Can We Talk About PBSR?


The Problem:


Ok, here’s the deal.

I’ve made posts about this before, and I’m sure I’ll do it again, but there is seriously something wrong with how PBSR functions.

Currently, your SR gain/loss is adjusted per game due to how you’re performing.

“But wait”, you say, “what do they compare your performance to so they can accurately measure your needed SR gain/loss?”

Well, a lot of people are led to believe that it’s every player who fills the role you’re playing in your rank. But I’ve been paying attention to my stats and gain/loss from them, and noticed that it’s not functioning that way.

What is actually happening is that you’re being compared to your teammates that game.

What does this mean?

Here’s an example:

Say I had four golds. The slots I had golds in were:

  • Eliminations: 25
  • Eliminations on Objective: 15
  • Objective Time: 2:00
  • Damage: 10,000

At the end of my game, I would get, say, 27 SR.

However, next game, if I were to get these exact stats but get all silvers, while the player who had gold had stats like these:

  • Eliminations: 45
  • Objective Kills: 40
  • Objective Time: 3:30
  • Damage: 25,000

I would only get around 20 to 23 SR.

The problem here, especially with DPS players, is that there is a constant competition between the teammates to do better than each other.

You can see this a lot. People will over-extend, knowing that they will probably pay for it, just to secure an extra kill for those beautiful gold medals.

“Ok”, you then say. “Well, just do really well and you’ll always get a lot of SR!”

Well, not performing as well as your teammate is not the only problem that makes you drop more SR.

Say you’re playing a game and you’re doing pretty average as DPS. Your other teammate(s) are doing better than you, but not by way too much.

At this point, someone leaves…

It’s a 5 v 6. A much harder fight. The chances are, your team is going to be too tilted or just completely give up. You no longer have any chance of surpassing your teammates in stats, and when you lose, you’re going to lose much more than you should’ve.

And that’s not all.

Say you’re playing McCree, and doing much better than the other DPS on your team, but you notice that there’s not enough healing happening, so you switch to Ana.

Your DPS stats will almost completely stop, but your healing stats will rise. However! The other DPS will eventually rise above your DPS stats, and you will most likely be unable to get higher healing stats than the other healer(s) that had already been healing. When the game ends, your stats will be lower than everyone else, and your gained SR will be much lower.
( Or lost SR will be much higher )

So not only does PBSR promote selfish play, it also promotes not switching or filling, effectively one tricking for a game as to not lose more SR than should be lost. And on top of that, leavers almost completely tank your SR as well.

So how do we fix this?


How to Fix the Problem


Well, instead of comparing your stats to your teammates, you could compare it to all the players in your rank, and not just those in your game.

What would this fix?

If this was how PBSR was calculated, teammates would not compete with each other for better stats. They would constantly be trying to beat their own stats, and always do better than last time.

But this isn’t the only way it could be fixed. Here’s the big one:

Instead of measuring performance on a game-to-game basis, your performance could be measured over a constant average of games.

If you do well, say, 8 out of 10 games, you’re doing well over an average of games, and most likely deserve to be higher, and the PBSR should slowly apply in your favor. If you don’t do well 8 out of 10 games, you’re doing badly over an average of games, and most likely deserve to be lower, and the PBSR should slowly apply against you.

Once your average approached a one-to-one ratio, the matchmaker would understand you’re where you belong, and your SR would even out to a constant SR, until it notices a change in your average.

If you started doing better, not only would the matchmaker notice a difference in stats between everyone in your rank, but it would also notice that you’re doing better rather than worse on average, and that would make it much easier for the matchmaker to notice that you deserve to rise. ( Or fall )

So what would this fix?

Think about it. There are some games where you just do very well, and some where you just can’t kill anyone.

When you do extremely well, does that mean you deserve to be much higher? Not necessarily. When you don’t do well at all, does that mean you deserve to be much lower? Not necessarily.

The matchmaker could get a feel for where you belong as an average, rather than a jumpy game-to-game basis.

And not only this! But it would also fix the leaver problem! ( Somewhat )

If the matchmaker was able to see past those not very well fought games, it would also see past games where there were leavers, and not rapidly shoot you down.

This would also advocate for switching and filling for your team, so you’re able to support your team in every way you can.


Tl;dr:


PBSR is too rapid, needs to be over an average of games and compare you to people in your rank rather than teammates.

Thank you for any replies to this thread. Let’s please get Blizzard to at least notice this. I’d rather much have a PBSR system like I explain here or no PBSR system at all rather than the one that is in place.

7 Likes

Can you prove that PBSR compares you to your teammates though?

1 Like

I already did in my OP.

Why would I get less SR if I got the same stats? Did everyone in my rank all the sudden rise in stats right before I finished the game?

What I mean is, you can get the exact same stats throughout games but get different amounts of SR. I’ve had games where I had nearly 50 kills but someone else had around 60. Although that’s well above average, I only got a normal amount of SR, rather than the above average amount that I should’ve received.

3 Likes

You making up examples is not proof at all. It is you literally making things up.

