Managing The Project

Chances are, you are good at managing yourself and your own time. That’s one of the things managers look at when they decide on whom to promote to lead positions.

But, managing a small portion of a project and managing an entire project require different skills. You are no longer just concerned with getting your slice of the pie done. You are also responsible for making sure that others get their work done, and that both the people you work for and the people that you manage understand where the project stands at all times.

Scheduling

How do you start? The first element necessary for building a schedule is the design document. The design document is the blueprint for the entire project, and lays out what the final product will be.

Determine the assets required by the design - number of levels, characters, kinds of animations, shell screens, fonts, weapons - the meat of the project. Once you have an idea of all the pieces that need to be assembled, start looking at dependencies. Most decisions about what needs to be done first will be common sense, but certainly not all. If levels can be tested with old character models, then new models could be done early or late in the process. Objects for levels may not ever be critical issues in development, so they could be stuck in anywhere in the schedule. Shell screens could be polished at any point in the project. It is at this point that talking to the engineering lead is critical.

The art schedule and engineering schedule have the greatest chance of clashing and causing conflict. This is common sense, really. Engineering needs to use your art, and your art needs to be supported by their code. These facts lead inevitably to a number of “chicken and egg” scheduling puzzles.

One of my schedules on a previous game had levels being completed first and characters later. Engineering, as it turns out, needed the character models to write climbing code. If we hadn’t caught that conflict, we wouldn’t have been able to test any levels that required ascending a ladder for several months after the level was done.

Sit down with the lead engineer and try to find potential conflicts in your schedules. Some will be more obvious than others, so devote some time to this and ask a lot of questions. If you have trouble poring over schedules face to face, email them to each other and schedule a time to talk about them later. If you have concerns and priorities, bring them up. Are you worried about a usable interface? Are your character models doing something they’ve never done before? Is there an effect or several effects that you feel are necessary to sell damage models? If you have a good relationship with the lead engineer from the start, then he will more likely be able to accommodate your need for tools, testing and scheduling. In the same way, you can make his job much easier when you can accommodate his needs and concerns. Individual items may require some debate and compromise, so be informed about your own needs up front and be flexible.

By the way, it is never too late to work out these conflicts, it just becomes more painful the longer you wait.

What do you need to plan in case things go wrong?

You know from experience that something always goes wrong. No matter what your level of experience, you will hit roadblocks that you haven’t encountered before, or some that you’ve dealt with before that you still have no solution for. Resist the temptation to schedule so tightly that any problem will throw you off.

When talking to engineers, they will tell me with a straight face that they schedule everything to take three times as long as it will really take. How often do engineers find themselves running behind schedule? At least as often as artists do. That should give you pause. If they schedule that way and still slip, what chance do artists have of doing better, with artistic temperaments and less well-defined problems?

The reality for most companies is that teams can’t schedule that loosely, so have good, solid reasons for what you do pad out. Where is the team least experienced? Where are you most likely to need to re-do art assets? Where is the project breaking new ground in technology or playing style? Look for areas that have the most overlap with other parts of the production team, and there are you potential liabilities.

No one can plan for every contingency. Three things will give you the best possible defense against slippage in your schedule:

1. Hitting deadlines regularly to reduce the impact of problems.
2. Being flexible and having contingency plans. What features can be cut if they need to be? What art needs the most attention, or least attention? What is the core of the art assets, and what is polish? And, …
3. Maintaining a cordial and smooth relationship with other leads.

Managing The Team

At the same time that you will be working out the technical issues of the schedule, you will be spending at least as much time and probably more on people issues. Be aware that the art lead will be managing up, down and across - not only people under him, but the managers above him and the leads on his level.

How will you keep up with the artists? (You could also pose this question as, “How will you know if you are on schedule?”) The answer is complicated by the fact that you will spend some time working on your own art tasks, and a lot of time in meetings outside the team room. The problem gets worse the more people you have on the team, and the more inexperienced people you have working under you.

At Red Storm we have tried a number of ways of getting information on individual and team progress. One was to ask everyone on the team to email his or her leads with a daily report. Everyone hated that, so answers ranged from helpful to downright insulting. I had at least one artist who just sent me the same note every day by hitting the re-send button in his email. We had some very good engineers who flatly refused write anything at all. It could have worked, but it didn’t. From there, we went to weekly emails, which were better received, but people had trouble remembering what they had done earlier in the week. We have also tried setting up an internal web page where team members could check in with updates. Lately, we’ve developed time sheets with task areas already typed on them, and empty blanks for hours spent. None of these methods were complete failures or complete successes, and they emphasize the need to have one-on-one contact with your team.

A balance of direct and indirect contact is important, not only in getting the data you need, but also in understanding what people are saying. The methods you choose for knowing if you are on schedule and on track should be developed early and tinkered with over the life of the project. In the same way, you should count on trying different approaches to maintaining contact with your artists. Some will naturally require more attention than others, and that time will shrink as you get later into the project. But cycles of turbulence are natural, so stay tuned for the next post which will look at team dynamics.

This entry was posted on Wednesday, April 23rd, 2008 at 6:03 pm and is filed under How To: Art, RESOURCES. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.