19 replies to this topic

[shopsmart]

Figured I would put this here as seems more appropriate. Now that we have tiers listed. What is the actual percentage of players in each tier? LIke is 50% of community tier 3, 20% tier 2, 1% in 5, etc...

[maxmarechal]

i suppose only PGI would provide an accurate answer ...
is it really important? tier level is only a statistic used by MM to try to provide a balanced team...doesn't reflect necessarily our attitude or teamplay on field..

Edited by maxmarechal, 23 September 2015 - 04:32 AM.

[RagingOyster]

I would imagine at least 30-35% are tier 3, maybe another 50% in 4-5, 10% in 2 and 5-10% in 1?

Just my guess.

[Torezu]

There's a fairly easy way to do this. Start a forum poll where people anonymously report their PSR tier. Look at the percentages. Adjust for forum users more commonly being higher tier than lower, and you have a general idea of the division.

[MechWarrior3671771]

Wouldn't it follow that they are 5 groups of 20%?

[Bilbo]

Fenrisulvyn, on 23 September 2015 - 07:52 AM, said:

Wouldn't it follow that they are 5 groups of 20%?

No. Would make more sense that 5 and 1 are lowest percentage, while 2 and 4 are a higher percentage and 3 would be the highest percentage.

[Torezu]

Bilbo, on 23 September 2015 - 07:55 AM, said:

No. Would make more sense that 5 and 1 are lowest percentage, while 2 and 4 are a higher percentage and 3 would be the highest percentage.

It really depends on what PGI is trying to do with the rating. If their purpose is to put people in skill levels specifically for the matchmaker, then yes, 5 equal tiers would work. A bell curve distribution, on the other hand, would make it easier to match Tier 3 players but very low numbers on either end would make for difficulty in matching those people.

[Not A Real RAbbi]

Bilbo, on 23 September 2015 - 07:55 AM, said:

No. Would make more sense that 5 and 1 are lowest percentage, while 2 and 4 are a higher percentage and 3 would be the highest percentage.

Agreed. Statistics 101. Give or take, 68% fall within 1 standard deviation of the mean. Another small percentage will be below or above but within 2 standard deviations (tiers 2 and 4), and a very few more will be more than 2 above or below.

But that makes some assumptions that are not necessarily true. May as well use some more arbitrary numbers.

[Tarogato]

Paul said that the distribution is not a bell curve. That only leads us to assume PGI is trying for a 20%/20%/20%/20%/20% distribution. Which is frankly... a terrible idea unless your only goal is "instant matchmaking".

[jss78]

You hear anecdotes about match-making being slower for high-tier players, though, so I'd doubt it's an even 20-20-20-20-20. You don't hear the same from low-tier players ... I might conjecture that it's something like 25-25-25-20-5.

Edit: math fail

Edited by jss78, 23 September 2015 - 05:26 PM.

[Spike Brave]

A lot of people like to compare MWO to BattleTech in balance and match making discussions. In this case, I think there is a great comparison. I think of it has the unit rating system so 5 would be green, 4 would be regualr, 3 would be veteran, 2 would elite, and 1 would be the ranks of legendary MechWarrior like the Black Widow and Kai-Allard Laio. It makes sense that tier three would be the largest group with the groups getting smaller as you step up or down the tiers.

[Spare Parts Bin]

I wonder what exactly they are trying to do with this Tier system?

[NeoCodex]

Ask Russ on Twitter. It's the go to place for questions like this. Just not here. He will surely give the answer. Go, ask him. I'm interested too.

[MechWarrior3671771]

"I wonder what exactly they are trying to do with this Tier system?"

It helps create more balanced teams than the old system. The grouping keeps very good players from playing with or against newer players, and leads to closer games with less 12-0 stomps

So matchmaker puts together teams from tiers: 1 2 3, or 3 4 5, or 2 3 4

[Tesunie]

It sounds like the tier system goes by a set "number" you earn, and places you into a tier based on that number.

Example: Tier 1 maybe 0-100. Tier 2 might be 101-200. Etc. If your (random number for example) PSR is 125, you'd be assigned T2 rank. If it was 99, you'd be upper T1. Etc.

This I think is how it's done, so it's how much you've "earn" your tier to be. There is no "set" percentage (such as 20% in each tier). I do suspect like others though that most of us fall within T3, or just outside it (T2 and T4).

Though, I wouldn't be surprised to find T5 to be largest, as those would also include people who DC from a match because "they didn't like X" (or just have a lot of technical problems). This would not necessarily be a statement of their skill, but a lack of effort playing the game or issues they have with the game. (Such as even poor computer performance. Some people play at 10 FPS or less!)

Edited by Tesunie, 24 September 2015 - 05:53 AM.

[Motroid]

Spare Parts Bin, on 24 September 2015 - 05:15 AM, said:

I wonder what exactly they are trying to do with this Tier system?

By the way how it is implemented in the game-client I assume it is for entertainment. It adds substance and flavour to the game and motivates the grind for some users.
I really don't think there is much more to it. No separate PSR for weight class? Sorry but I don't think this is very serious.

[Tesunie]

Motroid, on 24 September 2015 - 06:02 AM, said:

I really don't think there is much more to it. No separate PSR for weight class? Sorry but I don't think this is very serious.

That I'm finding to be my issue. The PSR looks great as an overview, but I don't always preform the same from weight class to weight class. Actually... not even from mech chassis to mech chassis. Okay okay. Not even from Mech chassis variant to mech chassis variant!

So far, I'm upper T4, and just a tiny gap away from T3. This is probably because since the PSR system has been rating me, I've been fielding new mechs to myself. Can you guess how that's been going?
Still. Having fun in the game no matter my tier ranking.

[Spare Parts Bin]

I want to make it to Tier 4 then 3, then we will see.

[Vlad Ward]

Tarogato, on 23 September 2015 - 03:50 PM, said:

Paul said that the distribution is not a bell curve. That only leads us to assume PGI is trying for a 20%/20%/20%/20%/20% distribution. Which is frankly... a terrible idea unless your only goal is "instant matchmaking".

It's neither, actually.

Tier deliniations are static and unaffected by population. That is to say, having X number of "PSR Points" will always correspond to the same tier, regardless of anyone else's rank. PSR points are gained on wins, and can either go up, down, or stay the same on losses, so overall players will trend upwards over a large number of games played assuming a 50/50 W/L and a moderate average score per game (250~300).

[Tarogato]

Vlad Ward, on 24 September 2015 - 09:06 AM, said:

It's neither, actually.

Tier deliniations are static and unaffected by population. That is to say, having X number of "PSR Points" will always correspond to the same tier, regardless of anyone else's rank. PSR points are gained on wins, and can either go up, down, or stay the same on losses, so overall players will trend upwards over a large number of games played assuming a 50/50 W/L and a moderate average score per game (250~300).

I get the feeling you missed the point. It doesn't matter how the system works, it's going to sort players into tiers in some manner and the way it distributes them can be described, whether it be 20% population per tier, or a normal distribution.

Actually, I just found exactly what Paul said, so I guess it is closer to a bell curve, but the actual thresholds for the tiers aren't "normal".