But the 25 credits make me bling bling with those dolla dolla bills yâall. Imagine all the cosmetics I would miss out on if I tanked in open queue. Probably like⌠one skin.
I only saw 3-3 a couple of times, one of the reasons thereâs more dps play is because 3 dps was generally feeling a stronger setup than 2-2-2. That and Orisa sucking, and me not having learned to play Sigma.
Hm, interesting. It seemed reasonably popular in my games, but I did six-stack a fair bit so we also went up against six-stacks.
At the highest levels it didnât appear to be played as much.
I ran into a ton in diamond and the matches that pulled master players into the lobby. I did not see it much in plat.
Could be a region thing I guess.
Once it is next to the comp tab, can we please, pretty please, next time we get data put this to bed?
The polls,.blizzard comments, and their data all pointed in the same direction, and I donât expect it to be any different after it ends up next to comp.
That Mode was pure garbage in my eyes, I canât imagine how I was able to play any comp back then when there was no role lock. Everyone was playing worse thst in normal Quick play to be honest.
Iâll take role queue over this mess ANY day.
1 Like
Put what to bed? I am not following.
There has been a VERY long set of threads over which (roleQ or OpenQ) has been more popular.
On one side there has been the results of polls from various forums -
Which was discounted originally from having not enough people,
Then when they did, discounted because they were from only one forum,
and when that was rectified then discounted because âobviously the people who have been hating on it have been leavingâ, and so we went out and showed that the rate of people leaving the game has become less AFTER roleQ went liveâŚ
So the arguments against it because âall of the forums everywhere were just filled up with Mercy mains, and they answer because they are angryâ.
So then we got the actual player numbers from the test by Blizzard, and they matched the poll results almost perfectly. (I know, it is almost like science works)
But THAT was discounted because it is in arcade, rather then next to comp in the selection screen.
Now we will be able to get numbers which it next to comp, and I am putting it out there, that there isnât any more complaints about the popularity ratios between the 2 when we get those numbers right?
Note through all of this the other side hasnât given one scrap of evidence that openQ was the more popular mode.
I went into it looking at it from a curiosity point of view, honestly wanting to know the results, and now it has been this running thing teaching people stats, etc.
Iâm sick of it, since people have stopped making ANY kind of reasonable argument around it, so, having one last definitive piece of evidence so we can ptu the whole mess to bed would be nice.
It is history, but history I would like closed.
1 Like
Side-by-side Open Queue doesnât even exist yet so of course thereâs no evidence of that.
But if weâre talking about Pre-Role Queue, that had a far higher peak number of players than Role Queue for one.
Non-response bias is a thing.
Wasnât rectified that I recall.
By looking at⌠Google Trends? Which barely showed what you claim, and a search term as generic as âOverwatchâ doesnât say much. Overwatch is also the agent system in CSGO for example.
No the argument was: the forums arenât representative of the playerbase, and thatâs exactly what we saw in polls with Support coming out on top, Damage sometimes even the least popular.
Now one argument you could use against that argument is that the sample size was too small in those polls (136 respondents in one, totally 300+ in both), which would be a valid argument and Iâm going to do another poll later.
For Quick Play, the charts may show something close to the truth as discoverability isnât an issue. For Competitive Play, thereâs the discoverability argument.
Thereâs still at least two additional factors to account for:
- Oneâs Open Queue SR isnât shown on the landing page of oneâs profile, reducing motivation to grind it;
- Lootboxes/coins incentives arenât rewarded to tank players as they are in Role Queue. These factors should be taken into account if we get the numbers.
Yeah, I did too.
Many of the arguments were reasonable, but there were definitely unreasonable arguments from both sides.
We grabbed a bunch of ones from Reddit, other forums in different areas, and the OW datascience discord.
Again, a drop there would have been noticeable, there wasnât one.
Right, but we then grabbed a whole bunch of polls from different areas of the net, which all showed similar results.
The meta poll was VERY statistically powerful. It SHOULD have put the entire thing to bed.
You gotta be kidding that THIS is what people will cling to?
I didnât see ANY arguments that the other side gave that openQ was more popular. Literally only one side was bringing ANY evidence to the table, and they did it in quantity.
