Thanks to your participation in the current Ladder Test Season we have been able to identify and correct a few issues that were preventing players from losing ladder points in certain scenarios.
After looking at the data we have collected over the past week, we have also decided to make a change to the scoring matrix that will now cause players to lose ladder points in match-ups where they were favored to win and did not.
We believe the changes outlined above will help us move towards a ladder algorithm where an emphasis is placed on winning games, rather than just playing games.
We will continue to make changes to the leaderboard as needed throughout this test season. Thank you for continuing to help us test our algorithms as we work towards building a better ladder system.
Clarification: These changes do not only apply to high MMR players. It applies to players from all ranges. You can be favored to win a match at any MMR.
If you’re favored to win a match at any MMR, and you lose, you will lose points.
Clarification: These changes do not only apply to high MMR players.
Is this some kinda backhanded slight to the larger base of players? Seems kinda rude coming from a dev. I guess takes a bold person to try to pull that off. I mean why the special mention would have been more than clear had it been put applies to all players
Clarification: These changes do not only apply to high MMR players.
Is this some kinda backhanded slight to the larger base of players? Seems kinda rude coming from a dev. I guess takes a bold person to try to pull that off.
I addressed it because I had a few players message me, very concerned that this change would only affect high-MMR players, thus making it harder to climb. I just wanted it to be clear that this applies all ranges.
I mean you’re basically just sorting by # of wins with this point system anyways. Why not sort by mmr and let the ppl that care about points the option to sort by wins. I went out of town with my gf after grinding points for like a week and fell from rank 700ish back down to 90%. It’s kinda dumb. Looks like the cap for points is like +3 per game, how about making this scale to the amount of points on the other team? Eg; if you queue into someone with 100 points and win, you get their points – your points /2 or something?
Good its an actual test then. But people are not going to respect any other than an mmr type leaderboard. And if people dont respect the leaderboard nobody will make premades and no teams will form.
So lets try the new system. But it is almost certain it will be a failure. The points system is fundamentally flawed. Small tweaks to how it is implemented will not matter imo
Character: Henry rank million/Duke Henry
Necromancer/Casual Warrior
[Team] Best WvW guild of all time. EASILY.
Seems kind of weird to change this halfway through unless its retroactive. Players who grinded out hundreds of points under the old system are still going to win if this change only effects games from now one. Not that most people had a chance before anyway but now there is really no point if you were trying for the armour.
Seems kind of weird to change this halfway through unless its retroactive. Players who grinded out hundreds of points under the old system are still going to win if this change only effects games from now one. Not that most people had a chance before anyway but now there is really no point if you were trying for the armour.
Its a test for the good of the game.
Character: Henry rank million/Duke Henry
Necromancer/Casual Warrior
[Team] Best WvW guild of all time. EASILY.
MMR based leaderboard system failed horribly lets not go backwards.
It didnt fail at all. The leaderboards reflected the best players in the team game and in the solo game across the leaderboards. The problem was only the lack of permenent decay. Which they could of fixed
Character: Henry rank million/Duke Henry
Necromancer/Casual Warrior
[Team] Best WvW guild of all time. EASILY.
Are you kidding? So opinion is the guy at the top of the leaderboards with a grand total games played of 45 was the best solo player? Ok, well, you Sir. are one of the few that thought so.
MMR based leaderboard system failed horribly lets not go backwards.
Compared to this, everyone wants the old one since this one is so bad there realy isnt a point in fixing it.
And it was overall great – like ELO simply works, the only problems were:
-no real decay
-big volatility
The grindsboards dont have those problems – but they simp!y DONT work at all
Are you kidding? So opinion is the guy at the top of the leaderboards with a grand total games played of 45 was the best solo player? Ok, well, you Sir. are one of the few that thought so.
Ok you really are the dense type.
Yes, that guy and a few other happened. But it was a (overall) rare exception. And a bit of permanent decay and tweaks would make the old ladder even accurater.
Current crapboards are vice versa, they 90% dont work – just take a look them.
MMR based leaderboard system failed horribly lets not go backwards.
They were better, MMR leaderboards just needed a little adjustments.
-Decay needed to be reviewed.
-Put a max limit amount of MMR that you could get or lose in one game.
- Make adjustments in the volatility of the MMR (your skill level won´t change much in 1 week to 1 month of not playing, especially because there are very few balance patches).
-And increase the minimum games required to put you in the leaderboards, may be form 10 to 50 games, no more than 100.
