Civ 5 postmortem in Game Developer magazine March 2011

[Buccaneer]

It had been assumed that there were going to be serious issues after we had heard about the significant resource reductions last summer. You cannot pull something like this off without critical resources. I had said all along that one cannot blame the testers because they did do their job; but alas, only the high priority fixes/changes were all the programming team could do. Besides, can you imagine the firestorm at CFC if they did delay the game by properlly extending the development cycle? That would not have worked (even if 2K allowed it) because most people (esp. here) would rather have the game NOW (or on the announced released) regardless of its state.

 

[Zyxpsilon]

Frankenstein and beta testers are not critical positions. Lead MP programmer, that's a critical position. The fact the add for an MP programmer has been on Firaxis website for many months says lots about Critical Positions Were Still Missing Entering Production.

That's one aspect of development which is mysteriously undefined from a standpoint of minimal assets necessary in a project of this magnitude.
Indirectly, Beta-testers must work for free... so there isn't much presumed seriousness when the time comes to sift through relevant input on the Devs (programmers or not, btw) part.
Sure, these aren't critical positions - i entirely agree. It's the rational assumption in a tasking or duty perspective.
But we have to question coordination from leads and THEIR decision to trust (consider, ignore or sweep away under the carpet) some or all external inputs -- this isn't a failure that falls on contributors' shoulders at all. To the contrary, Frankenstein team was an essential provision and a big STOP sign at a street corner.

Devs drove the bus right through the intersection and there was no Police to catch them - on staff.
Conclusion? An unstable release & an unfinished project. Need we be reminded of the tragic Summer layoffs at Firaxis; one of which most probably a MP programmer you're mentioning.

 

[Drawmeus]

One thing that I feel is important to get in this discussion for context - just as it is in the GDC talks thread - is that Game Developer Magazine is literally not intended for general consumption. That's not to say I think anyone specific is wrong in what they've said in this thread, it's just something I feel needs to be taken into consideration.

When they say "We ran into these problems", they mean it to inform other game developers that things went wrong with the development process and cost them time/resources overcoming them, not that their product ultimately suffered from certain flaws.

When they say they had problems finding a MP programmer, they mean that they had a setback because they couldn't find one early enough and that other studios should probably prioritize it quite a bit more and quite a bit earlier if it's crucial to their product. They DON'T mean they were uncomfortable with their ultimate MP solution.

I'm not saying they're happy with the MP they delivered to the end-user, mind - I'm saying that it's literally a different conversation.

 

[Dale]

Beta-testers must work for free...

The 2K team in California who test all 2K games would like a word with you.

But yes, some of the Frankenstein team worked for free, but they were more than just testers too. As Dennis alluded to in the GD mag article, the Frankenstein team is more of an "external design team" than just testers. Yes, testing is one thing we did, but not the only thing.

 

[elthrasher]

(see the future development team hard at work below)

Those dudes look serious. By the time Civ7 comes out, they will be in their 20s or 30s. I'm really looking forward to what they come up with.

 

[Venereus]

Civ6 will disappoint a lot of longtime fans. Some of them will turn away from the series. Some of them will hang around to tell everyone how crappy Civ6 is over and over. They will be accused of ignoring the finer features of Civ6 and frequently be told they just wanted Civ5.5. Eventually Civ6 will get patched and some of the harsher critics of Civ6 will come back to the fold and begin to enjoy the game. Others will say Civ5 was the pinnacle of the series and it's only downhill from there.

Now who the hell wants to go through all that again already?

Actually, at this point I think the only way I'll ever enjoy Civ 5, will be when I'm finally using the Civ 4.5 Mod, wich should obviously be Civ 4 BTS on hexes and nothing more.

 

[Rooftrellen]

Wasn't Civ4 a bit....unpolished at later ages/techs, as well?

In fact, BtS was all about expanding on the game in the age of guns, tanks, and corporations...beyond the sword.

I saw something with Sid himself talking about "the first 15 minutes." It is logical that someone that feels the 15 first minutes at the most important of all would try to make those early techs more interesting, then expand on the later parts of the game though expansions, after those first 15 minutes already have your interest.

I know that before BtS, I could hardly finish a game of Civ4...not sure how CiV is much more of a failure. Maybe even less so since I can win though war by taking all capitals now...not taking over...dang...60%...of...the...world......and...needing...500....transports...to....get....that....last...city.

 

[Zyxpsilon]

The 2K team in California who test all 2K games would like a word with you.

But yes, some of the Frankenstein team worked for free, but they were more than just testers too. As Dennis alluded to in the GD mag article, the Frankenstein team is more of an "external design team" than just testers. Yes, testing is one thing we did, but not the only thing.

It's really none of my business to determine how they managed to pull a rabbit out of the hat at no cost or any other form of compensation... but when they offer facts or summary out to the media while blaming a lack of resources in critical positions to justify a stiffy product at release there could only be two explanations;

-- They simply ignored Beta-testers input for some reasons.
-- There was a conflict on design principles that lead them to revise staffing components.

Both of which didn't stop employees from losing their jobs before the September deadline enforced by 2K.

Now, considering the structure of such projects (as in scheduled solidly within capacity) -- how can appropriate measures be taken once the boat is simply sinking since the crew abandons ship?

There's responsability and defined tasks but there's also a clear pattern to follow for optimal results;

1) Concept
2) Coding
3) Testing
4) Quality control
5) Loop back to 2 when absolutely necessary

What i'm getting from the post-mortem article is that 4 & 5 were somehow skipped to meet a strict schedule. I simply can't believe that at that stage seats were empty or that actual people could have been either replaced or paid to keep the boat afloat until Coding has been wrapped.
They stopped paying under contracted principles.
While Frankenstein team couldn't intervene for the simple reason that they weren't expected to do so.

At least, that's how i perceive the internal chaos that it probably was; late summer, bad news hit, back stabbing knives thrown across the project "Parts" & sections.
What a mess.

 

[Dale]

It's really none of my business to determine how they managed to pull a rabbit out of the hat at no cost or any other form of compensation... but when they offer facts or summary out to the media while blaming a lack of resources in critical positions to justify a stiffy product at release there could only be two explanations;

-- They simply ignored Beta-testers input for some reasons.
-- There was a conflict on design principles that lead them to revise staffing components.

Both of which didn't stop employees from losing their jobs before the September deadline enforced by 2K.

Now, considering the structure of such projects (as in scheduled solidly within capacity) -- how can appropriate measures be taken once the boat is simply sinking since the crew abandons ship?

There's responsability and defined tasks but there's also a clear pattern to follow for optimal results;

1) Concept
2) Coding
3) Testing
4) Quality control
5) Loop back to 2 when absolutely necessary

What i'm getting from the post-mortem article is that 4 & 5 were somehow skipped to meet a strict schedule. I simply can't believe that at that stage seats were empty or that actual people could have been either replaced or paid to keep the boat afloat until Coding has been wrapped.
They stopped paying under contracted principles.
While Frankenstein team couldn't intervene for the simple reason that they weren't expected to do so.

At least, that's how i perceive the internal chaos that it probably was; late summer, bad news hit, back stabbing knives thrown across the project "Parts" & sections.
What a mess.

If it helps you sleep at night thinking that, well go for it.

 

[Leif Roar]

I think the main Software Engineering insight is that they didn't use Software Engineering.

Isn't it always?