Tuesday, September 11, 2012

Agile Development Revisited - Rebuttal

Ripard Teg wrote a post yesterday that I'm going to have to disagree with. In COTW: Agile Development Revisited he is complaining about the speed at which the overhaul of ship balancing is occurring:
And yeah, I'm forced to agree. Best information says that frigates and destroyers will be done for winter, with the addition of one new destroyer for each race and the new logistics frigates being major features of the expansion. With the mining ship updates, that's roughly 40 ships rebalanced in 2012. Which means that at this pace, CCP will be getting around to balancing T2 battle-cruisers sometime late in 2015 and Black Ops, other T2 battleships, and maybe the T3s the following year. So as much as it sounds like an exaggeration at this level of resources, "a half decade" to get the re-balancing done is pretty much on-the-nose.
I started composing this post prior to the release of this dev blog which basically proves my point, but I'm going to write it anyways.

I work in an Agile software development shop, doing 2 week sprints and about 10-12 sprints per release. Hell, I even took scrummaster training and am a scrummaster of an eight person team and we've been doing Agile/Scrum for about 3 years. One thing that becomes apparent if you do it all right is that initial stories (i.e. pieces of work) take more time and effort than subsequent stories in the same development area. There is always ramp up effort to learn the existing architecture, determining the known unknowns, and making a task plan. Then there are refinements after the first couple stories as earlier designs and ideas get proven out or thrown out altogether.

Subsequent stories can build on the groundwork and lessons learned of previous stories. Backlog grooming for future stories can take these lessons into account as the picture becomes clearer. The end result is that no matter where you start, early stories take a lot more effort than later stories as the variables get nailed down.

The dev blog Ship Balancing Winter Update kind of proves this out, lucky for me. Instead of only the logi-frigates and new destroyers we are getting all 16 cruisers revamped, putting on the step of the all important battlecruiser and battleship classes.

So why was the initial ramp up of the first few frigate so long? Well, the plan of how to go from tiers to roles must of have been controversial and had a lot of false starts and revisions within the planning of CCP itself. How to make ships distinct without covering the same roles is a difficult proposition as I can attest to in my ponderings. But now that the framework is mostly set in the 24 frigates, it will be easier to take that model as a base and apply it to the other classes. Plus the devs have a better feel for what works in balancing and what does not work with the players and ships. Less trial and error, more forward development.

When I first started writing this post, I was going to say in reply to Ripard's post that "yes, the winter update is not a lot for ship updates but I expect the rest of the classes (crusiers, BCs, BSs, Tech II) will be a lot faster". However the dev blog shows that the acceleration is already begun.

I predict that the summer update will see all 24 battlecruiser and battleships done and discussions for what Tech II ships need love to have begun.

1 comment:

  1. Well put, I too work in an Agile development team, but I am too far removed from the process to speak authoritatively about the subject. Glad to see CCP pushing out more work than expected, the ship changes will make a whole different game when they are done.