You know, I have noticed a new genera of games emerging within indie circles that we should all be interested in, because it stems very much from roguelikes. Not sure what we would call these, but they seem to have several aspects to deliver on a core experience.
These are games like Spelunky, The Binding of Issac and Faster Than Light that some people attribute as roguelikes, but when you analyse their mechanics you find are some what different in some reguards. Core features include
Loosing is a core feature: In many games today, you are never going to die and start from the very beginning again. In these games this is not only important but beneficial to the player. Permadeath is required for this game experience. More on why later.
Randomised content: Randomised content is one of the most important features, because it increases replay value. These games are designed to be played through over and over, thus randomising keeps things interesting.
Item upgrades: These exist purely to make life hard for the player. Without upgrades you would have to balance the game for the players starting abilities, so any play through would have a good chance fr victory. With item upgrades, it gives the chance to miss items, forcing the player to collect them or become underpowered. They also force the player into interesting situations to gain these upgrades, and help to increase replay value. Once you have beat the game with the shotgun and jetpack, there is incentive to try again but on a run where you don't pick these up.
Expirmentation: Basically presenting the player with situations without an obvious outcome, and letting the player figure out what they want to do to solve it. These include putting enemies in pots, so the player must learn that these present a threat and be catious in the future or pretty much any event from FTL. These help force a curve onto players, as they will only progress after the knowledge gained from a death.
Unlockables: In a game where you are expected to play over and over, unlockables are there to make the game feel rewarding. Without them the player starts to wonder why they keep trying. They are the final incentive, so have a lot of them.
Now on dying. Because the randomised upgrades to the player and the inability to back trace, there will be times when a player is far too under powered to beat a section. In this case, permadeath is an ideal solution. It forces the player back through the earlier levels to attempt to gain better gear. If you didn't restart but just respawned, players would try much more to get through with inadequacy gear, getting more and more frustrated, and quit more often than restart.
As you can imagine, these games achieve many, many hours of gameplay, so if you want a game that people just keep playing, think on these mechanics.
This made me think a bit about
my own game in development. As it stands, when you die, you just restart the level from the beginning. The idea behind this is that you'll eventually memorize the enemy patterns enough that you can beat it with your current selection of upgrades, but it doesn't really let you experiment, since it doesn't throw you back to the shop. It also possibly starts an unstable equilibrium, since if you manage to BARELY scrape by a level, you'll likely have less cash from that level to buy upgrades, which in turn will make your life harder in the next level. I've recently added a Hardcore mode, which essentially means permadeath, which by your sound logic allows for more experimentation, and ends the unstable equilibrium once it starts to tip too far away from your favour. The problem here, however, is that the enemies are NOT random, quite the opposite actually. Every single enemy is placed individually and deliberately, so as to create a challenge for the player to overcome.
But i digress, what i'm thinking is this: Would permadeath be beneficial in a game without randomization? Damage is fixed, enemy placement is fixed, point value is fixed, there are no random powerups, the only thing that actually varies randomly is enemy firing intervals. The idea behind that is to minimize the players dependency on luck (luck-based missions should crash and burn), thus making player skill far more relevant. However, although there's no significant randomization, the enemies do increase in variety and threat level as you progress, sometimes in rather unanticipated ways, which could likely screw over a player just as much as a spiteful RNG. If you get hit by something unexpected while still being in reasonably good standing from the previous levels, does that mean you should start everything over?
This also makes me think that permadeath should remain optional, but enabled by default. Some people don't care much for running through the same levels over and over because there's one particular thing that messes them up, especially not when the previous levels don't change at all. However, fun thing is that Raptor: Call of the Shadows is almost exactly like this, and it's a huge inspiration for me. It has levels with fixed amounts of enemies (which varies by difficulty), and if you forget to save (which in and of itself kind of makes permadeath pointless), it's permadeath for you.
It's still a very entertaining game. My game doesn't really support saving, but potentially gets a free pass by virtue of being relatively short. Hell, another game with permadeath, The Binding of Isaac, is not all that short, you can't save in it either, AND it's often mercilessly difficult. It has random levels, however, so every time you start a new game, you actually get something new. My game does NOT feature that, which is a fairly strong counterpoint to permadeath.
So... I guess what i'm trying to say is that i'd like some external input on it. The player obviously needs some incentive to start over again, but what would that even be? The notion of a better playthrough? Not so sure about that.