Hello Warriors! Welcome to this installment in our ongoing series of State of Balance, where the Fight Team analyzes the results from the just-completed season, and discusses where we are headed in the future.
In this blog, we will share data on the Win Rates and Pick Rates of our heroes with you, the community. We will also use the look at the data to discuss our perspective on certain balance issues, and where we see it heading over the coming seasons.
(To know more about the data we present here and how we gathered it, please read more about our methodology for State of Balance. Note that we have changed the way we measure averages listed in the Win Rate columns – the values are now weighted by the amount of matches that occurred within a single matchup instead of using the average. This should better reflect the experience of matchmaking – as match-ups that are more likely to occur are given proportionally more weight.)
Now, let’s get to the data!
DUEL – WIN RATES
Note: in this State of Balance, we use the version of the Tier List that was posted on 2019-May-31, which was the final version of the Tier List that was posted during ‘SAKURA’ (Year 3: Season 2). All the Tier List images included are taken from the version on that date.
Raider, Shugoki, and Lawbringer all received significant improvements over the past 2 seasons, and all now appear in the remainder of the top 4 for Full Population and Platinum And Above. In the Tier List, we note that only Raider is marked S Tier, while Lawbringer is lower at A Tier and Shugoki is lower still at B Tier. Internally, part of the theory to explain this difference is that Full Population players take much longer to identify what moves can be safely nullified (by a roll) – for example, Lawbringer and Shugoki’s Unblockable Heavy Finishers could be safely backrolled during Year 3: Season 2 – but Full Population players usually seemed to attempt to defend these attacks, and therefore lost to the Unblockable, or the Feint > Guardbreak, or Demon’s Embrace. Top Players at the level of the Tier List didn’t take these unnecessary risks. With Year 3: Season 3, we made a number of changes meant to reduce the safety of escape tools, with the expectation that Full Population players and Tier List players will now both have to play against the same attacks, and we hope that this will somewhat increase the order of heroes in the Full Population data and the Tier List.
Hitokiri enters near the top of the Full Population and Platinum And Above lists. Hitokiri, however, ends up slightly lower in the Tier List, in the A Tier, which means that at the highest levels play, Hitokiri is considered to be a viable hero.. This is a pattern we’ve noticed in the past, where the Wu Lin and Black Prior entering high in both Full Population and Platinum And Above lists upon their introduction, but then normalize to expected numbers over the course of the following Season. We will keep a sharp eye on how Hitokiri performs as Season 11 progresses.
Jiang Jun and Black Prior still appear near the top of each list, but are no longer at their very tops. This indicates that while they are still strong, they appear to have settled back to their expected positions – Jiang Jun went from 63% Win Rate for the Full Population for Year 3: Season 1 and was down to 57% for Year 3: Season 2. Similarly for Black Prior, the hero went from 55% win rate in his first season to 53% in Year 3: Season 2, which follows the normal downwards trend of newly released heroes in their followup seasons. This seems to be par for the course for newly released heroes.
The important thing to note is that these are the reasons we try not to overly nerf newly released heroes: as players get used to them, they tend to trend downwards in their win rate, and as such moves that felt like they were overtuned end up being fine in the long run.
Warden is an interesting case. Warden ends up dead last in the Ranked Duel Full Population, which is very contrary to what the Tier List tells us – indeed, according to it, Warden is the best duelist in the entire game. Even in the Top 4% data, Warden’s Win Rate isn’t as high as the Tier List claims it should be. Other heroes that are very high on the Tier List, such as Berserker, also end up very low. A theory is that the Tier List represents consistent top level tournament-level play and as such even players in the top 4% do not always play at that skill level, even though they are much stronger than most players.
DOMINION – WIN RATES
Jiang Jun and Shinobi are still S Tier, and Nobushi has risen to S+ Tier. We’ve taken steps to reduce how safe these characters are at the start of Year 3: Season 3, and expect them to no longer be such near-mandatory picks at top-level play. Our expectation is that this will allow other team compositions to emerge, and more variety to be shown at top-level play. While the Full Population data shows that Nobushi and Shinobi are not performing at such levels, we find that prioritizing meta-defining heroes at top level will have a healthier impact on the game as a whole.
