34
\$\begingroup\$

Level Design Creep is what I've started to call what happens when you build a series of levels and notice that every level is a little bit better looking or designed than the last, or when working together with several people with different skill levels. Left alone, as the player goes through the game the level quality would go up and down according to the order the levels were built. It seems like we're spending more time on this than we should be.

Besides cutting out the really good and really bad levels, is there a way I can plan for or manage the production so that the detail doesn't "creep up," and we have to revisit the same work over and over?

\$\endgroup\$
2
  • 16
    \$\begingroup\$ Don't cut out the really good levels!! Cut the bad ones if you can't save 'em, but don't bring down the quality of your game just to have a more consistent quality level... \$\endgroup\$ Commented Jan 24, 2013 at 6:09
  • 1
    \$\begingroup\$ well obviously the problem is you are becoming more skilled as you create levels. I have to kind of grind my teeth when I call that a problem however. I suggest that you go back to your older levels and just refine them a little. As you didn't become a master overnight I am sure they are not that much worse than your current levels, so you probably only need a minor amount of tweaking (adjusting the set dressing most likely). It's probably not a huge problem either way, as your players go through the game they differences should be rather subtle and not noticed until they replay it. \$\endgroup\$ Commented Jan 24, 2013 at 18:58

1 Answer 1

40
\$\begingroup\$

This happens in almost every game -- as the artists become comfortable working with the tools, and as improved tools become available during production, the later levels to be constructed are almost always built to a higher level of quality in a shorter amount of time.

To cope, you want to do all of the following three things:

  1. Assume that you'll have to revamp the first few levels built by each artist and level designer. For a game with one artist/designer team which is making ten levels, budget to do extra work on the first two levels toward the end of production, to bring them up to meet the quality level of the later-produced levels. If you have more teams which are working independently, then you'll need to revamp more levels.
  2. In addition to the above, plan for each team's first few levels to take longer to build than the later ones. My experience has been that budgeting for 2-3 times the normal production time for the first two levels from each level designer/artist is in the right ballpark.
    • The idea here is that allowing extra time for the first few levels lets the designers/artists go through the whole learning process within the confines of just one or two levels, so you only have to revise those few levels at the end, instead of having to revise everything because they were still discovering new techniques while building their later levels.
  3. Bearing in mind that even when following the two recommendations above, the first few levels you build are still very likely to have a slightly lower quality level than the others, so don't let your team build your game's first level first. The first level you build should be somewhere in the middle of the game. Having a slight dip in your general "level quality bar" can be catastrophic if it's the first thing the player sees. Much better to put it midway through the game, so the player won't hit it until he's invested in the game.
\$\endgroup\$
6
  • 4
    \$\begingroup\$ Also, pre-production isn't just a fancy word for "free lunches at meetings every day." The time should be spent proving out your pipelines and generating schedule data for Trevor's great notes above. Especially #3. \$\endgroup\$ Commented Jan 24, 2013 at 5:49
  • 11
    \$\begingroup\$ +1 for option 3. Building levels in a different order than they will be presented not only will help you distribute the level design quality across the entire game, but can also help you as a director view the game as a whole product. \$\endgroup\$ Commented Jan 24, 2013 at 6:56
  • 3
    \$\begingroup\$ +1 for #3. Having your first level be the least appealing one could indeed be catastrophic. \$\endgroup\$
    – Leo
    Commented Jan 24, 2013 at 7:51
  • 1
    \$\begingroup\$ You could even have the first-designed level be an optional bonus level somewhere in the middle of the game. Nobody really minds too much if bonus levels aren't so perfect, or differ in style from the rest of the game, because, hey, they're optional! \$\endgroup\$ Commented Jan 24, 2013 at 9:20
  • 4
    \$\begingroup\$ Another +1 for option 3. Another important reason to build middle levels first: the early levels of a game typically don't stress all the systems of the game (players gain new abilities in one form or another as they go on), so if there are problems with your more advanced systems - be they the 'smarter' AI or better player traversal or what-have-you - you'll find them earlier than you will if you build your levels in linear order. \$\endgroup\$ Commented Jan 24, 2013 at 19:15

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .