I wrote previously about the effects of turret plating on the professional metagame in League of Legends. It’s been obvious that turret plates have shaped the priorities of pro teams as they try to get their hands on as much of that gold as possible before the plates disappear at 14:00. One of the key ways to pick up that gold is to take the Rift Herald and use it before the 14:00 mark. This has led to increased focus on killing Rift Herald, and especially killing it early.
In 2018 across LCS, LEC, and LCK, Rift Herald was killed in 89% of games. In 2019 to date, Rift Herald is being killed in 98% of games in those leagues, a dramatic increase.
The perceived value of Rift Herald has increased, but exactly how valuable has the early Herald kill become, and how is that playing out in leagues’ and teams’ priority levels?
How valuable is Rift Herald?
A Herald charge onto an outer tower is guaranteed to break two turret plates (worth 160 gold each), and it can lead to even more plates if the accompanying push isn’t well defended by the opposing team. Factor in the time the team gains in taking down the rest of the tower’s health in later pushes, and the Rift Herald’s value seems pretty high.
But you rarely get something for nothing. Objectives are frequently traded for something else cross-map. Anecdotally, the most common trade for Herald is a dragon.
According to the latest early-game rating (EGR) model, 320 gold at 15:00 is worth ~4.0 percentage points of win probability. Comparatively, holding one more dragon at 15:00 is worth ~7.5 points of win% — almost double!
If a team is looking to take the Herald pre-14:00 in trade for a dragon, they need to strongly believe either that a) the Herald is going to generate much more than two plates’ worth of value, or b) the dragon is less valuable than average due to its type and/or the two teams’ compositions and win conditions.
For example, if my team’s win conditions revolve more around early gold generation, controlling the map through aggressive rotations, and snowballing, then I may place higher value on using Herald to crack a tower and kick start my rotations, and less on stacking dragon buffs to gain more mid and late game power.
If a team fight breaks out around the Herald pit and the team that took the Herald gave up one extra kill (worth +/-450 kill gold depending on bounties), that should not be considered a win unless the resulting Herald usage highly outperforms the base expectations. If, on the other hand, the Herald was being used as a pressure point to bait the opponent into a fight, and the team can actually win the fight at Herald, then of course that’s a massive win!
TL;DR – Trading Herald for a dragon or a kill could be situationally worthwhile, but under default/average conditions the dragon or the kill will each be worth more.
Who values Herald the most?
Korea has always been heralded (LOW-HANGING-FRUIT PUNS!) for its macro play, its clear understanding of setting priorities on the map, working around objectives, minimizing risk, and controlling the flow of the game. It’s unsurprising, then, that the LCK has led the way in taking early Rift Heralds and using them to maximize gold from turret plates.
(Notes: The data below reflects games since the start of 2019 in LCS, LEC, LCK, and in some cases MSI. Data from LPL is not available, and LMS is not included due to some inconsistencies in data availability.)
Here are a few things that stand out to me from this data.
- MSI was a very fast-paced tournament, and that’s reflected in the Herald timings. Heralds were killed and used earlier at MSI than they have been in LCS, LEC, or LCK.
- LCS teams have been slowest to take Heralds and use them.
- Since LCK teams kill the Herald at least half a minute earlier than LCS or LEC teams on average, they have more time to set up a good usage. Average time between kill and usage is higher in LCK than in the other leagues, suggesting that the Koreans are more intentional about maximizing their Herald drops.
It would be great to have access to more granular data on gold, because a logical next step here would be to measure how much gold was generated within the 30 or 60 seconds after a Herald is used. I could count the number of times Herald leads to a tower being killed, but that would misrepresent the situation, because Heralds are so often used to break turret plates without actually finishing off the tower completely. In future, hopefully more data will become available to extend this analysis.
Herald Usage Per Team
Rift Herald timings are broken out for each individual team in the tables below.
Up front, I want to point out that some of the best teams in each league put noticeably lower priority on taking early Rift Heralds. Look at Team Liquid, G2 Esports, and Griffin. Each of these had the latest average Herald kill time in their league. That doesn’t mean that those teams don’t value Herald in general, just that they don’t go out of their way to take it before 14:00. (All three teams actually rank reasonably high in overall Herald control.) A deeper analysis into these teams could be very interesting, to see what they are doing in that 10:00 to 14:00 window when other teams seem to be working around Herald.
Some other notes:
- Setting aside the LCK Challenger teams, there are actually two LCS teams tied for “earliest average Herald kill”: CLG and Echo Fox share that honour with Hanwha Life Esports.
- Fast turnaround times between killing and using Herald are certainly not the hallmark of a successful team. The shortest time gaps belong to teams like CLG, Echo Fox, Excel, Afreeca, and Origen. These are not all bad teams, of course, but on average they aren’t the cream of the crop, either. Decisive action can be a positive attribute, but using Herald too quickly might also come from being too frantic, or desperately trying to force a Herald charge before 14:00 rather than being disciplined and intentional.
- Echo Fox are the only team in these three leagues to have two Heralds expire on them. :( As a note, I’m not 100% sure what shows up in the data if a player is interrupted while Herald is channelling, but I believe that this still gets counted as the eye being popped.