For a few months now, I’ve been playing Quarries of Scred (Noble Kale, 2014) which causes me frequently to scream at the screen. Nowhere near as much as NaissanceE (Limasse Five, 2014) did, of course, but pretty much every time I die in the game it is because I am crushed to death by rocks. And it seems like it was my fault.
Quarries of Scred is a game that offers procedurally-generated challenge and if you die, just once, that’s it for the level. No health, no extra lives. Just you versus the environment. Will you collect enough minerals to escape – or wind up dead after one wrong step?
When we talk about games that impose permadeath or similar aggravating conditions such as the sparse checkpointing of NaissanceE, we usually reference the power of consequences and how they make us feel. But have you heard of the “Peltzman effect”?
In 1975, Sam Peltzman published a paper in the Journal of Political Economy titled “The Effects of Automobile Safety Regulation” in which he noted the expected drop in traffic-related deaths as a result of safety improvements did not materialise. Peltzman suggested that by making cars safer, drivers felt safer and this was reflected in their driving. The safety benefits of regulation could be cancelled out with the smallest increase in risky driving.
This is sometimes known as the Peltzman effect or theory of risk compensation: improving safety causes an increase in risky behaviour thus negating some or all of the benefits of those improvements. It’s not just restricted to car drivers, though. There have been renewed calls to abandon bicycle helmets for the same reason. It’s obvious that if a cyclist were involved in a collision then that cyclist would have a better chance of survival wearing a helmet than not. However, evidence is mounting that helmets are increasing the number of accidents.
It doesn’t just work one way though and it’s possible to take advantage of this behaviour. If we heighten the perception of danger, then we can encourage people to take more care in whatever they do. If we make roads seem more dangerous then it should reduce accidents. This is the idea behind “shared space” traffic engineering where the boundary between pedestrian space and the road is deliberately obscured.
I remember Amanda Lange writing a few years ago about how the “incredible threat of failure” was more important than the actual possibility of failure. Players sit up and pay more attention when they think they might die. Permadeath games and the like don’t even bother with the fake-out; if you make a mistake, boom, goodbye, thanks for playing.
Instead of reflecting on how this makes players feel, we should consider how it makes them play. The theory of risk compensation suggests that they will play better. How many times have you felt like a “lazy player” in Half-Life 2 (Valve, 2004) with quicksave at close hand? We can’t contemplate life without the quicksave, but the resulting collapse of risk threatens the athleticism of the game. It’s only when you switch to multiplayer do you start to see competition for good play: you don’t learn that stuff in single-player.
Quicksaves, regenerating health and other sorts of developer generosity are well-intentioned but often diminish the experience. These systems make no demands, so the player sleepwalks through the game. Let’s call it: some of these shooters are just walking simulators with a bit of colourful pizazz. There’s nothing to lose, so there’s nothing to play for.
This isn’t to say all games need extraordinary challenge, but how risk compensation affects player behaviour is not something that should be ignored. Tricks to simulate risk might work but it’s all too possible players will figure them out after which point they begin to relax. Suddenly, it’s a walking simulator.
See, if I could save the game in Quarries of Scred, it would be abysmally boring. But I can’t. And that makes all the difference.
Download my FREE eBook on the collapse of indie game prices an accessible and comprehensive explanation of what has happened to the market.
Sign up for the monthly Electron Dance Newsletter and follow on Twitter!
That’s two nickels in the “walking simulator” swear jar for today!
Still, a nice write up of a common problem in games. It might be worth considering the opposite problem too: overly conservative play.
I notice this especially in some of the less lethal rogue-likes. If the penalties for death are high (such as a total restart) and there exists a low-risk strategy (such as baiting enemies in single file, then running back to a fountain to heal) there is a strong incentive to pursue that strategy. Even worse, players might be encouraged to ignore any fancy interactions/combos that could allow them to play better, but would require practice, in favor of repeating the same consistent strategy. These strategies quickly become boring/low skill and can make the player think the game is boring.
Games like Nuclear Throne get around this by forcing the player to move quickly to get important item drops (ammunition, health packs and experience chips).
However, there is something to be said for games with extensive check-pointing. I was playing Shovel Knight recently. The (standard) game is reasonably liberal with check points, all the player loses is a little money (and the money is easy to farm for if necessary). However, the short walk from the checkpoints allows for more experimentation and risky play.
To make punishments like permanent death work without introducing too much frustration or tedium, I think these games must NOT have a “baseline” or “safe” strategy. You must force the players to take on some risks on a regular basis, so your game doesn’t become a case of repeating the same set of moves 100 times and don’t mess up.
Yes the usage of “walking simulator” was very deliberate 🙂
But for sure, this stuff is an absolute minefield. Permadeath can easily lead to a focus on safe play. Scred requires a little more attention as you build up cash “in the bank” but you feel free to experiment in the early stages.
Getting all this right is the meat of game design. Procedural generation and short duration obviously negotiate some of these hurdles but longer play means it gets me tricky.
I’ve experienced a strange kind of play-influence based on, of all things, game running speed.
