DSDM Design Methodology (RAD)

The Viceroy · December 20, 2000, 04:48

OK, I guess this topic isn't everybody's cup of tea .. but I wondered what other people think about the DSDM approach to design which seems to be used by activision (knowingly or not ??)

Basically, DSDM (Dynamic System Design Methodology) is based around the idea that the deadline is unmovable, but the functionality can be reduced .. So if you set Dec 8th 2000 as a deadline, and you realise your not going to make it, you chop functionality and acheive your date, rather than let the date slip.

In my opinion, you should be conservative with your release dates, so for example .. you set your release date for March 01, let your boss know it will be finished in Dec/Jan .. and cut yourself 1 to 2 months to get the functionality you wanted in there .. ??

What do you think ?

------------------
"Wherever wood floats, you will find the British" . Napoleon

Colonel Kraken · December 20, 2000, 11:34

Sounds about right. It's too bad they had that design philosophy. But from other posts, it seems that if the design team had had its way, the product would have been much more polished. It seems the high-ups at Activision were the ones responsible for putting the vice on how everything was done.

By the way I love your quote. That is classic!

The Viceroy · December 20, 2000, 11:46

Unfortunatly it is a design methodology created for the modern business man .. not the hard working customer

I started the thread on the hope that somebody from activision may feed it back .. well, you never know ??

------------------
"Wherever wood floats, you will find the British" . Napoleon

Jen Dragon · December 20, 2000, 15:16

Look at avault.com and you will realize just how many patches are available for games the week they hit the shelves. You know it takes time to make a patch so they pressed a knowingly bad copy. They couldnt wait the 2 extra weeks to press and have a version one more patch closer to finished. Interplay just released Star Fleet Command 2 without its online Dynaverse as advertised in all the magazines and on the box. The patch is in the works not to mention the patch was on Avault befor the game hit the shelves for the part that was released, this is getting completly out of hand.
On the other side if they are a publicly traded company and earnings are less then expected, well just look at the NASDQ* they have to have something to sell.
And its the publicly traded companies that bring us eventually the good games. Some exceptions. So really I don't know what to do about it, because if I see a game I've been wanting for a year in one of my favorite flavors I can't resist and they KNOW IT. So I dl 4 or 5 patches and keep up. Sorry for the length of this post its a hot topic for me.

Alpha Wolf · December 20, 2000, 22:27

I always love that question at an interview:

Project is due tomorrow and you know it doesnt work, what do you do? Install or delay?

Guess we know Activisions expected answer.

------------------
History is written by the victor.

Pyaray · December 20, 2000, 23:16

Well, the question is never quite as simple as you have stated. There has to be an "end point". If there were no deadline dates, no game would ever ship. There is always "one more feature" we could add if we had just two more weeks. Creating a game is a creative process. An artist friend of mine once told me a famous (I believe) quote that "art is never truly done, you just stop working on it at some point." There are a few companies that believe that the ship dates are meaningless, but very few and far between. That's because for the majority of companies, business doesn't work that way, and they would go out of business trying to do it.

I would also point out that part of our pressure to release on a schedule is consumer based. This poll is from Apolyton:

CTP2: Delay if buggy?
No, release it today, fix the bugs later! 7 / 4%
No, keep the schedule 11 / 7%
Yes, if it is in the same state CTP1 was on its release 64 / 41%
Yes, until there are no bugs! 71 / 45%
Dont know/Dont care 2 / 1%

Now while it's quite clear that 45% of you all said to hold it till all bugs are gone, 54% found that some level of bugs are acceptable. Was CTP2 more or less stable than CTP1 upon it's initial realease? I don't know, that's difficult or maybe even impossible to quantify. Did we try our best to eliminate every bug we knew of? Yes, we did. We did the best job we could possibly do given the constraints we had. I also realize that this thread isn't about bugs, but this poll is along the same lines. If people wanted us to ship ignoring bugs, how many of those same people would have voted on less features for shipping earlier? (Please, don't do a poll on this.)

In the end, everyone always wants more. It is impossible in this industry to please everyone.

I hope this addresses the issue to some degree. I have worked at 4 game companies now, and 3 of them were large companies. It really isn't a simple problem with a simple answer. There are many issues involved. And while I may not agree with all the decisions made, I can certainly see why the decisions were made, and can't say that they were necessarily wrong. And the people making those decisions aren't stupid by any means, otherwise they wouldn't have the power to make those decisions in the first place. And that is my take on it from an insiders view.

