I would suggest it appears to be clear that every game needs some kind of love: the little details that gives us a smile when we spot them. That one fascinating experience we want to tell our friends about. Often things that seem to be unnecessary for the game itself.
This time, I want to share some thoughts about the struggle of taking enough time for adding love to a game.
Especially when you have limited resources this can be a problem. In most cases this means limited money and therefore also time. I never experienced a game perfectly finished to existing deadlines. There were always features and work that couldn't fit into the given time. Because of that each and every feature has to be prioritized. For instance, this can be done with questions like:
What does the game need to function?
What are the basics our users expect?
What are the requirements for our selling partners?
Imagine you had to decide between customizable controls and letting your players having a random snowball fight.
There is a long list with necessary things that have to be done and that's why adding love has to be a consequent decision. One could believe that if you work through prioritized lists constantly one day you will get to the lower tasks. Let me just tell you that this won't happen.
We were certain that Net.Attack() needed a lot of those love, so in the beginning we made it a part of our vision. As mentioned before it is not a surprise that this didn't lead to including those tasks into our plans. A long time then love only found it's way into the game, while we had a little planning free time or when we were a little ahead of our plans. After a while we realized that we wouldn't reach our goals this way and needed to find a better process.
Currently we use some kind of task pools for different categories which we give a specified percentage within our 3-week plan. So we mix a large part of prioritized tasks with a smaller one of those pool contents. In practice this lets us complete 1-3 "unnecessary" features each sprint, such as:
A Bootscreen:
Or the DPS meter:
For us this method gets the job done without leading to long discussions about feature values or time management and is probably a project learning for the next production.