1. Whether or not the blue side advantage is affected by players' skills; e.g. perhaps Diamond/Challenger players are less likely to be affected by this advantage, and by extension, in the competitive scene the blue side advantage is minimal.
2. Whether or not this effect depends on the champion you play. I hope the previous post convinces you that the blue side advantage is mostly due to camera angle (since One For All Mirror Mode removes all other factors); so there is a possibility that champions with long range skillshots are affected by blue side advantage more than champions with short range.
My answers to these two questions are:
1. Blue side advantage is affecting high level players (Diamond/Challenger) just as much as low level players (Bronze).
2. There are some differences between champions and how they are affected by the blue side advantage. Short range champions are slightly less likely to be affected by the blue side advantage. Rumble, in particular, is affected by this so much that he wins 57% of the time on blue side and only 41% on purple side.
The data and the analysis:
1. Blue side advantage vs players skill.
For this, I pulled roughly 1.8 million games played in patch 4.6 and 4.7, all of which are ranked. The results are as below:
| Blue Side Win % | Purple Side Win % | Difference | Total Win % | Games Recorded | |
| Challenger | 59.75% | 49.35% | 10.41% | 54.59% | 17299 |
| Diamond | 56.53% | 46.92% | 9.61% | 51.73% | 1035553 |
| Platinum | 55.88% | 46.32% | 9.55% | 51.11% | 2756535 |
| Gold | 55.94% | 45.98% | 9.96% | 50.96% | 5656094 |
| Silver | 55.11% | 44.44% | 10.67% | 49.77% | 7744235 |
| Bronze | 53.10% | 41.27% | 11.83% | 47.18% | 2675229 |
First of all, if we only look at the chance of winning on blue side, it seems that at nearly 60%, Challenger players are even more heavily affected by the blue side advantage. However, a closer examination shows that Challenger players are so good that they win 55% of the time overall, so their blue side win rate is inflated. To counteract this, I take the difference of the blue side win rate and the purple win rate. This gives us a roughly 10% difference in win rate for blue side vs. purple side for all league players. We will be using this number to examine the size of the blue side advantage.
Also, although this chart does not show this, Challenger players have equal chance to play on blue side or on purple side.
Therefore, we see that blue side advantage decreases very slightly as the players' skill increases. However, blue side still wins roughly 10% more often than the purple side - which is roughly the same as saying blue wins 55% of the time.
(Comment: sorry if you are a statistics buff. I chose to not use relative risk or odds ratio because I feel risk difference is far easier to interpret)
2. Blue side advantage vs champion
This time, instead of having one row per league tier, I plotted the table to have one row per champion. I only chose champions that had at least 50,000 games to avoid looking at the potentially misleading numbers that may be produced by champions that are rarely played (yes you, Urgot).
Similar to our analysis before, we are going to look at the difference in win rate between blue and purple side - because some champions are so strong that they win ~55% of the time overall, similar to Challenger players in ranked. I sorted this table so the champions with the greatest blue side advantage are on top, and the ones with least blue side advantage are on the botttom.
