelist). Understandably, taking these Fighter classes held up getting Use Magic Device, but I did that with the next Rogue level.
Edwin at 13th level (Rogue 9/Fighter 2/Ranger 1/Duelist 1)
I would have kept the proper sequence and provided Edwin at 12th character level, but 13th level was special for two reasons. First, we went through our first Total Party Kill (TPK) and spent the next few sessions in the Realm of the Death Goddess, our first time on other planes of existence. Second, when we eventually returned to our home plane we realised that all the animated adamantine statues we were destroying in ancient ruined cities were worth a fortune. Thus, 13th level was the first point at which the characters could buy the exact magic items they wanted.
At this point, I'd decided that Escape Artist was a very important skill to have for a low Str melee character to have (by comparison with the Str 22 Ranger/Fighter/Tribal Defender). Edwin didn't get hit very often, but it was a real pain to have on way of getting free when opponents with Improved Grapple or Swallow Whole rolled a natural 20. My skill point allocation now aimed for maximum ranks in what I considered Edwin's "core" skills (namely Disable Device, Escape Artist, Open Lock, Search, Tumble, and Use Magic Device) by 19th character level. I had not yet given up on the Rogue 16/Fighter 2/Ranger 1/Duelist 1 multiclass, but the more I looked at my plans for epic character levels the more ideas I had for changes.
Edwin at 15th level (Rogue 10/Duelist 2/Fighter 2/Ranger 1)
Those changes have resolved themselves along two lines: maximising BAB and saves before 20th level (after which they run off character level and not advancement in the chosen class for that level) and maximising class abilities after 20th level. Thus, character levels 19 and 20 are going to be two Monk levels (for not just the very good saves but also sixth sense, effective unarmed combat, and Deflect Arrows) and most character levels from 21 onwards are going to be Duelist levels (because prestige classes gain a new class ability every level). This gives up an additional 4 dice of sneak attack damage, but it's balanced out by 2 dice of precise strike damage and a generally mechanically superior character design.
Interestingly enough, the new direction is moving back towards my original expectation that I would divide character levels evenly between Rogue and Fighter: it just turned out that the Fighter side is made up of four classes instead of one. It sacrifices the benefit of Uncanny Dodge (can't be flanked), but that was already balanced on a knife-edge against standard CR opponents anyway (because a single-classed Rogue of the appropriate CR for our party would have 4 Rogue levels more than Edwin, and thus trump Edwin's Uncanny Dodge ability).
As for equipment, Edwin now has items that give him dramatically enhanced manoeuvrability. This synergises well with an extra skill in the list of "core" character skills, namely Jump (which, with Tumble and Escape Artist, is in the Duelist and Monk skill lists). I've now planned to allocate skill points towards those three skills with the regular "fighter" levels, while the other four skills (except for Use Magic Device, which is limited by Rogue levels and can only go up by 1 rank at a time) will jump by 3 ranks every three to four character levels when Edwin takes a Rogue level. This doesn't leave any spare skill points for close to ten levels to come, but there's nothing I can think of to spend them on right now, so that's not a burden.
Edwin at 18th level (Rogue 11/Duelist 2/Fighter 2/Monk2/Ranger 1)
And everyone was kung fu fighting ... two months have passed since what I wrote above and sadly, the game has been put on hiatus until the GM finishes the coursework component of her MPhil program and finds out whether there's enough time and enthusiasm to run a regular D&D game again. However, I can state with confidence what Edwin's design would have look like at 18th character level. Taking the two Monk levels earlier rather than later became an attractive proposition, given a +6 periapt of wisdom and a substantial boost to Armour Class and Will saves. Other than shuffling around the character classes for the next few levels, everything else is as planned at 15th level.
Predicting Edwin's advancement after this point is more difficult because it really depends on whether or not the GM would allow players to assign character wealth to purchasing permanent enhancement bonuses to characteristics. If the answer was yes, Edwin would have a +5 enhancement bonuses to all six characteristics at 22nd level. Simply put, the cost of Epic Handbook magic items is so much greater than the cost of DMG magic items that the other option is to put the money available at 19th, 20th and 21st levels in the bank so that Edwin could have a +6 rapier at 22nd level. +5 to all six characteristics or a +6 weapon instead of a +5 weapon? You do the math.