Popular duo bottom lanes in each league
Patch 5.5
tl;dr Give up thresh if you are trying to climb out of lower leagues
Today's dataset: 500,326 ranked NA solo-queue matches played on patches 5.1, 5.2, 5.3, 5.4, and 5.5. Oh yes, we broke 0.5M.
With this much data, spanning across every league, the possibilities really open up for all kinds of fun slicing and analysis. Today I grouped by league and grabbed the most popular duo-bottom pairings in each. Then I sorted them by how frequently they were played, and captured the win-rate for each (all within that league). By poking around with this data we can potentially gain some league-specific insight into what works down bot.
The dataset has 66,015 matches played at bronze elo (MMR). Here is the table of bot-lane pairs for that league:
| Champions | pick rate | win rate |
|---|---|---|
| 4.13% | 56.57% | |
| 3.46% | 51.05% | |
| 3.22% | 54.94% | |
| 2.91% | 54.53% | |
| 2.68% | 55.84% | |
| 2.68% | 54.12% | |
| 2.45% | 54.24% | |
| 2.04% | 46.62% | |
| 2.00% | 46.85% | |
| 1.84% | 53.83% |
Jinx is the undisputed queen of bronze lane marksmen (markswomen?). There is no way to guarantee that if I grabbed ten times as much data this particular ordering would stay the same, but I do like seeing Jinx/Leona having the highest win-rate on this table. Of all these supports, I think Leona is the easiest to play, and that will be strong among less mechanically gifted players who are still learning the game. Categorically, Thresh's win rates here are lower than Sona/Leona. It may be best to leave him alone if you are trying to climb out of bronze.
The dataset has 78,548 matches played by silver elo (MMR) players. Here are their favorites in the bot lane:
| Champions | pick rate | win rate |
|---|---|---|
| 4.39% | 50.39% | |
| 4.26% | 55.23% | |
| 3.96% | 54.53% | |
| 3.48% | 56.91% | |
| 3.46% | 47.88% | |
| 2.97% | 49.01% | |
| 2.85% | 51.07% | |
| 2.80% | 53.39% | |
| 2.77% | 52.27% | |
| 2.54% | 46.22% |
Sona is gone, and Graves is picking up steam. Thresh's win rates are still low, and the highest win-rate on this table is now Jinx/Blitzcrank. I can confirm, having played many games in silver league, that Blitzcrank is a complete monster. This is some kind of butter zone where the Blitzcrank player's aim far outstrips their opponents ability to dodge hooks. Leona is still winning somewhat consistently, and may be a good choice for carrying out of silver.
In the data there are 72,848 matches played at the gold elo (MMR) level. These are their popular picks in the bottom lane:
| Champions | pick rate | win rate |
|---|---|---|
| 5.14% | 52.24% | |
| 5.12% | 47.86% | |
| 4.68% | 51.98% | |
| 3.75% | 53.75% | |
| 3.56% | 55.99% | |
| 3.41% | 46.37% | |
| 3.26% | 52.04% | |
| 3.26% | 56.28% | |
| 3.03% | 55.89% | |
| 3.00% | 55.58% |
A wild Janna appears. It is interesting to see the pick-rate numbers going up. There are fewer pairings picked overall; players are consolidating their picks more in gold league than they were in bronze or silver. Thresh is still having trouble winning when compared with the other popular picks on the table. He is super-fun to play (and I am a huge fan of Aphromoo), but in season 5 it looks like Thresh may not be the best pick for climbing. In gold, Graves and Jinx are the most popular ADs, and it looks like rightfully so, given the win-rates on this table.
I have 123,242 platinum games in my database. Here are the most popular bottom lane duos from those games:
| Champions | pick rate | win rate |
|---|---|---|
| 6.30% | 47.96% | |
| 5.41% | 51.79% | |
| 4.73% | 55.38% | |
| 4.73% | 52.41% | |
| 4.73% | 51.17% | |
| 4.07% | 55.26% | |
| 3.49% | 50.56% | |
| 3.37% | 46.48% | |
| 3.36% | 49.48% | |
| 3.12% | 58.03% |
I love these per-league analyses because the gradient as one climbs is so surprisingly smooth. Graves' popularity is on the rise, but Jinx is still having higher win-rates on this table. These data suggest that Jinx is the best partner for Thresh as one climbs through gold and platinum, but Blitzcrank and Janna are still outperforming.
Diamond league is best represented in my dataset, with 146,440 games. This happens because I am searching around at the top of the ladder for games played by master and challenger players, but most games that are played at that level seem to have the median player in diamond league. Anyway, here are the popular bot lanes from diamond:
| Champions | pick rate | win rate |
|---|---|---|
| 7.74% | 51.79% | |
| 7.33% | 48.40% | |
| 5.51% | 54.26% | |
| 4.33% | 55.42% | |
| 4.14% | 50.75% | |
| 3.97% | 52.61% | |
| 3.84% | 52.06% | |
| 3.44% | 48.06% | |
| 3.14% | 55.98% | |
| 3.00% | 45.68% |
There is Janna/Graves, and the pick-rates are getting quite high now when compared with the lower leagues. The meta really firms itself up at these higher elos. I like how Sivir pokes her head in now for the firs time, and with a respectable win rate. Jinx/Janna and Sivir/Janna have the highest win rates on this table. High skill players fight much more intentionally, and can make better use of Janna's ability to disengage, kite, and shape fights.
Given how the league system is designed, there will necessarily be fewer games played where the median player is master or challenger elo (MMR). So, the dataset has only 8,356 master level games. This data is bound to be much less significant than the already questionably significant data above. But, it is still fun to look at:
| Champions | pick rate | win rate |
|---|---|---|
| 5.14% | 52.24% | |
| 5.12% | 47.86% | |
| 4.68% | 51.98% | |
| 3.75% | 53.75% | |
| 3.56% | 55.99% | |
| 3.41% | 46.37% | |
| 3.26% | 52.04% | |
| 3.26% | 56.28% | |
| 3.03% | 55.89% | |
| 3.00% | 55.58% |
Take from that what you will. Ha!
The situation is similar, but worse, for data from the challenger league. I have 4,877 matches that had mostly challenger players in them. It is possible that this is actually about all the data there is for NA at this level for season 5, but given the design of Riot's data api it is hard to say.
Regardless, here are their bot-lane choices:
| Champions | pick rate | win rate |
|---|---|---|
| 9.29% | 49.67% | |
| 6.91% | 47.48% | |
| 5.39% | 49.81% | |
| 4.39% | 55.14% | |
| 3.38% | 53.94% | |
| 3.38% | 46.67% | |
| 3.38% | 55.15% | |
| 3.34% | 63.80% | |
| 3.34% | 50.92% | |
| 3.28% | 57.50% |
Hello Kalista/Thresh win rate. Hello Janna/Graves pick rate. Sivir also appears to be doing super-well on this table. The crust of NA solo-queue is a wild and crazy place.
Once again, an article ends. We made it to the bottom together, and I could not be happier. I genuinely appreciate your time, and I am glad that you find these articles valuable.
I learned a lot here, and I hope you did too. Did I miss something obvious in the analysis? Tweet me as hard as you can so we can discuss it. Get at me too if you have any ideas for analysis you would like to see. What would you do with a half million games?
Now it is time to go back to the homepage and read some more, or just click the new 'lover' link in the sidebar and think romantic thoughts. Be sure to subscribe to the rss feed so you can can be the first on your block to read the next post.
In any event, keep doing league and keep doing math and you and LeagueMath.com will be best friends forever.
No ram slots were harmed in the installation of the memory I bought to chew through the data for this article.
Peace.
