Observations on the level design of SMB.

[Sejez]

Now I know that a vast majority of you would praise this game to high heavens, and looking from the outside it can look like this game is the end all and be all platformer of all time, and even though I used to hate this game, I am quite indifferent to the whole thing now, the thing that used to bother me was that I usually love platformers, and I enjoy hard platformers very much, but couldn't for the life of me enjoy this game one bit.

My initial impression with this was quite horrible, when I had played long enough to make it to the normal ending, I kind of hated the game, and now that I have a few of the harder achievements unlocked, I think I am starting to have some grasp of why that is the case.

The game plays or should I say, teaches you to play in a completely different manner than most other platformers, instead of a slowly throwing you into more treacherous environments, it pretty much starts with, well here you have a few tutorial levels... got the controls? Good! And then it throws you of a cliff, and then you fail until you get it right, which in a game like this, where my character is a blob with physics based controls, I find it strange that it would insist on teaching you to play in this manner, a game where character movement is this finicky. Finicky in the sense that it takes an assload of time to get used to the characters movement, which leads to the subject on how this game teaches you to master its unwieldy mechanics.

I am of course talking about the bite sized challenges and infinite lives system, which has recieved much praise as being the correct way to make platformers, instead of the old system, which had a lives system and usually pretty long stages, and depending on the game, when you lost those lives, you were either kicked to the title screen or you could choose to continue from the beginning of the stage, losing all progress from those checkpoints up until that point.

The problem being that up until Ch 6: The End, you could pretty much rely mostly on muscle memory and luck to plow through the troublesome levels if you wanted, the reason being that levels were too short for any need to have any sort of consistency on your performance, and since you had an infinite supply of lives, a careless spray and pray method would eventually get you throught all of the levels, and the longer levels were considerably easier that the rest, so that you wouldn't suddenly get stuck in one place with no idea why that one level is so damn hard.

Which is evident when going back to replaying the levels again, getting the "Dead Boy" achievement is essentially a non issue compared to the "Brimstone Boy" because the levels on Ch: 6 are a complete joke difficulty wise compared to the ones in the earlier chapters.

Still I struggled like a drowning person trying to get to the surface when trying to beat 6-5 Omega on my first playthrough, which is strange because the level is much much easier compared a lot of the levels that came before, and I suspect the reason is that up until that point the game does not really force you to learn the mechanics properly, so you can pretty much half ♥♥♥ up until this point, and suddenly struggle on the last chapter, because suddenly the game eschews what has been its design foundation up until this chapter, and in my opinion while these levels are very uninteresting designwise, the game starts to becomes a lot more interesting to play, because now you have to learn to play, half assing jumps is no longer an option, and relying on luck will get you nowhere, the bad thing is that you are on the last stretch for the normal ending, and the it's first time the game actually asks the player to play properly.

This is the funny thing about relying on luck, it can make it seem that you have all the control in the world for a short while, but it has an tendency to run out really quickly, you can rely on luck if you are essentially using quicksaves throught a game, but try to marathon any game and you will see that luck is really something that only comes in short bursts.

So in the end my point is that I think the game would actually have benefitted from longer stages with a more gentle difficulty curve, not saying you shouldn't have ♥♥♥♥♥♥♥ difficult stages, I'm just not very thrilled about this "bite sized challenge" thing at all.

[Spike]

I think I understand your point of view, but it might as well be the players 'fault'. See, I never relied on luck, and got through the whole game without sudden spikes in difficulty. I found the difficulty curve very smooth, and I enjoyed gaining the skills required to finish all the levels. But I understand it could be a problem if you didn't replay levels that you got through by pure luck. As I said, that's a playstyle choice, and none that I would ever take.

The bite sized challenges allowed me to repeat things over and over again untill I could do it perfectly, so that's a pretty neat design decision imho. And by adding in the warp zones, I could still have a bit of that old-school feel, where the stages were longer and lives mattered.

It's a new kind of platformer, so I suppose some die-hard old-school fans might not be too thrilled about it. Oh well, to each their own.

[Sejez]

The system used isn't always a bad thing in my opinion, but it's a bit too lenient at letting players pass challenges that they necessarily don't understand well enough, say just as an example the levels at Ch: 5 usually have 3-4 very tough jumps for a beginner, now let's pretend it's my first playthrough and after a while I develop enough skill to where I have maybe about 60% chance to pass such a jump, which means that I have about 12% chance of beating the level, even when my understanding of the mechanics are rather poor.

