It’s actually the opposite of this. And this isn’t even theoretical, we’ve seen this happen in game. This is going to take a bit to explain though, so please bear with me:
The systems in place that people are complaining about exist to make the competitive mode less grindy. The reason for this is that we want to rank people as accurately as possible as quickly as possible. That’s the problem we are attempting to solve with any competitive ranking system, and people have put a lot of time and energy into how best to solve that problem.
Players, however, do not want to be ranked as accurately as possible as quickly as possible. They want to be ranked as high as possible. So they will find ways to circumvent these systems. And you can watch this dance in real time. I’ll give some high profile examples of this happening and the steps Blizzard took to deal with this later in the post. But first, a detour:
Why do competitive ranking systems push people toward a 50% win rate?
The ladder rankings are a testable prediction. We think this player is better than this player but worse than this other player. Every match, then, is an opportunity to test that prediction. Did the players we expected to win actually win? Did people perform as we would have expected them to perform given their current ladder ranking? Did they over-perform? Under-perform? All of this information is new information for us. So we factor that in to our future predictions. And the ladder rankings are adjusted accordingly.
Time for a non-competitive gaming example:
I sit down to take the GMAT, because I am applying to top ten grad schools and I need a score that places me somewhere in the top 5% of people applying to grad schools. It’s a very competitive ladder and the GMAT is the ranking ladder used by these schools to rank their admissions candidates.
I know (let’s assume I’ve done my research) that the middle 80% of test takers will get roughly half of the questions they face on the GMAT wrong. Only those test takers in the top 10% or the bottom 10% will deviate significantly from a 50% win rate. That does not mean, however, that the test is unfair. Or that I am unable to achieve a win rate higher than 50%. (You can, in fact, get every question on the GMAT correct. It’s just that only a tiny fraction of 1% of test takers have the skill to do so.) Likewise, you could get every question wrong. (Though this would also be incredibly unlikely.)
What it means is that the test is designed in such a manner as to most efficiently and accurately assess the relative skill levels of the test takers, and to do so across a wide range of skill levels. That’s why it is used by those top tier graduate programs to sort out their candidates. How does the test do this (the short answer is the same way the OW competitive ranking system does)?
It does this by selectively choosing the difficulty of the questions any individual test taker faces. Each question is chosen to test a particular hypothesis- we think that the test taker will get the next question correct roughly 50% of the time. Why do we think this? We judge based on the test taker’s past performance. So the first question any given test taker faces is one from the medium difficulty range (this is the equivalent of placing a new account in Gold). The next question they face will be whichever question we think will give us the most new information about that test taker- what question is most likely to upset our current assessment of this test taker’s skill. And that question will be one that we would expect them to get right roughly 50% of the time.
Consider the alternative. What if every question we give someone is one that we would expect them to get right roughly 80% of the time? We have now chosen to give ourselves far less information about this test taker. Perhaps we could still accurately rank our competitive ladder, but it would take much longer. This is the grindier alternative to the systems that are now in place. Their SR cannot change as much from one match to the next because we haven’t learned as much about them.
And that brings me back to where I began this post. I said at the top of this post that we’ve actually seen this happen on the competitive ladder in Overwatch. And that makes sense. Players don’t want an accurate assessment of their skill; they want to attain the highest rank possible and they will seek to circumvent systems that are in place to more accurately assess their skill when they can do so (in order to attain a higher rank). So, some examples (and it’s all really the same example it just happened in different regions and on different platforms):
- A group of high-rank players in the Oceania region began 6-stacking and playing during off hours. Why did they do this? They knew that this would give them the best opportunity to thwart the matchmaker. Because, again, the matchmaker is trying to find the match that would be most likely to cause us to reassess our current understanding of the ladder rankings. What match would give us the most new information about these players? The answer is, of course, a match that we would assess as a roughly 50/50 match.
So by 6-stacking, they force the matchmaker to try to find another 6-stack. By playing during off hours they limit the pool of available players as much as possible. And they essentially prevent the matchmaker from finding a 50/50 match. It won’t even be able to get close to a 50/50 match.
What is the outcome of this?
This 6-stack wins roughly 100% of their matches. Every time they do so, however, they gain some pittance of SR- maybe 1 or 2 SR per match. So they rank up. But it’s incredibly grindy. They essentially buy less accuracy in the rankings (because that’s what they want- to continue ranking up, regardless of whether they are the most skilled players on the ladder or not), but they pay for it in extreme grindiness.
-
Groups of OWL players did the same thing on the NA ladder. They stacked during off hours so they could grind to the top of the ladder.
-
A group of players on PS did the same thing- grouped during off hours and grinded their way to some ungodly SR (north of 5k if I recall correctly).
What do these examples show us- the current systems in place make the competitive ladder less grindy, not more. That’s point one. Point two is that if you did something like remove the matchmaker’s ability to find 50/50 matches, you would not only increase grindiness on the ladder, you would also remove the competitive integrity of the ladder as the rankings (particularly those at the top and the bottom of the ladder) would be less accurate. Point three, people have done this intentionally. They’ve tested it for us. We know it increases grindiness and decreases the accuracy of the rankings. Point four- these systems are in place in order to solve the problem of assessing the relative skill levels of the players on the ladder. That’s why they are employed across a range of fields, not just in online gaming.
I can go into greater detail if necessary.
) was notorious for this.