I'd say much the opposite; if anything, leveling up has been refined significantly and is a staple of modern gaming. It's escaped from RPGs to become the basis for a hell of a lot of social games, COD multiplayer, and everywhere else that RPG mechanics have crept in. It's been perfected into an excellent carrot to keep players playing, and, even at its worst, it's rarely handled more terribly than in the games which pioneered it. I definitely wouldn't call it "outdated."
I suppose the question the OP was asking was more of "should it be in modern games?" The answer from a developer making a microtransaction-laden Skinner box is "absolutely." The answer from someone who is interested in making a fun game is much harder to define. At their worst, leveling systems add in loads of grinding and arbitrary scaling of tasks; they "change" the game by making the numbers bigger but not affecting how the game plays at all. At their best, leveling systems add a mechanical sense of progression to the character, change the gameplay to become deeper, and make the player feel invested in his developing character. Which extreme it favors is dependent on the game, on the precise leveling mechanics implemented, and on the players' preferences themselves.
To speak about designing leveling systems from experience: there's a tactics game with pen-and-paper rules I've been developing for the past few years that I'm also coding into a computer game, and one of the many major decisions I made in development it was to add levels into the game. The major reason that I left them out in the first place was that I didn't want the game to turn into stacks of arbitrarily scaling numbers fighting against each-other as is typical in the majority of RPGs. However, as the game developed, I found that, without a convenient system to describe how powerful a given character was, I had to make arbitrary statements on how many advantages a character could have, how much they could spend on attributes, and what the caps on those attributes were. As it stood then, I couldn't make those arbitrary statements allow for a wide enough variety of characters that I was happy with, and so I would have had to let those arbitrary be modifiable via cumbersome house rules. It was a horribly clunky system.
So, I ditched it when I changed everything in the basic one-off skirmish game to a point-buy system. Now, level is an attribute of characters that controls how expensive a character is, his number of advantages, attribute points, and caps on attributes. It's much cleaner and more efficient than the original. Additionally, with the way that the point cost scales due to level and a host of other gameplay mechanics, it should be prohibitively expensive if not impossible to create a single "superman" character who is too high level for low-level characters to do anything meaningful against. It's also very easy to extend into a leveling system for more active character advancement in the single player campaign I'll be implementing much further down the line, and it also will be useful when I'm delving into randomly generated characters for that campaign.
This demonstrated to me that level is a very useful measure of character power even if you're not trying for a effective range during the whole game of more than, say, 5 levels or so, or not trying for going from 5 health to 5000 health at max level.
Ahri said:
I seriously think more MMOs need to incorporate an Elder Scrolls type leveling system, where your abilities improve as you use them. It gives you an incentive to experiment with different abilities, and also helps you to tailor your character to what you want it to be. The trouble with giving players a set number of abilities and very little variation between them is that everyone ultimately ends up the same, and you feel as though the choices you make are worthless.
I strongly dislike the Elder Scrolls leveling system. It directly rewards you for repeating the same action over and over again until you hit 100 in the skill, which is tedious and is thinly related to working towards interesting character goals. The basic (and good) idea behind it is that characters are rewarded based on their actions, but it also means that if you've got 20 different skills you've also got 20 different ways of leveling them. That means that instead of the developers focusing just on making combat interesting and fun, you've also got to make sure that those 20 different skill progression schemes are also interesting and fun. The Elder Scrolls does a pretty poor job of that; it rewards you for standing in front of an enemy healing yourself, for attacking with the worst weapon you can find to get more hits in, for continually casting buffs on yourself, and for smithing the same iron dagger over and over again. They have been doing better with each subsequent game, but it's still a deeply flawed system.
That's not taking into account the abysmal level-based enemy scaling they use, either.
I could see a hybrid system working much better - one which, say, after each goal accomplished and/or creature kill awards experience based on the event's significance to the three skills you used the most - so that you wouldn't be rewarded for sneaking into a shopkeeper's wall for three hours. But even so, unless it was very carefully designed it would still reward gaming the system over just playing the game. Tying character advancement purely to the player's choice and not his actions certainly has its flaws, but it also means that, not only is it a hell of a lot easier on the designer, players will choose the options that interest them most and hence they'll find character advancement to be much more motivating.