8 Likes

I agree with your points, however for people like me that don’t play much comp, I think it would be even harder
lets say it works like this: every 5 matches are like placement matches and your rank adjusts after that, it would require more matches to gain large amounts of sr

1 Like

Here we go again. Please don’t reply if it’s just going to be formatted like this.

You’re in diamond. You don’t have PBSR. Anyone below diamond who pays attention to their SR gain/loss can tell that sometimes they just don’t add up.

I don’t understand how it would be harder. It’s not like I said the matchmaker would group you with harder people, I just said it would pay attention to an average of games rather than game-to-game.

Humans are not very good at doing one thing exactly the same constantly. You will always have games where you do really well or really badly, but you will never have a constant amount of games where you do the exact same in performance.

Therefore, you should be measured so that that factor is taken into account.

I had another idea somewhat similar to this, but I think players wouldn’t enjoy this. I think players enjoy watching their SR change every game. It’s kinda weird and worrying to not see any change after a few games.

I’m making a point about your post. You have proved nothing, and you have based your entire argument on something you have not proven. Why do you care about my formatting? Shouldn’t my argument be key?

I’ve been in the competitive system for 8 seasons, clocking in over 700hours. I started in season 2. I’ve been through performance based SR, for a long time. Roughly half of my playtime was in plat. I also placed 2700SR this season. And from my personal experience, I think performance based SR is fine. So now we’ve both derived different things from experience. My viewpoint, from experience, is just as good as yours.

You need to prove yours is true, because now you have literally nothing backing your point. Not even experience, as mine contradicts it.

5 Likes

Your formatting was, “You don’t have proof”. That’s not an argument, that’s just you trying to trash my post.

Then you either don’t keep track of your SR gains, or you do very well at playing at your average consistently.

I never argued this, why are you bringing it up?

Then in that case, your experience is wrong as well, because mine contradicts it.

Seriously, if you’re just here to troll, leave.

I’ve already explained this. If you perform above your own average, but someone on your team performs even higher, your SR gain will not be much higher like you’d think it would be.

2 Likes

The gold medal theory for DPS isn’t true. I main Sombra on PS4 and I’ve gotten 30+ SR in games where I didn’t have any elim medals. Your enemies hacked, enemies EMP’d and offensive assists are super important

3 Likes

It was just a general statement that people could understand. There’s obviously a lot more that goes into your performance with every character, but the point still stands.

1 Like

No.

"The calculation of your SR adjustment after a match doesn’t look at your teammates, but instead compares you to the performance of other similarly skilled players with that hero across an enormous pool of competitive matches. So, we compare your Genji play to the play of other Genjis, Ana vs. Anas, etc. Since we’re comparing “apples to apples”, we shouldn’t see any kind of support specific bias in SR adjustments due to player performance. " – Scott Mercer Overwatch Forums

3 Likes

Again, I can get the exact same stats in a match and get different PBSR depending on how my teammates did. They’ve either changed it since that post, or it’s not functioning correctly.

That’s not my formatting. You’ve identified my point, my argument. I’m pointing out a severe flaw in your post. That is why you perceive it as trashing your post.

I did keep track of my SR gains. I stopped after the data revealed nothing.

You seem to have thought experience was valid here:

How else could you have derived your viewpoint, without data, unless you used your own experience? I’m making an assumption this is from your experience since you have literally no data anywhere. I’m sorry if such an assumption is incorrect, but my original point still stands: You have no proof.

That is exactly my point, which is why I state “You need to prove yours is true”

Why are you dismissing me as a troll? I’m pointing out flaws in your argument.

You have yet to prove it, however.

1 Like

Exact same stats isn’t the same as exact same stats per minute.

2 Likes

My data would require actual footage of gameplay, which I don’t have the resources to provide at the current moment. ( But maybe in the future )

You obviously don’t get the irony then.

Because you and people like you love to go post to post trashing everyone’s ideas with, “No proof!”

I’m not trying to argue. I’m trying to provide ideas for Blizzard to change up competitive.

Wonder why they haven’t changed it a whole lot yet? Posts like these get sent to the trash because of you and your posy who like to shout people down.

You never addressed my other point about an average of games. Why not?

You only want to point out anything that you can trash on. Leave.

Show me this data, and the analysis to demonstrate statistical significance. There are a large number of factors that can affect SR gain other than PBSR, so you will need a lot more than one game to show an effect like this.

Again, neither you, nor your pal have addressed my point about an average of games.

And again, I don’t have the time or resources to conduct an entire investigation for my claim, but I can tell you, as can others, that the change in SR per game is much too rapid to correctly determine where you need to be.

Kaawumba has done some pretty hefty data collecting and made some great analytical posts in the past. Just food for thought when they are asking you for data. I can’t recall ever having a PBSR SR gain/loss feel unwarranted.

He likes to make you believe that his data provides any argument, but I’ve already shown that he’s just analyzing games normally and selling it as something big in a different post.

The entire reason I made that point was to point out that experience is invalid as proof.

Because you have no proof to back up your points. Considering these ideas are directed towards blizzard, they could potentially change the matchmaker. I do not at all want any sort of changes to be baseless, and as such, I argue against any ideas.

Why is that relevant if you have no proof?

1 Like