37 replies to this topic

[EvilCow]

I experienced the same, it took a while to go through T2 but T1 is steadily increasing.

The system is not optimal, I don't think to even belong to T1. Ideally it should be something like:

T5 - 50% of players.
T4 - 26% of players.
T3 - 14% of players.
T2 - 7% of players.
T1 - 3% of players.

Tiers should be fixed in size, if somebody goes from T5 to T4 then somebody must go from T4 to T5. This way high tiers would actually mean something.

I would also add a T0 tier with like the top 0.1%.

Another idea is to give tiers benefits:
T5 - CB increase, XP increase (useful for new players).
T4 - Discounts on CB mechs (for players still growing in the game).
T3 - Discounts on MC mechs/mechbays (advanced players).
T2 - Discounts on MC colors/camos.
T1 - Discounts on MC decals/premium.
T0 - Actually getting some MCs for each week staying in this tier.

I think will repost this in the proposals forum.

Edited by EvilCow, 10 June 2016 - 05:40 AM.

[SkavenDC]

I am at the other end of the spectrum but noticed the same thing.

I really dug myself to the brink of going Tier 6. It took me 2 months to dig myself out of Tier 5 hell.

That was a little over 2 weeks ago and I am a few hairs away from Tier 3.

Perhaps it is better skill, or better quality games, but this higher Tier is definitely moving faster than the previous.

[Hit the Deck]

EvilCow, on 10 June 2016 - 05:30 AM, said:

T5 - 50% of players.
T4 - 26% of players.
T3 - 14% of players.
T2 - 7% of players.
T1 - 3% of players.

IMO it shouldn't look like that. The tier distribution should follow a bell curve (it was sometime ago according to a poll in the German subforum - yes, they can create a poll there!) in which the Tier 3 players are the most numerous. T1 and T5 should roughly be in the same size and they should occupy maybe only 5% of the population each (together they account for 10% of the pop). Maybe something like this:

• T1: 5%
  
• T2: 20%
  
• T3: 50%
  
• T4: 20%
  
• T5: 5%

But that could only happen if we have a steady influx of new players. Were that not to happen, the curve/distribution would be skewed to the upper limit.

Edited by Hit the Deck, 10 June 2016 - 06:04 AM.

[Wintersdark]

EvilCow, on 10 June 2016 - 05:30 AM, said:

I experienced the same, it took a while to go through T2 but T1 is steadily increasing.

The system is not optimal, I don't think to even belong to T1. Ideally it should be something like:

T5 - 50% of players.
T4 - 26% of players.
T3 - 14% of players.
T2 - 7% of players.
T1 - 3% of players.

Tiers should be fixed in size, if somebody goes from T5 to T4 then somebody must go from T4 to T5. This way high tiers would actually mean something.

I would also add a T0 tier with like the top 0.1%.

*sighs*

Dude, the tier system isn't supposed to "mean something" the way you want it to.

That isn't it's job.

The tier system exists solely as a matchmaking tool. That's it.

Your percentages above would not only flatly not work, but they'd make matchmaking vastly worse than it currently is.

Further, the single most important part of matchmaking? It's protecting new players from experienced players. With fully half the player base in t5, it would be a seal clubbing wonderland.

No, here's the thing.

The tiers need to be roughly even in size, in order to work at all. Small tiers just means they get grouped with off-tier players more often, and not necessarily closely ranked ones.

Hit the Deck, on 10 June 2016 - 06:00 AM, said:

IMO it shouldn't look like that. The tier distribution should follow a bell curve (it was sometime ago according to a poll in the German subforum - yes, they can create a poll there!) in which the Tier 3 players are the most numerous. T1 and T5 should roughly be in the same size and they should occupy maybe only 5% of the population each (together they account for 10% of the pop). Maybe something like this:

• T1: 5%
  
• T2: 20%
  
• T3: 50%
  
• T4: 20%
  
• T5: 5%

But that could only happen if we have a steady influx of new players. Were that not to happen, the curve/distribution would be skewed to the upper limit.

Elo followed a bell curve, that's it's nature.

The tier system is ultimately just a means of protecting less experienced players from more experienced players, no more, no less. It's not going to be the epeen measurement tool people want because as I said above small tier populations will just break the matchmaker more.

[Champion of Khorne Lord of Blood]

EvilCow, on 10 June 2016 - 05:30 AM, said:

I experienced the same, it took a while to go through T2 but T1 is steadily increasing.

The system is not optimal, I don't think to even belong to T1. Ideally it should be something like:

T5 - 50% of players.
T4 - 26% of players.
T3 - 14% of players.
T2 - 7% of players.
T1 - 3% of players.

