In philosophy of law, the contrast between rules and principles is sometimes (maybe "often" would be better) understood as being between being amenable to all or nothing application (rules are like this) vs having weight but not, in themselves, being determinative (principles are like this).
An example of a rule: do not drive faster than the marked speed limit.
An example of a principle: those who create risk should bear the cost of harm that results from that risk crystallising.
Some legal requirements can be stated with the syntax of a rule - eg a person who causes harm by negligent action, to someone to whom they owe a duty of care, will be liable to compensate that harm - but may be better analysed as complexes of principle (eg the negligence "rule" combines the principle I stated just above, with other principles about remoteness, foreseeability, community and professional standards and expectations, etc).
I'm not sure what's at stake, in the context of RPG design and play, in the distinction between rules and principles. In Apocalypse World, a distinction is drawn between principles and moves. Player-side moves are particular mechanical "widgets" and processes that either affect other moves, or else mediate changes in the fiction (as per Vincent Baker's work on clouds and boxes). GM-side moves are particular patterns of fiction-introduction, that help discipline and direct the GM's introduction of fiction.
Principles, in AW, are not directly methods of, or patterns for, adding new fiction. They guide the GM in their making of moves: they play a sort-of "meta" role.
I don't see what utility there is in denying that the AW principles are not rules. They set normative standards for the GM to comply with. Sometimes they might need to be "balanced" or "weighed" against one another - eg Look through crosshairs and Be a fan of the players' characters might sometimes come into tension (suppose a NPC is very precious to a PC) - but this is not true of all of them. Make your move, but never speak its name and Make your move, but misdirect, are particularly clear examples here: there is no reason that the GM should not comply with these in every moment of play.
Here is something.
A game should tell its participants how to organize their minds. What they should be thinking about. How they should be orienting to play at each moment. How they should be engaging with all of system, premise, their fellow players, the unfolding situations/obstacles, and their responsibilities at large.
Modern design has come a long way to formalize this for GMs. However, I'm finding it lacking in terms of formalizing this for players. Some games do this better than others. Some games (even games that I love) are a bit mixed. An example of that last bit is Torchbearer. While Torchbearer does a fantastic job of formalizing procedures and organizing how people should be orienting to those procedures specifically, there are two smell sections in the books that tells players to "Good Idea Fish" (lets call it). That
not engaging with the system is "the best play." I_hate_that. Its terrible advice imo. And it does this while simultaneously (or directly after) telling players to think about how they use their gear and their skills and their surroundings to defeat obstacles. It does this while it promotes all the rest of the game (describe to live including help, how important testing is, how important failing and gaining Checks are, working toward your thematic material even if it complicates your life). So Torchbearer has 99 % of its content that pushes toward "use the system cleverly because you need to for every aspect of play (for advancement, for currency management, to survive at all, for thematic heft <which also feeds back into all the latter>)...but then it has a pair of blurbs that orients players around the generalization of
the best play being to not engage with system (with really...the entirety of play being an exception to this...and the text laboriously, and correctly, telling you this).
So, truth be told, the best Torchbearer play I've run is when players surmount this principle and pretty much always orient themselves to play around
how can I best leverage system at each moment to generate my best lines of play and choose deftly from those.
When I think about PBtA games, there is sometimes a different issue that arises because of the way "begin and end with the fiction" is expressed. That is a great player principle. But you know what you need to be doing in the middle? Be very cognizant about how those clouds feed into boxes (moves, gear, HP/harm, etc); which then feeds back into clouds. Its not just the GM's job to be thinking about how moves trigger, how they might manifest, what a custom move might look like, how currency interacts with this particular declared action. Its for the whole table and that includes (perhaps especially) the player in question.
I can trivially say that the best players I've GMed for in PBtA games are mentally oriented to play in each moment not dissimilarly to how I am as a GM. They are thinking:
* What is the nature of this threat/obstacle/opportunity in front of me? What is it trying to do?
* Ok, I've got my character...how would this PC approach this to best surmount it or how would this threat/obstacle/opportunity cause this PC trouble that leads to further compelling play?
* Ok, I'm going to have my character do x which will trigger y move. Am I clear on prospective suite of type/kind (even if not exact) consequences? Let me talk about that exactly as I relay my cloud (fiction) to the table, then say "I guess that triggers y move, right(?)", then talk to the GM about consequence suite if I'm unclear and/or want/need to crystallize prospective fallout a bit. Ok, we've talked a little more about how clouds (fiction) feed into boxes (mechanics) which feed back into clouds (fallout/consequences within the fiction). Maybe now I ask "hey, if I loadout n, how will that help (or, better still, maybe I put forth how I think that would help)?"
* <Resolve chunk of play>
TLDR; Games shouldn't put forth a player principle (even if the rest of the system deluges you with the inverse) to either (a) avoid engaging with system or (b) not being responsible for the way clouds (fiction) relates to (boxes) which feeds into further play (either/or/both clouds and boxes).
Unless the game is basically a Freeform game exclusively, player principles should pretty much always begin with:
1) Engage with system proactively (not passively and do not try to "work around system in your play").
2) You're responsible for thinking about the fiction, the mechanics, and how the two interrelate.