Climbing out of gold is impossible, Change my mind

[Spartacus.3192]

> @"Shon.6419" said:

> Don't play when tilted( kitten this doesn't sound like me :D) or once you lose 2 games in a row take a break.

 

This right here was my number 1 rule when i play ranked.

 

Except when i finally gave up caring about rating and wanted to finish my legendary armor. Then it was just spam games for shards. LOL

 

[ezd.6359]

> @"Ben Phongluangtham.1065" said:

> Some interesting data:

>

> 29.7% of the players who placed in the top half of gold (1350-1499) finished the season in plat or higher. Highest Finish: 1843

> 7.5% of the players who placed in the bottom half of gold (1200-1349) finished the season in plat or higher. Highest Finish: 1802

> 1.1% of the players who placed below gold (0-1199) finished the season in plat or higher. Highest Finish: 1680

> Lowest placement to hit plat: 686

 

Well. Usually just after placements i have lower rank than my previous season. I think others see the same, my observation. If not, tell me i'm wrong. So i think we should not count 1350-1499 here, since they are plats from last season. Then 92.5% of gold or lower do not reach plat. Sad.

 

Can you check 1350-1499 again and tell us who was plat in previous season? Compare with this season list.

[GuildSimplyDavis.7852]

> @"Ben Phongluangtham.1065" said:

> Some interesting data:

>

> 29.7% of the players who placed in the top half of gold (1350-1499) finished the season in plat or higher. Highest Finish: 1843

> 7.5% of the players who placed in the bottom half of gold (1200-1349) finished the season in plat or higher. Highest Finish: 1802

> 1.1% of the players who placed below gold (0-1199) finished the season in plat or higher. Highest Finish: 1680

> Lowest placement to hit plat: 686

 

 

this because the population in pvp is very small ... the data speak clearly .. with a score of 686 statistically you end the season on plat ... this means more than half of the players in pvp reaches the plat ... and yet the average effective level is very low. in practice platinum is gold and the legendary has become platinum ..... yet despite having these data do not do anything to improve the situation regarding the pvp ..... why?

[Ivarian.9018]

> @"Ben Phongluangtham.1065" said:

> Some interesting data:

>

> 29.7% of the players who placed in the top half of gold (1350-1499) finished the season in plat or higher. Highest Finish: 1843

> 7.5% of the players who placed in the bottom half of gold (1200-1349) finished the season in plat or higher. Highest Finish: 1802

> 1.1% of the players who placed below gold (0-1199) finished the season in plat or higher. Highest Finish: 1680

> Lowest placement to hit plat: 686

 

Well, as I read these stats, it says that for the majority of the players (70% at least if you are gold), it is difficult to climb much higher than your initial placement. And I doubt we can say that the initial placement shows a good reflection of your level.

It would be interesting to have the same stats for players ending below the division they have been placed in at start.

Myself, I have been placed gold 2 at start, had a losing streak and spent the rest of the season to climb back to gold with the feeling that I was fighting against the MMS more than fighting against players of my level. Like if you must be able to carry the match by yourself if you want to increase your rating. And I am not at that level nor my class is.

[BeLZedaR.4790]

Consider what you are asking - to be plat - to mean you need to be better than all gold players.

(After all, you want to be considered a plat player, and assuming this rating has any accuracy at all, that should be the meaning of it.)

 

That means that if you are unable to carry games with players of gold skill, you are not better and thus should not be plat!

[AliamRationem.5172]

> @"Devilman.1532" said:

> > @"Ben Phongluangtham.1065" said:

> > Some interesting data:

> >

> > 29.7% of the players who placed in the top half of gold (1350-1499) finished the season in plat or higher. Highest Finish: 1843

> > 7.5% of the players who placed in the bottom half of gold (1200-1349) finished the season in plat or higher. Highest Finish: 1802

> > 1.1% of the players who placed below gold (0-1199) finished the season in plat or higher. Highest Finish: 1680

> > Lowest placement to hit plat: 686

>

> So its not impossible but highly unlikely. If your 10 placement matches don't go well youre pretty much boned. If you got stuck with a few matches of AFKers (who don't actually DC) total idiots no one could carry, the multitude of bots etc... How does this system make sense at all? How does it encourage people to keep playing when they know there is a 70% chance they will never get out of gold hell? If you placed in mid to low gold you have virtually no chance. You people need to scrap the 10 placement match trash system.

