What has happened to Tier VIII?

HowTheStoryEnds wrote:obviously the sample size is to small to be statistically valid

THIS and ONLY this!
There is no but to it, if you like to talk about those statistics and the MM go and play 1000 games in the same tank during the same patch under the same conditions and keep track of what tiers are played the most overall on the server and then compare it to your results in the 1000 games.

Unless you do that there is nothing to discuss about!

Nah, thats obviously too boring an answer, Blackchamber. Cant be that

Even if i quit playing for today after a loosing streak for today i have to agree with blackchamber. My 7 days stats got better topped yesterday even the fact that i started playing 2 new tanks and most of the time,play tier 6 which is higher than before. MM distribution is normal in my opinion. Yes I see tier 8's in a tier 6.

blackchamber wrote:

HowTheStoryEnds wrote:obviously the sample size is to small to be statistically valid

THIS and ONLY this!
There is no but to it, if you like to talk about those statistics and the MM go and play 1000 games in the same tank during the same patch under the same conditions and keep track of what tiers are played the most overall on the server and then compare it to your results in the 1000 games.

Unless you do that there is nothing to discuss about!

Not entirely true. Yes, statistics rely on a big number sample, but that doesn't need to be from ONE player only. Also, within statistics there's such a thing as trends. If you take 10 players and look at their last 100 matches (which is rather easy over a couple of days) you also have a thousand games. Since most people follow the same pattern of playing, it doesn't need to be ONE tank either. Just similar to what they were doing before. Sure, if 10 people suddenly started playing only arty where they never played arty before, their stats are useless, but that isn't the case here.

Last but not least, there's also human judgement and "gut feelings".
As I said, my stats went up, but I too had the feeling before this topic had started that I get dumped in a whole lot of being "lowest tier" lately.

blackchamber wrote:Unless you do that there is nothing to discuss about!

How is it not productive to discuss performance in a learning environment?

There has been a noticeably drop-off in my WR the past week. And even before this thread, I have noticed a fair amount of other students complain about the same thing. Stats do seem to back it up in this case, which rules out confirmation bias, so what is going on? Note that this question does not mean I am looking for a hidden agenda from WG trying to make us lose. I am simply trying to figure out what possible external change has caused the sudden shift in WR in several people in SGTA in the same week. It could simple be coincidence, but to dismiss it as such prematurely, robs me of trying to find a cause and attempt to adjust my gameplay.

There may well have been a subtle shift in the metagame that some of us have not automatically adjusted to as well as other players. After the accuracy changes some players adjusted to it quicker than others did. I think I benefitted from that overall, because I like fighting at range and I try not to subject myself to risky exposure too much. I think players that tend to fight at close range and attempt risky pushes to get in close may have had a harder time adjusting to the accuracy changes, since people shooting at them as they close the distance are more likely to hit now.

It is things like this that are not immediately apparent and I think it is good trying to analyze and figure out the causes of such a sudden drop in performance. If it comes down to just plain bad luck, then so be it. But I don't want to immediately dismiss any other cause, just because it *might* be bad luck and forego the chance to identify a weakness in my gameplay and improve myself.

Just saying  

This may not apply being that match maker gives max tier 9. but last night i played maybe 6/7 games in KV5 and every one had at least 3 tier 9's in the game each side.

IronClaw wrote:
Not entirely true. Yes, statistics rely on a big number sample, but that doesn't need to be from ONE player only. Also, within statistics there's such a thing as trends. If you take 10 players and look at their last 100 matches (which is rather easy over a couple of days) you also have a thousand games. Since most people follow the same pattern of playing, it doesn't need to be ONE tank either. Just similar to what they were doing before. Sure, if 10 people suddenly started playing only arty where they never played arty before, their stats are useless, but that isn't the case here.

Last but not least, there's also human judgement and "gut feelings".
As I said, my stats went up, but I too had the feeling before this topic had started that I get dumped in a whole lot of being "lowest tier" lately.

Maybe you dont need to be as strict as i suggested, but i did that for a reason. First of all the talk in this thread was about the MM distribution and i referred to that alone, not any other statistics some might have in mind(winrate etc.). Since none of us know the exact way the MM formula works my suggestion towards testing it was to rule out any possible factors. For example taking only 1 tank: Different types of tank (medium, heavy...) might be treated differently since the Matchmaking tries(at least i think thats the case) to put an eqal amount of each type of tank into the match. Again I only said it this way because none of us know the exact MM formula, so you have to rule out all possibilities that you can.

