116 replies to this topic

[MacClearly]

Wronka, on 16 January 2017 - 11:44 AM, said:

I just came on here to make a thread about tiers, so perfect timing this thread is here. I have to quote this post because it is basically what I wanted to post about, and I want to make sure it is seen.

I am a tier 1 player as well, however I am not as good as some of the true T1 players, you know the guys everyone knows about, the regulars. There is no way I should be grouped in the same tier as those guys. I've been playing for years on and off, so I get why I would sometimes play against those guys, but I should not be weighted the same as them for matching purposes. I am not a part of a big clan or want to be, I just play a few hours a week with a friend of mine and that is what I want keep doing. But recently, me and my friend both made it to tier 1. And this is when it really became obvious. I am not opposed to a challenge, we always look forward to trying to beat these guys, but the reality is, we lose 95% of the games we play against them.

I must say it gets really annoying if you happen to play on a night a group of the pros happens to be on. Since i'm tier 1 it is a guaranteed chance I will have to play with or against them, and that is when the game gets really boring, really fast. Losing 12-0 on a match the game thought was fair, but in reality was over before it started, kills it for me. It's equally boring to be on their team because the game is over so fast, it feels like I could have just powered down at start and we would have still won. I understand part of the problem is the low population so they have to open the tiers up a bit for faster matchmaking, but they should create more tiers so that I am not counted for the same as someone that is a top 10 player in the game.

Also, it is too easy to level up in tier. I am not sure what the exact rules are, but it seems anyone that is half decent at this game will eventually make it to tier 1, even if they do not actually become better at the game. I think each tier should have its own range as what is considered a good game and what isn't. For example, 400 dmg might be a good game for a tier 3 player, but it shouldn't be a good game for a tier 1 player. I think they need to raise the threshold at when your skill decreases and greatly increase when your skill stays the same the higher the tier you are. This in combination with more tiers I would imagine would at least have the chance for more improved games, even if the same players match up on most days when population is low.

The thing is that when you are playing with a bunch of great players against equally great players, you are not going to see many people getting 1200 damage games. Actually as I have gotten better and play against better players 700-800 damage games look a whole lot better in that context. Some of the best games I have had in group queue with my unit against another decent team has seen incredibly tight scores where not a single person was over 600 damage and almost everyone was around 400. Those games are generally rare but spectacular and well coordinated.

So if you actually set the bar at over a thousand for tier 1 players, you'd be bumping a ton of consistently good players back and forth....

[Trevelyas]

The Lighthouse, on 15 January 2017 - 06:32 PM, said:

I have a huge feeling that Empyreal is wrong with this one.

< graph >

I mean, this assumes if all tier 1 players actually keep playing the game.

...

I suspect large amount of tier 1 players have stopped playing the game. I must say PGI really pissed off a lot of old players, and it is definitely possible they lost a lot of higher tier players as time passes.

The graph above is correct for 'all' players for sure, but I don't think this is actually close to 'active' players which actually matter for the matchmaking.

I made that graph a year ago as a mock-up just to show the long-term consequences of what I believe PSR does to player accounts over time. So you're correct that it doesn't reflect the active player population at a given moment. I'm sure many of the max tier 1 accounts are inactive most of the time. If you took a snapshot of the tier distribution of players actively playing at a given time, I'd imagine tiers 1, 2, and 3 are much lower population than 4 and 5, with 2 and 3 being the lowest.

I hope Russ decides to show histograms of player tiers for 1) the entire population and 2) several snapshots of the players in-game and in the queue at different hours of the day. Those would paint a very clear picture of what's really happening.

Edit: I don't speak for EmpyreaL. I've been working alone on this PSR evaluation/suggestion stuff.

Edited by Trevelyas, 16 January 2017 - 04:29 PM.

[Accused]

How small is the population if games might not kick off for 4:30 minutes while looking for t4 players?

[Votanin FleshRender]

justcallme A S H, on 15 January 2017 - 10:59 PM, said:

Ye I am fair certain I've had mine go down in a win, last year.

Don't forget that post is from Sept 2015 & for at least 6 months of last year T1 and T5 have been put into the same matches. Yet PGI have just came out saying that hasn't happened - & we know 100% that it does...

What that means is I have very limited faith on what I'm being told right about now as there is no accuracy coming out the other end. There are obviously bugs somewhere.

It's true... I had a game just a few weeks ago where my son fell down the stairs just as mechs were powering up. By the time I'd made sure he was ok and calmed him down, I was dead from a DC farmer, but my team went on to win. = PSR movement despite me doing absolutely nothing.


Try it yourself... yolo into the enemy team for a few games until your team wins without you and see. You cannot lose PSR on a win.

Edited by Votanin FleshRender, 16 January 2017 - 07:10 PM.

[justcallme A S H]

Yeah look it might well be working fine now, or I'm tripping balls entirely. (this was last year, I haven't done much QP so prob Aug?)

As I said - If PGI didn't even know T5 and T1 were being grouped together, for over 6 months, what else isn't working as intended? :/

[Brain Cancer]

The Unstoppable Puggernaut, on 15 January 2017 - 04:25 AM, said:

They need to double the Ttiers, split people up nicer into them and re-work the brackets of who plays with who. Or if lazy split the T1 bracket to have T0 as well.

