How to eliminate Power Creep in persistent online games
In which Hax presents a thinly-veiled plot to undermine the Experience/Level system, and the reader berates his terrible choices for example names.

Unfortunately, this does not refer to protecting yourself online from incredibly dedicated stalkers. The best advice I can give for that is never to reveal your name, e-mail address or your presence on other online communities, and definitely never use Facebook or the like.

Instead, this article refers to what should probably be called 'Power Inflation'; a concept often employed by MMOs or other games with some degree of persistence as a fast and cheap means of providing long-time players with new goals (i.e. a reason to continue playing). The trade-off, of course, is that the player becoming more powerful is equivalent to all previously-released content becoming more trivial. If the content was well-designed, it would have been fun and challenging when it was released, which is lost if the player is able to inflate their power indefinitely as new content is released.

As a special note, I do not make the argument that any equippable gear from the older content becomes useless. This is primarily because I reject the logic that the gear becoming useless implies that there's no longer any reason to play that content. If the only reason you are playing a piece of content is because it gives the best gear, the content is shit. Do not accept shit content.

Before I begin, I would like to acknowledge a particular episode of the excellent web series Extra Credits. The scheme proposed in this article is somewhat different, though does incorporate their concept of 'incomparables'; properties which cannot be compared directly by means of looking at two numbers and deciding which is greater. To illustrate this idea to oneself, have a skim-read of the following item tooltip from Diablo 3, and try to 'spot the incomparable':


Amongst the typical-looking bunch of numbers and stats, what have we there? Allow me to annotate:


That effect, which the developers were kind enough to print in contrasting orange text, is extremely, unfathomably powerful1 in some situations. In others, trying to make use of it is reckless and ineffective. Of course, building the rest of one's loadout around it and developing a skilled risk/reward playstyle to utilise it can tip the balance towards the former, but the point is that its effectiveness cannot be compared to other such effects, for example:
  • "Lesser enemies are now lured to your Meteor impact areas"
    Aside from reducing the chance of Meteors missing the target, this can also be used to herd enemies away from being able to attack the player, which gives the player more scope to sacrifice survivability in favour of higher damage.
  • "You may have one extra Hydra active at a time"
    Hydra is a slower and steadier way to deal damage, where the advantages are that it can be placed next to hazardous opponents while the player keeps a safe distance, and it continues to deal damage on its own while the player is free to deal more damage by other, more active means. This has the potential to turn Hydra into a staple ability as opposed to a trade-off.
When the game already has other ways to complement those abilities without involving gear, the possible combinations start multiplying, and it becomes rightfully improbable that any particular combination will prevail in all circumstances. Yet, despite the destructive potential of having four explosions in one, if the developers release a bigger stick with higher numbers then I might as well scrap that item because I could sink some time into acquiring the new gear then proceed to do the same damage as before but with a safer strategy.

In any case, what follows is a relatively complete, ground-up model for including levels and gear in an MMO or otherwise persistent game, with zero power creep or inflation of in-combat numbers2. The system is focused around PvE ('player versus environment', co-operative gameplay), though does contain a PvP ('player versus player') addendum.

  1. All players have only one primary attribute.

    I gather that some games already do this; no more 'Strength' for players with melee weapons or 'Dexterity' for players with ranged weapons. There will be one such attribute only, which can be flavoured to be appropriate for the game, though the rest of this article will refer to it as 'Awesomeness'. An amount of this attribute will be written, for example, '4 Awesome'.

  2. Without equipping gear, each gear slot provides a baseline amount of Awesomeness appropriate to the player's level.

    For example, and for sake of argument, suppose the player has ten gear slots (including weapon, armour items, anything equippable). Each gear slot gives 1 Awesome per player level, so a level 1 player automatically has 10 Awesome. A level 2 player automatically has 20 Awesome, and so on.

  3. All players deal the same base damage (or healing) and have the same base amount of HP regardless of level or gear.

    Here, 'base' means 'without considering character specialisation'. If specialisations exist for tanking, healing and damage-dealing, they would augment the relevant values. That aside, as an example, players could all have 10000 HP and deal around 1000 damage or healing per second if they use their provided abilities as per their class design.
    The developer is free to tune different types of enemies and different encounters as needed, basing this design upon the amount of HP, damage and healing they now know that the players will definitely have.

  4. When a player has less Awesome than the baseline Awesome of the level of an enemy, the player deals less damage to that enemy, and takes more damage from that enemy.

    Exact scaling depends on the scaling of the baseline Awesome. To continue the example values thus far:
    • For every point of Awesome below the enemy, the player deals 1% less damage to it, to a minimum of 0%.
    • For every point of Awesome below the enemy, the player takes 1% more damage from it, with no limit.
    This means that, again with no gear, player who is under-levelled by 10 (i.e. by 100 Awesome) will deal zero damage and take double damage.

    As an important note, once the player's Awesome value matches the enemy, exceeding that value provides no further benefit. The player will deal and take standard damage, which will never change by gaining more Awesome. This ensures that the player will never out-level or out-gear the content in which that enemy appears.

  5. When a healer has less Awesome than a target player, the healer restores less HP to that player.

    Example: For every point of Awesome below the target, the healer restores 0.5% less HP to it, to a minimum of 50%.

    This means that healers also need to gain Awesome to keep up with the rest of their party, yet it is still okay for the party to gain Awesome in order to equal a stronger foe since the extra damage reduction far outweighs the reduced healing. As a courtesy, if some party members have excess Awesome for the content they're currently playing, it would be sensible to include a means of placing a temporary cap on one's own Awesome value to account for a lower-level healer. A simple slider on the UI would work quite well, and instanced zones for special group content could also cap players automatically so that nobody taxes the healer by exceeding the strongest enemy in the zone.

