Going “back to the basics” is the oldest coaching cliché in the book. But there’s a reason for that—basic skills are the foundation for success. In League of Legends, no skill is more basic than farming.
Killing minions, the little enemies that spawn around the map, gives you gold. Gold allows you to buy better items. Better items gives you the power to score kills, push towers, and win games.
A multiplayer online battle arena like League is about killing other players, but often you can’t do that unless you build an advantage first, and the best way to do that is some time alone in a lane with minions.
The marksman or ADC role is often the most farm-heavy in the game. An entire player—the support—is assigned to assisting the marksman boost his farm total. Bottom lane carries are tasked with building into four, five, or six item monsters.
We took a look at marksman in the bottom lane of the League Championship Series, analyzing their farm numbers, their foe’s farm numbers, and delving into just why they worked out that way.
The most basic farming statistic is Creep Score per Minute (CSPM), a measure of just how fast a player kills minions.
It’s not an ideal metric, as many factors can affect each players’ numbers. For example, certain champions are favorable farmers. Lucian was so popular this season in large part due to his ability to exit lane with a CS advantage. Teams also use different strategies. Evil Geniuses, for example, feeds late game farm into their marksman Johnny “Altec” Ru so he can carry. LMQ gives that farm to their mid laner instead.
Rookie marksman Robert Lee of Complexity led all marksman in CSPM with 9.17 this split, but that’s partially because he favored champions which traditionally dominate the lane, with 10 games on Lucian, seven on Corki, and two on Caitlyn. Even so, the average ADC would put up 5.43 CSPM, meaning Lee was still a farming machine.
In Europe, the league MVP Martin “Rekkles” Larsson put up another big year with 8.86 CSPM, far outsripping his local competition.
Other top performers include another rookie, Johnny “Altec” Ru, and the benched Gambit Gaming marskman Evgeny “Genja” Andryushin. Counter Logic Gaming marksman Peter “DoubleLift” Peng put up another solid season.
But not all CS numbers are created equal. A better way to measure performance is in the context of the lane. Farming isn’t a solo activity—you do it in a lane, against an opponent. Some players may not put up huge CSPM, but might have a talent for applying pressure that prevents their foe from farming.
Another way to analyze the data is to look at the difference in CS between lane opponents. We looked at these numbers in five minute intervals and at the end of the game.
This chart tells a slightly different story than the CSPM numbers, adjusting for lane opponent and revealing a little more about the play style of each player and team.
The biggest winner here is LMQ marksman Li “Vasilii” Wei-Jun, who goes from the bottom of the table in terms of CSPM to one of the top performers through 20 minutes. Li sacrifices last hits in order to pressure the enemy from freely farming, allowing himself to build a lead despite a lower rate—Li actually is the fifth lowest in CS at 20 minutes in the NA region, but he suppressed opposing CS better than anyone outside Cloud9’s Zachary “Sneaky” Scuderi. Li also lead all marksmen in his region at CS differential at 10 minutes, showing he’s a force in lane.
Li is also an oddity. Despite exiting the first 20 minutes with a CS lead, he ends matches with the biggest CS deficit of any marksman in the LCS. That’s probably due to a combination of things: LMQ likes to feed late game farm to their mid laner, and they like to keep their marksman handy for the numerous battles they like to engage in.
The two top farmers in America, LCS rookies Lee and Johnny “Altec” Ru, got their numbers in different ways. As stated above, Lee favors lane bully champions, allowing him to build big CS leads in lane as shown by the above graph. But Ru plays differently. His heavy CS phase is in the late game, when Evil Geniuses feeds Ru to go farm on hyper carry champions like Twitch, Tristana, and Kog’maw.
Team SoloMid is known for their strong laning, but Jason “WildTurtle” Tran has taken flak this season for underperforming in that regard. His CS numbers show it—he ranks last out of all marksman in CS at every interval through 20 minutes. While his overall CSPM of 8.25 is in the middle of the pack, that isn’t because he’s starting the game strong. Tran gave up more CS to his opposing laner at 20 minutes than any marksman in the LCS this split. Despite killing more minions at 20 minutes than all but two other marksman, Tran was behind in his lane most games.
In Europe, the Summer MVP Martin “Rekkles” Larsson was a farming machine. No player killed more minions in lane than him, and it wasn’t really close—his 179.25 average CS at 20 minutes beat Lee’s 178.68 and obliterated the rest of the pack, with Tran’s 173.43 in third place.
Erik “TabzZ” van Helvert, the star marksman of Alliance, led his league by doing things a bit differently than Larsson. He gave up 300 less minions in lane, so despite less kills overall he held a bigger lead over his foes.