Pyaray

Falconeer · December 21, 2000, 01:41

I think this is a model that just about every company uses now adays. Most companies now don't aim at a single highly polished great product to rest their laurels on. They want continuity. Set a release date for a product. If there are some features that can't be met by that date, don't include them. If the product sells then you have more capital to spend on including the features you left out in the next product in the series and the company feels better working on a product it knows has had success in the past. If the product doesn't sell then the company saves money by not having to pay its producers an extra period it would have if they'd extended the release date.

This goes on not just in the gaming industry but in just about all industries. It's a little more blatent though in gaming because they don't always change the names of the products, just add a 2, 3, 4, etc to the name!

Alpha Wolf · December 21, 2000, 02:36

Thats why i love that question, theres no right answer. ultimately it always comes down to acceptable consequences to sending out bugs.

My answer to that question is always; if you have to ask that late in the ball game, management failed very early in the process to come up with realistic project schedule. And from what i hear, the programmers at Activision did the best with what they had.

I would guess that the majority of people would agree that poor AI was CtP1's biggest problem. yet CtP2's doesnt seem much smarter, but they totally redid the UI. Seems like screwed up priorities to me. Does activisions management actually play these games?

------------------
History is written by the victor.

WesW · December 21, 2000, 02:48

As far as screwed up priorities, the biggest let down I ever had was when I played SMAC, and realized that they had done almost nothing to improve the AI routines from Civ2. I mean, the altitude concept for terrain and the government choices were well-done and interesting, but basic gameplay was totally the same.
SMAC was basically civ2 with another layer of features added on. Hell, there are probably more differences to gameplay between Ctp1 and 2 than there were between civ2 and SMAC.

The Viceroy · December 21, 2000, 04:47

I work for a large airline, and most of the time I have the freedom to make my own deadlines (lucky me) .. so I guess I have the perspective of a project leader rather than a programmer ... Certainly thanks to Pyaray for his thoughts on the issue .. Im not certain that it is a methodology focused on the customer .. Im more inclined to think that Falconeer hit the nail on the head when he talked about revenue .. An interesting perspective .. (as I just deliver my projects to the airport .. don't get revenue from it).

Call me Scrouge (actually, no don't please

), but Christmas is one of the major contributors to this process .. as its a cut off date for developers, keen to get the software out before Christmas ... but don't worry, im not suggesting we cancel Christmas

ha ha certainly not to please my love of well estimated project cycles .. BAHHH HUMBUG!

------------------
"Wherever wood floats, you will find the British" . Napoleon

Alpha Wolf · December 21, 2000, 06:04

I totally agreed that the xmas season was written in stone, but how long have they known this? As a project leader, I do the necessities first then the extras based on what time is left. I betting that a vote of which is more important, smarter AI, or new UI, that smarter AI would win hands down.

For some reason, i get the feeling that Activision has many pointy haired people running things. (hmmmm, do y'all get Dilbert outside of the US?)

------------------
History is written by the victor.

The Viceroy · December 21, 2000, 06:15

Yes we do

.. and yes, I certainly agree ..if you can't plan for a deadline set in stone ... Then customers have mercy on their soals.

------------------
"Wherever wood floats, you will find the British" . Napoleon

Pyaray · December 21, 2000, 08:15

Just a note about the AI. There was much more time spent on rewriting the AI than on changing the UI. We did focus our attention where it was needed most.

Pyaray

December 21, 2000, 09:59

Just one question,

how come that a person who plays the first game on impossible level, gig. map, 8 civs, and thoughest barbarian level wins straight??? (ok there was some fighting but this is my first game)

OK, perhaps I played earlier CIV incarnations a bit too much, but still I would say that impossible should be what it says - trully challenging ( and than you could have lower levels )

I believe that you have put a lot of effort in making AI better, and that this is not easy by any means... but the finished product (AI) is a bit too easy to beat.

P.S. perhaps you could advise some of the mod makers - (like Hrlan, WesW) not me- how to make AI better trough tweaking of the text files. e.g. how to attack more etc... (if this is legally acceptable)

December 21, 2000, 10:32

i dont want to go off-topic, but does a harder to beat ai means a better ai, no matter how that is achieved??

December 21, 2000, 11:15

Not neccessarily, (depends what kind of player you are) but i guess I expect to be beaten by the AI sometimes, or the joy of beating it when you really thought that you blew it, is much bigger. Ok previously i didn't know all the correct tactics, but still this is a new game so it would be nice for AI to be a challenge. (would make it more tense and fun i guess)

i still think this is a great game, but when i remember now when i still had tougher levels to beat... I really looked forward to it... unfortunatley not in this case...

but there is hope as difficulty is concerned in the mods...