There's no way I should be in T1. I've seen some nasty (well nice) players and they need a bumpty to be in top tier. If I end up having to game with them then no problemos but some of the people I've been viewing are using just the keyboard to aim. They need to be moved to safety for their own sake of enjoyment and others.

You are in T1 because PGI has designed the entire PSR system to feed players into the games of their long-buying customers as quickly and in as queue-filling a manner as possible, regardless of what this does to the inexperienced or lower-skill players in question. They hide this with the "carrot" of "Here you are, you're better because you made it to Tier (4,3,2,1)!"

Tiers do not split players by skill. They are only at this point a measurement of time served, a fact that nobody at PGI could possibly not understand. That means it's deliberate, and PGI is deliberately shoving poorer players into the wood chipper of comp-level players just to fill games, regardless of the effect on the less skilled players.

[Navid A1]

justcallme A S H, on 16 January 2017 - 10:50 PM, said:

Yeah look it might well be working fine now, or I'm tripping balls entirely. (this was last year, I haven't done much QP so prob Aug?)

As I said - If PGI didn't even know T5 and T1 were being grouped together, for over 6 months, what else isn't working as intended? :/

- Did you know that clan targeting computers were eliminating your crit chance completely? (yes 0% crit chance with TC)... for 6 months?

- Did you know that mechs with structure bonus in their legs took more damage from falling?... for god knows how long?

[Abner Osis]

Brain Cancer, on 16 January 2017 - 11:08 PM, said:

Tiers do not split players by skill. They are only at this point a measurement of time served, a fact that nobody at PGI could possibly not understand. That means it's deliberate, and PGI is deliberately shoving poorer players into the wood chipper of comp-level players just to fill games, regardless of the effect on the less skilled players.

If you have won 100+ games you should get thrown into the pot, that is long enough to figure out how the game works. All PSR does it serve as a distraction so someone can whine about not getting the game they wanted out of the queue randomizer.

This game is packed with loopholes for completely average players to get high damage entirely from build, i.e. the 65% or more of the queue in heavies and assaults. I see more clan lrm assaults, whose gameplay is comprised entirely of sitting behind a hill and waiting for R to light up, at t2/3 than I did in t5. Round the whole thing out with EZM snipers and builds that are viable only because internet latency displaces the unit and splashes hitreg all over.

PGI should just drop the pretense of the entire tier system and just let quick play be what it sounds like, a button to jump into a game without any expectation that the game should turn out how you like it, fair, etc. As it is, all PSR does is encourage abuse of the metric through build damage overkill or flaws/gimmicks in game design.

[MrMadguy]

Tarl Cabot, on 14 January 2017 - 10:07 PM, said:

Lol. I've been telling you, that something is wrong with my PSR rating, for so long. I'm minority, cuz I'm Tier 3! And I know, why. Because players with low skill don't go up from Tier 4/5, players with high skill go up into Tier 1 and I have been explicitly put into Tier 3, when PSR was implemented, and I just can't go back to Tier 4/5, simply because PSR is way too biased towards increasing. My stats are extremely terrible, but PSR simply refuses to go down. I've been matched with/against Tier 1s, but the fact, that PSR is too W/L biased, simply doesn't allow me to go down with my terrible performance. It really makes me wonder, why it has taken so long for PGI to finally realize, that PSR is broken? Simple thing - Tier 3 should be most populated tier, simply because players with average skill - should be majority of playerbase. If it isn't the case - then something is wrong with matchmaker.

Edited by MrMadguy, 17 January 2017 - 05:11 AM.

[justcallme A S H]

Navid A1, on 17 January 2017 - 01:18 AM, said:

- Did you know that clan targeting computers were eliminating your crit chance completely? (yes 0% crit chance with TC)... for 6 months?

- Did you know that mechs with structure bonus in their legs took more damage from falling?... for god knows how long?

But it's all working as intended, right?

[MacClearly]

Brain Cancer, on 16 January 2017 - 11:08 PM, said:

You are in T1 because PGI has designed the entire PSR system to feed players into the games of their long-buying customers as quickly and in as queue-filling a manner as possible, regardless of what this does to the inexperienced or lower-skill players in question. They hide this with the "carrot" of "Here you are, you're better because you made it to Tier (4,3,2,1)!"

Tiers do not split players by skill. They are only at this point a measurement of time served, a fact that nobody at PGI could possibly not understand. That means it's deliberate, and PGI is deliberately shoving poorer players into the wood chipper of comp-level players just to fill games, regardless of the effect on the less skilled players.

I don't buy this argument. As I have stated earlier I have multiple accounts and experience in every tier except for tier 1 which I am just cresting into. I have been however matched up with many players I see on this forum as tier 1 consistently so I am not playing against tier 3 players the majority of the time.

So while I appreciate your lack of trust in PGI it would not make sense from a business perspective to meat grind new players. While retention is magical, no business can afford to chase off potential new blood in the name of retention.

