A Better Dragon System
Dragon has seen a lot of changes over the past six years. Once upon a time, it gave gold and experience to every team member upon kill. Riot decided that this made it too similar to turrets and unable to compete with Baron's 300g (and Hand of Baron) late, and so decided to make the gold scale up to 260g—in order to place more value on late game dragons. However, this was still too similar to turrets for Riot's liking, so they changed it to a stacking buff mechanic called “Dragon Slayer”.
Dragon Slayer persisted for just under 2 years, and largely stayed the same. There was a sequence of five buffs, the first four of which were static boons and the fifth of which temporarily doubled the others and granted your team a true damage burn. With the advent of patch 6.9, Dragon Slayer underwent a radical change for the first time since it was introduced in patch 4.20: there were now Elemental Dragons (equivalent to the old first four buffs) that spawned in a random sequence up until 35 minutes, when an Elder Dragon would take their place (equivalent to the old fifth dragon).
For the sake of having all pertinent information in one place, here are the elemental buffs as they exist currently in 6.9: -Infernal Drake: +8/16/24% AD and AP -Mountain Drake: +10/20/30% damage to epic monsters and towers as true damage -Cloud Drake: +15/30/45 bonus movement speed out of combat -Ocean Drake: Restores 10% missing health and missing mana every 18/12/6 seconds -Elder Dragon: Enhances existing buffs by 50% and grants true damage burn. Lasts 120 seconds or until death.
It's easy to see Riot's aim with the elemental buffs, namely to allow teams to contest and prioritise buffs that cater to their respective play styles. A teamfighting team like RNG would heavily prioritise the Infernal Drake, enabling them to wipe the enemy team with superior stats in their signature late-game teamfights. A laneswapping, split-pushing macro team like CLG might prioritise the Cloud Drake or Mountain Drake for faster rotations and easier tower taking. A team that excels at running poke compositions like Longzhu Gaming might prioritise the Ocean Drake—both to deny sustain to their opponents, and to keep their own mana pools high and ready for skillshot pot-shotting.
Unfortunately, the random nature of the dragons spawning means that dragon priority and contesting feels more like a crap shoot than an actually meaningful macro mechanic. What happens when two teamfighting teams don't get any Infernal Drakes? They're stuck contesting drakes that they don't really want, purely by virtue of not wanting to give the opponent an edge and not because they themselves want the dragon.
Furthermore, the random elemental buffs can also potentially dampen the advantages gained in the draft phase through superior pick-ban. Imagine your team picks a poke/kite composition, meaning you're not very good at contesting early objectives. The enemy team picks a full scaling composition, so in theory you've won the draft: you can group up early and poke and siege your way through their structures before they have the opportunity to scale. Barring a colossal teamfighting blunder or something similar, your team should have rightfully won the game by picking smarter. Now imagine the enemy team manages to grab two Ocean Drakes because your jungler was more concerned with grouping, poking, and sieging. Remember, you have very little hard engage, so you can't contest the dragon very easily. Suddenly, your superior compostion has been destroyed by a completely random element of the game. The enemy team has all the tools to turtle and sustain and farm their way into a late game where they outclass you in every teamfight.
When randomness is applied on a micro level, like dodge or crit chance, it's okay because its impact is permissibly low outside of a few, very particular situations. When randomness is applied on a macro level, it becomes a problem: everyone in the game has to deal with it constantly, regardless of their position on the map or their willingness to partake in the RNG. Random dragon buffs is the equivalent of having jungle camps spawn in random locations.
So what's the solution? Obviously enough, to remove the randomness. The Elemental Dragons are a very good step in the correct direction. In theory, it enables teams to play around their own styles, and reinforces team identities in a way that not only feels good to play as a team, but also is entertaining to watch.
The best way to eliminate randomness is to have the dragons always spawn in a static order, and to have dragons de-spawn after a set amount of time alive (à la Rift Herald). This way, teams can choose to ignore buffs they don't want while strategically denying buffs that the opponent wants. Instead of being forced to run down to kill the dragon in order to deny it, a team can simply apply pressure in the correct areas and keep up good vision control while waiting for the dragon to die on its own. To counterbalance the removed possibility of stacking the same elemental buff, the buffs would all take the middle-ground: Infernal would grant +16% AD and AP, Mountain would grant +20% damage to epic monsters and structures, Cloud would grant 30 MS out of combat, and Ocean would grant 10% missing health/mana regen every 12 seconds.
As for the predetermined order, it should probably be Cloud, then Ocean, then Mountain, then Infernal. That way, the buffs would be most relevant to the window of time in which they are available. An early Infernal Drake is nice, but it doesn't really start to matter until the mid- and late-game. A late Ocean Drake is nice, but really only for tanks and carries who aren't building lifesteal—even then, it's not as valuable as an early Ocean Drake for teams looking for standard lanes and scaling.
There are other small nitpicks with the elemental buffs that could be considered. For example, instead of granting +% stats (and therefore being unequivocally the best buff), the Infernal Drake should grant +% damage against champions to encourage teamfighting. However, those aren't as important as the overarching problem of macro-level randomness.
To recap for those too lazy to read the entire article: random dragons are bad, but elemental dragons are good. Put them in a static order, and everything will be golden sunshine and rainbows.
Link to my blog here for similar articles and analyses.