Playing UFO:Enemy Unknown on the Amiga, a single turn could take 45 minutes, what with having to swap disks to get anything out of your backpack, or see the map. You either couldn’t save during a fight, or we didn’t know you could save during a fight.
Perversely, this slowness made you ultra-cautious. There was no way you were going to take risks that meant you had to do the whole thing all over again. Everyone was tooled up, and carefully covering every move. Fences and buildings were dynamited from afar, just to be on the safe side. This just made each fight even longer. Encounters would end up consuming an entire day or more. Move and cover. Move, and cover. Every corner was a challenge, every hedge a potential nightmare. Practically, a disaster. Thematically, perfect. Enemy: Unknown, and rightly feared.
Switch to the PC, where you needed hacks to slow the thing down to a playable speed. Save during missions? Yeah, fine. Suddenly it’s all mad dashes, and pistol armed scouts doing solo leaps into UFOs. Suddenly, it’s slapstick. A friend insisted his unit had invented time travel. They could see where the aliens would be in a few turns time (by saving, playing a couple of turns, then reloading). Aliens never picked up on that one. Must have been a lot of head scratching back in Cydonia. Anyway, different feel. No longer an enemy to be feared, but to be mocked.
CdrJameson – really interesting. You’ve cut to the heart of what risk compensation in games really means – amount of time taken. I couldn’t even play something like this now – the risk of losing so much time in a game is something I couldn’t endure. Mount & Blade led me into a nightmare cul-de-sac and I chose to cut my losses and never play again. So there’s obviously a balance depending on how much time you have to “waste”, your appetite for game risk and how much danger a developer wants to project.
So, could anyone play UFO: Enemy Unknown in the exactly same manner today? With disk swaps and the like. I’ve dabbled with old 3D games and struck by how freaking slow they felt and my instant reaction is: I have no time for that.
I’d never heard of it as the Peltzman effect; I’d rather know it as the red light effect. At every intersection, there’s a period where every light turns red. It’s not an unusual instance that only happens during maintenance or such, rather, it happens every time the lights change. After vertical signals turn red, and before horizontal ones turn green, every light in the intersection stays red for three seconds. The moment when the entire intersection comes to a halt. But, conversely, there’s no time when every light is green -unless a manufacturer is checking things- because that’s what makes the world safe. Because everyone prefers safety to danger.
When the world is filled with red lights signaling danger, the world is safer than normal. But if there was a world filled with green lights, then that would be more dangerous than any other.
So, I have to say… aren’t you thinking of this the wrong way? You’re thinking that “this is how we engineer people’s safety” and “this is how we make better games” but, shouldn’t it be “this is how we become better people”?
The world is filled with people who think that god is protecting them as they go through an intersection, but in reality, the risk has only been cut. Like that, it’s only a bit better than a world filled with green lights. Yeah… if you really don’t want to face any danger, then only walk on the sidewalk and never cross a street. But even walking the safe path, there’s no guarantee of safety. A car could swerve off the street, a snake could bite your ankle, a passing pedestrian could put a knife in your side. Even if you avoid all danger, there is still danger. Which is why someone should say this. Which is why I’m saying this.
The world is dangerous.
“The world is peaceful. It’s full of dreams and hope and salvation. People were born to work together and to love each other. Children have a right to be happy.” If you go around saying such things, is it any surprise that you’re easily caught off guard? Even if all these things are true, then counting on them without any suspicion or care brings suffering, not happiness.
Children in war zones -even without proper schooling- are more sound. At least, they’re more avaricious about life… That’s because, in their eyes, all the signals they see are red, not green.
(I totally admit to this shameless show of plagiarism)
Pretty far off topic but I wonder what numbers Peltzman was looking at (I’m not at work so I can’t access the JSTOR article right now). Looking at US motor vehicle deaths per 100 million vehicle miles traveled I see the number dropping like a rock from 5.30 in 1965 to 3.35 in 1975. As far as absolute deaths go the big news is that Americans drove a lot less in 1974 thanks to the oil embargo but that doesn’t tell us anything about safety.
It’s a great metaphor for what you’re talking about though; I wonder if the original example is a bit of a boiling-the-frog-slowly bubbameister.
Sandy hit on exactly what I was thinking of with relation to permadeath — the first thing I think of when I think of permadeath is nethack, where a lot of the big machers who inhabited the rec.games.roguelike.nethack group were obsessed with a style of play that meant they could win almost every time. (Mostly it involved writing Elbereth which prevents most monsters from attacking you, often letting your pet fight for you.) And it seemed utterly joyless. It wasn’t even a question of loss of time to permadeath; many of them had a quasi-moral insistence that the object of the game was to win, and so refusing to use this kind of strategy was wrong or at least stupid (unless you were trying for a formal conduct).
I guess you folks all know about the idea of punishment for failure–I think I saw Edmund McMillen talking about it in Super Meat Boy, but VVVVVV is a good example–in VVVVVV you’re going to die a lot but it always resets you instantly at the last checkpoint which is almost always right next to where you were. And it’s not just the amount of time; usually there’s pretty much one checkpoint per challenge, so failing at one challenge doesn’t make you redo a previous one.
