Let's get Tilted.
For this analysis, I creeped on a few thousand randomly selected players that had at least 100 games this season, and analyzed the affect of tilt on their performance. To do this, I broke up their games into different “sessions,” aka, a series of games they played in succession without taking a break of more than one hour. To familiarize ourselves with this data, lets take a look at win rate by game in session:
Dope. Although the effect is pretty subtle, we can see that as someone warms up, their performance improves up until game 3, then they begin to lose focus. After game 3, people appear to play more passively (place fewer wards and such). Couldn't say definitively why, but my best guess here is that after a couple hours you just start going through the motions?
Now lets look at tilt!
For a first pass, I took players that had an overall win rate of 50%, and looked at how their win rate is affected by their performance in their last 3 games:
Here we see that if you lost your last 3 games, your next game typically has a win rate of 47%. Sounds pretty tilted to me. Conversely, pop off and win your last 3 and you have a 53% chance of winning your next as well. Ok, so if you are tilted, you're probably gonna play poorly. But this begs the question, what can be done at this point? Uninstall? Just que up and hope you get carried? Well, they say time heals all. Let’s see if it even heals tilt. Below I analyze the affect of taking a step back and waiting at least 10 minutes before your next game:
* This data is only using players that had a win rate between 49% and 51% in their last 100 games.
Would ya look at that? If you lost your last game, taking at least a ten minute break drastically improves your chances of winning.
Also interesting here, we can see that if you take a break, you actually have a higher chance of winning your next game after a loss than a win. This is the matchmaking system at work, making your competition harder / easier after each game.
Ok, so with this new learning in mind, lets check that same 3 game chart we saw above, but limit to games where all 3 games happened in one session, aka no breaks allowed:
Yup, without breaks the effect of the tilt (and poping off) is much stronger, with 3 losses in rapid succession leading to an abysmal 43% win rate.
So to wrap this up, we’ve learned that if you are losing, you should take a break, but we also saw that your first game in a session typically has a lower win rate as you warm up. Is there an optimal amount of time to take between games to break your tilt, without losing your warm up? Yes there is:
Here's what I'm seeing:
If you just won your last game, taking a break is actually a bad thing. Keep that momentum going yo.
If you lost, the sweet spot is a break between 20-30 minutes. Refresh, go frolic through a forest, but don’t be gone so long you lose the valuable warmup you just received from your previous games.
So in conclusion, tilt definitely exists but it can be defeated just by taking a 20 minute break! Goodluck out there on the rift!
Final Personal Note on this:
I get pretty tilted, so even though I never saw the data until now, I realized a while ago that I need to take a break after losing if I want to play optimally. I don't mean to be preachy here or anything, but for me, I decided after every loss I'd watch a 23 minute online course on the field I dreamed of getting into, Data Science. This system provided me motivation to watch those videos that I was unable to muster up myself, and because of this little habit my life has been dramatically changed.
Let's be honest, if you're like me, League can certainly be a detriment to other aspects of our lives. But it doesn't have to be. For me, I found systemizing my league habits turned my love of the game into a powerful weapon for developing new skills. I'm sure its not for everyone, but wether to learn something new, or just to play better league, I'dd encourage you to try actively taking breaks after a loss!
Anyways, as always thank you guys so much for being here, and I'd love to hear any feedback you have! See you on the reddit thread :D
Oh, and if you liked this analysis, consider subscribing at the bottom of the page! I promise no spam, just each week's analysis :)