How do errors get into a product ?

[Agentorange]

So, let me say at the start, this is NOT a criticism of anybody, or a dig at people or a complaint.

I genuinely am curious how this happens. I've been looking at the starter pack errata thread and I've seen other errata thread. The starter pack thread ran to 4 pages of identified typos errors and so on. How does this happen ?

I'm guessing the authors spend a fair bit of time ( probably a lot of time ) on their work. revising, honing,polishing, thrashing out problems, working corrections into the text and so on.

Then the editor spends probably a hell of a lot of time going through the work the writers have submitted, casting an eagle eye over proceedings, fixing errors, reworking stuff as needed and so on

Finally the proof readers ( often several of them )  go through and they are deliberately going out of their way to find errors and issues....it's what they're meant to do. it's their job to find fault - so that fault can be fixed.

Yet with all those people looking at the text and the product, we get 4 pages of errata, how does that happen ? How do 4 pages of problems get past people who are actively looking for errors and problems so as to deliver a good product.

Like I said, this is not a dig or criticism I really am curious how these things creep through with so many eyes looking for them

[Scornado]

My personal experience is that most people are not good proof-readers.  As a Asperger's suffering pedant I consider myself quite good at it, but still miss things.  I wonder whether there are different types of proof-reading skill and a good team has a mixture of them.

Edited September 2, 2022 by Scornado

[Steve]

Yep, across the industry there are a lot of people who think they are good proofreaders, but being able to pick up a few errors here and there is very different from being a professionally trained and experienced proofreader (same applies for copy-editors). Right across the whole publishing industry (definitely not just RPGs) there is a strong temptation to save costs by using people who simply aren't as good at it as they think they are.

Having said that, no person is infallible, professional or not.

 

[Agentorange]

4 minutes ago, Steve said:

 

Having said that, no person is infallible, professional or not.

 

Totally agree, i think what surprises me is that the errors have got past multiple layers of people, each of whom is actively trying to spot the errors  and so produce a good product.

 

[Ali the Helering]

13 hours ago, Scornado said:

My personal experience is that most people are not good proof-readers.  

At a family wedding they found themselves singing "Lord of all hopelessness".  The marriage failed.

[RandomNumber]

When writing reports and papers at work, we often find by the end of the process that we're reading what we think is on the page instead of what is really is there. Typically we get someone who knows the subject, but has not seen the report, to do a read through of the final draft.  This flushes out a surprising number of typos and inconsistencies.

Edited September 3, 2022 by RandomNumber

[Joerg]

Eacj üass of proof reading comes with its own focus. Once you concentrate on the number of times the ball is caught, you will miss the gorilla.

Having multiple error hunts by different people could refine a document, but only if you pass on the corrected document for each cycle, otherwise the same "obvious" mistakes left in will distract from those not yet caught.

 

[Squaredeal Sten]

Random Number is right: Once you begin seeing what you think is on the page, a new pair of eyes is needed. 

[Bill the barbarian]

16 hours ago, Agentorange said:

I genuinely am curious how this happens. I've been looking at the starter pack errata thread and I've seen other errata thread. The starter pack thread ran to 4 pages of identified typos errors and so on. How does this happen ?

Hmm, human error goes a lloonngg way!

 

4 hours ago, Steve said:

Yep, across the industry there are a lot of people who think they are good proofreaders,

Don't know if I am any good as a proofreader (I like my chops as an editor, though some say I am biased). Still, I am told I am okay as a proofreader and error trapper. It is quite a mechanical process. Not at all warm and fuzzy. That said, I have found that these machines we use are mixed blessings! For all the errors and typos they catch, the ease of cut and paste seems to bring more errors into the edit. 

 

3 hours ago, Ali the Helering said:

At a family wedding they found themselves singing "Lord of all hopelessness".  The marriage failed.

The good lord works in mysterious ways!

 

2 hours ago, RandomNumber said:

When writing reports and papers at work, we often find by the end of the process that we're reading what we think is on the page instead of what is really is there.

Oh, so sad... and so true! (I would like to think the "In The Garden of Eden" episode of the Simpsons has some truth in it... google it!)

 

30 minutes ago, Squaredeal Sten said:

Random Number is right: Once you begin seeing what you think is on the page, a new pair of eyes is needed. 

Yep!

 

I have found to have a chance at catching all the errors I might find in copy, I have to do two good and long passes.

A third pass two catch stoopid arrows dont hertz udder!

Edited September 3, 2022 by Bill the barbarian

[soltakss]

16 hours ago, Agentorange said:

So, let me say at the start, this is NOT a criticism of anybody, or a dig at people or a complaint.

