Fog Creek Software
Discussion Board

Everquest - Bug Tolerance and Online Patchers

I know most folks on this board may not have time or simply do not play games.  But I thought I'd mention this anyway.  Everquest is a MMORPG (Massively Multiplayer Online Role Playing Game).  According to Sony Entertainment (the guys that "own" and produce the game) it has a subscriber base of 400,000+ accounts each paying $12.95 a month to play this game.

Now if you have never played an online game you might not realize that these games are "patched" every week or so.  This is done to fix bugs and add new content.  You start the game through the "patcher" program and it automatically downloads the patches. They now even have expansion packs to the game that can be purchased and downloaded online.

The playbase is made up of casual players and so called "uber" (German for "above" or "super" i believe) players.  These are hardcore players who play 12+ hrs a day.  They run their own websites for their guilds (team of players in the game).

So every time Sony writes a new patch, these "uber" players normally know if something is messed up.  This last patch seems to have a lot of bugs.  In order to understand the bugs you have to play the game.

Now these bugs seem to occur frequently, almost every patch introduces some "unintended" side effects that disrupt game play.

What's funny is these guys never stop playing even though the patches are full of bugs.  They continue on there journey through fantasy land, fighting the gods and the armies of evil-doers.

Now why is the bug tolerance level so high on Everquest.  Maybe it's because it's just a game.  I mean you would'nt put up with bugs in accounting software.  I think it's also because the playbase knows that Sony will fix the bugs and patch the game the next day.  Sony pulls in enough cash via the game, to constantly convert that cash to code through programmers.  The devs over at Sony read the "big-name" boards like the one i previously linked to and if there's a lot of noise on the board, they act.  Of course these "uber" players still complain about a lack of testing the patches before deployment, but they continue playing anyway.

Maybe all software should have an online patcher eh?

Dave B.
Wednesday, January 29, 2003


Wednesday, January 29, 2003

It's not just EQ, it's pretty much all MMORPGs. Dark Age of Camelot and Asheron's Call are just two examples of RPGs that regularly introduce bugs with each month's update and patches.

Most software companies would love to have customers as tolerant as game players. Oh sure, they will bitch and moan like there is no tomorrow, but they will also be quitting work early, sacrificing family time or skipping their bath just to play the game. Addiction. Pure and simple.

Mark Hoffman
Wednesday, January 29, 2003

Wow.  Evercrack, indeed.  That Slashdot article is an interesting read.

Why is the bug tolerance so high?  I would postulate several possibilities:

It's just a game.  (Hmmm; now you know that I've never played Everquest.)

You've invested too much to quit because of a bug.

It's increasing your play time, which isn't perceived as having the same costs as increasing your work time.

Fear of being left out while everyone forges ahead.  (Which is, of course, why MMORPGs are more addictive than single-player games; if you skip playing, say, single-player Diablo II for a week, you pick up where you left off, but if you skip Evercrack for a week you'll be left far behind.  This is why I never enjoyed playing Diablo II online, and why I never bought any of those online-only games; it requires a real time commitment, and I just couldn't get into the idea of scheduling my life around a computer game.)

The nature of continually patched software.  Unlike software with release versions, if it's always patched, you may assume that the bugs will be found and fixed in the next patch.  Hope springs eternal.

Diablo II, which I've played a fair bit of, has gone through around a dozen major and minor patches since its inception; some of these have nerfed character skills that were, in the designers' opinions, excessively effective.  But in my own limited experience, they never damaged the playability of the game.  It was still fun, and that's why it got slack in the patching department.  On the other hand, I say this as someone who hasn't made an online time investment of thousands of hours...

Wednesday, January 29, 2003

" I mean you would'nt put up with bugs in accounting software. "

I completely disagree with this statement. The buggiest software I run is Intuit's QuickBooks - an accounting package aimed at small/medium sized companies.  Here are my top pet peeves - and they are quite serious: It crashes everytime there is a network disruption of any kind. The multi-currency "feature" only works half the time (providing the wrong numbers in certain reports like the Balance Sheet, where the numbers really ought to be accurate - one would think anyway) and once you enable it, you can't disable it. The payroll "feature" is implemented in such fashion that if you stop using Intuit's payroll tables mid-year (like we did when they told us that in order to continue there was a price increase of 300% and we'd have to get the software as a subscription instead of a package) - then all of your reports used for taxes will be wrong.  If you actually use the application for a number of years and you have quite a lot of data, searching slows down significantly (10 minutes just for a customer report). Finally, if you want to import data into the program the current method is so buggy that you can import the same file three different times - and one time it will fail, one time it might succeed and the third it will crash the application.