Some surprises include Gambit Gaming’s now benched marksman Evgeny “Genja” Andryushin, who actually posted solid lane numbers and overall numbers. But they were also influenced by his champion picks—you’d expect a player with 21 games on Caitlyn, Lucian, and Corki combined to win his lane on average, putting up 4.96 CSPM in those games. That means Andryushin’s slightly above average CS numbers are actually below the average level of marksman play, based off his champion picks.
Former Curse Gaming marksman David “Cop” Roberson put up the best CSPM of his four season career at 8.20, but his laning numbers reveal one reason why Curse thought it necessary to make a change. Roberson was down by seven CS at 20 minutes on average, only above Tran and ROCCAT marksman “Celaver” in all the LCS last season. Even worse, more than half his games were on Corki, Lucian, and Caitlyn, champions that on average dominate the lane.
The reason? His famous passivity. Roberson has a reputation for playing extremely passive and safe, and that’s obvious in his CS numbers. Roberson put up 172.61 CS at the 20 minute mark, a respectable number—it’s better than Peter “DoubleLift” Peng’s 172.38 CS, for example. But while Peng was in the top tier at denying enemy CS, only allowing 167.04 CS at 20 minutes against him, Roberson was nearly the worst player in the entire LCS in that category. The average marksman put up 179.61 CS at 20 minutes against Roberson.
The champion graph shows two clear tiers of champions: the lane bullies and the late game hyper carries.
Lucian, Caitlyn, and Corki all dominated lanes thanks to their range and ability to harass the enemy carry, building up significant CS leads by the 20 minute mark.
Tristana, Kog’Maw, and Twitch, on the other hand, had a tougher time in lane, but make up for it with their incredible late game scaling.
Other popular champions not included on the chart include Jinx and Graves. In 10 Grave games, he went even in his lane. In 25 games, Jinx fell behind by 11.2 CS at the 20 minute mark. Late game, however, Jinx was a monster, with a 14.88 CS lead by the end of the match, only beaten by Corki and Lucian.
This season was a banner year for farming minions. The league saw a 5-10 percent increase in farming from the marksman role over last year. That’s in part due to metagame changes, but a large portion of it is simply due to improved player skill.
That’s evident in the numbers from North America. The Spring split average, 7.87 CSPM, jumped up to 8.40 CSPM in Summer. Why? Two rookie marksmen, Johnny “Altec” Ru and Robert “ROBERTXLEE” Lee. Both players put up huge CSPM numbers, outpacing the league by a wide margin. Just compare Ru’s 8.71 CSPM to “Yellowpete”’s 7.37 CSPM, or Lee’s 9.17 to Apollo “WizFujiiN” Price’s 7.32 CSPM.
Li also contributed. While the LMQ gunner’s 7.85 CSPM was at the bottom of the barrel, it still topped the 7.42 combined mark of Christopher “Zuna” Buechter’s 6.84 CSPM and Jake “Xmithi” Puchero’s 8.00 CSPM. Buechter is the only marksman in the past two years of LCS play to farm under 7 CSPM.
In Europe the story was different. Konstantinos “FORG1VEN” Tzortziou set the LCS record with 9.38 CSPM in the Spring split. Replacement Paweł “Woolite” Pruski did an admirable job replacing him, with 8.37 CSPM, but that’s not close to the king.
Farming is one of the most basic skills in League of Legends, and it’s a pillar of the ADC role. But it’s only one facet of what makes a player great, and this study is just scratching the surface of possible analysis.
The most obvious area of further study is supports. Some support players are better laners, allowing their marksman to freely farm. Some teams pick supports with strong laning capability, while others may favor supports with better team fighting. Tran’s shaky season, for example, may be because of his now-replaced rookie support Nicolas “Gleeb” Haddad. Or at least SoloMid thought so—they brought in Ham “Lustboy” Yung-sik to replace Haddad. But in the regular season, Tran was actually worse in lane with a new partner.
Another thing to look at is jungle pressure. The bottom lane is a hotbed for ganks, and some junglers focus it more than others. A carry’s CS numbers may be propped up by extra jungle attention over his own laning ability.
CSPM increases as a function of time. The average LCS match is just under 40 minutes long, but some teams end up playing longer than others. However, glancing at the numbers for this split makes it hard to find a trend. Plus this only affects the CSPM numbers—not the lane numbers through 20 minutes.
It’s possible that winning and losing has an effect on CSPM. For example, a losing team might benefit from minion waves pushing into their base, allowing them to safely farm as they stall out the late game. Or, they could lose more minions due to them constantly crashing into their tower.
The sample sizes involved make further study hard—the more you break things down, the less games you have to look at and that makes any conclusions dubious at best. Sivir, for example, was the most dominant farming champion in the split, posting a ridiculous 36 CS lead at 20 minutes. While it’s probably true she’s a good farming champion with her superior wave clear, DoubleLift played all three of those games, meaning his skill level alone skews the results.
Illustration by Jason Reed