33 replies to this topic

[CFC Conky]

SharDar said:

1643046013[/url]' post='6443520']
I really wish that matching was limited to +/- 1 tier. If the tiers are made up of an even distribution of players--20% in each tier--then everyone should be able to get a match in a reasonable amount of time. As a Tier 3 player, I am matched against Tier 1 players, and I don't think that's equitable. One of the issues is that the game is driven by Tier 1 players, whose opinions drive most of the decisions. While that is good for them, I don't think it is good for the game as a whole. Any game players population will comprise a vast difference in playing ability. Providing good gameplay for players at all levels is important to keep a game healthy.

On the other hand, since you can't stop people from creating alternate accounts just to club baby seals, perhaps all this discussion is futile.

Regarding T1 driving the game, I am assuming you are referring to the Cauldron. As I understand it, that group is composed of players of all skill levels. Also, the changes they have proposed and that have been implemented do not really depend on skill level/physical ability in order to use.

Highly skilled players will usually beat lower skilled ones, regardless of mech/weapon changes.

Sometimes you just have to git mo’ better.

Good hunting,
CFC Conky

Edited by CFC Conky, 24 January 2022 - 10:10 AM.

[Michael Abt]

Simplified, you can break down matches into two categories, T5 to T3 bracket and T3 to T1 bracket.

The lower bracket matches are usually decided by the team which has more of the good players on their side. In the higher bracket that is reversed. There the side with fewer of the average Joe players on their side has the advantage.

My conclusion is that T3 is the most difficult tier to be in. In one match T5 to T3 you are one of the top PSR players, dealing high damage, racking in kills, and you are deciding the outcome. Then the very next match you end up in T3 to T1, you are the weakest of the average Joes and get killed so fast you don't even understand what went wrong, and you can't learn from it.

[SharDar]

Michael Abt, on 24 January 2022 - 11:25 AM, said:

Simplified, you can break down matches into two categories, T5 to T3 bracket and T3 to T1 bracket.

The lower bracket matches are usually decided by the team which has more of the good players on their side. In the higher bracket that is reversed. There the side with fewer of the average Joe players on their side has the advantage.

My conclusion is that T3 is the most difficult tier to be in. In one match T5 to T3 you are one of the top PSR players, dealing high damage, racking in kills, and you are deciding the outcome. Then the very next match you end up in T3 to T1, you are the weakest of the average Joes and get killed so fast you don't even understand what went wrong, and you can't learn from it.

Yeah, I'd favor a system where you played +/- 1 tier, not 2 tiers. If they made all tiers 1/5 of the player base, then everyone should have enough players to draw on for matches.

On the other hand, it doesn't really matter since people will go slumming in their alternate accounts.

[justcallme A S H]

SharDar, on 24 January 2022 - 09:40 AM, said:

If the tiers are made up of an even distribution of players--20% in each tier--then everyone should be able to get a match in a reasonable amount of time.

On the other hand, since you can't stop people from creating alternate accounts just to club baby seals, perhaps all this discussion is futile.

The distribution is nothing like that, so no, it wont work.

And this whole "mass Alt account to club seals" is just absurd. There is no proof it happens, no high skill player I know does it, it's all just conspiracy from lower tier players to justify their poor games.

Michael Abt, on 24 January 2022 - 11:25 AM, said:

Simplified, you can break down matches into two categories, T5 to T3 bracket and T3 to T1 bracket.

The lower bracket matches are usually decided by the team which has more of the good players on their side. In the higher bracket that is reversed. There the side with fewer of the average Joe players on their side has the advantage.

My conclusion is that T3 is the most difficult tier to be in. In one match T5 to T3 you are one of the top PSR players, dealing high damage, racking in kills, and you are deciding the outcome. Then the very next match you end up in T3 to T1, you are the weakest of the average Joes and get killed so fast you don't even understand what went wrong, and you can't learn from it.

