My team that i mentor and was a part of (who might be browsing this forum) has a hard time starting off the year. They generally have many ideas but either take too long to produce something effective or are unable to produce an effective robot.
So my big question to anyone is how does your team handle planning your robot?
Decide on a job for the robot. A single job that makes you useful and reliable. The design should be simple enough that you dont have too many problems. You should get the design done within the first week and finish building by the fifth week. Another tip is to not over load it with features. When building the robot and wile designing you need to make sure it moves. The base of your robot is the most important.
May I recommend “Effective FIRST Strategies” by Karthik Kanagasabapathy, strategy mentor for FRC1114
My team usually starts the robot by switching on the breaker 
But in all seriousness my team starts out by building the whole base to see it together, no cuts made just yet. Then we have a few designs prototyped and see which one will fit our capabilities and strategy.
Focus on the slide titled “Simplicity & the Golden Rules” in the presentation linked below. It goes with the video linked by Ryan and is consistent with what Paradox is recommending.
http://www.simbotics.org/files/pdf/simbotseminarseries-strategicdesign.pdf
My personal interpretation of “3 functions at 10/10” is that each of the 3 functions is successful in at least 90% of the attempts to perform that function. This means your team will have to do a lot of testing and refinement. You are likely to have to go through 3 or more iterations of your functions, some taking a totally different approach from the previous iteration. This means that team members and mentors have to let go of their ideas and let them die if they prove to be ineffective. This means that your team makes mistakes and learns from them.