There's a principle that says that if you can't explain something to a 5 year old, you don't really understand it yourself. I don't know if I think that's true or not, but it's certainly true that figuring out how to explain things to others can sharpen your understanding of them.
There's a useful corollary for people involved in tabletop game design: the harder it is to explain a rule, the more likely it is that you don't actually need it.
I have an attraction to rules that are
neat in the sense of "what an interesting intellectual exercise in abstract simulation!" than
neat in the sense of "what an orderly and tidy thing that fits together well!"
But when I decided to focus my resurged efforts on Adventure Song in the starter classes and the opening levels of gameplay, I made the decision to specifically focus on putting them together in a basic guide package aimed first at playtesting and then at standing as an entry level approach to the game, rather than having a plan to make such an entry level thing.
At first, when I came to things that were essential but too complex to fit the "entry level" theme, I tried to brainstorm simpler alternatives that could be put as a placeholder rule in the basic package and then used as an optional variant afterwards.
But as I did this, I often found that the "optional entry level" version of a rule was the better one. So then, whenever I found myself having a hard time explaining my ideas, I'd start changing the rule around until I came up with something that answered the need but could be explained more concisely, or was less cumbersome to the players.
In the process, my skill system went from one where players have 3-5 independent pools of points to spend and the points are all derived by dividing individual attributes by 5 into one where everyone has one pool of skill points that's created by adding two attributes together. The off-puttingness of arithmetic tends to be a blind spot for me, but I figure that adding two low two digit numbers is less cumbersome than dividing 3 to 5 of them.
The idea that I found so interesting that it necessitated having separate points for physical skills, mental skills, and social skills is still represented in the system. I just built it into the cost of individual skills rather than making each type of skill bought as its own separate (yet repetitive) step of character creation. Another way this improves upon the "separate pools" scheme is that characters can spend their points as they see fit instead of having their points divided between three categories.
The interesting idea is still represented, but players are less constricted and the game is lighter weight.
A similar simplification happened in the combat system. I'm using group initiative (as was once the standard in D&D), because I think individual initiative tends to undermine the importance of teamwork, but at one point I was envisioning each round of having two phases during which each side takes a turn, with some actions happening in the first phase, some being broken across the two, and some always happening during the second phase. Some actions it would depend on circumstances when they happened (like attacking after moving).
The basic idea was that wizard-type spells should be slower than regular attacks or sorcerer-type spells, but the execution was complicated and kept picking up new facets.
So now? No phases. Winners go, losers, go... and a rule that "slow actions" (like casting) don't happen until the end of the round, with initiative winners first and then losers. Something like that I can explain in one sentence, and if the speed of actions don't change by circumstances, the only people who need to pay attention to this rule are the people who are going to be casting every round.
Other things that have been trimmed include a proliferation of character pieces that stack together independently of character class, the number of spell channels casters are expected to juggle, and the types and number of character resources the game uses for special abilities.