I think that whether the system is good or not, it is clearly heavily balanced towards promoting team work. This is fairly obvious as winning is the best way to make any headway in the game. In fact your performance in a loss has to be absolutely stellar to achieve a small gain in a loss which really points towards it being set up to discourage soloists. The caveat being however is also that damage is heavily favoured in the scoring and the system can be manipulated to a degree by it, but again this is still weighted towards winning.

An example of how this might be working is lrm boats. Lurming can easily allow a player to rack up 450+ damage which appears to be the threshold. So lurmers will always go up in wins and potentially will stay even in losses. However as you go up in tiers, the guy who lurms from a corner by himself, will struggle to meet the minimum and will be a part of a losing team especially if there is more than one player doing this soloist type behaviour. They will be greatly reducing their teams chances to win and will likely get more loses and more negative tier rating.

Generally however, a player who is consistently putting out a minimum of 450+ damage will be considered a strong contributing member of his team, and will go up in wins and stay steady in losses. If a player is able to do this as the competition gets better, it makes sense to me that, that kind of consistency means solid play.

This doesn't seem indicative of a system grinding up new players that are not performing what appears to be the developers main expectation for success which is team work.

[MacClearly]

Eindridi, on 17 January 2017 - 02:35 AM, said:

If you have won 100+ games you should get thrown into the pot, that is long enough to figure out how the game works. All PSR does it serve as a distraction so someone can whine about not getting the game they wanted out of the queue randomizer.

This game is packed with loopholes for completely average players to get high damage entirely from build, i.e. the 65% or more of the queue in heavies and assaults. I see more clan lrm assaults, whose gameplay is comprised entirely of sitting behind a hill and waiting for R to light up, at t2/3 than I did in t5. Round the whole thing out with EZM snipers and builds that are viable only because internet latency displaces the unit and splashes hitreg all over.

PGI should just drop the pretense of the entire tier system and just let quick play be what it sounds like, a button to jump into a game without any expectation that the game should turn out how you like it, fair, etc. As it is, all PSR does is encourage abuse of the metric through build damage overkill or flaws/gimmicks in game design.

This is an interesting view and fairly common as many people have echoed what you are saying. However it is not accurate.

If I a player has won 100+ games and lost 150+ games or more, all that means is that player has been in over 250 games. Doesn't mean they have caught on or acquired some basic skills to be successful, even moderately so, in the game.

Create an alt account and make sure you don't do any damage especially in the first five games. Now play some games and see what tier 5 has to offer. Then come back and say that those players should be in the mix with tier 1 and 2.

As far as your min/maxing argument, remember that the system is heavily weighted towards winning. So players who are good at playing as part of the team and do consistent damage will be rewarded by going up in tier and facing better competition. If a player is playing a strong build and putting out decent damage and killing the enemy, he is also going to be helping his team win.

This is a simple formula and players stuck in tier 5, to be there and stay there have to be incredibly bad at the game for this to happen. To have no tiers would make the game horrible not just for tier 5 guys but for everyone else as well.

[Appogee]

Navid A1, on 17 January 2017 - 01:18 AM, said:

- Did you know that clan targeting computers were eliminating your crit chance completely? (yes 0% crit chance with TC)... for 6 months?

Wait... WHAT?!

I have used a TC in almost every one of my Clan Mechs... purely for the improved crit chance!

Edited by Appogee, 22 January 2017 - 01:39 PM.

[Deathlike]

Appogee, on 22 January 2017 - 01:38 PM, said:

Wait... WHAT?!

I have used a TC in almost every one of my Clan Mechs... purely for the improved crit chance!

It was a bug feature for a bit.

[Rhialto]

Tina Benoit, on 13 January 2017 - 04:26 PM, said:

PS: I was looking at your signature like, "whoa that's really cool" and then after a moment I was thinking that the shark looks like it's swimming backwards due to the back text rotation!

haha c'est drôle car j'allais lui dire la même chose! On dirait qu'il recule...

So that's what could explain why it takes longer to find a match now... I missed that post, just got here from here.

[LordNothing]

ive had games where there were people claiming to be in t1 and others in t5. put people lie. potatoes lie about their tier to make themselves feel better, other players over exaggerate their epeen, trolls exaggerate for their own amusement, so meh. i dont have enough data to make judgements.

been avoiding quick play lately due to the absurd number of snow maps that come up in rotation. its winter and i live in alaska, so naturally when i go full escapist i like to go somewhere warm, like terra therma. at least in fp you have a chance to play other maps.

[Void Angel]

Tina Benoit, on 13 January 2017 - 04:17 PM, said:

Heya,

We've been hearing a couple complaints here and there about high tiers getting matched up with or against low tiers players and wanted to let you guys know what's up.

We do currently have it set that Tier 1 players and Tier 4 players can get matched together, however we've just recently reduced it (due to reports), so that it now takes 4.5 minutes before that will happen.
We've basically done this so that you won't be stuck waiting for a match for too long but you should be getting much nicer matches now compared to last couple days.

Let us know if the Matchmaker is feeling better!

Thanks for letting us know! This really needs to be linked to in Announcements, though - otherwise it gets buried and people don't see it for like a month...