| Champion | Blue Side Win % | Purple Side Win % | Difference | Total Win % | Games Recorded |
| Rumble | 57.34% | 41.71% | 15.63% | 49.54% | 53239 |
| Jayce | 54.91% | 43.35% | 11.56% | 49.14% | 90014 |
| Shaco | 56.77% | 45.23% | 11.54% | 50.92% | 93684 |
| Olaf | 49.93% | 38.43% | 11.50% | 44.14% | 59337 |
| Teemo | 57.10% | 45.72% | 11.38% | 51.34% | 136910 |
| Veigar | 52.42% | 41.09% | 11.33% | 46.71% | 57354 |
| Velkoz | 56.24% | 44.97% | 11.27% | 50.62% | 79882 |
| Brand | 57.35% | 46.10% | 11.25% | 51.76% | 76920 |
| Lux | 56.18% | 44.94% | 11.24% | 50.59% | 224541 |
| Malzahar | 58.26% | 47.06% | 11.20% | 52.60% | 86484 |
| Nidalee | 55.58% | 44.42% | 11.16% | 50.09% | 392071 |
| Draven | 56.93% | 45.78% | 11.15% | 51.46% | 186644 |
| Ahri | 54.16% | 43.02% | 11.14% | 48.61% | 119034 |
| Chogath | 52.00% | 40.88% | 11.12% | 46.42% | 68512 |
| Varus | 55.42% | 44.39% | 11.03% | 49.92% | 140731 |
| Lissandra | 52.74% | 41.75% | 10.99% | 47.18% | 54384 |
| Ziggs | 58.70% | 47.76% | 10.94% | 53.26% | 256552 |
| Volibear | 55.23% | 44.30% | 10.93% | 49.72% | 119082 |
| Amumu | 58.87% | 47.95% | 10.93% | 53.32% | 141434 |
| Syndra | 54.74% | 43.84% | 10.91% | 49.30% | 150155 |
| Elise | 50.20% | 39.30% | 10.90% | 44.68% | 103848 |
| Singed | 56.71% | 45.82% | 10.89% | 51.24% | 59682 |
| Talon | 57.47% | 46.59% | 10.88% | 51.96% | 74872 |
| Morgana | 57.71% | 46.84% | 10.87% | 52.23% | 648008 |
| Sivir | 54.84% | 43.98% | 10.86% | 49.26% | 159978 |
| Heimerdinger | 58.39% | 47.57% | 10.82% | 52.95% | 70838 |
| Hecarim | 52.35% | 41.54% | 10.81% | 46.91% | 57400 |
| Jinx | 57.07% | 46.29% | 10.78% | 51.73% | 440826 |
| Blitzcrank | 56.18% | 45.42% | 10.76% | 50.85% | 248811 |
| Aatrox | 55.13% | 44.41% | 10.72% | 49.69% | 74615 |
| Karma | 53.86% | 43.14% | 10.71% | 48.51% | 239518 |
| Caitlyn | 55.53% | 44.84% | 10.69% | 50.17% | 709054 |
| LeeSin | 51.97% | 41.29% | 10.68% | 46.62% | 490862 |
| Trundle | 53.94% | 43.26% | 10.68% | 48.55% | 118351 |
| DrMundo | 51.78% | 41.16% | 10.62% | 46.47% | 87591 |
| Rengar | 54.32% | 43.73% | 10.60% | 49.03% | 97593 |
| Nasus | 55.41% | 44.81% | 10.59% | 50.10% | 233418 |
| Shen | 50.55% | 39.96% | 10.59% | 45.29% | 87704 |
| TwistedFate | 53.22% | 42.64% | 10.58% | 47.93% | 125955 |
| XinZhao | 57.88% | 47.35% | 10.53% | 52.63% | 318665 |
| MissFortune | 58.97% | 48.44% | 10.53% | 53.68% | 149226 |
| Yasuo | 52.39% | 41.87% | 10.52% | 47.12% | 242814 |
| Annie | 55.05% | 44.54% | 10.51% | 49.75% | 207493 |
| Vi | 54.28% | 43.78% | 10.50% | 48.98% | 258399 |
| Darius | 54.72% | 44.22% | 10.49% | 49.43% | 143617 |
| Thresh | 53.66% | 43.18% | 10.48% | 48.53% | 747150 |
| Vayne | 50.83% | 40.35% | 10.48% | 45.59% | 379347 |
| Pantheon | 54.80% | 44.33% | 10.47% | 49.52% | 274628 |
| Graves | 55.09% | 44.62% | 10.47% | 49.81% | 193827 |
| Garen | 53.60% | 43.14% | 10.46% | 48.31% | 122017 |
| Diana | 52.68% | 42.22% | 10.46% | 47.43% | 74122 |
