Thumbs Down to D&D 3.5 Edition

I'm a long-time, hard-core D&D (and AD&D) player and student. As can be seen from other parts of my online site, I played AD&D 1st Ed. for a long time starting around 1979 (my first book actually being from the "blue book" basic D&D set). I was disappointed with the departure of Gygax in the mid-80's, and skipped 2nd Ed. entirely, my opinion being it was a shoddy hack on the 1st Ed. material. I was pleasantly surprised with the appearance of 3rd Ed. D&D, and felt that it was a thematic whole, that the rules fit together and were well-considered and thought out by the designers, and had an impressive playtest program. I felt like someone really cared in the design of 3rd Ed., and that's what I've been playing since 2000.

At this point I've read the details on changes brought about with the so-called 3.5 Edition, I've downloaded and read the SRD, and I've flipped through the 3.5 books. In short, I consider it to be an aggravating disappointment and I won't be buying any of the books. Similarly, my playgroup has decided to keep using 3.0 rules. Very much like 2nd Ed., 3.5 Ed. looks like a sloppy and poorly playtested set of half-changes to the prior ruleset - it generally makes me think of the new edition as "Third-and-a-Half Edition Rules, Second Edition Feel".

Here's an analysis of why my impression of 3.5 is so low.

No Playtester Credits

The time spent on playtesting is a huge indicator of the quality of a gaming product, and something I identified as a priority years ago (grtdnd3). My theory is that the decline of TSR from 1st Ed. through 2nd Ed. is directly traceable by the reduction and final elimination of the "playtest" credit in their products.

Now, 3rd Ed. had a totally impressive playtest program - a whole page in the Player's Handbook was devoted to "Playtester Credits" (PHB p. 286) with over 600 individuals named there. In 3.5 Ed., there are no playtester credits listed ("There are no playtester credits. At all.", says Monte Cook here), and that change is really a very bad sign. For all I know 3.5 Ed. did have playtesting and failed to give credit to those who did it - but as has been reported, the design changes in 3.5 Ed. were basically all made between October 2002 (when the project started) and Spring 2003 (when the books went to the printer), leaving precious little time to make playtesting a priority.

Either playtesting did not occur, or the publishers felt it was unimportant to credit those who did it: either way, the lack of a "Playtester Credit" is a major red flag for a haphazardly thrown-together RPG product.

Unnecessary Rules Changes

Most of the changes brought about in 3.5 Ed. are questionable at best, and often seem about half-thought out. Many contradictions have now been inserted into the new rules, where a change was made in one part of a book but not updated in another part (perhaps changed in rules text but left as before in as table or an example) - this is most disappointing in light of the promotions for 3.5 mostly selling it on the basis of supposedly corrected and cleaned-up rules.

It's striking how much this resembles the case of AD&D 2nd Ed. - in 95% of the text it resembled the prior edition, with a fairly small number alterations which could be taken as knee-jerk reactions by someone who didn't fully understand how the rules worked together in the first place. In the case of 3.5 Ed.:

Major Spell Alterations

Somehow, in the half-edition that is 3.5, the designers felt compelled to make more changes to the D&D spell lists than have been made at any time since original D&D in the 1970's. In general, the standard spell lists have been maintained pretty consistently, with minor tinkering, from original D&D through AD&D 1st Ed., 2nd Ed., and D&D 3rd Ed. In general this facilitated conversion work and use of older products because most NPC spell lists and spellbooks were directly transferable to the newer rules. However, with 3.5 Ed., more spells have been altered than ever before, basically bringing that tradition to an end.

In 3.5 Ed., a wide range of spells had their effects radically changed (darkness, polymorph, fly, hold, disintegrate, bull's strength, teleport, raise dead, etc.). Classic spells had their names changed for no particular reason (fly, polymorph, and so forth). A whole raft of spells had their school designator changed for the first time, not because it fit together, but to engineer a numerical equity between the different schools. Spell levels went up and down in different places for the first time ever, breaking back-compatibility. Specialist wizards are altogether changed when converting to the new edition.

There seem to be really just 2 spells that the community clamored for a fix from 3.0 Ed.: haste and harm, maybe heal as well. (Note that the fix to haste is largely to back out the 3rd Ed. change and return to the 1st Ed. rule). That might have been welcomed, but it only occurred in the midst of wholesale, unprecedented changes to the spell lists which break the possibility of clear conversions from classic materials.

(Edit 02/29/08: A very nice documentation of changes in 3.5 was compiled by Steven Cooper on his website. Note that the changes to spell lists were so comprehensive that he had to break out each spell list as a separate document. I count 445 changed spells in these lists!) (Edit 06/06/09: Steven's website is no longer with us; a copy of his documents can be found here.)

Fantasy Flavor Removal

One of the really fascinating changes in RPGs over the years is how the very criteria of how an RPG should be designed has shifted dramatically. In the early issues of the Dragon magazine (around 1980), the axis of debate was always around "realism versus playability", that is, that RPGs should simulate a fictional reality as much as possible, up to the point of the rules being unwieldy. Now, in the post-CCG era, the game design mantra is for "balance", the idea that all player creation options should be on an equitable power basis, as if the game were basically a competition between each player.

