Very low Bronze 5-- only gain/lose 5 points per game

For a source and the actual truth please see the dev blog I linked.

Idk why people say it’d take longer for a game to pop. Matches are based on mmr and mmr is equivalent to rank. If you can’t find a match now, more ranks won’t change that. The only thing that can happen is players at the low end of ranks will finally get a normal point gain to not discourage their eventual climb

Agreed, creating a new league below bronze will accomplish nothing. Players will just be stuck in their new black hole league and understandably want out. As you pointed out it would also create longer queue times as there are even more divisions to match, so more players in the same MMR range need to be searching for matches at the same time.

I’m not sure what the solution is, but there needs to be a better incentive for someone stuck in Bronze 5 than gaining 5 points per win. I’m curious, players in Bronze 5, are they awarded more points for wins and losses at the start of every new season like most ranks? If this is so, then that at least gives them a window to break out.

I don’t think it would hurt if players in B5 were awarded more points when they go on win streaks to help them climb somewhat faster.

2 Likes

What is making everyone think queue timers would be affected at the lowest ranks!!! Unless as of now, all bronze 5 0 points are treated same, idk…

If a 0 point bronze 5 can be distinguished from another bronze 5 0 points when one is lower than the other, then the matchmaking system won’t be affected with more ranks

The current matchmaking system balances teams off of ranks/mmr, and everything is clear until the bronze 5 0 points mark. At that point I have no idea if the matchmaking system can tell a bad bronze 5 player from another.

I’ll continue to stand by more ranks would help the game for more players than not

I already linked that 8 hours ago, the issue at hand was whether a new league under B5 would help.

If a dev answered my question on how the matchmaking system considers a bronze 5 0 points and another bronze 5 0 points who may be lower, then the argument would be put to rest. Until then you’ll get players who think more ranks=more split players which then means longer queue times

Once you go below Bronze 5 with 0 points, just start the players off on a journey to the deepest circles of hell. Maybe even invert the point gain so losses award you more “hell points” and the top 100 worst players get on some leaderboard plus portraits with special burning flame border.

People would make smurf accounts to get this “reward”, not kidding.

2 Likes

This stuff makes a lot more sense if people would be better at math and understanding how mmr/elo work.

MMR is a comparative value and the value of a rank is comparative to those around it. For the system to work, there has to be some non-0 value. Since the value has to be above 0 and can’t be negative, the relative value of lower (and higher values at the other end) can effectively become something closer to logarithmic rather than sequential as people may expect.

Since most of the progression works around ± 200 points, people think of progression on the ranked curved as being linear, but it’s not linear, it’s a curve. That curve has different bloats of population, the more dense a particular area, the less value specific points have in relation to the given ranks of the populace.
(which is why some games award less than 200 points on some stretches of each ranked division)

The overall MMR distribution is a matter of Rank, meaning one’s placement in the overall population. So when someone makes progress up the rank, they are moving up in placement over other people in the population. Rank 1, rank 200, rank 5,000,000,000 etc. When someone wins a game, the amount of ‘ranks’ over the rest of the players may be the same (such as moving up 1 position) but the visual display of that shift is very different if a rank 1000 player moves to rank 999 instead of a rank 2,000,000,000 moving to 1,999,999,999. The movement is the same, but the ‘feeling’ of adjustment is different.

Most ranks are divided by 5000 points, 1000 points per division with 5 divisions in each league. However, bronze 5 is set apart by 1000 points instead of 5000, so the rate of progress follows a lower scale, and mathematically, already makes its own ‘divisions’ to follow the same rate of change for winning matches.

if 200 points is on a 5000 scale, then that’s 4% progress per game. In a 1000 scale, the whole division would be 40 points if it were only bronze 5. If a win is 5 points, then the scale is 125 per division instead of 1000 and thus creates a different ‘league’ in its matching. The next ‘league’ up would then award a different value, but that could still be by 4% progress per win through those divisions. This in effect splits Bronze 5 into two parts, those below the league threshold (farther apart from bronze 4 than just 1 division) and those just below bronze 4.

That could mean that Bronze 5 is then split into a 125 league, another 125 league, then a 250 league, and then maybe 500 for the ‘real’ bridge to then finally connect back to bronze 4 after crossing the 1000 point gap that visual represents bronze 5.