In old school games the jumps are usually a lot easier, but on the flipside you better understand them very well because the game will throw a ton of them at you before letting you pass the level, let's say 10 or so jumps, with the same probability of 60% of success per jump you will discover that you have a whopping 0.6% chance of succeeding to pull of the same feat, needless to say, infinite lives or not, it won't work as a strategy to just try to rush trough it no matter how stubborn you are.

So while it might be a decent tool to train players, it is a terrible way to check if the player has become skilled enough to progress, add to the fact that while most of the levels in the game play nice and fair as long as you take the time and approach them carefully, there are at times where the game just throws a few kaizo style traps at you, and you die because the level is designed in a way that even if you are fairly observant you will die, because the normal modus operandi in this game pretty easily becomes to rush headlong, because it does not matter since your lives won't run out anyway, and it just works most of the time, instead of taking a more careful approach, and often you just keep hammering on the same problem until muscle memory kicks in and you just win even though you are none the wiser after the feat.

Which is a big reason why I was pretty miffed when fighting the end boss on Ch:5, my handling of SMB was still so bad that I had great difficulty hitting the moving platforms that are supposed to aid you to be able to dodge no matter what angle the boss comes from or what attack is being used. Add to the fact that the boss has a totally unfair hitbox which is massive compared to the size of the boss itself when it sticks it's head out from the ground, you can probably understand that I am feeling that the whole thing is unfairly rigged, and coming from the 4th boss that was basically like playing a simon says game, where you don't know the pattern beforehand, the game was really starting to grate on my nerves at that point.

Which brings me to the point about being the players job "not to cheat", is a lot to ask in some instances, a good example would be me playing Street Fighter, and I discover that the A.I. is dumb and slow enough to fall for a cheap move time and time again, now I could fight fair, but it is of my humble opinion that you should use any dirty tactic available that isn't cheating/glitch abuse, and leave the balancing and making the game challenging to the developers, and not the player. Let's say in this example I play that way to the final boss, and this guy is programmed in a way where he has perfect reaction time, and my cheap trick won't work anymore, I suddenly have to learn to play against a unfairly hard opponent because of a A.I. flaw, and then by sheer luck I then find another flaw in the A.I. that I milk to victory, it makes the whole ordeal feel totally pointless, and playing fair adds another irritating factor of limiting yourself to not using certain moves, which can be really annoying when you are simply trying to win, expecting a player to tie one hand behind his back isn't exactly what a developer should do in my opinion.

Well in the end I just feel that getting any achievement that does not end in "boy" is as fulfilling as a kindergarden diploma, I felt that in the end the game didn't give a ♥♥♥♥ if I got better at it, so neither did I.

So I guess it is a bit strange of me to finally earn the "Seneca Falls" achievement and just feeling, meh whatever.

[Starlit]

I'm with meaty boy on this. I had a fairly steady learning experience in the game and I made it unnecessarily hard on myself by doing 100% of each chapter before moving on. I will say that i found the step-up to rapture hard and actually gave up for 2 months at that time, (being dissapointed for not being able to 100% didnt help :D) I would never leave a level till I A+ it. This had an obvious downside but it usually meant that I understood stuff before moving on.

I would like to know what you mean by the game throws you off a cliff, I mean rapture is the first big step up or if you go light dark light dark, dark salt is the first big step. And even then its given you probably a few hours practice and there's nothing crazy, just longer levels that require more precision. Still I cant argue that you've obviously thought about this so I just wanted to say this.

If you aren't enjoying the game dont force it. Play a different game for a while, try something else within the game etc. I'm guessing thats why you have some of the boy achievements but doing the boy achievements never really seemed fun to me. Try out the other characters, try time attack on individual levels or just take a break.

I'm pretty sure most of the mechanics pop-up before 6-5 but if you didnt pick them up along the way its a hard and painful way to learn for sure, however, if you can do seneca falls you have done as hard as anything in the game. Anyone with the control to A+ 7-20 can 106% the game imo.

[Sejez]

The thing I mean is that, it's like getting a brand new car that controls rather unlike anything you have driven before, Meat Boy's movement are physics based, which isn't exactly easy to wrap your head around in the beginning. So up until somewhere around after I had beaten the normal game, I really had been struggling quite a bit with the controls, and to tell the truth it is a bit infuriating, when everyone is praising how fantastic they are, which to be fair, they are accurate, but they aren't easy getting used to.