About the gut feelings: They are exactly that! when talking about statistics I dont see how gut feelings help in any way. For example we were platooning yesterday(You, me and rancidpunk) and lost 5 in a row. Then lycantia joined us and i believe we won 6 in a row. My gut feeling tells me to platoon with lycantia all the time now!  

nilzatron wrote:How is it not productive to discuss performance in a learning environment?

There has been a noticeably drop-off in my WR the past week. And even before this thread, I have noticed a fair amount of other students complain about the same thing. Stats do seem to back it up in this case, which rules out confirmation bias, so what is going on? Note that this question does not mean I am looking for a hidden agenda from WG trying to make us lose. I am simply trying to figure out what possible external change has caused the sudden shift in WR in several people in SGTA in the same week. It could simple be coincidence, but to dismiss it as such prematurely, robs me of trying to find a cause and attempt to adjust my gameplay.

There may well have been a subtle shift in the metagame that some of us have not automatically adjusted to as well as other players. After the accuracy changes some players adjusted to it quicker than others did. I think I benefitted from that overall, because I like fighting at range and I try not to subject myself to risky exposure too much. I think players that tend to fight at close range and attempt risky pushes to get in close may have had a harder time adjusting to the accuracy changes, since people shooting at them as they close the distance are more likely to hit now.

It is things like this that are not immediately apparent and I think it is good trying to analyze and figure out the causes of such a sudden drop in performance. If it comes down to just plain bad luck, then so be it. But I don't want to immediately dismiss any other cause, just because it *might* be bad luck and forego the chance to identify a weakness in my gameplay and improve myself.

Just saying Laughing

Sorry, but i havent seen much talking about performance in this thread only about the Matchmaking.
It seems i have to say it again: You experiencing a drop in your winrate and having other SGTAlers experiencing the same thing over the time of one week does NOT equal being backed up by the statistics.
Hell, if you look at my winrate development over the last 2k games it looks more like a rollercoaster-ride then anything else!

To your question what the reason for this recent drop might be: In my opinion its down to the random nature of the game. A few days ago i had a run of 18 games in my T54E1 and won a staggering amount of 2! The reason for that? Plain and simple tomato team after tomato team.
Might it be because of the changes to the accuracy as you point out? Maybe, hard to tell, but why would this effect kick in just now? The changes have been in effect for a longer time already.

Most importantly dont forget the most important factor for your winrate guys: YOU
You cant expect that your winrate just rises and rises. At some point there is a skill cap and the winrate wont improve anymore, unless you up your gameplay!
Nilzatron your winrate for example is at 54% already. I am not saying you cant get it up, but improving on a already high winrate is obviously harder then improving on a low one. Also keep in mind that a lot of you dont have that many battles, which means a streak of derp teams has a bigger effect on your winrate then on mine for example(same goes for winning streak obviously).
My advice is this: If your WN7(or WN6) rating and your average damage output(according to your average tier) is better then the last 30 or 60 days, there is nothing to worry about!

[quote="blackchamber"]

IronClaw wrote:
Maybe you dont need to be as strict as i suggested, but i did that for a reason. First of all the talk in this thread was about the MM distribution and i referred to that alone

Ok, I agree that for that, just our few matches are too few to *prove* anything about the matchmaker. But that doesn't mean that when a lot of people suddenly have a similar feeling, there can't be an actual, factual, cause for those feelings.
Also, I am not saying there *is* a change to the matchmaker, just that it wouldn't surprise me if there was. I'm not the kind of guy to jump on some forum and shout "omg MM changed!" when I get 5 bad games in a row. I've had the feeling something changed for quite a while, but thought it was probably just me, or just coincidence. But other people have noticed the same and voiced it independently of me even bringing it up. Could still be nothing, could also indicate a trend.

Let me phrase it in the opposite direction: How many "evidence" or "witnesses" would you think is enough to warrant suspicion? As you said, we don't know how the matchmaker functions in detail, so all we can do is guess. But lets say you get put as bottom tier for 95% of your matches today. Just coincidence of course. And the next day, 95% again. Which is pure coincidence. And the next 4 days again. Your gut feeling now tells you that something weird is happening, but that is just that: your gut feeling. Then your friend tells you that for the past 7 days he's been bottom tier in all his matches or at least that is how it felt. Too small a sample to really conclude anything of course. 10 days later you have been experiencing this 95% bottom tiers for over two weeks now. There's a load of posts on several forums from people complaining that since a week or two they feel like they are constantly being put as bottom tier. But that is just two weeks of data, and not even raw measured data, just gut feelings and too small a sample to statistically say anything. ... or would you say a trend has been spotted?

