Jung.gg is not owned nor endorsed by Riot Games. The site does not reflect the views or opinions of Riot Games or anyone officially involved in its production. League of Legends and Riot Games are trademarks or registered trademarks of Riot Games, Inc. League of Legends © Riot Games, Inc

Key concepts to understand before using the site

  • Every jungle route is circumstantial. Although the #1 recommended route is a good start, it won't be the best approach for you every game!
  • The site isn't designed to revolutionize the jungle. A lot of junglers do a full clear? You know that. It's the small details that matter, like whether starting Blue->Gromp is more common than Red->Krugs or whether you should skip Wolves or Raptors.
  • Riot don't provide detailed jungle data, the site uses a probabilistic method to work out the most likely route based on the data. This works when you analyze 100s of games, but you may occasionally see an odd route with a low played champion.
  • The above reason is why we don't have all junglers, high elo players tend to play "meta" champions and therefore routes may be limited to these.
  • It's also the reason we can't provide win rates as the level of uncertainty is too high.

Every jungle route is circumstantial

If you looked at a site dedicated to the highest win rate builds and searched for Rammus you will probably find something like this: Ninja Tabi, Thorn Mail, Dead Man's Plate & Randuin's Omen. It will probably suggest to have x2 armor runes too. This is because Rammus, when playing a majority Attack Damage team, will have a high win rate given his kit making him unkillable whilst deflecting damage. Does this mean that if you went against a full Ability Power team you would build the same? Of course not!

You should approach Jung.GG with the same mindset. The routes are the most common for that champion, but certaintly not the most effective at all times. The best example I can provide is the following: At the time of writing this post, Xin Zhao's #1 route was his own side Red, Blue and Gromp camp then enemy Red, Krugs or Raptors. This is incredibly effective against an enemy jungler who is hard farming from their Blue side (Blue, Gromp, Wolves, Raptors, Red). Does this mean that if you went against a Lee Sin you would do the same? Of course not! Why? Because there's a good chance he started Red, and even if he didn't you'll probably run into him there and it's not certain you'll win the 1v1. The routes are intended as directional, giving you the slightly edge, not as "Do this and win".

The site isn't designed to revolutionize the jungle

A response I have heard occasionally about the site is: "Obviously I'm meant to clear my jungle camps? This site is for newbies.". This is both true, and untrue. The site works well for new players as it provides the very basics of the jungle. If you're new you might not even know where to start! Jung.GG should help. However, the site is primarily designed for junglers who want to take their game to the next level. By analyzing thousands of high elo games we are able to extract the tiny and often subtle differences that make their jungle clear 10% more effective. This might be whether to start Red or Blue, whether to skip Gromp or Wolves or which 2/3 camps should you take before invading. This by no means is meant to turn you from Gold to Diamond overnight, but instead provide you with some information to get your win rate from 50% to 55% and get you to the next level.

Probabilistic Methods

Riot does not provide accurate jungle data, that is why you haven't seen any sites like this before! We use a complex probabilistic method that takes all the information at a given time (1 minute intervals) and works out the most likely path taken so far, we then re-run this at the next minute and see whether what we assumed was happening is still the most likely. If you asked the model to review your last game it could provide guesses of your jungle path, but these will not always be correct. However, if you ran that same path over and over again the model will predict the path correctly more times than it doesn't. With the Law of Large Numbers (i.e. if we run 10,000s of games each patch) the path deemed most common is usually right. This does mean that you will see some occasional anomalies, especially with champions with low play rates.

Where's [XXXXX] champ?

If the model hasn't seen enough games to be confident of it's guess, we won't provide the data. Better to be upfront and say we're not sure then give you a guess. If you saw that Urgot jungle was meant to do Raptors>Blue>Krugs>invade enemy Wolves you would rightfully question it, and this will impact how much our users trust all other champions.

But which route has the highest win rate?

Riot has done an incredible job at creating a game in which most champions/strategies/builds sit around a 50% win rate. Due to this and our probabilistic models, we're unconfident that a win rate of 45% is actually 55% and this makes all the difference. If the game was severely unbalanced and there were routes that led to a 80% win rate and others that were 20%, we'd be happy to provide this information and say the win rate for one route is somewhere between 70-90% and another is 10-30% - but it isn't and there's no value in saying all routes have a win rate "somewhere between 40-60%".

Why are your Champion stats different to U.GG / Champion.GG?

Frankly, it's because we analyse less games. Given the complexity of finding the data on jungle paths we don't have the compute power to run hundreds of thousands of games a week. We keep the data up as it's usually in the right ballpark, and we provide some niche data on the percentage of games the jungler takes the first dragon but if you are looking for "Which Champion has the highest Win Rate?!" I suggest you use U.GG instead.