Tiers should be fixed in size, if somebody goes from T5 to T4 then somebody must go from T4 to T5. This way high tiers would actually mean something.

I would also add a T0 tier with like the top 0.1%.

Another idea is to give tiers benefits:
T5 - CB increase, XP increase (useful for new players).
T4 - Discounts on CB mechs (for players still growing in the game).
T3 - Discounts on MC mechs/mechbays (advanced players).
T2 - Discounts on MC colors/camos.
T1 - Discounts on MC decals/premium.
T0 - Actually getting some MCs for each week staying in this tier.

I think will repost this in the proposals forum.

I'd definitely like to make more cbills a match or spend less in cbills rather than MC stipend, at least if its just a reasonable level of MC unless you mean to have those benefits cumulative.

Still I don't think a matchmaker system that puts you in different tiers to make sure you get matched up in a decent match should give awards to certain tiers.

Besides, if you had the tiers where an entire 50% of the player base is T5 then T5 would be extremely diverse while T1 would be pretty much the same, ontop of entirely breaking matchmaker wait times for T1 players, and having T5 players matched with half the playerbase so new players can be stomped by people who are way better than them but still in the same tier. I kinda prefer our current bell curve setup where your tier is dependent on you rather than it being a big leaderboard when its just meant to set up your matches.

[Hit the Deck]

Wintersdark, on 10 June 2016 - 06:09 AM, said:

Elo followed a bell curve, that's it's nature.

The tier system is ultimately just a means of protecting less experienced players from more experienced players, no more, no less. It's not going to be the epeen measurement tool people want because as I said above small tier populations will just break the matchmaker more.

I agree that a system with ELO should follow a Bell curve.

Regarding the current system, the tiers may not break the MM *if* it uses the players' exact PSR score and not their tiers. It then works like the old ELO system here in MWO.

[EvilCow]

The bell curve would not work in this case, it is not just a statistical distribution of skills, the mass of players are casual players, the tier would be affected by a lot of factors:
- Pure skill
- Time spent in game
- Experience
- Style of play (not everybody plays for maximum performance)
- etc

The mass would be of new\casual\unskilled players. another problem, a fat T3 would expose most players to "top" T1 and this is exactly what is ruining games for T3s AND ALSO T1s.

It is ridiculous getting in games as T1 with so many clueless players, something is clearly not working.

[Champion of Khorne Lord of Blood]

EvilCow, on 10 June 2016 - 06:19 AM, said:

The bell curve would not work in this case, it is not just a statistical distribution of skills, the mass of players are casual players, the tier would be affected by a lot of factors:
- Pure skill
- Time spent in game
- Experience
- Style of play (not everybody plays for maximum performance)
- etc

The mass would be of new\casual\unskilled players. another problem, a fat T3 would expose most players to "top" T1 and this is exactly what is ruining games for T3s AND ALSO T1s.

It is ridiculous getting in games as T1 with so many clueless players, something is clearly not working.

But we need new players who are just starting the game to not go against T1s, once you've gotten to T3 you've had practice and you should understand the game decently by then. Your system doesn't really keep new players protected too well as you've basically consolidated all of current T5, T4, and half of T3, into one tier in your system so they'd be fighting T3s much more often if they were just fighting in their own tier.

Now of course we mention that if you are going with low percentages of population for T1, you run into not having enough active players in the queue at once in that tier for them to not always mix into another tier, so you get bleed over much more than currently, which already happens often enough.

All the factors you mentioned your tier system would be based off are what the current system is already based off of. Issue is that some T1 players either were carried hard, had massive quantity of games at minimally viable quality that allowed them to slowly creep into T1, or you just have stupid people who don't learn but happen to play decently often enough to get there. That or you are playing with some guys who just got back from happy hour.

You cutting the T1 population down to 3% of the game's population wouldn't cut out how many clueless players are in your matches, it just might cut down on how many clueless players are in T1, but due to tier bleed over you'd still have to dig all the way into T5 in your system to hit the T3 players that get injected into some T1 matches so that you can start a match currently.


Really you could just do something like suggest that match making only makes a 1+- spread rather than 2+- spread and trade wait times for closer skill levels without having to entirely redo a system.

[EvilCow]

For each tier except 1 and 5 you would have an upper tier that is about half size and a lower tier that is about double size. That would create a more uniform mix.

The exceptions would be T1 and T5:

- T1 would always be on top but with much slimmer T2 and T3 the quality of games would improve because achieving T3 would require some real work (upper 25%), right now T3 is everybody with their grandmother and grandfather included.