 

You're suggesting that placement is the determining factor in whether or not a player makes it into platinum. How do you rule out player ability as the determining factor? If placement is accurate and player ability determines whether or not a player makes it into platinum, would this not be more or less exactly what you'd expect to see - players who place closer to platinum being more likely to reach platinum? In other words, you tend to place higher if you're more skilled and you're also more likely to end up in platinum if you're skilled enough to play in platinum. Common sense, right?

 

Nothing in these statistics indicates an invisible wall that prevents players who can play at platinum level from reaching platinum rank just because they place lower.

[zoopop.5630]

> @"Ben Phongluangtham.1065" said:

> Some interesting data:

>

> 29.7% of the players who placed in the top half of gold (1350-1499) finished the season in plat or higher. Highest Finish: 1843

> 7.5% of the players who placed in the bottom half of gold (1200-1349) finished the season in plat or higher. Highest Finish: 1802

> 1.1% of the players who placed below gold (0-1199) finished the season in plat or higher. Highest Finish: 1680

> Lowest placement to hit plat: 686

 

sooooooooooooooooooooooo, anyone can make it to plat hence the horrible match making lol.

[otto.5684]

> @"zoopop.5630" said:

> > @"Ben Phongluangtham.1065" said:

> > Some interesting data:

> >

> > 29.7% of the players who placed in the top half of gold (1350-1499) finished the season in plat or higher. Highest Finish: 1843

> > 7.5% of the players who placed in the bottom half of gold (1200-1349) finished the season in plat or higher. Highest Finish: 1802

> > 1.1% of the players who placed below gold (0-1199) finished the season in plat or higher. Highest Finish: 1680

> > Lowest placement to hit plat: 686

>

> sooooooooooooooooooooooo, anyone can make it to plat hence the horrible match making lol.

 

No, it means that if you start off in gold, you only have a 37.2% chance to make it to platinum.

[spectrito.8513]

It's not impossible.

I ended up the season with 1595 rating almost at plat 2

playing exclusively Weaver S/D

[eksn.7264]

Can't people ever accept they belong to a certain rank? Why should you be owed a higher rank? Haven't played this season much, but other seasons I'd go on long losing streaks from plat 3 to plat 1 (sometimes down to gold...) and climb back up to plat 3 without taking the game seriously at all. If ONE teammate is playing badly doesn't mean you're better than your whole team and somehow in the wrong rank. Hell, there's at least one player per game who makes dumb plays but is in complete delusion about it and will never learn, maybe that player is you? Bad matchmaking luck does happen and GW2 mm definitely has many faults, but not being able to climb out of a rank just isn't at all an issue right now, you just belong there. Stop the delusion, accept that you just might not play perfectly every game and that your attitude probably sucks and start learning.

 

There are way more factors in YOUR control that are keeping you at your current rank than factors out of your control that are preventing you from ranking higher. And, let's be honest, if you play GW2 PvP for your rank, you're playing the wrong game.

[avey.4201]

I always got to gold 3, then ran into people who afk in spawns half the match, afk til they dc, or the only word they know is gg so before the match starts they walk circles saying gg, gg, gg, gg, then you drop to gold 2 where 2 people go home from start and camp home/feed deaths there, or you get the team mates who talk shit in team chat so much there's literally no possible way they got a single kill. I think a lot of match manipulation happens Anet doesn't catch or isn't strict enough on. I'll stay in WvW flying kites. When they do 1v1 PvP ranks might mean something, currently they only mean you paid or favored some friends/guildys into tipping the balance, or won the lottery many times in a row.

[AliamRationem.5172]

> @"avey.4201" said:

> I always got to gold 3, then ran into people who afk in spawns half the match, afk til they dc, or the only word they know is gg so before the match starts they walk circles saying gg, gg, gg, gg, then you drop to gold 2 where 2 people go home from start and camp home/feed deaths there, or you get the team mates who talk kitten in team chat so much there's literally no possible way they got a single kill. I think a lot of match manipulation happens Anet doesn't catch or isn't strict enough on. I'll stay in WvW flying kites. When they do 1v1 PvP ranks might mean something, currently they only mean you paid or favored some friends/guildys into tipping the balance, or won the lottery many times in a row.

 

A player who repeatedly bounces around between gold 2 and gold 3 but never makes it into plat isn't playing at a 1500+ level. They're playing at around 1400, and even that might be generous depending upon how far into gold 3 they climb and how much time they spend in gold 2 by comparison. You aren't as good as the highest rating you've ever achieved. You're only as good as the rating you can maintain.

 

I'm not denying that any of this stuff happens. We've all seen it. Losing streaks, players throwing matches at the first sign of adversity, straight up match manipulation, and being matched by the system against players who have no business being in the same match together. It happens. But it happens to all of us. So, if you can't manage to climb playing under the same rules as other guys who do manage to climb and it isn't an isolated incident (e.g. going on a losing streak the last day of the season), what does that tell you?

