Designing a Level System

 If you’re making an MMORPG (or any kind of RPG, for that matter), one of the very first questions you have to answer is, “How do levels work in my game?”

Of course, before you answer that you might ask, “do we even have levels in our game at all?” Maybe you don’t have “levels” per se, but if you are making  an RPG then we already know that the characters become more powerful over time. You need to represent that power numerically to the player somehow, whether it’s in the form of character levels, space ship hull armor, sword damage stats, or the number of magic spells in the spellbook.

Whatever the numbers may represent, something tangible makes some players more powerful than other players. So let’s lump all of that together as “levels” for now, because it turns out you have to ask similar broad questions regardless of the particular form of level you use.

So back to the question: “How do levels work in my game?” First off, you need to know how big a difference there is between levels. How much more powerful is a level 10 character compared to a level 9? Do they become 20-25% more power when they level, like EQ1? Are they 10-15% more powerful, like WoW? Or is it only 3-5% more power, as in AC1?

This may seem like an overly narrow place to start, but the power differential between levels has a lot of gameplay ramifications.

If level differences are large:

  • When a character gains a level, the player can often go different places and kill different things than they could before, so levels provide a good way to break up content and provide a feeling of progress.
  • Players feel much more powerful when they level up, so gaining a level is a meaningful accomplishment.
  • Players have trouble grouping with characters of different levels players because their power levels vary dramatically. This limits the opportunities for grouping, especially at lower and middle levels.
  • PvP between differently-leveled characters isn’t very balanced, or fun, even if the players are only a few levels apart. Again, this limits the opportunities for PvP play.

If level differences are small:

  • Players don’t get excited about leveling up, because it doesn’t make them feel more powerful or open up new venues of gameplay.
  • Players can group with a broader range of other characters easily, making it easier to find a group to play with.
  • PvP between differently-leveled characters is more exciting and accessible for a broader level range.

Breaking up Content

Levels are an important part of giving players direction: levels are how players understand where they’re supposed to be and what they’re supposed to do. They intuitively know that they should be in areas with similarly-leveled players and opponents. Levels also control exploration. When the power differences between levels are big, players are not able to roam very far. The less important levels are, however, the more characters can roam the world at will.

In AC1, level differences were originally so miniscule that a level 20 archer could often kill a level 80 opponent. How could they tell that they could do that? They just had to try it and see what happened. This was a lot of fun for a certain type of player, but very confusing and even “cheaty” for many others.

On the other hand, players who liked exploration loved AC1 for this reason, because they could meander through most of the enormous game world with impunity.

PvP Considerations

You might be thinking, “Everybody knows that if you want a PvP game, you want small level differences.” When the power differences between character levels are small, then player skill has a larger impact in the outcome of PvP fights.

Certainly this is the case for EVE Online, as well as for AC1′s Darktide PvP server. Both games owe a lot of their PvP success to the fact that relative lowbies can kill players much higher level than them. This is a powerful enticement for new players and helps keep the competitive landscape fresh for long-term players.

So  a small level differential is a good place to start, but it’s not a magic solution. You still need to give your players a sense of forward progress and accomplishment over time, and if levels aren’t the primary means then you must have some other way. And those other ways can also represent strong power differentials that can limit PvP.

For example, it is often said that  Warhammer Online has a shallow level curve, but that’s not the whole story. WAR does have a relatively shallow “character level” curve, but then your power level becomes measured by your items and equipment. There’s still “high level” and “low level” equipment, and it still has a meaningful impact on the outcome of battles. So you haven’t gotten away from the problem of power differentials: you’ve just mitigated it significantly.

You can also mitigate this problem by making levels increase the breadth of game verbs available to a player, rather than by increasing the potency of the verbs. In EVE Online, for instance, mastering a skill isn’t particularly hard — but there are a ton of skills for you to master. You’re gaining more verbs, which can affect your overall potency in PvP somewhat, but not nearly as dramatically as if you just kept jacking your damage-dealing statistic up directly.

Grouping Considerations

Grouping in PvE is another important consideration to keep in mind when you are designing your leveling system. Players can generally only group up with other characters that are within their relative power range. In most games, if you were 500% more powerful than your team mates, it wouldn’t be a lot of fun for the weaker group members – they’d feel like hangers-on rather than part of the team.

