[LoL] Classic and Dominion Balance

Posted on November 18, 2011


Taken from the League of Legends Wiki "Crystal Scar" page

While there may be balance differences between Summoner’s Rift and the Twisted Treeline arising from the difference in team sizes, the completely different playstyle of The Crystal Scar has created a great number of balance issues, relatively few of which have been addressed.  Is this a problem?  It first surprised me that, in the two months since The Crystal Scar was released, the community hasn’t protested these issues as much as I had expected.  Upon further consideration, the League of Legends structure naturally downplays such issues.

Taken from the League of Legends Wiki homepage

Variety and Simplicity

With the addition of Fizz, there are 87 champions, each with a passive, three basic abilities, and one ultimate ability.  Champions have between 8 and 10 basic stats (some champions don’t use resources) and can acquire several more through runes, masteries, abilities, and item builds.  When it comes to balance, the wide variation on a relatively small number of variables makes things much simpler for the developers compared to a system with fewer champions but more abilities and stats.  If any one champion is universally performing outside the acceptable range, there are a reasonable number of places where changes can be made, and those changes have little room to create any unintended “butterfly effects”.  The result is that oftentimes one single number tweak is all it takes to bring a champion to a good place and, conversely, devs have more freedom to think of creative solutions without worrying about an unthinkable number of side effects.

On the player side of the equation, I believe that we subconsciously understand that 87 champions can’t all be on the exact same performance level in any setting, Dominion or otherwise.  This has always been the case: even when the game was originally released, it had 40 champions.  These champions fill a finite number of roles, meaning that it is rarely too hard to pick up a new champion in the same role, e.g. dropping Ashe for Miss Fortune, Yi for Talon, or Shen for Rammus.  That isn’t to say that these champions play exactly the same, or even similarly in the most extreme cases, but at the risk of becoming stale, if your new champion doesn’t work for you, just switch things up again.

This versatility is, in fact, required of the player base.  Though there are indeed a myriad of champions, a few factors can force a player to play something other than his “main” champion:

  • Another player choosing that champion (especially common if the champ has been recently released or is part of the free rotation)
  • The champion being banned in draft mode
  • Team composition needing a champion of a different role

At level 30, ranked games require every player has access to 16 champions or more, meaning that a player has to actually own a minimum of six champions that are not in the 10-champion free rotation of the week.  Short of simply buying six 450-IP champions to meet requirements, that means ranked players have experience with at least six champions.  Almost universally more, since the grind to 30 takes at least a couple months, leading the player through many free rotations.  The point to take away is that players should never feel that “their champion” is underperforming or doesn’t deserve a nerf, since the system isn’t designed to give players the feeling that any one champion is “their champion”.

Excerpt from the League of Legends game client

The Customization Effect

In addition to a vast array of champion choices, League of Legends features five performance-modifying customization features.  Three of these are pre-game decisions: Runes, which are basic stat increases, Masteries, which mix stat increases and other effects such as increased damage against nonplayer targets, and Summoner Spells, which provide significant utility of both the damaging and non-damaging varieties.  The fourth feature is the order in which skills are ranked up in-game, and the fifth is a player’s item build.

The pre-game decisions can and frequently are min-maxed for individual champions and playstyles, and the bad options are usually obvious: if you’re playing a champion who doesn’t use mana, then don’t use mana-increasing runes, masteries, or the Clarity spell.  That being said, smart players will acknowledge quirks in their own playstyle and adjust accordingly.  For example, I have a tendency to use my abilities a bit too much, draining my mana at low levels.  I therefore make room in my builds for mana or mana regen-increasing options.  Stubbornly sticking with a general build always leaves me dry for mana and unable to help my team.

