devtrackers.gg
After several seasons of the same ranked format over and over, I started thinking of the ways that ranked could be changed in order to make it better. So, a few days ago, I made some dinner, got a few drinks, and sat down and brainstormed how I would change ranked if I was in control of ranked balancing at Respawn. This is my attempt at explaining my thought process and the system I eventually came up with. WARNING: LONG. TL;DR AT END
First, it's important to understand what the actual purpose of ranked is. There are a lot of people who are considered "kill grinders" or "ranked grinders" in Apex, and while professional players normally gravitate towards the top of the leaderboard, it is not uncommon or even somewhat expected that the top ranked player is someone who is not signed to a pro roster. Knoqd did this in Season 4, which led to him being signed shortly after, and the same can be said about Keon in Season 11 or 12 (my memory isn't perfect on that one); Funnnq is another name that recently held the top ranked spot and found himself on a Pro League team. Due to this, I would argue that ranked play is a system by which players can show their skills in order to make a name for themselves in the professional scene. Therefore, it would be beneficial for ranked to be somewhat designed to achieve that goal.
My belief is that in order to achieve this goal, ranked should hinge on three core statements. First, ranked should be competitive. In order for a proper ranked system to filter the best players, they must be going up against people of similar skill on a consistent basis. Second, ranked should be balanced. This can mean a number of things, from legend meta to gun meta to queue restrictions to the scoring system, but there should be a feeling of progress and a feeling of stability when it comes to ranked games. Finally, ranked should be hard. Ranked should not feel like a kill-fest to the top players, but should instead test their ability to rotate properly, take positioning, pick and choose which fights they take, and be punishing when mistakes are made.
I believe that ranked does not fully achieve all of those goals. It is obvious that watching professional match play is much much different from watching ranked. It is also obvious that these matches are some of the hardest and highest quality lobbies in the game. While I don't think ranked should look just like pro matches, I do think there should be some resemblance. Think about it this way: college baseball is how top prospects are scouted to the major leagues. While these two leagues have the same operating rules, there are differences: college athletes use metal bats instead of wood, among other things, to make the game easier for players. Ranked should be like college baseball: similar, but made easier for the sake of varying skill levels.
This leads us to the first major change. The current ranked scoring system is being thrown out of the window. Gone are the days of -70 entry costs, because we're going back to smaller numbers with a system more aligned to the current ALGS points system. A kill equates to one point, and KP is shared among teams irregardless of kill or assist distribution, unless a player is dead (in which they receive no credit for kills). Placing is bracketed in a much more forgiving way, due to the fact that it's ranked and not comp, but still relatively exponential. Below is the current scoring system against my proposed scoring system:
| Placement | ALGS Score | Ranked Score (and Bracket) |
|---|---|---|
| 1 | 12 | 15 (1) |
| 2 | 9 | 12 (2) |
| 3 | 7 | 10 (3) |
| 4 | 5 | 7 (4-5) |
| 5 | 4 | 5 (6-7) |
| 6 | 3 | 3 (8-9) |
| 7-9 | 2 | 2 (10-12) |
| 10-15 | 1 | 1 (13-16) |
| 16-20 | 0 | 0 (17-20) |
With this system, entry costs are also entirely out the window, as are rank thresholds. Rank thresholds are the RP values which are required to cross into a new tier, such as 15,000 for Masters. Instead, rank-ups are achieved on a split. Each season will now consist of three ranked splits, meaning one occurs roughly every month. During that month, players earn as many points as they possibly can. This gives a constant sense of progress, and removes the frustrating feeling of loss that players can feel when they die at the beginning of a match and immediately lose RP. Instead, you will simply just gain nothing. There is no backwards progression within a split.
After a player has earned as many points as possible and a split comes to an end, we get into the math of all of this. The basic idea is that the best players at the end of a split are promoted a rank, while the worst players are demoted a tier, but this has us run into a problem. The problem with promotions and demotions is simple: you cannot promote or demote in a flat rate manner or a portioned manner by tier if you wish to keep exclusivity in lower and higher ranks. Let's use an extremely oversimplified example.
In this example, there are 10000 people who play ranked Apex Legends. Using the current distribution of ranked players (prior to today's split reset), only 2.22% of players are in Bronze, while Silver contains 15.22% of players, Gold has 41.5%, Platinum has 32.57%, Diamond has 8.05%, and Masters has only .45% of players. So, by the example, that's not even enough players to fill a Masters-only lobby. How do you even go about promoting or demoting players without inflating the lobbies? My solution is to use a standard normal distribution to attempt to balance the amount of players that are within each tier. This means that we take being between Gold and Platinum to be considered "average," seeing as over 74% of the player base resides in that category. Using the chart that haunts every eighth grader
This also gives us our solution to promotions and demotions. Let's assume for now that the bottom 20% of players in each tier are demoted at the end of the season. For our example population in Masters, this is 42 players. Those 42 players move to Diamond, and the top 42 players in Diamond by accumulated points are promoted to Masters. The same is done on the tail end: the bottom 272 players in Diamond are moved to Platinum, and replaced by the top 272. Platinum sends 682 players to Gold and gets as many back. Now, because we are on the opposite end of the spectrum, the top 20% of Bronze is promoted (again, 42), and the bottom 42 players in Silver are sent to Bronze. Silver promotes 272 and Gold demotes the same, and, well, the 682 who would be promoted from Gold were already sent up to replace the 682 who were demoted from Platinum.