Why do we put up with it?  No real alternatives targetted to our market. The more "upscale" packages are way overkill (and at least three times the price).  The main "competitors" that are sold here in Canada are MYOB and Simply Accounting.  MYOB was purchased by Intuit - and we switched from Simply Accounting because entries used to randomly get "eaten" on the system. 

Given that the alternatives don't look that great and the cost of switching is very high in terms of set-up and time, we're likely to stay with the Devil we know until something substantially better comes along or until QB becomes unusable.

Lest you think we are unusual: we are currently in the midst of a project rewriting software used by a client for the last five years (It's a fairly well known cell phone repair and inventory package).  The software we are replacing is so terrible that the client has shipped their database back to the software vendor almost every week since implementation for data repair.  They can only search by first name, because there were a limited number of "searchable" fields and it was poorly setup.  The process to generate an invoice for their walk-in customers is so timeconsuming that the person at the counter usually uses the "quick sale" option - which records much less information and is useless for tracking.  It is slow and routinely crashes. It imports data into the client's accounting package (AccPac) - but the success rate is under 50%.  Sometimes the import overwrites previously written data eg replaces customer information with a completely different customer, or assigns invoices to the wrong client.  The software is so difficult to use that the users estimate that it takes more than two weeks to get back to speed after a vacation.  (And they even wrote their own manual because the one they had did not include the little "quirks" that they needed to know about in order to work around the problems).

After all that - the users are not that upset with the software.  This is "normal". 

The only reason we are rewriting it (ie creating a replacement custom app) is because the client wanted some extra features (which they were willing to pay for), and after a year and a half of little to no progress, the vendor still hadn't delivered...

My conclusion after this experience (and many, many others) is that users aren't actually that bug intolerant, especially if the software has few alternatives, is fairly well-known and/or the user has previous experience with the software.  If you've been using something for a number of years (cough Windows cough), and it gets buggier and buggier (anyone else having issues with Explorer?) - either there has to be an alternative that is light years better, or the software has to become nearly unusable before most people will switch.

And that's probably why most software is in such a state of bugginess.  People complain - but they don't quit using the software. 

Bah humbug
Wednesday, January 29, 2003

I'll take Bah humbug's points, and add one more little observation: I'm developing a new in-house system for a company that has software a lot like Bah humbug's redoing.  Terrible workarounds, bugs everywhere, fragile, etc.

We too are creating a new replacement, and it's slick, massively unit-tested, and has some nice new features.  We're shipping parts of it already, but are users switching? They are not.  Until we start providing them will real killer-app features, they're unlikely to do more than play with and 'beta test' our shipped release.

Wednesday, January 29, 2003

Some very interesting posts here.

Wayne, I have read that article and... well it seems there a lot of players who feel they have been "cheated" out of their "time" or their "money".  There are numerous rants like that one all over the Everquest boards.  It may be players like this don't understand the fickleness of software.  The "Why can't they get it right the first time because I'm paying $12.95 a month for this game and I play 24/7 and they are becoming rich off of me." syndrome.

Mark, I have'nt played the other games but I believe it when you say they are patched often.  It's the nature of software (the beast).

Kyralessa, I agree that time invested is a huge factor.

Bah humbug:  Very interesting post.  If I write a piece of software and a certain group of people decides to use it, they will stick with it even though there are bugs and the reasons are: time/money invested, fear of lost data, bug fixes are inevitable?, they become comfortable with the current software, they don't want to learn a new software package, they may not recognize or notice the bugs in the current software, any other reasons?

Dave B.
Wednesday, January 29, 2003

Off topic post!!!
Humbug, have you looked at ?
Start a new topic and tell me what you think. I am looking at packages, but this is not my area.

Doug Withau
Wednesday, January 29, 2003

First of all, that slashdot article was not particularly accurate.  Or rather, it was written from the perspective of someone who is extremely bitter about their experiences.  While you may think EQ is a horrible piece of junk from reading what people say about it on messageboards, keep in mind that the vast majority of those people complaining continue to plunk down their $13 a month, and continue to play several hours a day.  Why? Because despite all its problems, it is still an enjoyable experience.

As for putting up with the bugs, I think there is one significant factor to consider... there is no real competition in this market.  EverQuest currently holds the majority of players in this market, and that is because it is in most ways the best product available.  Several games were introduced that were barely playable for months after release.  Others have simply not reached the level of maturity of EverQuest, which has 3 years of added content and play balancing under its belt since release.  Even if people were so put out by the problems that they wanted to quit, there is nowhere to go for a similar play experience, so they stay.

Secondly, despite the problems with patches, it is actually a pretty stable game.  By far most of the problems are with play balancing.  Patches nowadays happen closer to once every few weeks, at which point the servers are usually down for a few hours.  That's usually the only time servers come down, which ends up being pretty respectable uptime.