And the game does seem to have a hurry to get you to do all the cool stuff so you feel like a badass, but more often than not, instead of having challenges build upon earlier ones, the game often threw novel challenges at me, when I was still trying to figure out basic stuff, which lead me to become rather irritated.

Something that becomes very evident when you compare the levels in Ch:6, to earlier ones. Is that they are longer than the stages in Ch:5, but as far as game mechanics go, they are even simpler than the stages in Ch:1, and are fully built around the fact that you are supposed to get a feel for how your character moves around, which for me seems to be a really strange stage theme for what is supposed to be the last world.

In fact there is so little going on those stages that I managed to get the "Dead Boy" achievement with less than 9 deaths, when my intention was to only check the stages, to compare them to the earlier ones, and I managed to die alot more on a few of the Ch:4 stages, because quite frankly, a few of the traps are positioned in a bit douchey manner, if you go back and study those chapters you will see that the difficulty takes a nosedive in Ch:6 if you are comparing the levels to the ones in Ch:5, or even Ch:4.

So the fact that the stage focus of the last world is to get a basic grasp on how to control your character, seems like an erratic design choice if you ask me, and it makes me feel like the game would have benefitted from a little longer stages earlier in the game.

But unless I or someone else can prove my point by creating a level pack that effectively prepares a person quicker for Ch:6, than the main game does and also supports my point of view, I guess it's only my word against yours, so unless I put my "design philosophy" to a practial test, there really isn't anything more that can be said.

[Starlit]

OK now this I completely get, I think in some ways doing light dark light dark, though harder, has no sudden burst in this way. However you may have missed the meat boy "mechanics crash course" which I think Team Meat expected people to do before chapter 6 :D There is one warp where the mechanics are "spelled" out like nowhere else

About to reveal the location of a warp, if you really dont want to see dont reveal.

Have you bandaged the warp off 3-5? It's basically designed to teach you that wall jumps are longer and how to s jump.

That warp aside, the end doesnt strike me as much worse than a+ on rapture so my guess is that you went for the complete first and go back later approach? I could be wrong but I feel like a+ on hell teaches most of what you need for the end with the exception of the first thing mentioned in the spoiler but most people do tend to go for just pass it first approach and devs should expect this

[Sejez]

Hmm tried the level, and well I didn't exactly mean that the game doesn't teach you to use the moves, it just does not throw enough variations of them to make sure it sticks, that is just one of the things about how humans minds work, you have to repeat things several times if you want something to stick, so developers are kind of supposed to constantly make sure they repeat things, and making sure it doesn't feel repetitive.

Something I thought about that would be a weird experiment, if someone spliced together the existing levels, so each world only had about 12 to 7 rather long levels instead of the normal 20, and left the last chapter as it is, I wonder how it would affect the end result, probably not a completely wise idea, considering the levels have been built with the short length in focus, but I'd still like to see how it would affect player skill, completion time and death count, compared to the current model, given that it would be run on a large enough sample group that hadn't played the game before, and the people were willing to subject themselves to play till they completed the normal course of the game or till they had accumulated a certain number of hours of playtime.

[Volnaiskra]

I'm not really getting what you're saying at all. You say that chapters 1-5 let you get by on "muscle memory and luck". I don't understand why you lump these things together, as they are virtually opposites of each other.

The only way to achieve muscle memory in anything is to do it over and over, methodically honing your actions until they are near-perfect. This involves a level of mastery where you have executed a series of actions so many times that they literally shift into a different part of the brain. To say that this doesn't involve "consistency on your performance" just makes no sense.

It's also completely different from luck, which implies a haphazard approach devoid of learning or purposeful repetition. I'm amazed that anyone could get through any of SMB using luck. I personally found pretty much the whole game to be an intensely (and frustratingly) purposeful process where the slightest lapse of concentration meant failure.

I also don't understand what you mean by SMB having "physics-based controls". SMB's controls aren't any more physics-based than just about any other platformers - unless we're really talking about the most primitive Atari 2600 games.

Are you referring to the inertia that makes SMB slide around a bit even after you've let go of the controls? Almost all platformers have that (A platformer without it would feel odd to most players). It's a little bit more pronounced in SMB, making you kinda feel like you're ice skating. I personally find this a bit of a cheap method to make the game more difficult actually.