I genuinely am curious how this happens.

For me there are various types of error, in fact there are probably lots more:

Typos: Misspelling of words, the wrong letter that changes a word's meaning etc.  These are usually caught by performing a spelling and grammar check. As I use Word for the layout of my supplements, I do an extensive spelling and grammar check before I publish. It is surprising how many people don't seem to do that.

Errors in NPC Stats: RuneQuest is quite labour-intensive in how various attributes are derived from Characteristics. I have an Excel routine that calculates Attributes from Characteristics, so I can copy and paste the statblock from Excel to Word, then tidy it up a bit. I also have a Leon who does most of my statblocks. I would recommend that everyone get a Leon.

Inconsistencies in descriptions: Sometimes authors write from memory, other times they check supplements for facts about a person, place or event. Sometimes they get these wrong, or have incompatible information. The best way to check this is by proofreading the whole supplement.

Poorly stated rules: Sometimes a rule looks good on the surface, but breaks down when taken to extremes, or when combined with apparently unlinked rules. The best way to get around this is to get a rules-lawyer to check the supplement. I check my own work for this kind of thing, as I have a rules-lawyer component to my personality.

NPC Names: As we all know, having to name all NPCs is a nightmare, which is why I resist it. I am much happier with Trollkin 1 to Trollkin #10, or Sverre's Norwegian #1, #2 and #3  who he brought to Convulsions many years ago. Naming an NPC and then using the wrong NPC name in a scene is very easy to do, and is best found by proofreading the entire supplement.

16 hours ago, Agentorange said:

I'm guessing the authors spend a fair bit of time ( probably a lot of time ) on their work. revising, honing,polishing, thrashing out problems, working corrections into the text and so on.

Then the editor spends probably a hell of a lot of time going through the work the writers have submitted, casting an eagle eye over proceedings, fixing errors, reworking stuff as needed and so on

Finally the proof readers ( often several of them )  go through and they are deliberately going out of their way to find errors and issues....it's what they're meant to do. it's their job to find fault - so that fault can be fixed.

Yet with all those people looking at the text and the product, we get 4 pages of errata, how does that happen ? How do 4 pages of problems get past people who are actively looking for errors and problems so as to deliver a good product.

My guess, and it is only a guess, is that they either use one proofreader for a supplement, or they use several similar proofreaders, who look for the same things.

Certainly getting someone to read through a supplement and identify issues is a good way of spotting errors.

I was once asked to check an article to see if it made narrative sense. While doing sop, I found some grammatical and naming errors and pointed them out, only to be told that I shouldn't be looking for those.

2 hours ago, Joerg said:

Once you concentrate on the number of times the ball is caught, you will miss the gorilla.

Yes, that one completely amazed me at the time.

2 hours ago, Joerg said:

Having multiple error hunts by different people could refine a document, but only if you pass on the corrected document for each cycle, otherwise the same "obvious" mistakes left in will distract from those not yet caught.

Yes, that is true.

I try to get around that by sharing the document so everyone can proofread the master document.

Otherwise, I go through each set of proofreads and make corrections sequentially.

 

[Squaredeal Sten]

In addition, something I was told decades ago:  Don't proofread on a screen.  Proofread in hard copy.  You will find more of the errors.

I don't know why that is true, some subtlety of the way humans work.  But it does seem to be true.

Got this from a training session at work.  Not just someone's bull session.

 

 

 

[soltakss]

6 minutes ago, Squaredeal Sten said:

Don't proofread on a screen.  Proofread in hard copy.  You will find more of the errors.

I always proofread on screen, as I hate going through hardcopy supplements, especially if they are double-spaced and, hence, twice as long.

[Agentorange]

3 hours ago, soltakss said:

Errors in NPC Stats: RuneQuest is quite labour-intensive in how various attributes are derived from Characteristics. I have an Excel routine that calculates Attributes from Characteristics, so I can copy and paste the statblock from Excel to Word, then tidy it up a bit. I also have a Leon who does most of my statblocks. I would recommend that everyone get a Leon.

 

 

 

It's not a  Hyperdyne Systems 120-A/2 Leon is it ?

I've heard those early model Leons can be a bit twitchy  😄

[Bill the barbarian]

4 hours ago, Squaredeal Sten said:

In addition, something I was told decades ago:  Don't proofread on a screen.  Proofread in hard copy.  You will find more of the errors.

4 hours ago, soltakss said:

I always proofread on screen, as I hate going through hardcopy supplements, especially if they are double-spaced and, hence, twice as long.

