How to Get Better at League of Legends: 6 Habits That Work

To get better at League of Legends, work on one thing at a time. Pick one goal you control, such as your CS at 10 minutes or no deaths before 10:00, turn it into an if-then plan, and check after every game whether you did it. Keep your champion pool small, and review one moment of each game instead of the whole replay.

Getting better, and climbing with it, is less about playing more games and more about what you do differently in each one. That sounds obvious, but the data says most players never get there. This guide turns the best evidence we could find into six habits you can start in your next game, with a sample week at the end.

Why most players stall

A study of 242,000 League matches found “no significant long-term improvement in the individual performance of most players.” A preregistered study of 411 players found no meaningful effect of what they reported as deliberate practice, while plain playtime did predict long-term success. Both point the same way: the hours count, but only when each game teaches you something specific. League throws so much information at you that, without a focus, the lesson of a game is usually “my team was bad”.

So the goal of the habits below is simple: make every game a small, measurable test of one thing you control.

1. Set one process goal, not a rank goal

“Reach Platinum” and “win 55% of my games” are outcome goals. You don't control them within one game, so they tell you nothing about what to do next. A process goal is something you do: “no deaths before 10:00 that give the enemy nothing for free”, “CS at 10 minutes of 70 or more”, “a control ward on every back”.

In a meta-analysis of goals in sport, process goals had an effect of d = 1.36 on performance, against 0.44 for performance goals and just 0.09 for outcome goals. Process goals also raised confidence. One goal at a time is the point: working memory holds three to five chunks, and a game of League already fills most of them.

Good first goals by role:

• Top and mid: zero deaths before 10:00 when the enemy jungler hasn't been seen.
• Jungle: be on the side of the next dragon or grubs with full health a minute before it spawns.
• ADC: a CS at 10 minutes that beats your own average by five.
• Support: buy a control ward every time you go back.

2. Turn the goal into an if-then plan

An if-then plan (“if situation X, then I do Y”) links the goal to the moment it matters. Across 94 tests, plans in this form had an average effect of d = 0.65 on reaching goals. They work because the decision is made before the game, when you are calm, instead of mid-fight.

• “If I have 1,100 gold and my wave is pushed, then I recall.”
• “If I have no ward and the enemy jungler isn't visible, then I don't walk past the middle of the lane.”
• “If I have to choose between a trade and a last hit, then I take the last hit.”

Read the plan once in champ select, when you can already see your lane opponent. That rehearsal is part of why it works.

3. Track it after every game

Monitoring your progress raised the odds of reaching a goal with an effect of d = 0.40 over 138 randomized trials, and more so when the progress was physically recorded. After each game, answer one question: did I follow the plan? Yes, partly or no. If no, write one word for why: gank, trade, tilt, other. Thirty seconds, every game. After fifteen to twenty games you have a real answer instead of a feeling.

A warning about the numbers: if you pick your weakest stat, it will often improve a little on its own, simply because a bad stretch tends to be followed by a more normal one. Compare your goal with a second stat you didn't train. If both moved the same amount, the plan didn't do it.

4. Keep your champion pool small

Experience on a champion pays, and it pays most in the jungle and least for ADCs. By iTero's analysis, the break-even point is around 20 games on a champion. League of Graphs finds that players with 50 or more games on a champion win about three points above that champion's average. Every new champion you take into ranked starts you below that line.

Two or three champions per role are enough. If your first picks are banned, take the next one from your pool, not a champion you have never played. That rule is itself a fine if-then plan for players who first-time a lot.

5. Win the first ten minutes against your laner, not the first blood

Games are surprisingly open early. A model tested on 64,556 ranked games predicted the winner with an accuracy of 0.62 around minute 6, 0.70 around minute 12, 0.78 around minute 18 and 0.86 around minute 24. A lost skirmish at minute eight is not a verdict, and playing as if it were is how leads get thrown and deficits get worse.

What does predict wins early is the difference with the other team: in a 2024 analysis of ten-minute data, the gold and experience differences stood far above everything else. Single events weigh little. First blood correlates with winning at only r ≈ 0.14, and the team that gets it wins only about 57% of games. So don't force the first kill; protect your farm, your experience and your life.

Not every death is equal. Analyst Philip Maymin separated “worthless” deaths, the ones that gave your team nothing, from deaths that bought an objective or a teammate's escape, and found the worthless ones among the strongest drags on winning. Cutting them is the most direct goal most players can set.

6. Review one moment, not the whole game

Watching an entire replay after every game is a lot of work, and most players stop after a week. Look at one or two moments instead: your deaths before 14 minutes, or the moment your goal broke. Players who reviewed their games as a timeline of actions improved more than those who only saw summary dashboards, in a 2024 study on game review.

Keep the review about the task, not about yourself. Across 607 effects in a classic meta-analysis, more than a third of feedback interventions made performance worse, mostly when the feedback pointed at the person (“I was the worst on my team”) instead of the task (“I died at 7:42 with 1,300 gold unspent”).

And set a stop rule before you queue. In play sessions of three or more matches, win rates drop by more than 10% between the first and the last game. Our guide on how to stop tilting covers how to build a session rule that you will actually follow.

A week of climbing, in practice

1. Day 1: look at your last 20 games in your main role. Pick one process goal and write it as an if-then plan.
2. Every game: read the plan in champ select, play, then answer yes, partly or no in thirty seconds.
3. Every session: stop at the number of games you set before you started, win or lose.
4. After a loss: look at one moment, your first death or the moment the plan broke. Nothing else.
5. After 15 to 20 games: compare the goal with your starting point and with a stat you didn't train. Keep it, adjust it or move on.

That is the whole system. It is slower than a new tier list pick, but it is the part of your rank that you control. If you want help with the bookkeeping, LazyElo's Focus page does exactly this loop: one goal per role, the if-then plan in front of you before each game, a check after it, and a verdict after fifteen games. The tier lists and scout help with the rest, and the CS per minute and vision score guides give you two good first goals.