But it's the same inertia as in a thousand other games, with a slightly different value for one single variable (I know this from experience because I've built a platforming engine for my own game). This single variable is then multiplied by some number whenever you're in mid air, so that your ability to slow down and speed up is hampered in the air, compared to when on the ground.

The only other physics I can think of are gravity (after you jump, you progressively slow down, until you stop for a moment, and then you fall progressively faster until you either hit a capped top speed, or the ground.). Again, all platformers have this, unless they're incredibly primitive.

This gravity is slightly altered in certain contexts, but it's still basically the same bread and butter. For example, when you hit a wall, the gravity immediately becomes much lower (this is probably done by multiplying a single variable whenever contact with a wall is detected). Similarly, when you're near a fan, the gravity is offset and/or is pointed into a different direction.

There really isn't much about SMB's mechanics themselves that are unique. They're well polished, and they're particularly fast, but that's pretty much it. The rest comes down to the level design.

Also, it's pretty clear that you like the game much more than you admit. Because only an insane person would spend 41 hours on a game that they "couldn't enjoy one bit". So you may as well come clean ;)

[Starlit]

@Volnaiskra, I quite like the momentum balance in smb, it makes little sense to be plowing at a full tilt sprint in one direction and then instantly turn like so many other games though ironically meat boy does allow this which most people dont realise, you slide for a bit if you try to pull back but holding nothing while grounded performs an instant stop, done with correct time, you can fall and instantly grab an edge you just walked over.

Anyway that point aside I personally feel the end is long boring and adds nothing fun to the game but thats just me, the levels while easy except the 6-5s and dark escape, all feel fairly bland but this is just my opinion.

Now what I notice is alot of people who dont go for a+ times or bandages feel a massive step up in difficulty at the end. the player who goes for A+ in rapture usually doesnt feel at all but I don't think you can blame the devs for your meeting a wall after taking the easy way out all through the game and not honing any of the more difficult skills.

One of the things that amazes me in this game is that most people think they should pass light and then dark and then go a+ and then bandage in that order. In hindsight it should probably go something like light forest -> light salt -> dark forest and then alternate light and dark with getting bandages and A+ when they seem achievable and then occasionally recapping but people who have only fluked a pass after 100 or so seconds, (I've actually seen higher) on the fans in hospital which are very forgiving, come onto 7-7 in cotton alley which is the next time you see fans if you go about this way and BAMMM!!! they realise they have no clue how to control them. I've seen this happen several times on twitch streams.

As for the muscle memory argument, this is a little off I feel, you use muscle memory on say 7-19x even just to pass it you need to perform the same timing 4-5 times at the end. You dont have to be able to do the level deathless for some degree of muscle memory to have taken affect, all you've done is increased you chances of "getting lucky" so that pattern might now be a 1 in 10 instead of 1 in 100 chance of coming off.

[Sejez]

Yeah it's true that is an big overstatement the "not enjoying it one bit" thing, while I felt a lot of things I still think enjoyment is quite low on that list, I felt quite a range of emotions when playing the game, perplexed, baffled, varying levels of irritation, dumbfounded so actually mostly different variations of confusion really, while I can't say that you can't enjoy being confused, it was at least never a clear cut emotion of feeling enjoyment, but still I felt something so I guess that counts for something, the strongest feeling I have had of joy in this game was when I was getting the "dead boy" achievement, I messed up a jump badly one of the levels, but I somehow managed to recover, that was the one time I actually felt good about myself during game, and I think it might be because it made me snap out of the trance like state I usually am in when playing the game to make a attempt at turning a seemingly certain loss into a victory so to speak.

The thing I trying to convey when I talk about the game being physics based, is that really feels like the game uses Box 2D physics, which most platformers don't use, maybe it's more poupular on modern 2D platformers idk really, I have toyed around with Box 2D on a few of my own projects, and it feels similar. Most platfomers I have played use a much cruder simulation of physics if you can even call it that. In this game the controls are really sensitive, and loose in the sense that accelration is a bit slow, it's accurate but it feels fiddly.