Alas, I must come down on my fellow proofreader's side, Simon (soltakks), on this question. When last I had an editorial deadline locked in stone by sponsors and advertisers and an elected exec, I killed so many elves in search of the editors's Moby Dick—perfection—that my eyes bled red ink! Happily on screen, while I do have all the foibles only a computer/human interaction can create (and there are oh so many), most folk (an Orlanthi All?) are happy enough with what few errors get through that they usually let them go. As a final bonus, screen editing allows one to leave the authors's text untouched while giving annotated notes to him or her about suggested edits.
Plus no dead elves!

 

I truly believe it is impossible to craft a perfect piece of writing... sans errors (factual, or believed, or ?) of any good length.  We may well try, but we shall fail!

Edited September 3, 2022 by Bill the barbarian
'cause it was easy to deit... er, edut... damn! Edit!

[Rodney Dangerduck]

4 hours ago, soltakss said:

Poorly stated rules: Sometimes a rule looks good on the surface, but breaks down when taken to extremes, or when combined with apparently unlinked rules. The best way to get around this is to get a rules-lawyer to check the supplement. I check my own work for this kind of thing, as I have a rules-lawyer component to my personality.

This is my major complaint with the rules.  A competent GM can deal with typos, and errors in NPC stats.  Many of the melee rules, and spell rules, are inconsistent: for example, is there a POW v POW roll?  Having a "rules-lawyer" do a proofread, especially one looking for loopholes, would, perhaps, have improved them.

[soltakss]

1 hour ago, Agentorange said:

It's not a  Hyperdyne Systems 120-A/2 Leon is it ?

I've heard those early model Leons can be a bit twitchy  😄

You know, it probably is, that would explain a lot.

 

[Rick Meints]

21 hours ago, Rodney Dangerduck said:

This is my major complaint with the rules.  A competent GM can deal with typos, and errors in NPC stats.  Many of the melee rules, and spell rules, are inconsistent: for example, is there a POW v POW roll?  Having a "rules-lawyer" do a proofread, especially one looking for loopholes, would, perhaps, have improved them.

Part of the difficulty in crafting a very well stated rule or rules stems from the author knowing how the rule is supposed to work. In short, they sort of mentally miss some of the awkwardness potentially there. They do not have the total ignorance/inexperience someone brand new to the system has as their starting point. With Gloranthan based writing this has an extra level of difficulty because most people with a decent knowledge of Glorantha usually also have a decent grasp of the RQ rules.   

[Agentorange]

13 hours ago, Rick Meints said:

Part of the difficulty in crafting a very well stated rule or rules stems from the author knowing how the rule is supposed to work. In short, they sort of mentally miss some of the awkwardness potentially there. They do not have the total ignorance/inexperience someone brand new to the system has as their starting point. With Gloranthan based writing this has an extra level of difficulty because most people with a decent knowledge of Glorantha usually also have a decent grasp of the RQ rules.   

That's interesting, puts me in mind of that famous quote by Heinlein: " when one teaches, two learn"

[Diana Probst]

The A to B structure that you are assuming isn't always what happens.  Assume several writers are working on something.  It's merged into a master document.  It gets checked by someone who has opinions on all of the writing, and marks them in, but doesn't know where everything cross-references.  Getting them to hold the entire document in their heads while checking each problem area is impossible.  Someone who knows prices adds in prices.  Someone who knows Kralorela or Innsmouth puts in some details which are true and interesting but were not known to one of the original writers.  Meantime the table of distances has been updated because a mapping error was solved in a different product, which means there is a new master distance chart.  However, that means another bit of writing vital to a plot has been upset which means three new paragraphs of text to add in a complication that keeps the timing on track.  Then the art gets flipped because the layout artist absolutely has to put a table in a certain place, or else it can't fit in the chapter, and you've now got people talking about left-handed warriors and an image that clearly shows a minor NPC with a patch on the wrong eye, but if you turn the whole page around the document flow is wrong.

It's not a simple process that goes along a measured or even measurable path.  It's multiple parallel paths, some of which are lit only by burning torches being juggled by people who would like there not to be 7 of these things please.  To have the process going from A to B would mean it took far longer for each book, so to stay in business, any company is going to have to write in a way that fits everything in as efficiently as possible.  That means trading time for the higher possibility of errors.

I was the final external writer and editor for the Culbrea book, and towards the end of it, I was still combing out errors.  I was just about able to hold it all in my head, at least for long enough to comment on the document so I could check them all, and keep the plots and the setting consistent, but it took me three months and a lot of writing to get to the point where I was that familiar with the document.  That's not available to most people.

[Nick Brooke]

When I get a draft text to review, I'm checking for several things in parallel.