What I find the most punishing is usually not redoing a challenge that’s actually hard — I might feel that I lucked my way through it the first time and should learn to do better — but redoing a challenge that I’ve pretty much mastered but that remains fiddly. GIRP is huge on that, when I was playing it I eventually didn’t have to worry about the first half or so if I was paying attention, but I needed to pay that attention, which made it especially vicious that the endgame had That One Thing that I only got to practice once after going through the rest of it.
This can be used to enforce a cautious playstyle — the Games That Exist description of La Mulana made it sound like that was what was happening. (That site seems to be gone, shame.) But enforcing caution isn’t always what we want, which is what Sandy said of course. Maybe GIRP is trying to tell me the opposite; I can be cautious at the beginning but I’ll have to let it go eventually so I might as well practice at the beginning when the time loss isn’t so great. Or maybe, as a lot of other suggests, Bennett Foddy just doesn’t like my face.
@mwm – I dunno. The Peltzman effect is more about the limits of improving safety, that the human mind will appropriate that knowledge and basically undo its positive benefits. So you have to be sneaky if you want to maximize the benefits of, say, a seat belt.
@matt – I’m not sure; I haven’t looked closely at the article as it seems to be a given forty years on. I heard about risk compensation in the late 80s I think and it was more about the seat belts not delivering on the their expected promise.
I guess I’ve never analysed my own response to punishing games like this. I’ve always been “well, this works for me” and “this doesn’t work”. NaissanceE pretty much pisses me off, but I think the game works better with that frustration. The punishment usually has to feel “just” in some way. I’m not sure what to make of variable difficulty modes, because although you can look at it as a choice of masochism, there’s also the shiny badge associated with beating something on HARD.
Oh wow, Games That Exist is gone. Alex Pieschel is not, though. He’s one of the editors of The Arcade Review.
Matt, I just quizzed Alex over email and he told me the old site is still available for reading but lives under a different domain name.
Thanks for the tip!
I went back to GIRP and boy my chops had decayed, though I eventually got back to the part where you drop down which is basically what I was thinking of by something I could do but was fiddly. Anyway. I’m not sure what I think the exact interaction of procedural generation and punishing failure is. With procedurally generated content, if you screw up you’re never getting that game back, but with preset content if you screw up you have to. do. the. same. bit. over. again. and. again. Well, you can see my bias there a bit.
Anyway, I guess there are a lot of effects this can have. GIRP seems like it’s encouraging caution at the beginning, but eventually you have to throw caution to the sea, or hopefully not. And I think there’s definitely an air of sadism or deliberately unfair punishment there, given that it’s explicitly a goal of the game to cause the player physical pain. Nethack too as I mentioned, or Spelunky I guess if we’re going to stick to platformers (I’ve never got far into Spelunky). But then there’s Probability 0 on the procedurally-generated side and Bit.Trip Runner on the predesigned side, where the forced scrolling means you can’t play cautiously at all, and the play sessions are short enough that you don’t lose that much by messing up (and Prob0 gives you several hit points). I guess they’re telling you “You’re going to learn how to do this.”
Oh hey tangentially related enough that I feel OK with putting it here instead of in the forum: I just had the greatest idea for a driving simulator. The Wages of Fear: The Game. I don’t know if you’ve seen the movie but it’s about driving trucks full of nitroglycerine across mountain dirt roads in order to explode out oil well fires.
The way it works is, you have a certain time limit to get the nitro to the oil well — but if your truck jostles more than a certain amount, boom, and the roads are very bumpy. And (here’s the tangential relevance) there are no saves except maybe for save-and-resumes. Because the whole point is that you aren’t trying to motor your way throw, blowing up five times before you find just the right amount of pressure to get over that rock. You have to crawl. The movie was tense as hell, and I don’t see why the game shouldn’t be.
Think of it as like Desert Bus but good.
Aw, it looks like some commenters on giantbomb already thought of The Wages of Fear: The Game.
Proc gen + permadeath => not repeating same bit again. Like you were saying with GIRP, I luuurve VVVVVV which is all about repeating and repeating and repeating. Dark Souls, too. But there’s a sense of strange loss when you fail a proc gen game in that the environment you were playing has been lost. You failed it. No second chances. I don’t think you get that feeling from Probability 0, though: it feels eminently throwaway as the depths feel generic and unmemorable. This isn’t a vote against a game, just pointing out it doesn’t evoke that same feeling.
There are lots of components flying around here. And seeing what happens when people figure out the psychological hooks that make people pay for F2P makes me wary of actually trying to figure out how punishing games actually work =)
Sorry to hear your latest game design that you will never make has actually been made!
But unlike GIRP and those other high-punishment games VVVVVV is about repeating the parts you haven’t failed at yet — once you get past something then you don’t have to do it again. Even in the Gravitron I was able to do it relatively quickly once I realized I didn’t have to stay alive a minute, I had to stay alive for twelve separate five-second segments. I’d classify VVVVVV as having low punishment for failure but very very high failure rates.
Good insight about the sense of loss and Prob 0. I wonder what else has this. Do people feel a sense of loss like that in Binding of Isaac? Are there short proc gen games that generate it?