| Sona | 57.63% | 47.17% | 10.46% | 52.36% | 154788 |
| Ezreal | 52.30% | 41.87% | 10.43% | 47.10% | 307483 |
| Swain | 57.20% | 46.82% | 10.39% | 51.98% | 61108 |
| Quinn | 54.77% | 44.40% | 10.37% | 49.60% | 64520 |
| Udyr | 55.57% | 45.20% | 10.37% | 50.31% | 335632 |
| Shyvana | 54.04% | 43.68% | 10.37% | 48.84% | 323633 |
| Akali | 55.84% | 45.48% | 10.36% | 50.62% | 152723 |
| Gragas | 52.31% | 41.97% | 10.34% | 47.07% | 79301 |
| Zyra | 58.70% | 48.38% | 10.32% | 53.54% | 105222 |
| MasterYi | 55.68% | 45.37% | 10.31% | 51.36% | 197507 |
| Nami | 56.18% | 45.90% | 10.29% | 51.04% | 156443 |
| Leona | 56.75% | 46.47% | 10.28% | 51.63% | 550002 |
| Ryze | 54.56% | 44.30% | 10.26% | 49.39% | 271738 |
| Braum | 58.22% | 47.96% | 10.26% | 53.33% | 93063 |
| Leblanc | 50.18% | 39.92% | 10.26% | 45.11% | 220880 |
| Ashe | 56.01% | 45.77% | 10.24% | 50.89% | 140299 |
| Tryndamere | 57.11% | 46.87% | 10.24% | 51.97% | 183790 |
| Renekton | 53.31% | 43.10% | 10.21% | 48.22% | 347800 |
| Malphite | 55.37% | 45.17% | 10.21% | 50.18% | 129486 |
| Irelia | 54.97% | 44.78% | 10.19% | 49.88% | 105519 |
| FiddleSticks | 57.72% | 47.53% | 10.19% | 52.57% | 119728 |
| Corki | 55.38% | 45.23% | 10.15% | 50.27% | 65902 |
| Khazix | 54.69% | 44.55% | 10.15% | 49.65% | 299274 |
| Tristana | 54.76% | 44.61% | 10.14% | 49.62% | 183186 |
| Fizz | 55.54% | 45.42% | 10.12% | 50.43% | 215390 |
| Vladimir | 52.22% | 42.17% | 10.05% | 47.16% | 74755 |
| Katarina | 56.57% | 46.54% | 10.03% | 51.52% | 180100 |
| Jax | 56.56% | 46.54% | 10.01% | 51.61% | 420357 |
| Orianna | 53.33% | 43.39% | 9.94% | 48.37% | 203849 |
| Warwick | 59.33% | 49.40% | 9.93% | 54.32% | 333202 |
| Nocturne | 53.00% | 43.09% | 9.90% | 48.02% | 245216 |
| Lulu | 54.31% | 44.44% | 9.87% | 49.29% | 193290 |
| Twitch | 55.07% | 45.21% | 9.86% | 50.15% | 337923 |
| Janna | 56.41% | 46.59% | 9.81% | 51.42% | 106945 |
| Soraka | 57.14% | 47.36% | 9.79% | 52.22% | 324054 |
| MonkeyKing | 57.40% | 47.71% | 9.69% | 52.52% | 269605 |
| Lucian | 56.07% | 46.38% | 9.69% | 51.24% | 612664 |
| Zed | 50.19% | 40.55% | 9.65% | 45.37% | 212769 |
| JarvanIV | 55.77% | 46.14% | 9.63% | 50.94% | 143599 |
| Riven | 53.96% | 44.52% | 9.44% | 49.23% | 215531 |
| Evelynn | 53.68% | 44.25% | 9.42% | 48.90% | 118944 |
| Kassadin | 55.96% | 46.60% | 9.36% | 51.25% | 121801 |
| Taric | 55.51% | 46.21% | 9.30% | 50.77% | 84396 |
| Fiora | 58.14% | 48.86% | 9.28% | 53.48% | 117399 |
| Alistar | 48.73% | 39.62% | 9.12% | 44.04% | 81495 |
| Kayle | 58.97% | 50.21% | 8.76% | 54.55% | 268477 |
By eye-balling this chart, I think it's clear that champions that are long range and/or heavily skillshot dependent (e.g. Jayce Nidalee Olaf Lux Varus Ziggs) have greater blue side advantage; on the other hand, champions with short range and few skillshots (e.g. Eve Kassadin Fiora Alistar Kayle) do not have such heavy advantage.