Raider and Hitokiri are going strong near the top of each set of data. These two were buffed or introduced during Year 3: Season 2, and are performing at all levels of play. Uninterruptible Stance timing on Hitokiri’s opening Heavy has our eye on it – and Raider’s damage output as well.
Lawbringer is especially outperforming most players’ expectations, and the improvements he got in Year 3: Season 2 seem to have been undervalued. In Dominion as well as Duel, we see Lawbringer performing very highly. With Year 3: Season 3, we nerfed escape tools and thereby buffed the Unblockable Heavy Finisher, and so expect Lawbringer to continue to perform above most players’ expectations.
Shinobi and Shaman are still under-performing in Full Population, while they enjoy success at top-level play. Shinobi and Shaman notably perform quite highly in the Tier List, while all Assassins underperform in Full Population settings. The Shinobi and Shaman seem to be played fundamentally differently between Full Population and top-level play – at top-level play, we see the Shinobi spending a fair amount of time at range using the Ranged GB, and top-level Shaman play often revolves around successfully performing ganks without allowing Revenge to pop - this efficiency seems to explain the difference in performance between the data sets for these 2 heroes.
Assassins are still underperforming globally. Reflex Guard makes it harder to defend as successfully, and Gladiator still suffers from a shorter-than-usual Reflex Guard.
BREACH – WIN RATES
Assassins, just as in previous seasons, are generally under-performing in Breach, as well as Dominion. The reasons are similar, and since Breach has less focus on 1v1 gameplay with a stronger focus on pikemen, Fortifiers, Commanders and objectives, Assassins typically find themselves in worse situations than in Dominion, where more external attacks are making things difficult due to Reflex Guard.
Dominion and Breach seem rather similar in terms of win rate and pick rate. There are a few outliers, such as Nobushi who is picked much more in Dominion than in Breach. This is most likely due to the fact that Nobushi is an excellent hero in the lane in Dominion, but since pikemen in Breach tank more damage, she is not as effective as in Domininon.
Note: There has not been a current community Tier List for Breach – so it’s not included in the ‘State of Balance’ for that reason.
PICK RATES
There are a few observations we can make regarding the Pick rates –
Warden and Orochi are still the two most picked heroes in the game. This has been the case for a long time, and we don’t expect this to change: Warden being the most accessible hero plays a lot into this, and Orochi is the most iconic Samurai hero.
Kensei, Raider and Centurion are also found high on the Pick Rate charts, and have been for a while, regardless of mode or skill threshold.
Hitokiri enters the roster as the top of the Pick Rate charts, just like the Black Prior in the previous season. Players seem to enjoy the Executioner fantasy, and Uninterruptible Stance seems to be a strong and valued property when picking a hero.
NEXT STEPS – A SHIFT IN THE META?
Escape from mix-ups was too safe – this was nullifying a lot of characters’ mix-ups. We also believe it helped create a large difference in playstyle between top-level players and Full Population players – contributing to a large difference in perception between the power level of certain heroes at different skill levels. In Year 3: Season 3, we standardized forward and backward strafe speed (to keep opponents close), and improved the ability for Feint>Guardbreak to catch rolling opponents (to ensure many mix-ups can work). We expect to see players have more success with their mix-ups, especially at top level play.
S+ Tier Dominion Hero (Nobushi), as well as Jiang Jun and Shinobi, all had their safest defensive options nerfed in Year 3: Season 3. Nobushi has been the focal point of the top-level play for quite a long time, and helped define the Meta. All three heroes were quite safe, and we hope that by nerfing these options, we open the door for them to have a more healthy playstyle, as well as open the door for the Meta to include more variety in competitive team composition.
We’ll be keeping an eye on how these changes affect gameplay. We hope to see Duels become more dynamic (especially when players get into their chains) since opponents won’t be able to freely escape. We also hope to see Dominion have more variety in team composition (with the nerfs to certain heroes). Going forward, we hope that the Testing Grounds can help us as we strive to bring a few more heroes into competitive viability.
CLOSING NOTE
We would love to hear more from you on the Full Population data, Platinum-or-Above data, Top 4% data, and the community Tier List. Please remember that we look at all this data, information, feedback, and a whole lot more when we make our decisions.
Keep sending us your feedback, and we look forward to sharing the next installment of the State of Balance at the beginning of ‘Year 3: Season 4’!
- The Fight Team
