System Matters Less
At no point in this piece of writing do I claim that System Doesn't Matter.
I've been thinking about the ways we talk about game systems, and how it's almost impossible to discuss the role they play without immediately finding yourself at loggerheads with people reacting to something you haven't claimed. The phrase "system matters" has endless baggage, and I'm not at all interested in relitigating any of it. What I am interested in is looking at why it matters more to some people than to others.
Let's define some terms. You may disagree with how I'm using them. That's your prerogative, but these are the definitions I'm going to be working with in this specific discussion. You'll also note, I'm sure, that there are no value judgements being made here.
For the purposes of this discussion, "modern trad games" or just "trad" are games that share lineage and ethos with a play style that developed and was popularised somewhere between the publication of Rahasia in 1980 and the publication of Player's Option: Skills and Powers in 1995. Yes, that's a broad time frame. No, I'm not concerned with that.
This is, for all intents and purposes, the dominant mode of play in the industry and is of course not just confined to D&D. The Storyteller System that emerged in 1991 with the initial publication of Vampire: The Masquerade and all the many iterations of White Wolf properties may claim not to be trad but the benefit of hindsight is that we can see they very much are. Basically every game that came between the mid 80s and early 00s probably falls under this definition of trad (with Amber Diceless being a notable exception).
The other term we're going to use is "OSR" or "old school". I don't mean this in terms of the marketing label or the multiple and fractured scenes that currently claim that label. The way I'm using it is as a matter of ethos and approach to play. These tend to have a much lower complexity/burden of rules than modern trad, much faster and less detailed character creation, and much more of an emphasis on discovering characters through play. (This style of play shares a lot in common with so-called "storygames", which I'm not touching on here except to say that this isn't surprising as both schools of play and design largely exist as reactions to the dominance of trad. They just approach things from a different direction).
In general, old school games care less about system than trad. And I think that a lot of our difficulties in talking about the importance of system come when people from these different schools of thought try to have a conversation. As we'll see in a minute, the priorities each school brings to the discussion are vastly different in ways that seem to go unacknowledged, and this is - in my mind - the source of friction.
To my mind the main difference between Trad and OSR is in where the focus lies. This spills out into how things are written, how products are built, how games are played - and how we talk and think about the role of System. These aren't entirely distinct categories - there is some overlap between styles, because of course there is - but let's ignore that for the sake of argument.
Let's take Trad as the starting model because it's been the D&D model since the 80s and, as established, everything else is a reaction to it.
Trad cares about characters more than anything else. It cares about the world, of course, and the GM preps that world ahead of play, but it cares mostly about character. Players spend a lot of time building characters. In games like 5e and Pathfinder we often spend hours building characters ahead of actually getting together and playing. We plan "builds" across multiple levels, write backstories, figure out how our characters connect to one another and to the world, all before we ever sit down as a group to actually play.
Often the things players decide ahead of time about their characters inform the world the GM preps, and play is largely about watching those characters navigate a world that has in a lot of meaningful ways been designed with them in mind. Character death is frowned upon if it's "arbitrary" or "isn't meaningful", because we've invested time and energy in these characters.
Because we care so much about characters we see products released to provide new character options, new ways for players to express and explore and prep their characters. Players are the driving force in this model of play, and everything the GM and the publisher does is in service of them. (My language sounds like this is a criticism but it's genuinely just an observation).
As a result of this, the way players interact with the world and with their characters is primarily through interacting with the game system. Everything they need to know about their character exists on the sheet and, for the most part, is mechanised in some way. If it doesn't exist as a discrete chunk of Rules that your character explicitly has access to, it's highly likely that you can't do it. (Thinking about my dates in the definition of trad earlier, we could even roll it back to the introduction of the Thief as a class in Greyhawk in 1975. The second we carve out a set of abilities that until that point all characters had been assumed to be proficient at and say "now only this class can do this", we've invented trad.)
In trad, system is player-facing. It's the key way we interact with the game and the game world. It matters in a big, meaningful way.
At the other end, OSR is focused mainly on the world. The GM preps the world, the players turn up with characters who are largely blank slates at the start of play. Characters don't get prepped in the same way, they don't exist in the same way prior to play. We roll 3d6 down the line, buy some equipment, pick a handful of spells, and roll into the dungeon. We might have a name and a vague idea of who we might be but level 1 characters are, largely, an amorphous bag of integers waiting for someone to hang a personality onto them.
In this mode, play is a matter of exploring a world that wasn't built with you in mind, exists and operates independent of you, and doesn't really care about you unless you force it to. Players learn about their characters in the moment, watching them grow and develop through play. Their character sheets are largely empty, and the way we interact with the world is by prodding and poking and questioning and reacting. Because we don't have feats and skills, we can just attempt things that there may or may not be explicit rules for. Often nobody at the table knows how something will be adjudicated until it comes up in play.
This manifests itself in play style and in the manner of productisation. The rules are largely GM-facing, and tend towards procedures rather than rules: here's how you build and populate a dungeon, here's how you check for encounters, here's how you check for reactions, morale, distribute treasure, etc. Publishers tend to release adventures and settings, not player options, because the GM is their customer, not the players.
In play, it manifests as different priorities. Death isn't "arbitrary" and it isn't the result of the GM planning an "unbalanced" encounter, because balance - a facet of a tightly-built, characters-as-protagonists system - largely doesn't exist. Where trad expects that players will win most (if not all) combat encounters, OSR makes no such assumptions. The world and the things that populate it exist, it's up to you to assess whether or not you can handle any given challenge, and you live or die by the consequences of your decisions.
As a result of this the actual system part of the game - the mechanical nuts and bolts, the things characters can do and have done to them, the Rules As Text - become less important. There's a reason that trad games tend towards being 250 pages books that they to cover every eventuality in the rules and old school games don't, and it's entirely about the approach to character. That's also part of why modern OSR games that have moved on from being Just Retroclones tend towards rules light, because system matters less here.
And all of this is why we often find ourselves unable to talk about the purposes system serves. Trad wants a rule for every occasion, because the rules have direct impacts on the efficacy of characters. We can't just have a conversation with an NPC and roleplay it out, because some characters are better at talking to people than others and we need to know how to adjudicate that regardless of whether the player is themselves good at talking to (or pretending to talk to) people. In trad, the rules are the game.
OSR play, on the other hand, largely exists in the gaps between the rules. When the dice come out it's to elide something that we can't just play out naturally. We can have a conversation with an NPC and let that occur without ever touching dice; for the most part we can't figure out how combat is resolved just by talking, so we turn to the dice. (Though even with combat, you'll frequently see GMs recognise that the party has turned the tide of battle in their favour well enough that the other side can't possibly win or escape, and simply narrate how it all finishes up rather than playing it out round by round. This is, largely, something that can't happen in trad because of concerns about balance - because, again, system is more important there).
I don't have a pithy conclusion here. That's the post. Anybody who claims that "system doesn't matter" is just being inflammatory. But it's worth acknowledging, I think, that the extent to which system matters is very much dependant on context.