One particular champion on that chart is Rumble. Rumble currently has slightly below 50% win rate overall. However, Rumble actually wins 57% of the time on blue side, and only 41% on purple side. This is roughly a 15% difference for Rumble on blue and purple side, far more than all other champions which hover around 9% to 11%. I believe the reason here is that Rumble's ultimate is a hard-to-land skillshot that has very long range. Due to the camera angle, it is far easier to aim Rumble's ultimate from the blue side than from the purple side.
Some implications of this:
1. You can make a champion look a lot stronger just by playing that champion more often on the blue side. Suppose a champion became FotM and he is always first picked - making this champion to be always played on the blue side. Then this champion's win rate will increase by 5% just by the blue side advantage. So, for example, if everyone decides that Ashe is strong and always first pick her, she will win ~55% of the time instead of her normal ~50% win rate. making her look a lot stronger than she looks. This hasn't happened yet and no champion has achieved 100% blue side appearance on ranked solo queue - even pre-rework Kassadin "only had" a 52% chance to play on the blue side.
2. It is demonstrated that blue side wins 10% more than purple, so if you can somehow "dodge" all purple side games without penalty, you will be climbing up in rating like crazy. So, for example, I can see that a player intentionally troll at champion select when he knows he is playing on purple (at champion select you already know which side you will be playing on), hoping that someone else on his team will dodge. This will give him a better chance of playing on blue side instead of purple.
3. The competitive scene is affected by this as well. So far, Riot has dodged the question by claiming that the sample size from LCS games is too small to see its actual effect, but I think it's pretty clear now that blue side has an advantage after three seasons of LCS. If we believe that competitive games, similar to solo queue, also have 55% win rate on blue side for two teams of comparable strength, then playing on the blue side in competitive LoL is roughly comparable to playing as White on chess (http://en.wikipedia.org/wiki/First-move_advantage_in_chess).
Some implications of this:
1. You can make a champion look a lot stronger just by playing that champion more often on the blue side. Suppose a champion became FotM and he is always first picked - making this champion to be always played on the blue side. Then this champion's win rate will increase by 5% just by the blue side advantage. So, for example, if everyone decides that Ashe is strong and always first pick her, she will win ~55% of the time instead of her normal ~50% win rate. making her look a lot stronger than she looks. This hasn't happened yet and no champion has achieved 100% blue side appearance on ranked solo queue - even pre-rework Kassadin "only had" a 52% chance to play on the blue side.
2. It is demonstrated that blue side wins 10% more than purple, so if you can somehow "dodge" all purple side games without penalty, you will be climbing up in rating like crazy. So, for example, I can see that a player intentionally troll at champion select when he knows he is playing on purple (at champion select you already know which side you will be playing on), hoping that someone else on his team will dodge. This will give him a better chance of playing on blue side instead of purple.
3. The competitive scene is affected by this as well. So far, Riot has dodged the question by claiming that the sample size from LCS games is too small to see its actual effect, but I think it's pretty clear now that blue side has an advantage after three seasons of LCS. If we believe that competitive games, similar to solo queue, also have 55% win rate on blue side for two teams of comparable strength, then playing on the blue side in competitive LoL is roughly comparable to playing as White on chess (http://en.wikipedia.org/wiki/First-move_advantage_in_chess).
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.