You catch my drift of course. I totally agree that *probably* there's nothing changed at all, and its just people adjusting to a new patch, the matchmaker dealing with a lot of Tier 9 and 10s that were dusted off after the amount of arty has dropped. But you can't just put every argument down with "not enough data, put on your tinfoil hat!" either

IronClaw wrote:
Let me phrase it in the opposite direction: How many "evidence" or "witnesses" would you think is enough to warrant suspicion?

Way more then we have! Also some real evidence and not just people stating their experiences.

IronClaw wrote:As you said,we don't know how the matchmaker functions in detail, so all we can do is guess. But lets say you get put as bottom tier for 95% of your matches today. Just coincidence of course. And the next day, 95% again. Which is pure coincidence. And the next 4 days again. Your gut feeling now tells you that something weird is happening, but that is just that: your gut feeling. Then your friend tells you that for the past 7 days he's been bottom tier in all his matches or at least that is how it felt.

Are you stating facts or is this just an example?
Also keep in mind people in this thread have been talking about 2 different things at once. One part talked about the Matchmaking the other part about the winrate or a mixture of both!
Lets stick to this 95% bottom tier example(?). Do you realise that this means, someone else had to be top tier 95% of the time(or multiple people top tier at a higher percentage then average), which leads to the question why is this happening if its not random? The only real answer for that would be that you are treated differently from the Matchmaker, which for me leads to the question: Why the hell should the Matchmaker treat you differently? There is just no logical reason to do that at all.

Ironclaw wrote:There's a load of posts on several forums from people complaining that since a week or two they feel like they are constantly being put as bottom tier. But that is just two weeks of data, and not even raw measured data, just gut feelings and too small a sample to statistically say anything. ... or would you say a trend has been spotted?

And how many posts are there where people praise the Matchmaking because they have been top tier for 10 straight matches in a row? My guess: 0 Conclusion: Nobody is top tier at any time in any match!(Hax! Cheat!)
Of course people complain only about being bottom tier, but they have done it before the last few weeks and they will continue doing it. And i would say no there is no trend spotted, because of a few dozen(even if its hundred) people reporting on the forum that they are 'experiencing' unfair matchmaking out of 100000+ players online on a daily base.

Just to add my two cents...

Over the past 3 days I have been grinding the SU-76. The first 3 battles were straight tier 3 battles, the next 2 were a mix of T3/t4, every battle since then has been a tier 5 battle, that's a run of 29 straight battles, 4 of which I was the only T3. Clearly not a sample size that is statistically valid, but 29 straight battles in a row, no variance?

So I switched to my trusty Cruiser MkIII, 1 tier 2 battle, 9 tier 3.....

KV1S 8 battles, all either t7 or t8....

I'm not complaining and I know that this is not a statistically valid sample but it has contributed to my declining win rate as I have failed to influence the battles enough to change the outcome, except in the KV1S.

nilzatron wrote:And even before this thread, I have noticed a fair amount of other students complain about the same thing. Stats do seem to back it up in this case, which rules out confirmation bias, so what is going on?

The problem is that you can ask any time if people feel currently trolled by MM and you will probably always get a number of confirmations because feeling trolled by MM is an omnipresent state for many players. So I agree with the argument that it takes a much larger sample of games, to conclude something about ingame changes.

Another thing is that many people seem to be more emotionally engaged when loosing than winning. E.g. having had 8 losses in a row yesterday inspite of doing well myself in most of the games almost also made me believe something must be wrong with the MM and I decided to have a closer look at the next results. Today playing felt "quite ok", I checked the result list and was surprised to see 10 consecutive wins. Didn't really feel as good as the last losing series felt bad. Winning is expected, losing the unpleasant surprise.

But at least here are some good news for people who got their XP stolen by a poor performing team:

Patch notes 8.8:

- when losing a battle, players, that recieve a “battle hero” award, epic achievement or platoon achievement, will recieve as much XP, as if they actually won (1,5x)

And for those who still think, something is going wrong with the MM:  

- fixed some of the MM problems, that emerged in 0.8.6 and continued thru 0.8.7

More about patch 8.8: http://ftr.wot-news.com/2013/08/17/8-8-patch-notes/

forum.worldoftanks.eu/index.php?/topic/284339-i-have-given-up-trying-to-play-today/page__st__20#top

those players are alot better than me and they to complain about always being low rung.