So most games restrict groups to a certain power range. But this drastically narrows the population of characters who are available to group together at any given time. In a game with a lot of group-centric content, the inability to find a group can be deadly – especially for players who join the game after the majority of the players have reached maximum level.

Of course, there are nifty mechanics like Sidekicking and Mentoring to help fix this: they temporarily alter a player’s level so that they can group with friends at all levels. This is an elegant fix that neatly side-steps the question of power levels, but keep in mind that it may require some serious design planning to fit into your particular game.

Making Leveling Up Fun

Finally, we come to the most important consideration: fun. Players get more excited about leveling up if they get significantly more powerful when they do so. When they get new verbs, new places to explore, and new things to kill, they tend to find leveling up to be pretty rewarding. If leveling up doesn’t give them these things with reasonable frequency, leveling up becomes rather dull … or worse, a grind.

In my experience, by far the best way to keep leveling up feeling fresh instead of repetitive is to introduce new gameplay verbs as often as possible.

You might also consider exactly when players earn their new power: in discrete chunks or smoothly over time? In Asheron’s Call, players didn’t need to reach new levels to get more powerful. As soon as they earned any XP, they could “spend” that XP to improve their skills. At low levels especially this provided immediate and continuous advancement, but at a price: when you actually leveled up, it was extremely anticlimactic. The level didn’t mean much of anything in itself.

Your choices here  will really depend on the game, but if you do have explicit character levels, then you should make it fun to reach those levels. Otherwise you’re just missing out on an obvious chance to give players fun.

Choosing The Right Leveling System

As always, it comes down to: “What experience do I want my players to have, and why?”

Are you making a game for mathematically-inclined nerds? They tend to have a broader perspective of overall power, and weaker level differences may make the most sense. But if you’re aiming for young people (tweens, say), you almost assuredly want strongly differentiated levels to act as a carrot and keep them playing. In fact, all the issues I’ve talked about here can be answered by figuring out how your target audience might think.

And one final  thing to keep in mind: if you are having trouble making a leveling system that will appeal to all of your demographics, then your target audience is almost certainly too broad.

This entry was posted in Design. Bookmark the permalink.

10 Responses to Designing a Level System

  1. Pingback: Mike Gowen » Hot Sidebar Action: October 21st

  2. Eugene says:

    I am seaching for some idea to write in my blog… somehow come to your blog. best of luck. Eugene

  3. Tesh says:

    I’ve been pondering what Mark Rosewater calls the “power band” of design elements. In his case, he’s referring to MTG cards, of course, but I think it can apply to levels as well. A narrow power band, where the top level isn’t hugely different from the bottom level, makes for a more skill-intensive gaming format. You’ve aptly described the power concerns in the RPG format.

    The concept of horizontal vs. vertical expansion is especially interesting. The “vertical” power band might be fairly narrow, but the horizontal breadth of abilities, specialities and such would give players even greater opportunity to define themselves by their skills and interests. This is appealing to me as a way to make a game world relatively open, with skill-based PvP, yet have plenty for players to do.

    We see some of that with WoW’s talent system, but it seems to me that pushing things even further might make for stronger design. Guild Wars leans more in that direction, with a narrower power band, but a wider horizontal breadth. Interestingly, it also gives players more customization options, with skill hunting, dye and heroes.

    In the end, the game really does need to focus on what it wants to do, and you’re right, it can never be everything to everyone. I do tend to think that more MMO innovation will come from the GW lineage rather than the EQ/WoW line, mostly since a narrow power band will tend to appeal more to a casual audience. Players don’t get too far ahead of friends, or too behind their enemies.

    Of course, that’s mainly approaching the DIKU game type. Puzzle Pirates has a completely different take on things, where “levels” are defined by player skill, and players of any skill can play together and wind up contributing. Of course higher skill is more effective, but the dynamics are such that the PvE system tunes itself to whatever your current group dynamic is. It’s a different sort of game, but it’s a very interesting one.

  4. Babs says:

    One way or another, players will be defined by levels of something mainly because living things like to define their universe. We’re list makers at our primal cores – this is dangerous, that is not; this is better, that is worse. So you’re right, Eric, we’ll either define ourselves by actual level advances or by steadily advancing gear/abilities.

