First, I should say that I certainly understand why some people don’t get wound up about small math errors. Some of my favorite source books for my campaign are from different game systems altogether, particularly Ars Magica. When I plunder one of those books for ideas, I have no idea or even interest whether the game mechanics behind those ideas are rock solid or completely rotten.
For most of my DMing career (starting back in 1st edition AD&D), I almost completely improvised all encounters using just the average PC levels, ACs, hps, and THAC0s. My notes would only say something like "this is a bad-ass ogre," and when it came time to roll the dice I would fudge the stats on the fly to fit the situation.
But for me 3e is different, because there is a solid game mechanic framework underlying everything. I take a wonkish delight in seeing how all the little parts fit neatly together. From an aesthetic perspective, I appreciate good craftsmanship for its own sake. But I also realize that not everyone groks clean, elegant game mechanics.
But from a practical perspective, I also want to have confidence that the stats in my books are solid and reliable. Because sometimes, I need to quickly pull a monster out of those books and drop it into a game.
Just two weeks ago I ran an out-of-the-book encounter that resulted in a PC death from a claw attack that did 1d8+6 points of damage. Only afterward did I realize that the attack should have done only 1d8+4 -- the designers had incorrectly used 1-1/2 the monster’s Str bonus for damage. What was the different between 2 points of damage? As it turned out, one dead PC.
Does that mean I freak if a stat block has a wrong skill bonus? No -- as almost everyone recognizes, skill bonuses are one of the most complex pieces of the d20 system and also one of the least important for running most encounters. Knowledge (geography) +6 as opposed to +5? Whatev. So a glitch here and there is OK.
But if the glitches are here and EVERYWHERE, that’s a problem. What's worse: a lot of little errors or a few big errors? Neither is really very good. It suggests a level of carelessness that might run deeper than just math mistakes. Maybe these monsters haven't been properly playtested. Maybe some fundamental aspect of a special attack hasn’t been thought out thoroughly.
Note that John doesn’t even look at skill bonuses in his review -- almost all of the game mechanics that he calls out are important numbers in running the encounter: AC, hit points, save DCs. What’s the difference between 2 points of damage? See anecdote above.
And keep in mind this is only one guy, looking at this book one time, for one week. Sometimes, there are mechanical problems that run deeper than just the surface math, and can only be found in a playtest.
So based on John's review, the sheer number and type of errors in the MM III is, to me, deeply troubling. But also troubling is who is making these errors. I think Wizards of the Coast should be held to a higher standard than most other publishers. Different contexts demand different standards.
If you invite me over to your house for steaks and you accidentally overcook my porterhouse, no problem -- I'm not gonna complain and I’m not going to hold it against you, even though I may not like it. But if I take you out to a nice restraint, pay $200 a plate, and get overcooked steak, I'm sure as heck going to be ripped.
So if I buy a $5 pdf that is riddled with errors, that's still not cool but doesn’t really get me very incensed, either. But If I'm buying a $35 hardcover book from the industry leader, I expect that book to have excellent production values, including editing. Do I expect perfection? No, not at all. Given the complexity of the game system, some errors are inevitable. But to accept simple math mistakes on 38% of the monster entries? That seems a bit much, to me.