Like, when one side brings proof over and over again, and the other doesnât bring anything, it is a pretty strong signâŚ
As It is, I expect that OpenSR will be shown, but, lootboxes? Seriously? If that is the best argument against it, then, it is over.
What polls are we talking about? Around the time of RQâs introduction, there was between a 60-40 and 70-30 split.
Now sure, thatâs a minority after OQ. Weâre months down the line from there, so itâs not hard to anticipate that those 30-40% with a preference for OQ are diminished.
Diminished to something like 12%. Who the hell thinks letting maybe 15% of their base leaving can be counted a success, just because a majority have a preference for a feature?
Even if it was as little as 5%, itâs still a very substantial loss over a change that didnât need to be made, or at least didnât need to be implemented in a way that removed the preference for, what at the time was at least 30% of players.
Right, I ALWAYS called roleQ a tyranny of the majority situation.
I never said it wasnât, and was pretty clear that THAT was a problem.
The arguments were around the âof the majorityâ part.
The people who hated roleQ, REALLY hated roleQ, which is one of the reasons I am happy to see openQ exist.
1 Like
In previous discussions, youâve never actually linked those.
You just said, âwe didâ with no links.
Reddit has already been proven to be non-representative. I posted the images in the discussion with Venus. Will need to go and grab them.
There was a sharp drop. I donât believe Google Trends of so generic a search term are a reliable way to assess this.
None of which you ever linked in discussions with me. You just said, âwe did this.â
The meta poll?
Cling to? I donât know, but theyâre factors. Somewhat minor factors, but still factor and should be taken into account.
Well, I donât even think most of the reasonable people on the Open Queue side ever claimed Open Queue was more popular. They just disagreed with the claim that Role Queue was, based on invalid data.
The proof that weâve shown time and time again exactly why it was invalid.
Itâs a factor. You donât think people play Tank in Role Queue purely for the coins? I think youâre mistaken, just look at the comments above by MoreHeroes. He literally said heâll stick to playing tank in Role Queue for the coins (even though he has 30K banked), or something like that, and thereâs many more like him.
I did, the big list of polls included some not originating from here.
You SAY there is a sharp drop, but then say, but, the method we use to measure it canât see that dropâŚ
My argument is, it showed every other change, like, new hero releases, events, etc, and it showed drops perfectly well in apex, etc.
Like, again, I am bringing evidence, and you donât.
I did the maths showing the variance of results in the polls.
They will become âMajor, the reason we can dismiss everythingâ factors⌠you just watch.
Nope, they argued that openQ was far far more popular, and because of that the evidence was wrong.
very tail wagging the dog thing.
If we STILL get similar stats from the next time, after it is put at the main thing, is this the argument that you will make that it is completely invalid data?
Or would THAT finally be enough to show you the relative popularity of the queues?
Because having the poll predict ALMOST perfectly the ratio of popularity of the queues when that data was released, should have been a HUGE sign that they were valid.
Like, all of the arguments have been âthese things could possibly be invalidâ but, each one is invalid in another way, but all agree almost perfectly on intensityâŚ
Which, given the chances that say, the underlying data (reality) is that they are a fair view of what is really happening is FAR FAR FAR more likely than they are ALL wrong, and wrong in a way which gives the same answer.
Like, Bayes would have some pretty pointed words here.
Iâm having the same arguments around lockdowns working for places with high trust in their governments in my day job.
In places with high trust, the result from lockdowns mirror how much of a lockdown pretty much perfectly.
In places with low trust, they donât, because people donât follow the lockdown (and we can see that from mobility of people)
It is like, the people who donât want to believe it argue that every single country which it is working has some unique thing which gives the same results.
Even though their âuniqueâ thing isnât working in other places which also has it.
You have some basic data which you can point at, lots of examples, very similar explanation, and people arguing against it because they donât like what it tells them.
Itâs like, well⌠that would be a VERY strange for there to be a bunch of exceptions, ALL being exceptional in a way which gives very distinct data which totally agrees with each other.
âIts 5g!â, um, you KNOW we rolled out 5g here just before it, and here the lockdown worked perfectly right?