3.5 Ed. continues this movement. Whereas D&D originally attempted to simulate and bring to life fantasy novels and mythological stories, the current priorities are that if any such fantasy events are "unbalanced", then they must be done away with. Can mythic wizards paralyze opponents with a word? Yes, so 1st Ed. had hold person; but 3.5 Ed. finds that too potent, so it is reduced to only a momentary pause. Can magic-users cast their foes into darkness? Likely so, but the 3.5 Ed. version of darkness creates "shadowy illumination" so no one is at too much of a disadvantage. Strength-type spells are reduced from hours to minutes, even though those are exactly the kinds of effects priests should give as blessings before a knight goes on an excursion, for example. Fly is reduced to a few minutes, even though witches in legend can soar all night.

Similarly, there's a worrisome tendency for game abilities to be transformed in a way that look more and more like computer video game effects, instead of personalized fantasy myths. For example, in all prior editions a cure spell required a touch on the recipient, so we could at least imagine that the effect was a healing ministration, perhaps with some useful herbs or attention to a specific wound. However, 3.5 Ed. has introduced the idea of mass cure spells where an aura instantly cures everyone in a certain area. No doubt this will be highly useful in a game situation, but it doesn't resemble anything in any myth or legend.

Another point is that magic items in general continue to be made more and more generic. In early D&D one of the great points of excitement was when a new magic item was discovered, since they could only be found in a treasure as a climax from a dangerous encounter. Now player spellcasters can make any item in the rulebooks. (A very nice, tight set of rules for magic crafting are provided, but they have the side effect of players no longer being interested in magic treasure for their carried gear). Even more worrisome, I have no doubt but the "invent your own new items at will" movement will continue, so that at some point magic items will be entirely a la carte from a menu of abilities, and no one will have any interest in any item they get from an adventure unless it precisely matches the exact set of abilities they are most in need of.

In short, as the D&D game "balance" is made more and more competitive, as the unique elements of fantasy thematics are steam-ironed out of the ruleset, the ability for players to suspend disbelief in the campaign is made more and more difficult.

Breakdown of "Open Gaming"

To me, the keenest disappointment of 3.5 Ed. is that it fails to capitalize on the promise of the d20 System and the Open Gaming License (initiated with the advent of 3rd Edition). In fact, in many ways it seems to retreat from the commitment which Wizards of the Coast seemed to make in 2000. This is probably not surprising since the primary motivators of the OGL (Ryan Dancey, et. al.) along with all the 3rd Ed. designers (Jonathon Tweet, Monte Cook, Skip Williams) have been flushed from WOTC. Note that this issue is completely separate from whether any third-party companies happen to be finding corporate success.

In 2000, a strong link was made between the philosophy of the Open-Source Software movement (GPL) and the OGL. It was suggested that in making the D&D rule mechanics available to all players, writers, and publishers, that the larger community would be allowed to experiment and come up with new and better rule mechanics - and that those popular modifications would be worked into future revisions of the game. Here is what WOTC asserted as part of their "d20 System Concept:

Frequently Asked Questions" (

Using the OGL, anyone who wants to can "fix" that content and publish the fixes, returning the variation to the core d20 system known by hundreds of thousands of gamers. And if those "fixes" are more popular than the variations, pretty soon, the "fix" will become the standard, and the variation will become a curiosity known only to a few people...

In fact, one of the biggest groups affected by this force will be the Wizards of the Coast tabletop RPG Research & Development team. When the time comes to make a new edition of Dungeons & Dragons, they'll have to make a very persuasive case to the market to adopt any changes to the core rules they want to make!

Well, that totally hasn't happened. Just the opposite - the changes in 3.5 have almost entirely originated with the in-house designers, and almost never resemble now-common house rules or points which have been publicly called out for fixes. The changes in 3.5 have all been surprises, causing players to generally respond first with "why'd they do that?" reaction (with a very few exceptions, such as for the haste and harm spells). Meanwhile, many commonly-identified trouble spots in the rules have gone untouched. Not much of a "persuasive case" was made in WOTC's advertising for 3.5 Edition (

The revised Player's Handbook received revisions to character classes to make them more balanced, including updates to the bard, druid, monk, paladin, and ranger. Spell lists for characters have been revised and some spell levels adjusted. Skills have been consolidated somewhat and clarified. A larger number of feats have been added to give even more options for character customization in this area. In addition, the new and revised content instructs players on how to take full advantage of the tie-in D&D miniatures line planned to release in the fall of 2003 from Wizards of the Coast, Inc.

The changes in 3.5 haven't been a product of the community (e.g., no evidence of community feedback can be found in the blurb above); they look like almost arbitrary house rules by the designers, novelty changes just for the sake of forcing more new rulebook sales. This might be distasteful in and of itself, but it's downright discouraging if it demonstrates that WOTC has no intention of there being a public give-and-take on improvements to D&D via Open Gaming, as it originally asserted there would be (and really should be).

So apparently the point of the OGL and d20 System for WOTC is entirely just to offload the unprofitable adventure-module business to other companies that are willing to take it up. And at the same time, WOTC has added more restrictive content oversight to the d20 System License, inserted suspicious product identity claims into the OGL in the SRD, and attempted to take back certain terms from Open Gaming (such as "d20") which is an act expressly not allowed by the license. It leads me to believe that the Open Gaming idea no longer has the commitment of WOTC, and that its promise will not really come to fruition as originally advertised with the 3rd Edition of D&D.

October 5, 2003