That’s just a guess of progression as the whole bronze 5 might be in smaller divisions than even that; based on the amount of population in “bronze 5” it may need even more leagues than just the 3 I hypothetically suggested. With the system as it is currently, the formula distributes those “1000 points” into however many ‘leagues’ it may need regardless of the visual label given. The game could take that lump of below-bronze 5 and toss a new label at it, but they can’t keep dividing up each league below the threshold with a specific name.

Bronze 5 having a cap that divides the available value is a consequence of having a set value to divide around the players. If the value were ‘set’ and linear as people expect, than that would cause a bloat of numbers at the other end of the spectrum. Granted, the visual for that wouldn’t matter so much since Grand Masters don’t display points, but rather they do adopt the set rank in the system, but the scope of how far one gets at the top from how from one is from the bottom remains the same ‘distance’ regardless of how numbers get compressed at one end, or stretched at the other. (gm already does experience stretching, it’s just not as visually evident because people don’t feel ‘punished’ at that end) However, the different between having a sink or a push influences the midpoints of progression where most of the populace of the game divide up the most amount of the displayed mmr points.

If below-bronze 5 were to keep with ‘200 points a win’ the rest of the scale would then be pushed outward and start awarding thousands of points a win instead (200 being 4% of 5000 means that a 5-point value for a 125 point scale could cause 5000 points to be the new 4% of the adjusted scale upward)

So in that way, the 200 points still becomes meaningless compared to other players then winning 5000 points, but it also then means that those players would have to sit through the game counting out those thousands instead.

Ranks try to map out ~30000 points of available mmr across a population of millions of players so by it’s design there has to be compression somewhere otherwise the game may as well have millions of points shown for mmr. Having a compressed system is ‘better’ as games don’t have an infinitely growing population (and if anything, players will leave a game eventually) so there needs to be some sort of direction for the points to flow, otherwise it ends up ‘flooding’ the system. So in most cases, it becomes ‘better’ to have a compressive sink at the low end of the scale instead of somewhere else.

Granted, if ‘progress’ is generally somewhat the same in a given league, it could be possible to forgo numbers entirely, and just have devolving ranks instead; Gm becomes league 1, M becomes league 2, D becomes league 3 and so on and so on so the game just adds the sequential number of leagues, and then has stars (or something) that cover about the same 25 steps progress from one division to the next with the system then adding new leagues each season players fall further and further away from league 1.

But again, so long as a game has a populace divided by millions, there’s going to be compression in the display of those ranks somewhere.

3 Likes

Dear Xenterex, I think less than 5 people in the HotS world have enough IQ to comprehend your post fully. I am certainly not one :smiley: :smiley:

Consider that you are after all, writing to thousands of people whose main argument is “this is stupid nerf it, blizz sucks!!11111”

Pearls before swine. But great post, I now know more of why there is a “hidden” league under Bronze 5, we just don’t see a visual league rank for it.

1 Like

So you don’t want more ranks added then😭

The issue with that is you’d end up having lots of rainbow games. It’s like top GMs having 4 low masters on their team nearly every game, only in reverse. So it’ll be like 4 Bronze 5s having to carry a Mud 3. Since the goal here is to make B5s feel better, it would probably be counterproductive to give them a visibly lower rank than their teammates. Plus, from a design perspective, it would feel so ugly to create an entire rank that may only contain like 10 players.

The debate is gonna come down to would you rather have little progress trying to climb out vs being in a visual lower rank and gain regular points

The mmr-rank being equal means you’d get the b3 and low b5 matched up anyway, title of the rank changing means nothing

The scenario posed is basically what LoL did by adding iron and another rank tier. Mmr won’t change, but visuals will. Matchmaking will still work the same as mmr didn’t change. Lower ranked players will be lower than bronze so the risk is they’d be sad, but if it’s a small subsection in the new lower tier, bronzes may actually feel better they’re not the lowest tier anymore

Keeping in mind that for the people in the lowest ranks nearly every game they would visibly be the worst on the team. If the ranks are kept consistent, there may be a 2-3 division difference between such players and their teammates.

Meanwhile, the tighter tethering of displayed rank and MMR grew out of plats complaining they had more golds on their team than the enemies. So, if I were a B5, I’d probably not take the extra ranks deal.

Ranks are an abstraction as is and there’s already 25 of them for a vast collective of people that don’t care what they represent, they just want to feel better for themselves out of magical ignorance, and blame anything else that doesn’t suit them by lumping it into overgeneralized fluff to demonize.