BTW one can not 'fix' random, the mere fact they mention a fix of any kind I'd an admission that mm here is a lot of things but true random.



1 game ok, maybe even 2 but not almost every game of the past 2 weeks. The teams I've been getting have been incredible and my opposition always is at least blue. This is seriously off the hook.

My jagdpanther is now at an appalling 43% with an average damage of 1453. I'm starting to get to the point that it'll be more profitable for me to start playing the lottery since the chances on those are somewhat of the same.

blackchamber wrote:

HowTheStoryEnds wrote:obviously the sample size is to small to be statistically valid

THIS and ONLY this!
There is no but to it, if you like to talk about those statistics and the MM go and play 1000 games in the same tank during the same patch under the same conditions and keep track of what tiers are played the most overall on the server and then compare it to your results in the 1000 games.

Unless you do that there is nothing to discuss about!

Statistically speaking, 30 matches in the same tank would give you something like a 95% confidence interval
Which given that you can only be in one of 5 states is pretty good.

(eg, a tier 7 tank could be facing opposition of tiers 5-9)

It might be interesting to do 30 platooned, 30 not platooned.

Ive been riding a wave the last 2 weeks.. despite this everything is going up! I hope it will continue to do so overall as I learn. This said I have noticed being lower in the battle lineup and winning less games in random matches. that may just be normal.

Because of this discussion i started a new page on my WoT excel and wrote down my tier, the battle tier and which tanks i used.
So for yesterday 31 battles. Tanks: VK36.01h, Hellcat and Cromwell result:
15 x top tier
8 x tier +1 (so tier 7 max)
8 x tier +2 (so tier 8 max)

win or lose is not relevant for this topic in my opinion and yes I know my yesterdays winrate "dropped" to 56,2% but as blacky mentioned on that high level you it's hard to always improve.

still not really statistically relevant so I just throw some numbers in the discussion

While I haven't played since sunday, I don't think there's anything going on with the MM. Sometimes you have streaks of good luck, and sometimes you have streaks of bad luck. For instance, two weekends ago I played well over a 100 battles and ended up with a win ratio of less than 40% even though I performed at my usual level (WN7 rating of about 17-1800). One weekend ago, I played 41 battles in my Cromwell and ended up with a win ratio of 72% and performed well above my usual level (luck with the MM played a major role) ending up with a WN7 rating of roughly 2550 for those battles.

But I'm a firm believer in that the good and the bad luck cancels each other out in most cases. If you play consistently you'll keep your 60 day win ratio close to your usual performance, and if you continue to improve, you'll improve your 60 day win ratio.

blackchamber wrote:

IronClaw wrote:
Let me phrase it in the opposite direction: How many "evidence" or "witnesses" would you think is enough to warrant suspicion?

Way more then we have! Also some real evidence and not just people stating their experiences.

and

Are you stating facts or is this just an example?

Please read my post a little more careful. That was an example indeed, used to explain the question I posed: "If you feel the data plus comments from users is not enough to detect a trend, then how much DO you think is enough?"
Is 100 matches enough data for you? 1000? 10000? 10000 x 10000 users? And if 10 users stating their feeling is not enough for you, then how much would be? 1000 users saying they feel something odd is going on? 10000? An entire server? Two million users?

The point of my question is: Don't be too easy to dismiss people's feelings as "only gut feelings, no evidence, too small sample". If a lot of people are saying they notice something, there's usually some cause for it. (although not always the cause the people think it is of course)
"Where there is smoke, there is fire"

Again, what I said are the following things:
1) I do *not* claim that something has happened to the MM or that I have any evidence or proof.
2) I do claim that there seems to be some sort of trend going on. Many people are reporting the same feelings. (And some back it up with smallish samples of evidence)

My own feeling (and I've spoken that one out to people on TS long before this topic popped up) is that the MM seems to aim for a full tier-spread more often than it used to. (No clue if this is due to a change in the MM, or due to differences in how many people play in which tiers, or if it's just coincidence)
On a busy night with triple-digit players in all tiers from 4-8 you would expect the MM to have no trouble creating matches with no tier-spread other than that from stuff like light tanks etc. I'd expect to see a match with all tier-6 tanks plus maybe a few lower tier scouts when I press "BATTLE" on my hellcat.
But instead I get game after game that have a full spread +2 or -2 tiers. Additionally I feel I am at the bottom of that spread more often or not, but that could easily be just a feeling and I'm willing to dismiss that.
And this full-tier spread is a problem in my eyes. Sure, I can damage tier 8's in my hellcat. I can damage Tier 7's in my M4. And it can be fun, but not game after game. And it's fun raping some tier 6's in my IS-3, but I prefer fighting against other tier 8's.

