referees have the power to judge player actions, that is their job
You can call it enforcing the rules, but there is some leeway there
How does this engage with the content of
@soviet 's question in any way? This seems a complete non sequitur.
The schematic of soviet's comment is as follows. The claim:
Systems that principally and rules-wise constrain GMing is "oppressive."
Ok. What is presumed to follow (otherwise what is the point of the claim, particularly with respect to the thread-at-large)?
Systems that principally and rules-wise constrain GMs curtail GM agency via oppression.
Ok. Don't agree in the slightest because this requires both (a) a significant drill-down of a textual analysis of a game in question and (b) actual experience to backstop that textual analysis and the claim being made as a result (neither of which is available here). But lets go with it (as soviet is). So if, in fact, oppression curtails GM agency, how is it that
GM having veto authority over all actions doesn't likewise curtail player agency via oppression?
Its an interesting question and "its their (GM's) job" isn't engagement with that question because the only thing that follows from that (which engages soviet's question) would be "its their (GM's) job <to curtail player agency via oppression>." Oppression as a result of
constraining the stuff a participant can do which therefore constrains their ability to shape play is doing different work in both situations, but at the level of evaluation of "agency" that is immaterial.
As an aside, I've spent the majority of my GMing life running Pawn Stance dungeoncrawls and hexcrawls where the creativity is mostly before play in the enormous prep load and assimilation of that prep (though some of it is responding to interesting NPC Reaction Rolls and Wandering Monster "hits" in interesting situations). I've also GMed a metric f-ton of Narrativist games. The in-situ creativity and cognitive load required to run Narrativist games is titanically more than that of in-situ creativity in dungeoncrawls and hexcrawls (and Metaplot-heavy, Setting Tourism FR games of which I ran one for 7 years). Its not even comparable. The former frontloads the overwhelming amount of cognitive load during prep while the latter (Narrativist games) burdens 90 + % of the cognitive load upon GMs (a) in-situ and (b) that load is a direct byproduct of the systemitized constraint placed on GMs and their specific duties in terms of game-coherent situation framing and consequence handling.
So the claim is wrong in every way it can be wrong. The cognitive load in GMing those games (during the actual play) is just profoundly greater (because the other type of game frontloads the cognitive load during prep and assimilation of material to see play) and its married to the constraint. So if someone were to make the claim "GMing Narrativist games feels oppressive and overwhelming because the cognitive load during play is just so damn much...", ok...sure. Absolutely. That is an empirical claim that is trivially testable and trivially true. Likewise, if someone were to make the claim "GMing games with heavy prep models (metaplot, setting, dungeon and/or hex mapping/keying/stocking with table population et al) feels oppressive and overwhelming because the pre-play cognitive load of coming up with that content and then the material load of assimilating and executing it is just so damn much...", again...ok...sure.