Hmmm, interesting question. Here are my five things you shouldn't do for the jam.
1. Don't worry about changing your ideas midway through the jam. My final games are often very different from what I originally envisioned, but by actually playing the game, I was able to experiment, tweak and sometimes even redo stuff. The game that I 'won' the jam with was completely different when I started it, it was almost like angry birds meets cut the rope, but by playtesting and tweaking my idea, I was able to make something way more fun.
2. Don't worry about going out of your comfort zone. In my opinion, jams are an awesome opportunity to do something new, something that you've never done before. Especially for this longer jam, you've got time to experiment. Don't overburden yourself or anything, but consider trying something new, perhaps an artstyle you haven't used before, or a new mechanic.
3. Don't leave all of the polish to the end. It's fine to do a bit at the end, but I'd argue that it's important to polish throughout development as well. Polish is really important, and I think it's a bad idea to leave it all to the last minute. Polishing throughout ensures that your game feels good, and if it doesn't feel fun, even when polished, then maybe it's time to make changes. I also tend to forget what things do towards the end, which can make it hard to polish.
4. Don't forget to take breaks. In the afternoon and evening, I'll start losing focus, and I find that a quick powernap helps. I also think it's important to step away from your work every once in a while to get a fresh perspective.
5. Don't let the scores get to you. This was my gamedev win, ever, and it's taken many games to do it. Honestly, I don't expect to win again, I'll just be proud of whatever I make. I hope that everyone else will share this thought, and know that your score doesn't matter.
And here's a sixth, bonus one: Don't forget to have fun!
Anyways, I hope that helps!