You can easily have T2-T4.

Hence T2 is actually the harder one as you will spend more time playing against T1 but you're skill level, really, is only slightly above average.
At least in T3 you'll end up in a decent portion more of T3-T5 matches.

[Michael Abt]

justcallme A S H, on 24 January 2022 - 01:27 PM, said:

The distribution is nothing like that, so no, it wont work.

And this whole "mass Alt account to club seals" is just absurd. There is no proof it happens, no high skill player I know does it, it's all just conspiracy from lower tier players to justify their poor games.



You can easily have T2-T4.

Hence T2 is actually the harder one as you will spend more time playing against T1 but you're skill level, really, is only slightly above average.
At least in T3 you'll end up in a decent portion more of T3-T5 matches.

Even in a T2-T4 match the basic logic is the same. I'd argue that it also falls into my definition of the T3-T5 bracket for the simple reason that the team which has less players making rookie mistakes has better chances.


Statistics, and how they can deceive you. My current tier is the result of my choice of mechs, not my skill level, as I have demonstrated and posted in that other #lightsOP2 thread. If I choose to play strong mechs I'd be back into T1.

[dubstep albatross]

SharDar, on 24 January 2022 - 09:40 AM, said:

I really wish that matching was limited to +/- 1 tier. If the tiers are made up of an even distribution of players--20% in each tier--then everyone should be able to get a match in a reasonable amount of time. As a Tier 3 player, I am matched against Tier 1 players, and I don't think that's equitable. One of the issues is that the game is driven by Tier 1 players, whose opinions drive most of the decisions. While that is good for them, I don't think it is good for the game as a whole. Any game players population will comprise a vast difference in playing ability. Providing good gameplay for players at all levels is important to keep a game healthy.

Although I have no hard data to support my intuition, my intuition is that the distribution across tiers is not uniform. I believe it is right skewed, with tier 5 being on the left. In other words, as you increase in tier there are less players in that tier. It would almost have to be this way because the friction to advancement increases as one moves up and starts to be out of their element. Even more so given the matchmaking system. I think advancing into tier 3 is much easier by virtue of the tier spread than advancing from tier 3 to tier 1, even if you play at your tier level.

With that said, I believe the intention of the current system was to create a normal-like curve, centered on tier 3. The problem they were trying to fix was that everyone apparently floated up into tier 1 eventually. Take a moment to consider the implications of a population mostly centered in tier 3. That would mean that tier 3 players would mostly be able to match with tier 3 players and that players on either end of the curve would have a bulk of tier 3 players to fill their games. In any event, I do not think the matchmaker has achieved this.

The reality of being in tier 3 is that one can inch into tier 3 by being just better than average. The other end of tier 3 is a different story. The topic of how "wide" each tier is and the true stratification of each tier has been widely discussed and acknowledged. The range of players in tier 1 is referenced constantly, but that range also exists in tier 3. I would argue that it is a starker difference. A tier 4 player poking into tier 3 could be in for quite a shock. A seasoned tier 3 player poking into tier 2 is not going to experience much different of a game under most circumstances.

Another very grim meathook-reality of being in tier 3 is that when population is low, one is a filler for tier 1 games/players. Plain and simple. This has to be true if we accept that the higher the tier, the smaller the active population, and that the goal of the matchmaker is to provide timely matches for all players, thus irrespective of tier. The whole premise behind the tier spread allowance and gradual widening of it is to achieve this goal.

Why do I say tier 3 is filler for tier 1? If a tier 3 game is seeded, it is likely to contain tier 4 players, so the matchmaker would not want to pull in tier 1. In fact, it would prioritize tier 4 and tier 2. It is much more likely that a tier 1 player seeds a game and the matchmaker pulls in tier 3. If a tier 2 player seeds a game, again, it will pull from tier 3 and tier 1. If it happens to pull in tier 1, it will not pull in tier 4 -- thus tier 3 is a filler once again.