And while you can say muscle memory and luck are at odds, it so happens that often you mess up even when you are playing the level as if it was a song on the piano, and while it usually it means you lose, it occasionally leads to a positive result instead, and sometimes you just get really lucky, it's a weird kind of feeling when you manage to pass a level while you know that you weren't quite in control half of the time, a few lucky guesses, you clearly mess up and manage to dodge things regardless, or in some cases like with the rockets and the maw things, you don't know well enough what is going on, there are a faint idea of what you should be doing but everything is a guesstimate at best, or a blind guess at worst, but somehow you still manage to survive the onslaught, get lucky with a gravity orb and touch the goalpost (bandage girl in this case), and you are like "well that just happened, oh well I won anyway" without any real sense of accomplishment.

To be honest I have mostly played trough the game with curiousity as the main driving force to get through the whole ordeal, I'm not saying that it was a complete waste of time, but it was driving me up the wall that I felt that the game always felt weaker than the sum of it parts, and I really want to know why that is the case. So in that sense this game really is/was a novel case, it's not really unusual for me to play bad games at length just because I'm curious for one reason or another, not saying that this game is bad, my gut feeling is heavily leaning towards average at the moment. There is things I think it does right, and there are things I feel it does wrong. But I suspect that it really is the short levels that really pull the experience down for me, that and the fact that I think that the 4th and 5th bosses really are terrible bosses, while the others are fine in my opinion.

[Starlit]

The bosses are pretty terrible, and 4 is a travesty. there's a story there which is basically that Microsoft over pressured them on the release date and kept pushing things forward. Watch indie game the movie and you'll get an idea of why the game has as many bugs as it does.

I think the best example for most players of the luck you describe is 7-16. The one with a single gravity ball that you ride across, for most people this is the most annoying, luck based performance in the game.Even now I dunno if i can even reasonably reliably do that level though I did some practice and improved enormously so maybe, but it's probably the only level along with its dark counterpart that gets this honor.

The levels that most frustrated me were
1.7-19x They never teach you that not running can be good :D but really I had the bug that makes the blocks close so fast its barely possible, I've fixed this since Millay's run brought it to my attention and now this level ain't so bad.
2. skyscraper (easier than kid warp but 3 lives and long respawns makes it worse, especially since I refused to slow down until my will was eventually broken, its one of the only times in the game where you really 100% have to slow down first playthrough and I think literally the only time I left go of sprint on my first playthrough.)
3. Kid warp
4. 5-8x (nothing like waiting 17-20 secs for your shot at dying, which I did an unfortunate amount on my first run.

5. Anything with an optimal time like design, 4-2, 3-12x, 3-13x, 3-13, 5-8 etc. Levels where you have to wait are annoying, this game usually got this right.

On another note I do dislike the design of one or two bandages/ warps. 2-18, and 3-16 come to mind, bandages that require you to wait in order to spot or that you can see but cant tell how to get because the game has a hidden passage thats actually wall. One of the most attractive things in the game for me is that everything is visible, theres no move a screen over and its a wall of spikes like in fangames. or work a long time for a checkpoint that was actually only on the top option of the fork, so that you go too far.

[Sejez]

Originally posted by Sentry:

The bosses are pretty terrible, and 4 is a travesty. there's a story there which is basically that Microsoft over pressured them on the release date and kept pushing things forward. Watch indie game the movie and you'll get an idea of why the game has as many bugs as it does.

Not really about the game directly, but that really sounds like it goes totally against the concept of "indie spirit", the whole situation pretty much reeks of irony, the whole quality of the game being compromised because you have Microsoft breathing down your neck thing, does not exactly scream independent if you ask me.

Not that there is anything wrong with it, just seems like indies are being cut a lot more slack than larger publishers/developers, when in many situations they are pretty alike.

[Volnaiskra]

I see what you guys are saying about some levels being so tough to master, like the ones with numerous fans. I guess I remember a few where I finished without really feeling like I knew what I was doing.

But I think these levels are in the minority, and overall I feel like the game *wants* you to master its controls. After all, SMB leaves red gloop everywhere (so you can better retrace your steps), and the way it shows you all your attempts at the end is almost like the way a golfer watches videos of himself to improve his swing.

I just can't stomach the idea of longer levels. I seemed to die so many times on most of the levels that it didn't show them all in the replay, because I'd gone over some capped number of deaths. Longer levels would just multiply the number of deaths. Unless of course the levels were made a bit easier, but then it wouldn't have that raw intensity.

I don't know anything about Box Physics. I use Clickteam Fusion (like I see you do, Sejez). But when making my own game Spryke, I modelled the mechanics on SMB (using a custom-made engine in Fusion - no Physics extension). I was able to replicate all of SMB's mechanics no problem (though I adapted it to fit with Spryke's own look and feel of course, plus added a few new mechanics).