- T5 would be the mass of new/casual players, I agree that there would be differences is skills within that 50% but it is not much different that playing with a T+1 and T-1.

What percentage would you give to new/casual players? less than 50%? I think that if you take the bottom 50% of the MWO players and form games with just them you would get some good and fun games.

A large T5 gives casual players a reason to STAY in the game without being slaughtered by competitive ones too often. It would help retention.

Edit: Percentages may change, it is just an example, but I think that tier sizes should be strictly decreasing.

Edited by EvilCow, 10 June 2016 - 06:54 AM.

[MOBAjobg]

You're right, indeed.

[Brain Cancer]

Dakota1000, on 10 June 2016 - 04:08 AM, said:

At Tier 1, you are essentially supposed to be the big fish of the pond, no one is in a tier above you and you can be matched with people 2 tiers below you. So if you are in Tier 1 odds are higher than you are the best player in the match in terms of average performance than if you are in tier 3 where you can be right in the middle, top, or bottom. As long as you are in T1 and carrying you go up and never down.

The big fish were in the T1 tank as soon as the rankings started. They've been stocking it with smaller fish for the sharks ever since.

[Chagatay]

I noticed this as well but it really doesn't matter.

[Y E O N N E]

My T1 bar filled up way faster than my T2 bar. Took a little over a year at T2, but T1 was three months. That's with me only really playing around the weekends.

Ryllen Kriel, on 10 June 2016 - 02:14 AM, said:

Level up at least three Locusts variants, that will stabilize the experience bar for your tier!

Level up three Mist Lynx.

[Chagatay]

easy peasy...in fact that is what I have piloted 99% of the time since being in T1 and guess what.....the bar is 60% full...

[jss78]

Lukoi Banacek, on 10 June 2016 - 03:59 AM, said:

Minor consideration, but with more reps, more experience, learning more of the nuances of the game, you might have actually gotten a bit better at the game, and as having fun often also contributes to success (better mood and energy can improve performance) your anecdotal evidence be true for good reasons.

I would offer up that its no more exciting at tier 0 anyway, so just play the game to enjoy it, and the speed at which your bar fills will likely go up anyway.

Good point. Most people learn from experience.

Most people who've been playing for years should be "real" Tier 1 players, unless they're unusually bad at learning, or possibly do stuff like play suboptimal 'mechs for lore reasons.

Now I agree the PSR probably has some of the "XP bar" effect going on, but "most people going up in time" is not proof of this, but something that should be expected even in a perfect system.

I'd also observe that there's no shortage of Tier 2-5 founders on the forums. I'm going to assume that most of these people have played the game a fair deal.

Edited by jss78, 23 July 2016 - 06:20 PM.

[Brain Cancer]

There are plenty of people who never should be anywhere near a T1 game that rank plenty high enough to end up there.

Games played or time spent playing never equals skill here any more than it does in an MMO where you get that top-level guy who actually was PLed to max and then wipes your dungeon run seven times in a row with his utter lack of actual ability.

[Satan n stuff]

Kalleballe, on 10 June 2016 - 02:02 AM, said:

Probably a lot of T1-3 having their tier reset due to inactivity. I was matched against better players entering T3 5monts ago than now in T2.

I've been away for months at a time and I've never had a reset.
Maybe if I keep at playing for a few weeks this time I'll actually reach T2, but the game gets stale pretty quickly if you've been playing it for years.

[L3mming2]

EvilCow, on 10 June 2016 - 05:30 AM, said:

I experienced the same, it took a while to go through T2 but T1 is steadily increasing.

The system is not optimal, I don't think to even belong to T1. Ideally it should be something like:

T5 - 50% of players.
T4 - 26% of players.
T3 - 14% of players.
T2 - 7% of players.
T1 - 3% of players.

Tiers should be fixed in size, if somebody goes from T5 to T4 then somebody must go from T4 to T5. This way high tiers would actually mean something.

I would also add a T0 tier with like the top 0.1%.

Another idea is to give tiers benefits:
T5 - CB increase, XP increase (useful for new players).
T4 - Discounts on CB mechs (for players still growing in the game).
T3 - Discounts on MC mechs/mechbays (advanced players).
T2 - Discounts on MC colors/camos.
T1 - Discounts on MC decals/premium.
T0 - Actually getting some MCs for each week staying in this tier.

I think will repost this in the proposals forum.

most of my maches i still get mached with T2's and T3's ... so whats the use of making T1 smaler then it already is...