Link them so we can assess their validity. The pattern I notice is a lot of you rarely if ever actually provide the sources you based your arguments on.
You realize, someone could be searching for their new wristwatch and have that affect that data?
Also, yeah new hero releases would show sharp inclines, because content creators, etc. would be searching about it.
I donât believe Apex Legends is comparable, its success and sudden drop in players happened over such a quick period - and also âApex Legendsâ is less generic of a term.
When it comes to less drastic decreases in playerbase over time, your graph was very much less clear. Maybe you should also post the graph youâre talking about so we can actually assess it and see for ourselves.
The peak number of players in Overwatch was higher pre-Role-Queue than post. There you go. Iâve brought you evidence. Your graphs will show that, once you provide them.
If they do, then you have a solid reason to tell them theyâre wrong.
Strawman argument, no quotes provided. Most of us werenât claiming Open Queue was far more popular. We made no claim of a majority either way.
If when theyâre side-by-side and we see that Role Queue is still more popular by a massive margin across all regions averaged, then Iâll happily admit Iâm wrong.
What I wonât accept is you for example cherry pick one region like NA like weâre seeing now, and use that to claim ânobody plays Open Queueâ.
Show what calculations you did to arrive at this. How did you factor that it was in the Arcade and therefore had lower visibility, into your calculations? Did you factor this in at all?
To repeat though, if the devs give us another comparison and it shows Role Queue is still way more popular for Competitive, Iâll happily agree that yes, Role Queue is therefore more popular for Competitive.
Itâs like Megadodo said. Invalid data doesnât suddenly become less invalid the more you have of it. Response biases can be repeated if proper polling techniques arenât used.
I did, rememeber the big poll thread, where I linked a whole bunch of them, BECAUSE people argued that the poll results didnât say it.
And yet, we see the same results in TF2, and basically everything else. Like, the effect is pretty well known. And we ALSO know that Overwatch, the game dominates the search results, because we have the baseline BEFORE it existed, when it was almost never searched for.
YOU made no claim of Majority, but I went poll looking after people did claim it, because it was interesting that Blizzard was saying one thing, and the forums were saying the other.
I didnât make the claim, no one played open queue, I DID say that roleQ was more popular, and it is, in every region outside of South Korea, in both comp and qp.
I didnât, since it was a meta result on the polls. They asked people prefence.
But, data which people have distinct, issues with saying it is too small of a sample for instance, absolutely is removed by more of it.
Saying that this one poll was a special case is ALSO removed of it.
.
Yes , I AM aware of selection Bias, but, different areas would have different selection bias, which was so small, as to be in the noise.
Secondly, the results from Blizzard data release ALSO showed the same result, which also goes a long way to show that the polls were not effected.
AND we also had the search result data, which ALSO showed that there wasnât a massive drop.
AND we had Blizzards own comments.
And, what did we have on the other side? Again, a TOTAL lack of evidence.
And what are we going to have when they release data from when it is placed next to comp and shows the same stuff?
The same arguments.
Even though it will be ANOTHER piece of evidence, strangely pointing in the same direction, with roughtly the same intensity.
Again, Bayes would have a LOT to say about the chances of that happening.
I donât think people argue against it in good faith anymore.
1 Like
But thatâs not the issue. The issue is theyâre not representative as far as role distribution.
We saw the exact same selection bias in all areas.
I still have no links from you, so nothing more to discuss.
Their comments, that now said theyâre surprised at the success of Open Queue, and their wording âsome players like Role Queue, some players like Open Queue.â With no claim of a majority either way, as opposed to previously, is very interesting.
Again, that would be SUPER unusual. And I already linked stuff back then, MULTIPLE times. This is the tactic of making one side do all the work over and over again, and I think it shows again, not arguing in good faith again.
AND then you will argue that Reddit, AND the discord would have the same bias. Even though, you donât have anything to back that up.
You can go link to counter evidence for once.
There HAS been a poll which pointed in the other direction, oneâŚ
lmao NA PC wanted NOTHING to do with open queue damn. personally I didnt play it because i knew it was going to be a pointless clustertruck.