[avey.4201]

> @"AliamRationem.5172" said:

> > @"avey.4201" said:

> > I always got to gold 3, then ran into people who afk in spawns half the match, afk til they dc, or the only word they know is gg so before the match starts they walk circles saying gg, gg, gg, gg, then you drop to gold 2 where 2 people go home from start and camp home/feed deaths there, or you get the team mates who talk kitten in team chat so much there's literally no possible way they got a single kill. I think a lot of match manipulation happens Anet doesn't catch or isn't strict enough on. I'll stay in WvW flying kites. When they do 1v1 PvP ranks might mean something, currently they only mean you paid or favored some friends/guildys into tipping the balance, or won the lottery many times in a row.

>

> A player who repeatedly bounces around between gold 2 and gold 3 but never makes it into plat isn't playing at a 1500+ level. They're playing at around 1400, and even that might be generous depending upon how far into gold 3 they climb and how much time they spend in gold 2 by comparison. You aren't as good as the highest rating you've ever achieved. You're only as good as the rating you can maintain.

>

> I'm not denying that any of this stuff happens. We've all seen it. Losing streaks, players throwing matches at the first sign of adversity, straight up match manipulation, and being matched by the system against players who have no business being in the same match together. It happens. But it happens to all of us. So, if you can't manage to climb playing under the same rules as other guys who do manage to climb and it isn't an isolated incident (e.g. going on a losing streak the last day of the season), what does that tell you?

 

I've been plat 1 or 2 once that I seen back when I could collect achivements, never cared about rank, I typically start late in the season if not last few days/week of it, between gold 1-3 is where I see all these matches was my point, silver is typically casual I like it there, gold is fun until the window lickers arrive "even losing is a lot of fun when it's 5v5", I mostly only sPvP to warm up for WvW, I've never encountered sPvP enemy's that push me as hard as those in WvW where I have to work for my kill. sPvP rankings are based on matches with random pugs, it's not guild groups/partys such as tournaments, its not about how skilled you personally are, you may beat your head on the wall long enough to go past the cesspool in gold, or you may get carried. Rankings have no real meaning when your ranking random pug matches.

[Bort.8647]

Bad players arent only on your team. They are on the enemy team as often as they are on yours. If you are hard stuck at a specific rank after 100s of games, you are probably very close to the rank you belong at. The ranking system doesnt work if everyone can climb to plat.

 

[Exedore.6320]

> @"GuildSimplyDavis.7852" said:

> this because the population in pvp is very small ... the data speak clearly .. with a score of 686 statistically you end the season on plat ... this means more than half of the players in pvp reaches the plat ... and yet the average effective level is very low. in practice platinum is gold and the legendary has become platinum ..... yet despite having these data do not do anything to improve the situation regarding the pvp ..... why?

 

Lol, no. The 686 rating making it to platinum was an anomaly - likely a player who is normally platinum and intentionally lost all placement matches to show that it's possible to climb out of bronze/silver all the way to platinum (if you actually belong there).

 

And what kitten math did you do to get that over 50% of players reach platinum? If you assume a normal population distribution centered at 1200 (bottom of gold), less than 30% of less than 50% of the population made it to platinum.

[Ivarian.9018]

The argument that personal skill is the only determining factor to climb up ranks is interesting.

 

Let’s follow that logic:

 

In a 5v5 situation, it means that plat level players starting in gold division for example are going to carry 4 other players each match they win. Among these 4 other players, we can safely consider that some of them will end up at a higher rank than their true skill level. Then, when one of these players will play his next match, if only personal skill matter, he will make his team fail!

 

From this very simplified example, we can say that in order to climb up ranks, your personal skill (or I would rather say your “strength” level reflected by your build and skill) must surpass your current rank PLUS the handicap of playing with players of “strength” level below their rank or they will bring you down with them.

 

In a situation where class balance and PVP population is healthy, the handicap on players to climb up while they improve their game and skill through experience is relatively small. In a situation where class balance and PVP population is bad, the handicap to the “strength” level required to climb up becomes a limiting factor making rank progression a frustrating experience for players trying to improve their game and expecting their rank to go up .

 

I am sure that the MMS algorithm has some mechanisms to circumvent this handicap factor but I tend to believe that the current situation of PVP in GW2 has exceeded the capacity of the MMS to put up fair matches a majority of the time.

[eksn.7264]

> @"Ivarian.9018" said:

> The argument that personal skill is the only determining factor to climb up ranks is interesting.