- Reset the leaderboards every month or 2 months but not the MMR. That makes easier to track the progress for people that used to play a lot with a very low %win ratio but got better. Also because even if you win a lot it’s harder to track the progress if you have played many games because the more games you have, the less you will notice in you win ratio when you win or lose.
But right now the leaderboards are really bad. There are even 4 people in the top 25 with less than 50% wins and one of them even less than 40%. And this win ratio only takes in consideration the games from this season so games from the past can´t be an excuse.
Wait MMR leaderboard reset without resetting MMR seems blaintly unfair. So lock people in only certain postions within the leaderboard based on only one season?
Wait MMR leaderboard reset without resetting MMR seems blaintly unfair. So lock people in only certain postions within the leaderboard based on only one season?
It wouln´t be unfair if they adjust the decay properly with permanent decay. They would have to keep playing to keep the position and every season they would have to play the minimum games before being put in the new season leaderboard. 50-100 games (the minimum should be adjusted).
Are you kidding? So opinion is the guy at the top of the leaderboards with a grand total games played of 45 was the best solo player? Ok, well, you Sir. are one of the few that thought so.
There were very few cases like that comparing it to what we have now. Using http://www.gw2score.com/PvP you can see that there are 275 players with less than 50% wins in the actual LB, from those 275 there are 13 with less than 40%.
In the old leaderboards at the end in solo queue there were 66 with less than 50% wins and just 1 with less than 40%. In team queue 69 with less than 50% and 3 withe less than 40%.
From those numbers in the new leaderboards there are 50 players with less than 50% wins in the top 250. In the old leaderbords in solo only 5 and in team queue only 2.
Is this test purely for how the ladderboards are supposed to rate players or is there any chance this is attempting to find more ways to not let premades face pugs?
Toons: Foreseer, Geikamir, Rapscallion, Specimen, Scythian, Zeau, Ärtifact, and Replica.
Wait MMR leaderboard reset without resetting MMR seems blaintly unfair. So lock people in only certain postions within the leaderboard based on only one season?
So the #7 player under the previous system with a sub 50% win rate is complaining…I wonder why?
Wait neither system counts win percentage. I play for points as the leaderboard is designed if they designed it for win percentage i would play for that. But it is not, so a large win percentage is not optimal for a point based leaderboard. Win % can be manipulated as easily as points hence problems with last system
The new system penalizes losses more. So, a low win percentage will mitigate your ladder points to a much greater extent.
The old leaderboard was “designed for win percentage” too, one could just overcome a low percentage easily by spamming more games. Since a loss was usually a zero all it meant was time and it did not affect your score. Now, with negative scores, more time played might equal lower score, i.e. you can’t grind your way out of kitten created by a low (below 50%) win percentage as easily.
I assume these changes will be retroactive of course?
You’re assuming this because you believe the system is holding a record of the exact composition of every single battle roster you’ve been a part of for the last few weeks?
Could be, but somehow I doubt the system is documenting tens of thousands of matches with the precision of a major league baseball season.
“You keep saying ‘its unfair.’
I wonder what your basis for comparison is…”
- Jareth, King of Goblins.
No i assume the game retains the odds of victory, since this change required no patching it must have already been available. Composition would not be needed after odds are calculated. And since system already used these odds, assigning new values would be a simple matter.
As to the win percentage matter yes you can the idea is keep your MMR moving down, not sideways and not up. always down. This assures maximum points. out of the total of games available although i doubt anyone could play them all.
The old leaderboard was not win % based, it was MMR based there is a differnce.
Wins and skill are two different matters, if you wanna win based system that’s easy most wins —-- wins But that seems to be be why everyone is complaining, So skill based system, gl on that since there is no measure of skill. So i get the feeling everyone wants a win % system that should be easy, although the already small community will become even smaller as full team will be required to achieve the top ranks,
I’d like to point out that the only people advocating for this ladderboard are those with lots of time and no skill. Which I guess is fine, if we want a ladderboard that shows who plays the most instead of, well, those who literally lead the board.
The MMR based leaderboard was definitely more correct just had some flaws that needed to be bumped out. I don’t know why it was thrown out entirely.
So long as team queue and solo queue remain merged, judging by win percentage seems flawed to me. There’s so many factors that can affect your win percentage, and while points are a definite increase in leaderboard accuracy, the way the game calculates intended losses vs. preventable losses seems very unclear to me.
If the game already categorizes players based on MMR, why isn’t MMR the basis for the leaderboard?
Throw out the MMR totally, that way everyone plays everyone make ladder based on Win% for both team and solo, with a nominal minimum game requirement say 1K or so. That would be fair to all.