I have used data collected in matches to infer population levels at time of play (use of SteamCharts numbers do not really capture the population levels because not everyone uses Steam and, as far as I know, we do not know how many players are playing when not using Steam). I can associate a qualitative feel of the game quality that correlates to inferred low population levels.

The reality of the entire situation is that the current active MWO population is insufficient, at certain times certainly if not all the time in general, to support the wide range of player skill levels.

[ScrapIron Prime]

dubstep albatross, on 24 January 2022 - 03:19 PM, said:

With that said, I believe the intention of the current system was to create a normal-like curve, centered on tier 3.

I recall ASH posting a graph to that effect. Its not a perfect bell curve, but T3 has the most population in it.

And yes, you're correct that T3 players can and do play against anyone. If they're the top dog in T5 matches, they run the field. If they're low totem pole in T1 matches, they're the fodder. You can kind of tell which one you're in by how people move and what kind of builds you see around you at the beginning, even if you don't recognize names. (Doesn't stop you from being used as cannon fodder in some games though, trust me. )

[justcallme A S H]

Nah I've never posted that.

Maybe an example of what it could look like, but not that.

I recently alluded to, rough, distribution. In that, it's limited what Cauldron can say abd what I posted was what other members have before.

[ScrapIron Prime]

justcallme A S H, on 24 January 2022 - 06:35 PM, said:

Nah I've never posted that.

Maybe an example of what it could look like, but not that.

Hmm. Then I wonder who did. I know I've seen stats... or something that looked like stats. (Now I'm worried. )

[justcallme A S H]

ScrapIron Prime, on 24 January 2022 - 07:28 PM, said:

Hmm. Then I wonder who did. I know I've seen stats... or something that looked like stats. (Now I'm worried. )

The limit of what can be said is pretty much HERE

[Tarl Cabot]

justcallme A S H, on 24 January 2022 - 11:57 PM, said:

The limit of what can be said is pretty much HERE

And the math, using avg 20K players from Jarl's list. As a reminder, if a player does not play 10 games in Quick Play they do not show up in the Leaderboards nor Jarls. No multiplier for Tier 5

X=T1+T2= 3076.92 (without T5)

T3=3.5(X)=3.5*3076.92= 10769.23 (without T5)
T4=2(X)=2*3076.92= 6153.84 (without T5)

----

Lets see what happens when T5 = T1+T2

X= 2666.66

T1+T2 = 2666.66

T3 = 3.5*2666.66 = 9333.33
T4 = 2*2666.66 = 5333.33

T5 = 2666.66

(chuckles) Had a few missteps before I figured out which way to approach it but I am sure I screwed something up here...

And again, it is only an approximation since T5 is an unknown, then even then the 3.5 for T3 and 2 for T4 is likely an approximation at that moment. And knowing PGI and the last time they released a graph before the PSR reset, they had also included players who had just played few games in the previous 3 month period. Even for players who had played at least 10 games to show up on the Leaderboard, many players didnt stick with the game. T5 is simply a revolving door.

Edited by Tarl Cabot, 27 January 2022 - 05:24 AM.

[justcallme A S H]

Ye it is close but, not quite, issue is around clarification of data etc

[Horseman]

SharDar, on 24 January 2022 - 09:40 AM, said:

I really wish that matching was limited to +/- 1 tier.

For a while after the reset, it was. The matches were tighter, but matchmaking took longer. Casuals kept screaming at PGI that they don't care about matchmaking quality, they want quick matchmaking, "JUST FIX IT NOW!!!" and got what they asked for...

... just not necessarily what they wanted, but it's not like the consequences shouldn't have been ******* obvious. Some people just lack in pattern recognition.

[MW Waldorf Statler]

Don't ask what the matchmaker can do for you and the team, ask yourself what you can do for the match and the team

Edited by MW Waldorf Statler, 27 January 2022 - 09:21 PM.