 

Of course it's not the ONLY factor. Everyone knows matchmaking gives you a fair share of "predetermined outcome" games. Across a few games, you might be stuck in a bad luck loop, but if you TRULY belonged in plat, there's no chance you'd stay in gold a whole season. People just need to accept that they don't play well enough to climb.

 

 

[Ivarian.9018]

> @"eksn.7264" said:

> Of course it's not the ONLY factor. Everyone knows matchmaking gives you a fair share of "predetermined outcome" games. Across a few games, you might be stuck in a bad luck loop, but if you TRULY belonged in plat, there's no chance you'd stay in gold a whole season. People just need to accept that they don't play well enough to climb.

>

>

Talking about players who belong in plat does not really answer the question of how difficult it is to climb out of gold. It is not impossible of course and obviously players who ended up in plat managed to get there somehow. What I believe is that the current state of PVP plays against a good representation of rank vs skill where players are held back / carried in gold more than what they would be if all area of PVP were healthier in a fairness perspective (balance, queue, population, etc.).

 

After, it is more of a statement than anything. PVP is the way it is and trying to climb out of gold can be frustrating even for plat level players I'm sure (which I am not :D).

 

 

[kito.1827]

> @"Ben Phongluangtham.1065" said:

> Some interesting data:

>

> 29.7% of the players who placed in the top half of gold (1350-1499) finished the season in plat or higher. Highest Finish: 1843

> 7.5% of the players who placed in the bottom half of gold (1200-1349) finished the season in plat or higher. Highest Finish: 1802

> 1.1% of the players who placed below gold (0-1199) finished the season in plat or higher. Highest Finish: 1680

> Lowest placement to hit plat: 686

 

stats - yay!! \o/

 

it would be interesting to see the average rating span of matches too. so highest of the 10players of one match to lowest of the 10 players rating.

i have always been wondering about that.

[Liberis.9573]

> @"Ben Phongluangtham.1065" said:

> Some interesting data:

>

> 29.7% of the players who placed in the top half of gold (1350-1499) finished the season in plat or higher. Highest Finish: 1843

> 7.5% of the players who placed in the bottom half of gold (1200-1349) finished the season in plat or higher. Highest Finish: 1802

> 1.1% of the players who placed below gold (0-1199) finished the season in plat or higher. Highest Finish: 1680

> Lowest placement to hit plat: 686

 

I am the 1% hahaha

 

Ranked placements sometimes put me in bronze or silver, but I immediately climb to plat every single season. I get stomped for 10 matches because of bad teammates, and then I get to farm even worse noobs for 20 matches lmao.

[Exedore.6320]

> @"Ivarian.9018" said:

> The argument that personal skill is the only determining factor to climb up ranks is interesting.

>

> Let’s follow that logic:

>

> In a 5v5 situation, it means that plat level players starting in gold division for example are going to carry 4 other players each match they win. Among these 4 other players, we can safely consider that some of them will end up at a higher rank than their true skill level. Then, when one of these players will play his next match, if only personal skill matter, he will make his team fail!

>

> From this very simplified example, we can say that in order to climb up ranks, your personal skill (or I would rather say your “strength” level reflected by your build and skill) must surpass your current rank PLUS the handicap of playing with players of “strength” level below their rank or they will bring you down with them.

 

Your logic is flawed.

First, you assume that the plat player doing the carrying is going to get the same 4 teammates each time. Without a team queue, that's unlikely to happen. With a win, the player who is underrated will have gained some rating and in the subsequent match [in an ideal world] will have slightly higher ranked teammates and opponents.

Second, you only look at a small sample set of games and view rating rigidly. Not all games can be carried. However, the player who is underrated by the system will win enough games over time through their own contribution (carrying) to come out with a net rating gain.

 

So in the end, skill does determine your rating. Matchmaking quality will govern how quickly you can stabilize at your proper rating.

[AliamRationem.5172]

> @"Ivarian.9018" said:

> The argument that personal skill is the only determining factor to climb up ranks is interesting.

>

> Let’s follow that logic:

>

> In a 5v5 situation, it means that plat level players starting in gold division for example are going to carry 4 other players each match they win. Among these 4 other players, we can safely consider that some of them will end up at a higher rank than their true skill level. Then, when one of these players will play his next match, if only personal skill matter, he will make his team fail!

>

> From this very simplified example, we can say that in order to climb up ranks, your personal skill (or I would rather say your “strength” level reflected by your build and skill) must surpass your current rank PLUS the handicap of playing with players of “strength” level below their rank or they will bring you down with them.