Personally, I've been kicking arse & taking names recently, or at least subjectively it feels like I have. WN7 going up, w/r going up. Several new tanks started, one or two old favourites retired. About 40/60 platooning & solo, I'd say.

But then again I'm probably shite enough that I get the soft, cotton-wool coated end of the matchmaking shaft.

If what Ding says is true... I dunno. It boggles my tiny mind that the MM would be able to sort a match by tier & by "ratings" & get 100,000 people into a match within, say, a minute. I'm not disputing that that's the case--if Ding says it, I'm down with it, I s'pose--but the logistics of it kinda turn my head a little.

Khul wrote:It boggles my tiny mind that the MM would be able to sort a match by tier & by "ratings" & get 100,000 people into a match within, say, a minute.

As a "professional programmerer" I can tell you that technically that's not an issue at all.

Anyway, for the sake of statistics, I've started keeping track, just because Black's comments got me thinking about my personal "gut feeling vs facts".

15 games tonight
11x 2-tier spread (not counting any light tanks)
3x 1-tier spread
1x all same tier

Of the 14 "spreaded" matches, I was:
9x bottom tier
3x middle tier
2x top tier

Will keep tracking for a week or two

IronClaw wrote:
On a busy night with triple-digit players in all tiers from 4-8 you would expect the MM to have no trouble creating matches with no tier-spread other than that from stuff like light tanks etc.

I don't think the MM is intended to maximize equal tier matches because it would severly limit the variety of tank combinations, game scenarios and thus make the game less interesting. E.g. at the moment a tier 5 only matchup is limited to a combination of theoretically 35 different tanks. Practically it will be even less because there are a couple of unpopular tanks which are rarely seen and only played to grind through. Encountering the same small set of tanks all the time would be rather boring after a while.

IronClaw wrote:Anyway, for the sake of statistics, I've started keeping track, just because Black's comments got me thinking about my personal "gut feeling vs facts".

A more significant alternative compared to tracking personal matchups might be to simply watch the XVM list at the start of the game. If skill based MM really exists, you should clearly recognize a tendency towards a rainbow like order of colours, i.e. red guys on top, followed by yellow, green and purple on bottom level.

Just imagining for a moment the consequences of a skill based MM on the player base: Power users (probably most of them paying premium members) being annoyed by MM penalties for the sake of supporting newbies and learning-resistant players not able or willing to master the game (how many of those are paying members?). Not to mention an endless shit-storm like discussion about the fairness of the MM skill calculation including I-want-my-money-back requests. WG devs so far strictly denied any plans to introduce skill based MM and I think they have good reasons not to do so.

What I would really like to see is the MM acting like this:

When creating a match it would do the following steps:
1) Select 30 players according to the current matchmaking rules. So you get 30 random players from the queue, with a rough similarity in tank types (to avoid 15 arties, 25 TD's , etc)
This will result into a random mix of good players, bad players, experienced players, inexperienced players etc, just like we have now.
2) But instead of also randomly dumping these people 30 into two 15-player teams, only considering the tank balance, I would like the MM to also consider the rating of the players in each team. So instead of having the chance that all unicums end up on team 1 and all tomatoes end up on team 2, I would like the matchmaking to even it out. If there's 2 unicums in the random 30, put one on team 1 and the other on team 2. If there's 7 tomatoes with 42% WR, put 4 on one team, 3 on the other, and so on.

That would still lead to random people in a match, so the bad players can still play together with the good players and learn, but it would avoid (most of) those boring 15-4 and 2-15 steamroll matches.

IronClaw wrote:What I would really like to see is the MM acting like this:

When creating a match it would do the following steps:
1) Select 30 players according to the current matchmaking rules. So you get 30 random players from the queue, with a rough similarity in tank types (to avoid 15 arties, 25 TD's , etc)
This will result into a random mix of good players, bad players, experienced players, inexperienced players etc, just like we have now.
2) But instead of also randomly dumping these people 30 into two 15-player teams, only considering the tank balance, I would like the MM to also consider the rating of the players in each team. So instead of having the chance that all unicums end up on team 1 and all tomatoes end up on team 2, I would like the matchmaking to even it out. If there's 2 unicums in the random 30, put one on team 1 and the other on team 2. If there's 7 tomatoes with 42% WR, put 4 on one team, 3 on the other, and so on.

