Lane efficiency is a new statistic that measures how well teams manage minion waves throughout the game, both in terms of maximizing their own farming and preventing their opponents from farming. It is reported as a percentage, with values above 50% showing strong performance and below showing weakness.
Full details are below, including the formula, important context for interpretation, and reasons why I chose this approach instead of some alternatives. To see the lane efficiency statistic in action for the NA LCS and EU LCS, head to the findings article.
The formula for Lane Efficiency is simple. For a given team, the calculation is:
minion kills / (minion kills + opponent minion kills)
Minions only include lane minions, not neutral minions (usually referred to as monsters) that live in the jungle.
This approach yields a percentage that represents the team’s number of lane minion kills as a proportion of all the lane minions killed in the game. For aggregate statistics, like you’ll find in the team stats pages on this site, what you’ll get is an average of the team’s lane efficiency number from each game.
Lane efficiency is a useful statistic because it helps us see how well a team manages minion waves, which is a crucial part of controlling the map as a whole. I’ve reported jungle control (JNG%) in the team stats tables for a long time, because it shows how well teams manage the neutral portions of the map. Lane efficiency is the other core piece of that map control puzzle. (We can also talk about large neutral objectives, like dragons and Baron Nashor, but those are already very well represented in existing statistics.)
Causes of high and low lane efficiency
Teams earn high lane efficiency by playing in all three lanes as much as possible, applying pressure on all parts of the map so that they have a player present to catch as many waves as they can, while forcing their opponents to lose CS to tower shots or damage from their own minions. And of course, lane-efficient teams have to be good at simple last hitting!
Teams with low lane efficiency may spend a lot of time playing on just two lanes. They may pull a player out of lane to move around the map, either openly or through fog of war. For example, a team that is trying to siege a tower may bring four or five players into a single lane to push the wave, chip the tower, and threaten a tower dive. But if that team didn’t get the other two lanes pushing first, they may allow the other team to push the lane into their tower, where they will lose a lot of CS.
Naturally, it’s easy for the team in the lead to move between the lanes and catch all of their lane farm, keeping the waves pushed forward. The team that’s behind may have to spend more time trying to break sieges with flanking movements, or backing off from their towers to avoid initiations or tower dives.
Range between high and low lane efficiency values
Compared to jungle control, there will be less range between high and low lane efficiency numbers. Jungle control numbers often range between 45% and 55%, while lane efficiency is usually within 48% to 52%. You can see this in some actual values within the findings article.
This difference comes from the fact that jungle monsters are neutral, meaning that it’s possible to gain “double value” from stealing monsters from the enemy’s half of the map. It is, in theory, possible to have 100% jungle control. But it is not possible to kill your own team’s minions (unlike Dota, where denying minions is part of the game!). So the only way to get 100% lane efficiency would be for your opponent to avoid ever attacking minions!
For future added value, it’s also possible to break lane efficiency out into specific time periods, say the first 15 minutes, or minutes 15 to 30, to see how effectively teams manage the lanes in the early, mid, or late game. This won’t be reported in the team stats tables, but may be an element of analysis for focused articles of infographics.
Alternative methods and why I chose this approach
One alternative option for this stat was to calculate a simple ratio: minion kills divided by opponent minion kills. This would have been the simplest method, creating a range of scores where anything above 1.0 was good and below 1.0 was bad. In practice, though, I found very little difference in actual values once they’re averaged across several games: most teams fell between 0.98 and 1.02, which is not very compelling analytically.
Percentage difference approach
For more depth and variance, I also could have used percentage difference, a formula I use in the GSPD statistic. The downside of percentage difference is that it’s a bit less intuitive to read. You get a mix of positive and negative numbers, which is intuitive enough (positive is good! negative is bad!), but the actual meaning of those numbers is a bit more obscure. What does “+5.1% lane efficiency” mean, exactly? To understand the actual number, beyond just understanding its direction and magnitude, you have to reverse engineer the actual formula, which means searching back to this introduction article… I don’t think this is a complicated enough statistic to warrant that runaround.
Relative efficiency vs. absolute efficiency
When this idea was first forming, I considered making lane efficiency an absolute measure rather than a relative measure. By that, I mean that I could have created a formula that determines how many lane minions spawned during the game, then compared the team’s lane CS to that perfect number to get a percentage. This would have been a bit more complicated to set up, but it would have generated a pretty interesting little measurement of how close teams get to “perfect” lane management.
The problem is that League of Legends is not played in a vacuum. This isn’t a solitaire game where you sit in your lane and simply maximize your CS, especially once you reach the mid and late game. No, this is a game of battling against your opponents and trying to do better than them. Sometimes it’s the right call to sacrifice your minions in one lane while you get something more important done elsewhere. It’s about both maximizing your own farm and trying to minimize your opponents’ farm. This is the same logic for why we report CS difference at 10 minutes as a core statistic, and only look at raw CS at 10 minutes (and compare it to “perfect” CS) as a context piece.
Go forth and analyze!
If you haven’t already, check out the findings article and see this statistic in action. You’ll find the stat appearing soon in the team stats tables, as well, and it I hope to eventually add it in retroactively for past seasons.
For questions or comments, leave a comment or reach out on Twitter.
Photo courtesy flickr.com/lolesports