>

> In a situation where class balance and PVP population is healthy, the handicap on players to climb up while they improve their game and skill through experience is relatively small. In a situation where class balance and PVP population is bad, the handicap to the “strength” level required to climb up becomes a limiting factor making rank progression a frustrating experience for players trying to improve their game and expecting their rank to go up .

>

> I am sure that the MMS algorithm has some mechanisms to circumvent this handicap factor but I tend to believe that the current situation of PVP in GW2 has exceeded the capacity of the MMS to put up fair matches a majority of the time.

 

Everything you describe as a barrier to accurate ranking should be applied evenly to all players on both teams and factored out. Is any given player on the other team expected to carry? No more or less than you are. Is poor matchmaking due to insufficient population a problem specific to only you? Is poor class balance? If these factors apply to everyone, then how do you explain why some players climb while others do not? Further, how do you explain players who are stuck at a certain rating tier, make some changes, and then either climb or fall dramatically? If matchmaking doesn't work and individual skill doesn't matter, these things couldn't happen.

 

These factors will certainly have a noticeable (and frustrating!) impact in any short-term snapshot you care to take. However, over the long run it's pretty hard to explain why some players climb and others don't as luck over skill. For every blowout match you win due to poor matchmaking you're just as likely to have a blowout loss. And so is everyone else. What matters is making enough of an impact to win more than you lose in the matches that are within reach. And that's something that can easily be determined by a single player on a few (or even just one!) critical plays.

[AliamRationem.5172]

> @"Ivarian.9018" said:

> > @"eksn.7264" said:

> > Of course it's not the ONLY factor. Everyone knows matchmaking gives you a fair share of "predetermined outcome" games. Across a few games, you might be stuck in a bad luck loop, but if you TRULY belonged in plat, there's no chance you'd stay in gold a whole season. People just need to accept that they don't play well enough to climb.

> >

> >

> Talking about players who belong in plat does not really answer the question of how difficult it is to climb out of gold. It is not impossible of course and obviously players who ended up in plat managed to get there somehow. What I believe is that the current state of PVP plays against a good representation of rank vs skill where players are held back / carried in gold more than what they would be if all area of PVP were healthier in a fairness perspective (balance, queue, population, etc.).

>

> After, it is more of a statement than anything. PVP is the way it is and trying to climb out of gold can be frustrating even for plat level players I'm sure (which I am not :D).

>

>

 

Is there some magic number of players that belong in plat? The stats indicate that nearly 30% of players who place within 1-150 rating points of plat actually make it into plat. What should that number be, in your estimation? It stands to reason that there will be fewer players in plat 1 than in gold 3 or gold 2. That's sort of how a rating ladder works, right? Is 30% indicative of a problem? If so, how?

 

You also conclude that matchmaking does a poor job of representing skill. How do you come to that conclusion? If luck in the form of being either "held back" in a loss or "carried" to a win is common (and it is), how do you figure that this somehow applies unevenly across a large number of games? Does the matchmaking algorithm choose favorites somehow?

 

I'm not saying matchmaking is perfect, but that you're looking at the wrong matches to draw your conclusions. Over a large number of games, you're as likely to be carried to a win or held to a loss by circumstances beyond your control as anyone else. Thus the games you should be focusing on are the ones you could have won.

 

To put it into numbers, let's say that you play 500 games with 40% of your games blowout losses and 40% blowout wins. This would indicate that 80% of your games are determined by luck/poor matchmaking. Super frustrating, right? Yet everyone else is dealing with the same problems and their rating is relative to yours. So ignore that 80% and focus on the 100 games you could have won. How many of those did you lose or win? That's the question that determines whether or not you climb the ladder. The third or so of players who climbed from gold into plat made critical plays that tipped the balance in their favor in that small subset of games often enough to climb while those who didn't never made it.

[Exedore.6320]

> @"AliamRationem.5172" said:

> Is there some magic number of players that belong in plat?

Since the population is assumed to be a normal distribution - yes. Though it's a percentage of the population, not a fixed number.

[AliamRationem.5172]

> @"Exedore.6320" said:

> > @"AliamRationem.5172" said:

> > Is there some magic number of players that belong in plat?

> Since the population is assumed to be a normal distribution - yes. Though it's a percentage of the population, not a fixed number.

 

My point was more that, without more info, why assume that ~30% is inappropriate? We know the number is going to be fewer than 50% or else you'd end up with more players moving up than staying or moving down, which doesn't make sense under the basic assumption that the ladder narrows as you climb from gold into plat. In any event, that number by itself doesn't indicate a problem and certainly doesn't indicate that matchmaking is broken, making it too difficult to climb.