So you need a theme—

This comes up rather a lot, so I figured I should put my advice in one place. 

What do you do when your initial idea for a game does not include a theme? (Or if you have to retheme?)

Interrogate your design. That is the one and only step, but don't worry: I'm going to unpack that into useful tips. (More about this concept here and here.)

Ask interesting questions. Interrogating your design involves asking interesting questions. If the answers you get aren't interesting, chances are the questions you're asking aren't either. You don't just want any theme; you want an interesting one. But where do you start?

Identify one element. Decide what one resource or card type is. Which is to say, don't theme your whole game first; theme one bit. (If this exercise doesn't yield a good direction to build out a theme, reset to this point and start over.) The element you theme should spark ideas for the rest of the game. I.e. the element you theme first should be interesting. How do you make a single element interesting?

Find the movement. I'm assuming that your initial idea for a game included enough mechanics to play at least half of one turn. (Yes, I said turn, not round.) Because you're moving forward with the design, that means there must be something in your idea that excited you. Within that idea, which pieces moved physically? Of those pieces, which piece moved in the most interesting way? Theme that piece first. Why does that piece move that way? Find the most satisfying explanation and you'll have a good start on your theme. Not every game has to have pieces that move in thematic ways, but this is a good exercise to generate theme even if you end up moving away from your initial concept. 

Build out your theme. Now that you have one element, figure out what the rest are. This should be fairly easy if your initial element really resonated with you. If not, you can always return to the previous step and continue brainstorming. Identify each element in a way that makes your initial element more interesting. If your design shifts focus to a new "most interesting element," build out the theme around that element instead. You also need to decide who the players are playing as in this stage. That includes the characters' motivations—why they want to be performing the actions of the game. 

Find the relationships. As you are building out your theme, establish how the elements are related to one another. If the names of your resources could be easily swapped out for their colors and players wouldn't notice, your resources don't have thematic relationships to each other. It is easier to remember that wheat makes flour than that brown makes white; so if your playtesters aren't using your resources names it is likely because they aren't really thematic. (Made up fantasy names can be some of the least thematic bits of a game. Use sparingly.)

Identify the boring bits and troubleshoot. Play your game and focus only on the theme. Did you care all the way through? Which parts were you less invested in? Did you care about your player character? Did you care about the NPCs? If you built something, did it feel important to you? It's easy at this stage to blame the boring bits on the unrefined mechanics. Approach this stage with the aim to make the story your game tells better. You may end up essentially retheming the game at this stage, and that's ok. If a mechanic no longer works in your game, you would cut it and replace it. Think about theme the same way. Keep what's interesting and ditch what isn't. One way to know if you're on the right track is if your elevator pitch gets shorter and punchier. (Test your elevator pitch. There's plenty of online forums that will give feedback. In the process, you are covertly testing your theme.) Be aware, this step is an ongoing process as you playtest and tweak your mechanics. For more practical tips to make your theme more interesting, read my posts on resonance

Hopefully, at this point you will have a not just a theme, but a thematic game.