I also looked at various other platformers for reference, such as Rayman Legends or Thomas was Alone. It's been a while since I did that research,but from memory, most share the same basic features: acceleration, gravity, inertia, variable jump height, etc.

They just all have different levels of those attributes, and SMB is a bit exaggerated in some areas (eg. in acceleration, as you've noted). The acceleration and inertia in other games tend to be more subtle.

[Sejez]

I guess the experience of intensity is completely different for me, I really didn't feel any most of the time when playing Super Meat Boy, and I think it all boils down to the concept of loss, I won't lose anything so I don't care, and it just feels that it is just a matter of time until I beat any level, which I have for the most part done (don't feel too compelled to go bandage hunting at the moment, since the unlockables are pretty uninteresting to me), I think 6-5 came the closest to make me rage on my first playtrough, and it's among the longest levels too, so go figure.

It's completely different when you are playing an old arcade game, say Metal Slug 3 for example on a single credit (credit feeding through the game will destroy any tension obviously), and you are on your last life, when you manage to get to the final boss, you are close to victory but you have everything to lose (which in this case is the progress up until that point), you feel the need to perform and give 110% of yourself because you don't want to start over again, that is what feels intense. This game feels more like something you do to pass time, like Solitaire, but again that's only my opinion. :)

[Sejez]

I have been watching some gameplay videos from older platformers, and boy does it make me appriciate the level design in this game a lot more, sure it might not measure up to the old classics, it might not even be what I consider to be good, but most of the titles I would have considered mediocre as a kid, probably would get a failing grade nowadays from me, and there are people that love those game too, meh, guess it shows how cynical I have grown over time, this game might not even have what I consider good level design, but if you remove the handful of companies that could make games properly back then, almost everything of what you have left really is worse, there is no denying that.

I guess there's a good reason companies like Sega, Capcom, Konami and the rest were so revered back then, and why I can still to this day go and play and enjoy their games, but I struggle to remember 90% of the other crap I played trough as a kid.

[Volnaiskra]

Yeah, it's definitely a different type of difficulty compared to the games we played as kids. Like you say, there's less riding on it if you know you won't have your entire progress wiped when you fail (though I guess there are the Dead Boy etc. achievements for that).

Design-wise though, I actually think many of those old games were kind of 'impure'. The 3-lives-and-then-you-restart model wasn't a game design decision, it was a business model decision: physical arcade boxes needed players to spend money, die fast, and then move on so that the next person in the queue could do the same. So games were designed around that goal.

That model of gameplay carried on far beyond arcade machines of course - into consoles, Amiga and so on. Even Doom and Doom 2 on PC had a lives system. That's what people grew up on, so for a long time, no one knew any different.

But it can't be stressed enough that this system was far from ideal. It wasn't born out of a desire to make great games, but out of a desire to squeeze as much money as possible out of customers. The same thing is happening today with free-to-play mobile games. Candy Crush is an incredible money-making machine, but anyone with a bit of objectivity can see that its gameplay has been compromised as a result.

I had never heard of Metal Slug when I was a kid - I only saw it recently when some friends played it on an arcade emulator. They reminisced breathlessly about how amazing the game was when they were kids, but I couldn't see the big deal. It didn't seem very exciting to me, and they completed the whole thing in like 20 minutes.

Then I realised that they had played with unlimited lives on. Without constant restarts, it became apparent how little content there really was, and, for me at least, the gameplay failed to excite on its own, when it didn't have the looming threat of "game over" to amplify it.

The same can be said for something like Ghosts and Goblins, which was notoriously difficult. But the reason it was difficult was because enemies spawned right next to you, and the controls were dour: you were slow, and had the jump height of an ant. Without the threat of death, it wouldn't be a fun game to play. In other words, it was all stick and no carrot.

It was inevitable that game designers would eventually evolve their thinking past the coin-op model of the 70s. Which is where games like SMB come in.

There are various advantages to a game without "game overs". For one, players don't need to play as cautiously. They can experiment, get wacky, or try crazy daredevil antics. It can be all about the the gameplay now, and not just about the coin.

Ironically, games can also be harder now. I can die 200 times in SMB in one sitting. I never would have tolerated that level of difficulty in Double Dragon because I would have been out of coins after 10 deaths, and would have gone home, too frustrated to return.