posted this in the thread Jeff Kaplan posted the %'s break down of each tier.
But the RAW numbers and REALITY is that people from silver to platinum can be in same game together (1000 rating difference). The difference between a 1900 player and 2100 player is indistinguishable no different than a 2400 rated player and a 2500 rated player or a 2800 rated player and a 2900 rated player.
But 78% of the player base is in silver, gold or platinum - which all can be in same game together. Its insulting to our intelligence to tell us that 78% of 30 million people are near equally as good at ANY subject in life let alone something thats an actual activity such as a game.
For someone like me, I could care less about my rating icon. I care about equal games though. EVERY game EVERY day theirs 1-2 players on EACH team where im just literally mind blown that this person is considered my equal (by rating) and is in the same game as me (due to this inaccurate rating system).
Have you ever heard of a bell curve? It’s the most common type of statistical distribution. Overwatch’s ranked mode distribution mimics it extremely well.
Ok but Blizzard has other competitive games (mainly Starcraft) and it is not on this ranking system.
This game throws 78% of its player base, regardless of placement matches, new account, old account, etc within in close enough ranking to play with each other.
If I bought Starcraft right now, and only played 10 hours of it, I would not be placed in middle tier (silver-plat) automatically. Id have to climb up.
Overwatch is not forcing a bell curve. Bell curves are natural, and very common things in every game. Using the example you have, Starcraft has a bell curve for it’s ranking tiers as well. That’s just how skill works. The average player will be just that, average.
It’s not fair to try to force a ranking on someone that they don’t deserve/belong in just because it looks nicer to have it even. The majority of people won’t be as absolutely terrible as bronze, or as absolutely godtier as Masters+. It’s just not the case. Not everyone is willing to put in the work to be really good at the game, and the ones who do are the ones who are in the top tiers.
First of all, what reasons listed exactly? I don’t really see one besides “quality of games is bad” but that is not because of the bell curve whatsoever.
Secondly; how? How would they boost a bell curve in a system based around how well you do, unless they’re doing it on purpose?
And in that case, why? It doesn’t really do much.
Personally, I don’t like that idea of so many people being in the same spectrum. I think the system should discourage duo-ing beyond a 500, maybe even a 300 SR difference.
but putting someone whos at the bottom 20% of the ladder vs someone whos the top 20% of the ladder is pretty gross. (1900 vs 2900)
the point of a bell is to define it to brackets ie, 5~10% get to play together, 10~18, 19~30, etc. fair brackets to elicit good quality games. when you can take from huge ranges it defeats the purpose of having the bell in the first place.
otherwise whats the point, because clearly the 1900 and the 2900 are in your 2500 rating game anyway. same skill, doesn’t matter. have at it.
it will matter none whatsoever that the 1900 gets decked or plays mercy and tethers one person, or that the 2900 plays dps and stomps your whole back line. those things have no impact whatsoever in game, after all blizzard says both teams have a fair 50-50 going in despite someone being a 3rd wheel and one person being a normal person at a dwarf convention.
Bad matchmaking, people at same rank but with highly varied skill, placing people in the middle rather than low.
These factors lead to toxicity and a mismatch between those playing for gold weapons and those playing to see how far they can go.
It is easy to push a higher bellcurve, and the biggest reason is blizzard wants people to “feel good” about their ranking.
Why do it? If 70% of players think their rank is fine, 20% think they are amazing, and 10% think they are low, then egotistically speaking blizzard believes the general happiness should be good.
I disagree with this sentiment, but I’ll leave those reasons for now as this is already long enough.
That’s not a problem with there being a bell curve. That’s a problem with skill being really, really hard to quantify.
Humans are weird, different people are good at different things, and they can be good or bad depending on arbitrary things like exhaustion or even just the temperature of their rooms. It’s impossible for any system, or even just any person to be able to factor in this stuff.
If a silver player played like a gold for a good couple of games, how is the system to know that that person didn’t just improve a bit? If a plat played like a gold, how is the system to know they are better than they were?
which is why the gap between SR spreads needs to be tightened from 1000 to a more fair 250.
BUT THAT PREVENTS BOOSTERS SO ITS NOT HAPPENING.
but if you “want a good time with your friends” you have quickplay/arcade just for that.
“but we want to play competitive together” TOUGH, you two who are 1k+ SR apart can not duo together because it defeats the purpose of the competitive nature of the game.
You’re acting as if they are definitely pushing a bell curve, when there is no evidence to support it.
Furthermore, I’m not even sure if your hypothetical reasons make a lot of sense, considering most players (in my experience) think they’re better than they are and that they’re playing with morons. Wouldn’t they try to push for higher, like for diamond?
they dont force it, it just happens to work that way in its entirety, if everyone were to start aimbotting and played another 1000 games each the curve would still exist in its entirety, skill does not outweigh the flat gains and losses that would be experienced by everyone. someone must win, someone must lose, and for every diamond players 8 more have to be below him in bronze/silver etc. you would NOT experience a larger diamond pool and the vanishing of bronze/silver. the top 10% would still be diamond.
someone must be in each respective category, you can not have an enlargement of the top without a much larger bottom pool of players.
go look at the korean rankings, same bell curve, it isn’t pushed up to plat/diamond despite them being able to kick our butts. same average 2k rating 2350 season high.
this is exactly how it functions, someone must fill the ratings, they do not disappear just because everyone else is suddenly really good.
this is exactly how it functions in every other online ranking system like dota or league.
…what? lol well yes 100% of players who play competitive, are in some ranking, but 78% of that 100% are able to play together and defined as “similar skill” when its impossible for 78 out of 100 people to be equally good(or bad) at something…
if you picked 100 people and told all 78 to make 5 free throws in 10 attempts… not all 78 would make all 5 and some of the 78 would make more than 5. But in Overwatch, the current rating system says that all 78 are equally as good (or bad, however you want to look at it) and can make generally all 5 shots…
Actually it’s more like a handful of the people will miss all the throws, a handful will land all 10 or at least most of them, and the majority will land a couple but not all of them. Some people landing four, some landing six, etc.
And I’m pretty sure that’s how that experiment would end up.