Maths errors in RPGs

[mattcolville]

To turn your question back round to you - who proofreads programming text books?

As I said, editors. An editor can proof the English instruction set. But proofing the non-English result?

 

[mattcolville]

You might be surprised to find that most people who edit (or, if you like, fact-check) physics course books have a master's degree in the appropriate subject. You're making the argument that only the person/s who generate a stat block are qualified to validate it, which is not unlike saying that scientists shouldn't be held to peer review.

You've missed my point; I said what you just said. Yes, a editor can edit a physics text just as he can edit the PHB. But when a physicist publishes his original research for peer review; only other physicists in his field can judge the validity of his work.

Too often in the RPG field, "game designer" and "writer/editor" are treated like the same thing. In a narrative, cinemtic, rules-light system where the design isn't anywhere near as important as the writing, ok, sure. But in a game like D&D where the rules are WAY more important than the writing, you need a designer validating the result of the instructions.

 

[Herremann the Wise]

To turn your question back round to you - who proofreads programming text books? Clearly the code examples must be *absolutely spot on* otherwise it undermines the whole point of the book. Clearly there are proofreaders who understand the domain well enough to verify the correctness of the code examples.

There is no reason why stat blocks couldn't be considered in the same light.

Cheers

I disagree.

With code, you have a compiler that can at least check if it runs. (This of course does not mean that it will run correctly, but it will run).
There is no specific tool for statblocks - although PCGen most likely comes closest. The other thing is that an example of code in a programming book is normally far less complex than a complete statblock and yes, most programming books still have mistakes that make it through to the keeper.

Best Regards
Herremann the Wise

PS: In Australia, it is normally considered poor to say Math instead of Maths unless of course you do it with an American accent - and even then, you will most likely cop a comment of some nature. There are many "Americanisations" that have made it into Australian culture but funnily enough this is one where the Mother Tongue will always take precedence.

 

[dcas]

You might be surprised to find that most people who edit (or, if you like, fact-check) physics course books have a master's degree in the appropriate subject. You're making the argument that only the person/s who generate a stat block are qualified to validate it, which is not unlike saying that scientists shouldn't be held to peer review.

That shouldn't be a surprise at all. If I were in the textbook publishing business, and I wanted to publish a math (or is it "maths?" By the way, "math" is short for "mathematics" as we generally don't speak of "mathematic," certainly not as a subject of study) textbook, I think I would hire someone proficient in mathematics to proof it, at least the equations and whatnot. Someone who could proof both the text and the equations would be ideal. I would rather not be embarrassed by a high-school kid finding an error in my proof that the square root of 2 is irrational. Likewise, if I wanted to publish a D20 supplement, I would hire someone proficient in D20 to proof it. I would rather not be embarrassed by a high school kid etc. etc. etc.

 

[Plane Sailing]

I disagree.

With code, you have a compiler that can at least check if it runs. (This of course does not mean that it will run correctly, but it will run).

There can be a world of difference between code which compiles and code which does its job properly though! (speaking as a long-time coder...)

 

[MerricB]

I disagree.

With code, you have a compiler that can at least check if it runs. (This of course does not mean that it will run correctly, but it will run).
There is no specific tool for statblocks - although PCGen most likely comes closest. The other thing is that an example of code in a programming book is normally far less complex than a complete statblock and yes, most programming books still have mistakes that make it through to the keeper.

That was certainly my experience when learning my craft - programming textbooks are often riddled with wrong code.

Cheers!

 

[prosfilaes]

You've missed my point; I said what you just said. Yes, a editor can edit a physics text just as he can edit the PHB. But when a physicist publishes his original research for peer review; only other physicists in his field can judge the validity of his work.

The job of an editor of a physics text or a new roleplaying book has to include checking the numbers. It's not about original research; we're not complaining (on this thread, at least) about overpowered feats or underpowered races. We're complaining about getting the numbers to add up and the characters to come out right. In D&D, that's less than trivial, but anyone who knows the game should be able to do it, and why would you hire an editor in this business who didn't know the game? If you can't do that, how can you spot text that's unclear?

At Steve Jackson Games, I know that at least some of this is done by the playtesters. I earned my free copy of GURPS Japan, back when playtesting was more open, by going through and checking all the stat blocks. (For a while I was afraid I wasn't going to find anything; Hunter Johnson did pretty good work getting that all correct on his own.) But I certainly expect that when Steve Jackson or Dr. Kromm edit a book that they check all this stuff.

 

[RFisher]

That said, I understand that companies like Steve Jackson Games and WotC don't have the kind of money to throw at proofreaders that companies like Random House do.

Yeah. But in my experience, SJG manages to do much better with even fewer resources than WotC does. & I'm not even talking about math(s).

To be fair, WotC isn't alone here -- and some companies simply dont provide errata.

Yeah. I like TLG. I understand they are a small company & give them a lot of slack for the amount of time it takes them to get products out & even on the quality. Their complete ignoring the errata for LA Essentials, though, drives me crazy. (& again--no math(s) involved) They even have an "LA Errata" link on their site that takes you to a page without any LA errata! At least link to the thread where your customers were trying to help each other out.

Expecting a proofreader to validate statblocks would be like expecting a proofreader to debug C++ code.

Except that no matter how overly-complex you might think D&D has become, it ain't got nothing on the dark corners of C++. (^_^) (& I like C++!)

 

[dcas]

There can be a world of difference between code which compiles and code which does its job properly though! (speaking as a long-time coder...)

Not to mention the fact that errors can be introduced after the code compiles. . . .

I don't have a lot of D20 books (and I don't play 3e so errors wouldn't really bother me anyway), but are math errors in them more pervasive than coding errors in programming textbooks? (I'm not talking about crap like "Learn C++ in 7 Days!!" but serious stuff like The C++ Programming Language.)

There's another thing with RPGs vs. programming. An RPG, even if it is complex, doesn't have to be designed so that avoiding errors in stat blocks is difficult. For example, HackMaster is a very rules-heavy game, very complex (IMHO), but it is designed in such a way that it's not hard to create an error-free stat block, because skill levels are not so closely tied to class levels and ability scores, and talents (feats) are not tied to class level.

It's hard to see how this would work with programming languages. It's hard to get code right no matter what language one happens to be coding in.

 

[delericho]

I think there are two reasons we're more aware of stat-block errors in D&D than other games and/or in D&D books by WotC as opposed to d20 books by other publishers. The first is that d20 is a lot more mathematically transparent than many other games - as mentioned before, with Shadowrun or Storyteller games you don't know how many XP the character has had applied, so can't determine whether the stat-block is correct, whereas d20 has clear rules for these things.

The second reason is that any given WotC book has many many more eyes looking over it, such that any mistakes are more likely to be found.

Personally, while I understand that no book will ever be flawless, and that mistakes are inevitable, I am rather concerned at the rate of errors - John Cooper's review of "Scourge of the Howling Horde", for example, gives errors for 19 out of the 20 stat-blocks in that product, which is especially problematic when that is an 'introductory adventure'. (However, I believe SotHH is regarded as a particularly bad example from WotC.) I'm inclined to feel that Wizards really should do better, especially since it seems Paizo regularly do a better job with fewer resources (although, it should be noted, some of the very best RPG editors in the business on staff).

I also feel that Wizards could probably save themselves a lot of grief by developing (perhaps internally) a really good stat-block generation tool. It's really easy for a human being to miss-count bonuses to get stat-blocks that are off by 1 or 2. But computers don't - that's exactly the sort of thing they are best at. (It would also mean that if the devlopers later decide to apply, for example, a 5-point Power Attack across the board, generating the revised blocks becomes trivial.)