Part of the issue is that sort of mindset seems to frustrate one’s ability to actually attain the victory that they want.

People toss out terms like “equal” without bothering to think through what would actually be equal, so the amount of ‘ranks’ tends to become arbitrary since the visuals are just an abstraction as is. If people wanted to be better informed about mmr/ranking as is, then they’d put in a bit more effort to understand it (usually indicated by inquiry instead of assumption, accusation, and complaint) However, since [their] usual go to is (as arFrog put it) “this is stupid nerf it, blizz sucks!!1111!” [they] are not generally the sort to want to understand the visuals in front of them anyway.

There’s two key complications to an mmr system for people that want things to be “equal”

  1. The game doesn’t match based on probability, it uses fractions. What we might think of as ~45-55% might be 5/11 to mmr; that is to say, out of 11 games, it could expect someone to win 5 of those.

  2. People don’t have the patience/interest to play out those 5 out of 11 games before seeing an adjustment to their mmr, so what they want is already in conflict with what the system displays and creates inequality for the matching

  1. Matching rating systems are a prediction of who is going to ‘win’. However, a section of the player base are essentially competing to see would will lose instead (intentionally via trolls and unintentionally via players initially placed too high; the ‘k’ value of elo is always an assumption) That sort of conduct undermines the system’s ability to create ‘equal’ matches and it has to create larger and larger spaces between the comparative value of players that it has a positive confidence in their capacity to win compared to those it has a negative confidence (or uncertainty)

In a comparison of groups, something is always going to be a ‘worst’, but people don’t like that reality, so they demand more and more abstractions until they don’t understand what it is supposed to be, but they’re still unhappy because they don’t have what the they want. If a system compels people to ‘start over’ to get out of being at the bottom of the hole, then there’s only so much interest in catering to them as they just feel like a sunken cause.

“I spent x on this game, how dare it force me into bronze 5 where I can’t get out”

The amount spent doesn’t have an ‘equal’ bearing on their expectation of the value it should have for them, so their expectations are their problem regardless if the rank is ‘bronze’ or ‘wood’ or ‘iron’ or whatever; it’s a problem of depreciating value for people that demand ‘equality’ but don’t appreciate that value in it. The more seasons that go by, the larger the gap will became because of the difference in direction between winning increasing number of matches compared to not at different relative values that aren’t ‘equal’.

And granted, that’s a big thing with business anyway; the best ‘deal’ is when both traders feel they got the better end than the next person because ‘value’ is subjective anyway. (ie $5 doesn’t have the same value to everyone) So then it becomes a problem of ‘quality’ it trying to appease a system that either frustrates people out of their own ignorance, our frustrates them out of their own greed.

But fixing any sort of problem rarely comes down to actually being ‘equal’ because that is a value of perception, and not so much one of objectivity.

2 Likes

There’s enough players in low bronze 5 to cause them to complain on the forum. Making 2-3 ranks below bronze could probably keep enough of those former lower bronze from complaining they can’t climb from their low point gain. Iron in LoL is actually smaller than masters so only a very few players would complain they’re lower tier. And even less would complain they can’t climb (the new low bronze 5) as the players in the new low bronze 5 would be much less than before. It might’ve been a good decision to make a select few feel bad (iron tier) to boost everyone else’s confidence. In fact, from a reddit post, iron players tend to be chill and not care about ranked. So this means you’ll get the players who don’t complain in the lowest rank, and leave the former bad bronze 5 players in a higher tier assuming the bronze tries and won’t still end up in the lowest tier where no one cares about winning. Everyone wins for the most part

You’d still get complaints, but I think there’d be way less complaints. Instead you’ll get complaints bronze is paired up with lower ranks but the matchmaking would’ve paired them up anyway in the current algorithm. Knowing blizzard they tend to do less than more with the exception of the anomalies which they will stop at the end of the year

Complaints indicate dissatisfaction, not necessarily that something is problematic. Some systems have ‘problems’ as an incentive for people to change, but they’re the sort that’ll just complain and demand the system changes around them instead.

Visually, if the game ‘still matches them’ (which is an assumption) then it doesn’t matter. If they’re put into a ‘different’ league and still matched, then it fuels the complaints about ‘rainbow matching’ and just leads to ‘more complaints’

If the ‘complaint’ is the same regardless of the outlet taken, then there’s not as much incentive to act on it accordingly. As I wrote before, the system as is already makes different ‘leagues’ as is, so either the whole collective is given a ‘new label’ or they just keep the current one and still acknowledge that it is the ‘worst’ in a system that’s still going to have depreciating value.