    This system of partial reduction to effectiveness for the case of healing another player also works quite well as a general system for PvP, including damage in particular. The complete details for PvP integration follow at the end of this article.

  6. Equipping gear boosts the Awesome for that gear slot to a fixed value.

    Specifically, gear does not have properties such as '+5 Awesome'.
    Instead, for example, a level 10 player could find a piece of gear with simply '15 Awesome'. Being level 10, their gear slots already provide 10 Awesome by default, but equipping that piece of gear will cause that gear slot to provide 15. This extra power gives the player closer to equal footing against enemies beyond the level 10 baseline, while providing no benefit at all against lower-level enemies, thus curbing power creep.

    As the player gains levels, the extra 'reach' provided by the item naturally fades. Once the default amount of Awesome provided by the gear slot exceeds the Awesome value of the item, the higher default takes over.

    Notably, even if the player does gain levels beyond the usefulness of their 15 Awesome item, they are still free to keep it equipped to take advantage of secondary effects.

  7. For each gear slot, secondary effects provided by gear must be incomparable.

    Secondary effects are strongly encouraged for almost all gear; to support this, the game itself is encouraged to provide a wide selection of secondary effects. The best types would be small enhancements to the player's abilities (or indeed large in some cases, as per the orange-text Diablo 3 weapon items), though using the usual side-values such as critical chance/damage and attack/cast speed is also fine as long as the actual numbers used are always the same for that gear slot. If the player finds a similar item with the same secondary effects but with a higher Awesome value, the new item would be objectively better for a while, but the player will eventually out-level that item too. At this point, both items are now equally effective, and the player is free to use either based on appearance.

    For example, if the player can find gloves with +10% critical damage (regardless of Awesome value), all gloves of the same quality/rarity with critical damage need to be +10%. This can co-exist with a hat which provides +15%, again as long as all other such hats with critical damage are +15%. It sounds like a harsh specification, but it is necessary, and I did say that providing quirks to class abilities was a better option. The vast amount of possibilities for the developer and the resulting vast amount of combinations for the player are both good things.

    To get a glimpse of this vastness, I will think of an example right now. Imagine a piece of gear which modifies a healing spell so that 20% of the healing occurs over the course of the cast time, followed by the remaining 80% at the end - designed as a slight increase in utility, since it gives the target more room to stay alive while the heal is incoming. Actually, it's a lot more than that:
    • The casting action is now a channelling action; effects which improve or trigger from channelling will now apply where they previously did not.
    • The ability itself now has a healing-over-time component; again, effects which improve or trigger from healing-over-time will now apply too.
    • The player also now has an interesting strategy available when high-volume healing is not required; they could cancel the ability just before it finishes, forfeiting the 80%, but effectively pocketing that 20% for free. Not only that, it would become one of the very few cases where a bonus to casting speed would improve output without affecting resource expenditure.
    That simple enhancement is now starting to look potentially overpowered, even though it doesn't actually increase the player's total output at all. This should illustrate the point about magnitude of possibility, as well as the joys of emergent gameplay.

That's it! The focus of gearing is now on secondary effects, though it still has a means for the player to enjoy some extra power for a bit, at least in the sense that it enables the battling of opponents beyond their depth as opposed to obliterating their equals. The player can acquire powerful gear to push their Awesomeness beyond the level cap and conquer the toughest content in the game, without making any previous content obselete due to raw numbers becoming higher. Damage and HP values all have the same baselines and totals at any level of content, which is a great help to players and developers alike. Most importantly, content and gameplay are no longer slaves to progression.


As mentioned, the same basic model can be extended to include PvP as a separate progression path. The idea here is to take the healing component of the PvE model, in that it already represents the notion of targeting another player, and apply that to damage versus enemy players as well. Essentially, attacking an enemy player is the reverse of healing a friendly player.

To do this, we introduce a new main attribute for PvP which runs entirely alongside Awesomeness, and which I will name Badassery.
  • Empty (or PvE-equipped) gear slots provide a base amount of Badass in exactly the same way as they do for Awesome.
  • When attacking a player with equal or lower Badass value, you deal full damage.
  • When attacking a player with higher Badass value, 0.5% less damage is dealt for every point below the target, to a minimum of 50%.
  • Extension: When healing a friendly player, both values are compared (i.e. Awesome and Badass), and the most deficient difference is used.
    While maintaining Badass gear would be advisable for healers anyway due to the PvP damage reduction it provides, this ensures full separation of PvE and PvP gear without any need for the game to decide which heals restored HP lost due to PvP.

Players with high Badass have a definite advantage in the form of PvP-focused secondary effects and reduction to incoming damage, but now cannot go around destroying low-level players in a single hit. Low-level players are less effective, but can still make a meaningful contribution to any given PvP battle, as opposed to having acknowledge their avatar's worthlessness under the typical present MMO model3.

1 My source for this claim is that the item in the screenshot belongs to my own account, and I used it to clear the toughest content in the game on my own (at the time, of course; it was a while back).

2 Blizzard has recently applied something to World of Warcraft which is being referred to as a 'squish', basically compacting all damage and health numbers down by a factor of 10 or so to make them more readable after years of inflation. They have done nothing to solve the underlying problem, so those numbers will re-inflate again over time, but the recognition and the relief effort is a start. Somewhat prior to that, Blizzard themselves once mistakenly said that it wasn't a problem, and that introducing magnitude prefixes (e.g. 'kilodamage', 'megadamage') would be considered.

3 In the context of MMOs with both PvE and PvP present, of course. PvP-only MMOs (such as Planetside 2) are able to avoid a lot of problems by making the PvP action follow a specific format, with gear progression to match that format.
Comment thread »
No comments!