As everyone enjoys Hexakill: Twisted Treeline, it felt like a good time to give some insight into how we make Featured Game Mode's by analyzing our approach to the previous mode, Ascension. Hopefully I can give you guys some insight into our iterative development process, what we take away and learn from each mode, plus some stories about things that didn't make it.

We start each Featured Game Mode by trying to define what kind of experience it should be. There are design pillars which help guide this, but we often want you to "discover a new or interesting way to engage with your favourite champions in League of Legends". Of course we still want modes to feel like LoL to some degree, so you can expect to level up, gain power and fight other champions, but we’re not afraid to bend the rules. Sometimes there’s an overarching theme we’re working towards with a mode (e.g. Shurima) and other times it’s a fun mechanic or rule change that just wants to be turned into a mode (e.g. One For All, URF). The golden rule is that regardless of its shape, a mode has to be fun!

With Ascension, we wanted to highlight the Shurima event and bring some of the pervasive lore themes like “power corrupts” through with fun gameplay. It needed to feel scrappy and intense so we started drifting towards a Team Deathmatch (TDM) kind of feel. The Crystal Scar's circular shape made it a strong initial candidate for that kind of gameplay. The tower pads nicely doubled as spawn points enabling a strategic choice for players when entering the battle (and also neatly solved design issues around TDM spawning in the LoL ecosystem). It wasn't going to just work 'as is' though, so we ripped out the towers and minions to fast forward through the laning phase of regular LoL, enabling you to get straight to the fighting! With those changes, daily playtests began to have a running theme emerge: a constant 'struggle for power'.

Two problematic design issues became evident during prototype playtests that I want to zoom in on:

1) Snowballing

Ascension initially felt very snowbally, and while we want to reward #bigplays, it doesn’t feel great to make a single mistake and never be able to recover. We tested and eventually kept the following 'anti-snowball mechanic' in Ascension; all players on both teams had exactly the same gold & XP at all points throughout a match. This meant that at any point in time, you always had an equal chance to win/lose any teamfights. There was no mathematical 'snowballing' possible in Ascension, making the Ascended buff the only actual statistical advantage on the map. Even if the other team was stomping, you had as much chance to win the next Xerath fight, Ascend, and make a comeback. A side-effect was that even when losing badly, you could still get valuable skirmishing practice right up to the end of a game, due to never actually being 'behind'. We were pretty happy with how this solution played out across Ascension's lifetime.

2) Lack of 'strategic choice'.

There was also an early lack of 'strategic choice' in the mode. Original prototype playtests felt too sparse and barren for a TDM, which led to us shrinking the map into the tighter, more action-packed jungle area of the Crystal Scar. This increased the action, but still felt like a directionless TDM. Our team internally dubbed it "Murderball Simulator 2014". The prevailing meta was to just group as 5 and roll around the map killing anything that moved. :/ While murderball was fun in short bursts, it was too monotonous, and we'd tire of it after only a few games. There were no peaks or valleys in the tension curve across the game life. It needed a higher objective layer to float above the TDM, giving purpose to your rampage (rather than killing everything in sight being a means to its own end). We eventually settled on the Relics of Shurima & Ancient Ascendant Xerath combo that shipped, but here's some of the ideas that didn't make it:

• Transferring the Ascended buff player-to-player based on the killing blow. This was one of the earliest prototypes and just led to a snowbally comp style with monotonous gameplay. Badtimes. :(

• Capturable towers that provided an area of vision around them while held. These were basically ignored as the vision wasn't valuable enough to directly affect the scoreboard. "Hmm, I could grab some vision... or maul someone's face off with Udyr... face mauling please!"

• Transforming the 4 remaining teammates on an Ascended team into 'sand soldiers'. We wrote a whole new kit for these 'sand soldiers' based around helping the Ascended track, hunt and lock down kills. Turns out that taking away your ability to play your favourite champion is a terrible idea. The kit tried to walk a middle-line of doing everything, and ended up being good at nothing in particular. "Why am I this sand soldier guy trying to slow guys and shield the Ascended? I could be doing this as Lulu... and WAY BETTER!"