So having Bronze 5 being the ‘void’ doesn’t change that there’s going to be a ‘void’ unless the system overhauls the entire rank again as per what I concluded with on my previous post.

I’d be inclined to think you didn’t read that part, but want to reply anyway. And that’s pretty much a hefty chunk of the ‘bronze 5’ problem; some people don’t want to know their particular situation, they’re just upset because they’re not granted the rank of the ‘best’ as is. So trying to appease that sort doesn’t fix their problems, and arguing on that behalf is inconsequential.

IF the problem is that the display isn’t ‘equal’ for everyone, then the only way to fix that would be to create so many ranks to perfectly map for all the ‘compression’ the system would have anyway.

But if the system is going to have compression, then the visual flares don’t matter so much. At the point, its better to just try to inform people and if they want to be informed, they can use that to influence their perception of ‘progress’ but if it’s ‘progress’ some people want, they it helps if they even know which way to hitchhike before complaining that they ended up going the wrong way.

At which point, that’s still a matter of people being willing to get informed so they can adjust their orienteering skills.

With your reasoning about having a “void” there wouldn’t be a for or against more ranks though. In fact, if the devs chopped off bronze or any other rank to have less ranks, you’d still have a void except now even more players would complain about being stuck in a “low silver” rank

You’d try to say “actually you’re still same as always and you’re just lower than what we can visually display.” But you’ll still eventually climb!

I did find an old post you made to someone else though

With points being representative of a player’s rating, that’s the standing in relation to all the rest of the playerbase. If you’re the 2,000,000th rated player and win a game, you could go up 1 ranking by winning a game, and then be the 1,999,999th player; the if 1000th player wins a game and moves up to 999th, the progression is the same (1 rank) but the movement in relation to the rest of the populace is vastly different; one player still has 1,999,9998 ranks ahead of them.

With the distribution of players at bronze 5, they have that many more ranks to climb (per mmr) in line. While a UI change could help map the progression better visually (as a league within a league within a league) the effect of change in the length of progression would be similar.

If there were a bypass to that, then that leads to stat inflation being interjected into the system if players can ‘skip’ and then lose and get kicked by down: higher ranks just get that much further ahead if the mmr isn’t representative of player ranking, and just a collection of points.

Part of the issue of the game’s previous leagues is that there was stat inflation before as initial placement and seed-borrowing from other leagues allowed players to jump in too far, and then fall down, thus creating a disparity in the upper and lower ranks with mmr distribution. Remapping Mmr distribution to represent the player rating thus causes a kickback (visually) where the point inflation is offset with a point sink at the low end. ie, less points shown per game.

Losing a game still causes a similar placement change, but the numbers mapped to represent that narrow so there doesn’t become a negative infinity in regression.

In that regard, it may be easier to visually understand if the league distributions were representative of a populace rating (Like grandmasters, but as a range for the whole system) so people could see if they ‘climb’ by moving up rating ranges (being 2,000,000th or 1,900,000th) but the slope of the bell curve is still a similar climb, it’s just how people see it represented by a ui.

Idk I just think more ranks at the lower end would be good but oh well. In the end everything would be the same but I’d think customer satisfaction would go up. It worked for LoL at least

when a post was made 2 hours ago in the same topic you’re already posting in, it’s not an ‘old post’. I would also hope you didn’t go lurking through my post history to ‘find’ that here. Esp since it was posted at the topic as not ‘to’ someone :confused:

It’s also the same post that you replied to before and indicated that you hadn’t read at that time. The question you have posed 3-4 times now would have been ‘answered’ if you ‘read’ my ‘logic’ through the first time you went over the post :stuck_out_tongue:

The issue of players complaining about the bronze thing is that they aren’t at the top of the game, so long as they aren’t considered the best (cuz they think they are) they’ll complain about being in any rank less than what they expect. Part of the concern isn’t that they complain, it’s how they complain. The medium is the message after all.

1 Like

You posted that block in February and I found it by googling “make more ranks below bronze heroes of the storm”

More ranks below bronze might not fix anything, but I think it’d just feel good. Would you have objected to the new ranks made in LoL too?