That would still lead to random people in a match, so the bad players can still play together with the good players and learn, but it would avoid (most of) those boring 15-4 and 2-15 steamroll matches.

good idea. question: influence of a unicum in a tigerP is bigger on the outcome of battle than unicum in PzIV. should this also be balanced? (so unicum tigerP vs unicum T29 or unicum PzIV vs unicum Crusader)

CountOfTuscany wrote:
good idea. question: influence of a unicum in a tigerP is bigger on the outcome of battle than unicum in PzIV. should this also be balanced? (so unicum tigerP vs unicum T29 or unicum PzIV vs unicum Crusader)

That wouldn't be possible under the idea I proposed, since the total players in the match (30) still get randomly selected. So that could include only two unicums, one in a TigerP and one in a PzIV. However, the MM could give some rating according to player stats + tank they are in.

Lets assume a battle including tiers 5, 6 and 7.
Unicum in Tier 7 = 30 points
unicum in Tier 6 = 20 points
unicum in Tier 5 = 10 points
Or something like that. It could then try to make up the difference in points in the team that gets assigned the T7 unicum compared to the other team that got the T5 unicum by giving the first team more "tomato". But tbh, I don't think we really need to get that far.
If the MM just ensures that it doesn't put all the unicums from the 30 players in team 1 and all the 40% WR guys from the 30 into team 2 like we sometimes see happening now, it'd already be a vast improvement.

Ok, since the discussion subject changed from the tier position in battle to MM system, I have a thought. When you set up a formula to make teams equal by some values for their team members, based on their skill, type of tank, etc. Somewhere the randomness of battle is lost, is it not? Don't get me wrong, but the battle map is always same (you can learn all maps). The tank base is mostly same, as majority of players use same preferred tanks in their tier. So the majority of the randomness come only from the player base itself, does it not? And the more systematic you are the less randomness is in it. At the extreme you can meet same opponents over and over again.
I personally don’t like those 15:0-4 battles to much too, but sometimes opponent is simply better, have better idea, better setup, more luck. Most of these theories about "MM hates me etc." are just power of suggestion nothing else.

Anyway, its just thought, so don’t feel opposed or something . And by the way, I personally like to be lesser tier in battle if I drive solo, as it takes a lot pressure off, when you don’t have to carry your team or choose the exact expected way of play, if you are not the top dog. In toons it is opposite, as a top tier platoon can cut through enemy like knife through butter, if the enemy is low  . Or in other hand, if they have top platoon too, the clash between us is what makes the most of fun in this game.

At first I want a definition for a unicum/good player because the winrate says nothing. Believe me i got a not too bad win rate (classified as greate player  ) but you prefere me to be in the enemy team in a tier 8 battle in my centurion or when looking on my stats with them any tier 7-8 ... trust me.

So is it WN6/7 then? Efficiency/Performance (old/new/wot-news,wtfIdontCare Site with own efficiency/performance formula)? Kiev armor rating? TWR? There are also discussions about what really identifies a good player between unicums (at the moment I am doing a bit of research on the whole stats topic possibly more to come on this later in this forum so i read thru many forums). What is more important? Recent (7 days, 60 days) or overall ratings. Should they be specific for the tank used or taken as general?

The main problem I see in this is as the time needed to actually "match" the battle would explode by taking more and more factors into account ... even if anybody could decide what factors matter her for the matchmaking. Also you might have noticed that when you are in a platoon the matchmaking takes longer than when you are playing random solo. The reason for this is simple the MM tries to match platoons as far as I know (ftr and other sources said at least so). So you see how long it takes to calculate 2/3 persons based on their number and their tier spread. So think about what happens if there are even more complicated calculations in the background for matchmaking. Click battle, fly to brasil pick your coffeebeans, fly back, roast them, grind them and so you made a coffee while waiting for the matchmaking. If you think about how many factors the matchmaker must take into account now each one added i think will make it much worse.

And how often do you meet unicums in tier 6-8? I on some days don't play many battles 10 sometimes 20 but i think if I meet ONE (unicum the purple guys) in several days or a week it is often. This all sounds to me like you click battle and are up against the next 3-4 unicum tripple platoons in the enemy team when you are mid to low tier in a tomato team.

just my 2 cents hopefully I don't annoy anybody with it

Greets
Ronin