Nick

i guess my question stems from the desire to see more games implement a "fail-forward" philosophy, like Pyre, or to lesser success, L.A. Noire. Dark Souls can reset you to the nearest bonfire, but in Dishonored or Prey, i generally load a game if die or fail a quest or, at best, hope the designers implemented a specific outcome for failure. because in most games, failing means getting punished by seeing less "content", so the incentives are generally cut and dry.

Savescumming is not a problem. Folks want to play the game they want to play regardless of devs' intentions. It's better as a designer to allow players to be self expressive in that way. It's why we should be putting in "cheats" and other accessibility options.

I feel that players who are like "savescumming lessens your experience" are just being silly. Games are not achievements of any sort of note in general. Any community that builds up around playing games as achievements (not to be confused with "Achievements" that were introduced with the Xbox) will come up with a set of rules that may or may not be supported in the game that a player must follow. Like speedrunning, score chasing, etc. But unless the game is specifically about that it shouldn't be built around that.

Sure devs can put in things to motivate players to play the game a certain way (ie Achievements,) but I think folks gotta let go of the idea that playing a game in one way is somehow lessening the experience for themselves.

I don't have the intuition to know the answer to this but I'd be curious to see the results of an A/B test where one version of a game has significantly longer quick load times.

This is just a hunch, but artificially increasing the time it takes to quick load—or even just the load time in general—could help mitigate save scumming for some players. I don't think it would solve everything, but I know that for certain games with lengthy load times I'm more inclined to play conservatively just so I don't have to see a minute long load screen if I die.

Being unsuccessful while remaining in game is (hopefully) better than being unrealistically successful while stuck in loading screens.

Rather than a binary either-you-have-save-scumming-or-you-dont, by introducing artificial friction, you might be able to convince more players to value their actions and save scum less frequently.

one small thing as a designer is that you can spell out some of this, even just as a loading screen tip, saying "sometimes allowing a failure to play out will be more interesting and cool in the longterm", just in text. It's kinda like campaign books for a tabletop RPG, suggesting different ways for the DM to handle things but letting them choose which one applies. Putting it in text might let them make better choices here, since it shows that the designer at least tried to make it cool when you fail sometimes, which definitely isn't something most players would assume otherwise!

I feel like the conjunction of achievements and quicksave/quickload existing motivate player behaviours that accelerate every player's greatest skill - the ability to optimize the fun out of a game. I specifically recall that, all the way back in 2007, I destroyed my own experience of Half-Life 2 episode 2's ending by trying to get the achievement "Neighborhood Watch", which requires near-perfect play, which I was only able to achieve though 'savescumming'. My knowledge of this objective and the ability to "groundhog day" the same handful of seconds repeatedly let me handle this task and completely broke the illusion of the strider assault, as I saw them reduced to puppets triggered by player proximity, rather than experience it as the thrilling nigh-impossible assault it was meant to be.

Personally, I've never been a fan of quicksave/quickload. I've always prefered suspend functionality to address point A (The convenience of being able to leave the game at any point without losing progress), where you can save the game at any time but resuming that save deletes it, and failing afterwards returns you to a conventionally placed checkpoint, last autosave or manual save. Point B feels like a matter of level and systems design - if the game doesn't let you explore options from a given save point, then something feels flawed. As for point C... I have no argument against point C, that's what savestates and quicksave/quickload are truly great for and I have used them and appreciated them in that capacity and that capacity alone. I think they're fantastic for return trips to games you've played before, especially.

I think what my point boils down to is that I wish that quicksave and quickload were unbound by default, and that the purpose they serve is better suited to suspend functionality and game design, and should be a much more clear opt-in system.

fire emblem has had this feature for the last few installments in the series, but of course it's different to rewind discrete turns in a strategy game vs rewinding actual seconds of time in other genres

I -love- when games give you built-in acknowledged redos that they're clearly designed around