Mike McNertney
Wednesday, January 29, 2003

As an addendum... you have to take anything said about games like this on messageboards with a huge grain of salt.  It really seems to bring the worst out of people.  As a post I recent saw said, "An EQ player could be frustrated if Absor hand-delivered them a million dollars in cash.  [Complaining] and actual game issues only have a slight relationship." (Absor is the head customer service representative, sort of).  Anyone who has been a member of the EQ community for a reasonable amount of time would agree with this sentiment.  It is strange, because at the same time the customers seem willing to accept whatever the game throws at them, they also have extremely high demands for what the game should produce, often asking for infeasible solutions, or solutions that contradict other requests they have.

Mike McNertney
Wednesday, January 29, 2003

Dave B., other reasons for sticking to buggy software unless it becomes unusable or unless an alternative emerges that is exceptionally and obviously better:

- expectations about the behaviour of "computers" (or "technology", or "software" - depending on your audience) are low.
-difficulty of assessing whether an alternative will in fact be better.
-many people will choose the "easy" route now even if changing the way they work would be more efficient/better  (Laziness? Fear of being replaced as a result of process change?)
-lack of alternatives
-resistance to change (especially true if the users have little input into/control over the procurement of the new product)
-change is feared as being very expensive (especially true of large / custom / used for years packages)
-feeling of inadequacy when using the software, making the user reluctant to criticize it (especially true of people who are not very technically minded - one of  the users I interviewed as part of the user needs analysis for the project we are working on now explained that he felt that most of his problems using the system were a direct result of him being "incapable" of understanding the system - a bizarre version of "it's not you, it's me").  If you get a bunch of these users working together, they can become absolutely convinced that they are the only users who can't take the bugs in stride or make up work arounds - so they try harder. Then there's an element of pride in terms of being able to work around the system that may come into play too.  That makes it more difficult for the group of users to make the decision to switch because a) switching is admitting failure (giving up) and b) people become hesitant to be the first one to give up on the software altogether.

That's all I can think of off the top of my head. Anyone else have theories on why users are bug tolerant of software?

Bah Humbug
Wednesday, January 29, 2003

I'm currently in the process of totally rewriting a company's internal billing system, and I was stunned at the poor quality of the existing software when I came on to the project(which was completed last year and is now being replaced). 

The application is a mod_perl website fronting a database, and the provious programmers seemed to take the maxim of perl being a write-only language to heart, so the code basically cannot be saved.  The database design is totally denormalized, and at the same time also full of tons of many-to-many relationships that only ever get instanced as one-to-one relationships by the business logic.  Billing reports for a single customer take hours, and sometimes even days to generate.  And the reports are so crudely formatted and constructed that people in accounting end up cutting-and-pasting data into excel so that they can massage the numbers into a customer invoice.

But it is internal software for the company, so there is no alternative (commercial packages don't handle the issues in this vertical market) and it is what people are used to.  Invoices go out late, and extra people get hired to compensate for the inefficiencies, and life goes on.  If you're getting paid to use a bad piece of software,  it takes a lot of the pain out of using it,

Colin Evans
Thursday, January 30, 2003

Another possibility as to why users stick with buggy software:  They actually think the click-wrap license means something.

I recently read Cem Kaner's _Bad Software_, which opened my eyes.  Not being versed in contract law, I actually believed that those click-yes-to-proceed-or-take-it-back-to-the-store licenses were valid and legal.  Turns out they're generally completely unenforceable in court, but I wonder how many there are out there like me who read that big AS IS and figure that if the software doesn't work, it's their own fault for clicking "Agree."

Story:  I now manage my money on my Handspring Visor with SplashMoney, an excellent program, but I used to use another company's similar program.  I used that program (which shall remain nameless) for the 30-day trial, liked it, and forked over my $19.95.

Unfortunately, only after the cash was forked over did I discover an annoying bug:  Whenever one of my accounts was at a round number total, usually about $50, the total would be wrong; I think it usually read $49.10 instead.  Kind of problematic since I like to keep my checking account at around $50 all the time (and transfer in from savings just before I need it).  I even tried manually correcting the total--and that caused the program to crash.

I contacted tech support several times, and I always updated to the latest version as they directed.  The problem just wouldn't go away.  Finally I told them I wanted my money back, because it appeared to be a bug that wasn't going to get fixed any time soon...and what good is a money management program that doesn't give the right totals?

To their credit, they replied promptly and said the money had been refunded to my credit card, and it was.  But I guess most people don't realize that, clickwrap notwithstanding, a defective program is a defective program, and if it has such serious defects that it doesn't do what it promises to do, you _can_ and _should_ ask for your money back.

Thursday, January 30, 2003

*  Recent Topics

*  Fog Creek Home