Now, you may be thinking: It seems a little unfair that in order to be promoted into Silver, Gold, and Platinum, you just need to be in the top 20% of players in that rank, but for Platinum to Diamond, you have to be in the top 8%, and for Diamond to Masters, you have to be in the top 3%. And to that, I say this: ranked should be hard. And, by being selective at the top, you create higher-quality lobbies all around.
You may have also noticed that rank divisions are gone (Plat 2, Silver 4, that stuff). Yup. Also, Rookie is escaped by achieving a fixed amount of points (e.g., 250), and upon reaching said threshold you are placed into Bronze.
So, now that we have a much simpler scoring system with no rank multiplier, placement multiplier, or entry cost, partnered alongside queues that are exclusively for a specific rank and a promotion/demotion system that allows for only the cream of the crop to move up in rank, as well as a scoring system that (in my opinion) offers more balance between kills and placement while still being forgiving, it's time to focus on the overall health of the lobbies. This is probably my most ambitious idea, but here it is: role queue.
With the introduction of classes at the beginning of Season 16, the meta shifted drastically, but there hasn't been a lot of experimentation with said meta. In order to encourage such experimentation, there should be a loose role queue based on which legends a player intends on playing. Each player must rank their preferred class on a scale of 1-5, with an inverse point value (e.g., rank 1 is weighted at 5 points). Thus, when a player queues, they are matched with teammates who compliment their preferred classes as opposed to overlap with them. This is accomplished by creating teams that value each class at a weight of at least 7. Let me show you another example, using Players A, B, and C. Let's assume that the players ranked the classes this way in their preferences:
| Player A | Player B | Player C |
|---|---|---|
| Skirmisher | Skirmisher | Assault |
| Recon | Assault | Skirmisher |
| Assault | Support | Controller |
| Controller | Recon | Support |
| Support | Controller | Recon |
This team weighs Skirmisher at 14, Assault at 12, and Recon at 7, but Support and Controller fall at 6. This means, assuming all these players were solo, the matchmaker would not put them on the same team. However, Player B now has their ranks like this:
| Player A | Player B | Player C |
|---|---|---|
| Skirmisher | Controller | Assault |
| Recon | Support | Skirmisher |
| Assault | Assault | Controller |
| Controller | Recon | Support |
| Support | Skirmisher | Recon |
This gives us a value of 11 for Assault, 10 for Skirmisher and Controller, and 7 for Support and Recon, allowing for this team of all solos to be arranged as a team.
For the purposes of a duo, they may always party with each other, but the queue may be longer while trying to find a complimentary third. As far as I am aware, there is no possible way for a duo to rank their roles in such a way that makes it impossible to find a third. For trios, the weighting system is determined prior to allowing them to queue together. Also, to avoid abuse of the weighting system wherein a trio makes their ranks work but then picks within the same class anyways, players cannot pick from their bottom rated class in ranked. This will give us more diverse team compositions and could lead to shifts in the competitive meta.
Note: I could take or leave role queue. It was an idea I've had for a while and as much as I would like to see it implicated into the game, I don't really know how other people feel about it. Feel free to tell me how good or bad of an idea you think this is.
So, to recap The TL;DR:
New ranked scoring system that is much closer to ALGS scoring
No more entry costs
Promotions and demotions are given at the end of splits
Easier to get out of Bronze, Silver, and Gold, while harder to get into Diamond or Masters
Ranked lobbies exclusively have players of the same rank
Ranks have a fixed portion of the player base contained within them
Players are encouraged to play their main role, with matchmaking prioritizing diverse team compositions
Predator remains the top 750 players in Masters, determined by the amount of points they earned in the previous season
I believe that these changes accomplish all of the goals I previously listed. It makes ranked competitive by having similarly-skilled lobbies with an incentive to promote into harder lobbies, not just higher ranks. It makes ranked more balanced by having a scoring system that is much easier to figure out and does not have a heavier focus on kills or placement, needing both to succeed and maintain your spot at the top, as well as making sure that teams have a diverse spread of legends and one legend isn't over-utilized. And, it makes ranked harder by ensuring that in order to promote and be at the top, you must be able to consistently earn more points than your opponents while fresh talent is always on the rise.
I also firmly believe that this would stop the smurfing problem, seeing as it would take a minimum of 5 splits to be able to get to Masters, meaning that a player would have to be consistently good for the course of well over a season to make it from Bronze to Masters in the minimum amount of time.
I know this is a very long-winded post, but let me know what y'all think. Cheers!
External link →