The in-game decisions generally follow a framework but have the distinction of being reactive choices.  For instance, carry champions generally need to build up their AD/AP as fast as possible so their high-scaling abilities outpace the effective health growth of the enemy team.  In my last match as Miss Fortune, a ranged AD carry, the enemy team was comprised of five AD champions and no tank.  I eschewed my normal build to fit in Thornmail during the mid-game, giving me enough armor to survive a few tough encounters and taking advantage of the 30% damage return to batter enemies during teamfights.  In other matches, I delay learning my weaker, area-of-effect, slowing ability to max my targeted and passive damage abilities as fast as possible.  Guides are based around the premise that you’re playing against a balanced team with reasonable mechanics.  The minute you notice the other team is divergent from the norm in some way or another, you should re-evaluate your champion’s basic skill and item frameworks.

Looking at team composition from another angle reveals that it is a sixth form of customization, but one that individual players cannot easily control.  Your teammates won’t necessarily choose who you tell them to choose, and it’s even harder to influence the decisions of the other team (impossible in blind pick).  In-game, the performance of the other players is yet another variable that each individual must adapt to.  Aggressive players might have to sit on top of their towers because of a bad matchup, or defensive players might have to constantly babysit a less skilled player.  Ignoring these effects is the surest way to lose a game.

Bottom-line: with all these influences, it’s uncommon for a champion to be so imbalanced as to earn “broken” status one way or the other.  A champion is merely the most visible of the decisions a player makes, but each choice has an impact on how the champion performs.  Of course, champions also have their ‘flavors’ – maybe you always lose against Gangplank because the best players prefer to play him.

Generally speaking, complaints on the forums are made from people who had a few bad games and are looking to vent their anger.  Players with a more even distribution of wins and losses against a given champion are significantly less likely to post about their experience because nothing has indicated to them that anything needs to be changed.

Excerpt from League of Legends game creation screen

Two Different Games

Compounding the customization effect is the fact that Summoner’s Rift and the Crystal Scar are two different games prioritizing different styles of play.  My Summoner’s Rift Nocturne build prioritizes my damaging Q, then my ability-blocking, passive attack speed W, then my CC-applying E.  In other words, as much damage as fast as possible.  My Crystal Scar build is the exact opposite because inhibiting enemy action plays a far greater role in this mode.  E allows me to perform this task while W allows me to protect myself against enemies attempting to do the same.

Illustrating a different situation, I enjoy playing Xerath in Summoner’s Rift.  One day I forgot I had selected the Crystal Scar and was locked in before I realized my mistake.  That was by far my worst game of Dominion to date, but at no time did I think “Xerath needs major buffs”.  The first thing that ran through my mind was instead “Xerath isn’t a Dominion champ”.  His design thrives in lanes and falters in open 360-degree spaces, and there is simply no way to have him perform equally well in both settings.  If buffed for for the Crystal Scar, Xerath would become far too strong in Summoner’s Rift.

The abundance of champions is what allowed Riot to simply port their entire catalog over to the Dominion game mode.  Rather than give each of the (at the time) 83 champions modified abilities, in effect forcing players to half-relearn their champions in addition to the new map, Riot was able to take the Summoner’s Rift balance and drop it into the new map.  The logic isn’t unreasonable – the roster is so huge that if it’s balanced in one game mode, it should for the most part remain that way in a new environment, and adjustments can be made to any champion that falls too far away from its peers.  It’s far more efficient to fix one or two, or even ten or twenty oddities rather than try to predict how all 80+ champions will perform and adjust them before players ever get a chance to try them out.

– – –


If we treat every build and team composition of every champion as a unique experience, even after paring away the obvious bad choices, there are so many options available to players that it’s unrealistic for one person to be able to say with any credibility that a champion is overpowered or underpowered.  The vast majority of these options are apparently (and understandably) within Riot’s acceptable performance range, meaning that circumstance plays a huge role in how champions appear to be balanced.  This is true in both game modes.  While the community must rely on anecdotal experience, Riot has access to global data and the real averages.  Each patch brings new champion adjustments based on these numbers, preserving the invisible balance.

Posted in: LoL