• Is this a typo? These are the easy bits. Markup the fix, and explain why if it looks dubious.
• Is this sensible? What's the point of presenting "conversions to miles / yards" with decimal-place accuracy, when all the initial numbers are estimates? Aren't these three paragraphs in the wrong order? Why mention Sorcery at all? And so on.
• Is this RuneQuest? Have you inadvertently used your own house rules e.g. for how shields work, without cross-checking vs. the core rulebook? Isn't this bit a good opportunity to show how Rune affinities work in play? Have you considered assigning an Honour hit for bushwhacking an opponent?
• Is this Glorantha? Does it present things in a way that feels weird to someone who knows the setting as well as I do? Are we making things hard to approach for readers who are new to the setting? Is the text needlessly multiplying entities, or referencing obsolete terminology, or abbreviating things in a way that is actively misleading,  or making mountains out of mole-hills? My honesty also makes me highlight cases where terminology I prefer, but that's officially deprecated nowadays (e.g. "Sheriff of Apple Lane"), has somehow crept into the manuscript.
• Is this wise? Wouldn't everyone's lives be easier if you left out that one weird bit about Elmal that we know will spark idiotic flame wars somewhere? Does the text (inadvertently or otherwise) present all Lunars as Chaos-tainted, or turn a blind eye to ethnic cleansing, or encourage psychopaths and bigots? If the draft manuscript encourages players' worst instincts, I'll write a small personal screed explaining why this is a Bad Thing, and propose alternative text.

Whoever's editing (Jason, in this case) receives all of my feedback (about twenty pages for the RQ Starter Set, not counting SoloQuest markup), plus what anybody else sends back. They have to process all of that, and don't necessarily agree with everything I say,* and might put tricky things on the back burner (or cross-check with other reviewers and team members). If I've suggested a change that'd be tedious to implement, it's totally OK by me for them to shrug and say, "I can't be arsed to fix that bit, it'll have to do." At the end of the day, we are publishing a silly elf game, not an operator's manual for a medical device or nuclear reactor. Adding more expensive proof-reading overhead will increase the cost of all our books; missing an occasional error is entirely acceptable, and we can always fix it on the forums / in errata once it surfaces.

I have better things to do with my life than review all of my proposed markup vs. the finished product. I do check whether any stuff I thought was unwise or un-Gloranthan has crept in, because I know there will be flame wars about that nonsense, and I need to keep my powder dry. If a lot of stuff I wrote is passed over in the final product, I might find another way to share it (c.f. the Smoking Ruin playtest notes in The Duel at Dangerford). But then again, I might not. I am a whimsical and mercurial creature.

___

* I know this will come as a shock to some of you, but I am not always right.

[Akhôrahil]

The RQG rulebook also picks up some unusual bugs because it uses an amount of copying and pasting from earlier editions, which can work but is also a great way of introducing edition-dependent errors. Bladesharp is a good example - the proviso about Dullblade makes no sense, until you realize that it did make sense in RQ3 where the text comes from. This kind of thing can be very hard to catch, as it's dependent on the interaction of the old rule and the new one. (This corresponds to how in software testing, bugs that depend on the interaction of two functions are much harder to catch than a straightforward error in the basic function.)

Edited September 6, 2022 by Akhôrahil

[Agentorange]

21 hours ago, Nick Brooke said:

___

* I know this will come as a shock to some of you, but I am not always right.

mmmmm.....but are you right about not always being right ?

[Nick Brooke]

48 minutes ago, Agentorange said:

mmmmm.....but are you right about not always being right ?

I'm not sure I understand your point. You might find this discussion thread on Twitter illuminating: link.

[Agentorange]

2 hours ago, Nick Brooke said:

I'm not sure I understand your point. You might find this discussion thread on Twitter illuminating: link.

Sorry, it just reminded me of that famous philosophical ( logical ? ) problem  The Liar paradox

"Everything i say is a lie"

I've actually found the whole thing very interesting and informative. Easy to see why people go for a PDF release first and then allow the time for feedback and error spotting before proceeeding to a print run.

 

[Bill the barbarian]

On 9/2/2022 at 3:35 PM, Agentorange said:

I genuinely am curious how this happens. I've been looking at the starter pack errata thread and I've seen other errata thread. The starter pack thread ran to 4 pages of identified typos errors and so on. How does this happen ?

Just watched a film that addresses the same question about a couple of books with a little greater import than the RQ G core books... I know, I know... blasphemy... and speaking of which...

 

 

Errows in copy and piste have been around for awhile!
Not to mention retconning and out and out changing of text for the editor or proofer's own agenda or...

Edited September 29, 2022 by Bill the barbarian