    What I’m observing these days more so than in the past is that online games are increasingly clique oriented. They’re becoming hugely guild-based, or raid-based, or at the very least something less than universally social – unless you’re talking about tween/teen games. You get in, get your feet under you, join a group of like-minded individuals (if you didn’t already subscribe with your guild) and have at it. Interplay between these cliques occurs closer to the end-game (forming raids/alliances) rather than at the beginning (the classic RPG dynamic). How to make leveling matter for individuals vs. individuals in the same manner that it matters for groups vs. groups is a challenge. Most of the games out there have, say, 50-100 options per class, of which only 10-20 are really used because groups are less about playing the MMO than beating it. I think that changes how devs should think about levels in general, too.

  5. Eric says:

    Tesh – I think having a lot of breadth is a good idea for a lot of games. But it can definitely change your target audience. Think of the breadth of knowledge you need to be competitive in WoW versus the breadth you need in Magic. The difference is phenomenal. This keeps M:TG from picking up the same audience as WoW. I think in the end it depends on who you’re aiming for.

  6. Eric says:

    Babs – Interesting point. I guess here using some extra breadth can really simplify hitting more audiences; if 20 of your 100 skills are useful in raids, but 20 more are useful in PvP and 20 more are useful when crafting, you’re getting somewhere. But as I just pontificated above, that probably means fewer people will be willing/able to learn your game.

  7. Tesh says:

    True, Eric. There’s also the whole “Limited vs. Constructed” difference in MTG, where the audiences overlap in places, but there are definite fans of one or the other. It actually almost reminds me of the “Guild Wars vs. WoW” game play argument. There’s more breadth in GW, and the “character build” mechanic certainly feels like a Limited event… and the audiences are different between GW and WoW. To my mind, GW and MTG (Limited) are more skill based, where WoW and MTG (Constructed) are more about time and investment. There’s certainly skill involved, but with the WoW game largely based on “the game starts at 70″, with concurrent itemization and content focus, there’s definitely a huge time investment being made. Contrast that with GW’s “instant level cap” PvP characters. The playstyles are definitely different, and will cater to different audiences. I’ll admit, I tend to lean heavily to skill testing games, partly because I have a “Casual” player’s tight time constraints, but a “Hardcore” player’s interest in skill and design.

  8. Django says:

    It’s definitely a fact that most player want to feel a noticeable increase on leveling (new skills is of course the easy way to do this) but recently I’ve been somewhat put off by the number of ‘things’ that MMOs feel a need to include. When I’m playing just about any MMO now I find that I have 3+ skill bars full of ‘things’ (skills, gear, potions, etc.). Do we think this is really necessary?

    Limiting the choices to a lower number to reduce complexity and screen usage will simply cause the general FotM (players only use the ‘good’ ones).
    Removing the choices completely, or just mostly, is going to take that simple ‘POP!’ out of leveling.

    I figure part of this comes down to if ‘things’ hit the widest targetable audience and that’s the intended plan or if something I’d consider a more elegant solution would be capable of capturing the same audience. (We could just be dealing with needing a smaller scope on your target audience, but I haven’t played a major MMO release in a while that wasn’t trying to target everyone possible)

  9. gattsuru says:

    Another alternative is to have levels mean extremely little or nothing. Early after City Of Heroes released, levels 40-50 in themselves didn’t really add much; there were no new powers other than stocking up on the largely useless pool powers, and sidekicking meant that a player from level 40 could help in a level 50 mission with ease. In S4 League, levels offer little more than a token amount of currency and access to flashy but entirely aesthetic clothing choices. The levels exist in order to warn other players of your personal experience. Because a level 1 player has 5 hours worth of use of every weapon type, they’re often better equipped than a player who’s been around for a while.

    This works for S4 League because, while the player character starts with every weapon and ability, the player himself or herself needs time to develop his or her understand of the game’s mechanics, and other players and areas start to generate new concepts and challenges on their own. It’s still an RPG, with stats, levels, equipment, and areas reached based on level, but the levels themselves don’t do much.

    I’ve not seen many other games try that. I’d be interest, though; from S4 League and Phantom Dust, it seems like a good opportunity to keep the interest of both hardcore and tween players, and move rewards closer to the gameplay itself.