Are you kidding? So opinion is the guy at the top of the leaderboards with a grand total games played of 45 was the best solo player? Ok, well, you Sir. are one of the few that thought so.
Old leaderboards were without a doubt more accurate in reflecting player skill. You just deny it presumably because you can play enough to be considered a ‘top player’ by the new jokeboards.
I didn’t deny anything and i was very happy with my position on the old leaderboards. As far as my current standing idk what it was until someone pointed it out to me on this forum.
I didn’t deny anything and i was very happy with my position on the old leaderboards. As far as my current standing idk what it was until someone pointed it out to me on this forum.
This whole thread is covered with posts of you implying the old system failed horribly and the new one is better. And you saying that you didn’t know your own leaderboard standing after grinding hundreds of ranked games just points to more denial.
No, hundreds that’s laughable ,more like thousands and i said the old system was replaced because others complained, as they are now. Honestly i ground ranked games for a year before i even could access the leaderboards matter of fact before i even knew there were leaderboards. Why are you making this personal? oh yea pvper everything’s personal, my apologies.
No, hundreds that’s laughable ,more like thousands and i said the old system was replaced because others complained, as they are now. Honestly i ground ranked games for a year before i even could access the leaderboards matter of fact before i even knew there were leaderboards.
People complained about two things on the old leaderboards:
1. Non-permanent decay
2. The volatility of new player’s MMR (it moved too quickly in the few first few games which lead to people getting lucky in their placement matches and ending at top lb placement)
Both of these things were easy fixes that would take any competent developer team less than a day to fix entirely.
I haven’t seen anyone defend the current system or the old one for that matter, To me the best would be a win % system with no MMR at all and a minimum or games played,
Throw out the MMR totally, that way everyone plays everyone make ladder based on Win% for both team and solo, with a nominal minimum game requirement say 1K or so. That would be fair to all.
1k is too much. If the season lasts 1 month, that’s 33 games per day, 8.25 hours daily on average.If it lasts 2 months it ´s 4 hours daily, that´s still too much.
Just a simple little question you may have overlooked: Aren’t high MMR players favored to win more often? (In fact that’s the only way to get high MMR is to win often…)
So, you’ve simply encouraged the same practice that currently gets people to the top of the farmboards.
Step 1: Play constantly.
Step 2: Win around 45-50% of your games.
Step 3: Glorious Hero’s Armor???
A real tournament (not just leaderboard) should have a ceiling. No more than, say, 15 points per day. Or per 12 hours. Farm the leaderboard all you want, the tournament should have a separate leaderboard. I work 10 hours a day, sometimes 12. I come home I do 5 games tops before its bed time i barely make it into top 500 with almost 60% win rate, just to see myself out of top 500 the next morning. And by the time i come home im so deep in the leaderboards that I ask myself if I can compete with the 16 year olds that nolife in the game :/
TL;DR – The leaderboards are ruining my life by making me asocial LB farmer T_T. I want to see my family more often without losing my god kitten progress xD /s
Let the QA flow through your blood: 1 person out of 10 in the pvp dept. Every player is still holding pitchforks to fix lb which the previous system did better (prove a screenshot that didn’t). I’ll lay it out plainly (which I HOPE, your team discussed). Since we are going full yoloq, assume I am a high ranked player* I have an exceptional chance to win, get paired with 2 people that don’t know what they are doing, derank, gg. (Obviously picking diff games, but this kitten is pathetic
assume I am a high ranked player* I have an exceptional chance to win, get paired with 2 people that don’t know what they are doing, derank, gg.
Odds are calculated over the team, not individual players. So if you are paired with 2 unexperienced players and 2 average players it will be like 1 + 1 + 2 + 2 + 3 and the other team might have five average players so 2 + 2 + 2 + 2 + 2 and the odds are 9 – 10 so you won’t lose any points if you end up losing the game. Of course those values are made up but i hope you still get the idea.
Just a simple little question you may have overlooked: Aren’t high MMR players favored to win more often? (In fact that’s the only way to get high MMR is to win often…)
So, you’ve simply encouraged the same practice that currently gets people to the top of the farmboards.
Step 1: Play constantly.
Step 2: Win around 45-50% of your games.
Step 3: Glorious Hero’s Armor???
yes, the only way to get to high mmr is to win more than you lose… but to maintain high mmr you only need to win 50%. unless youre constantly getting poor matches against much worse opponents.
JQ: Rikkity
head here to discuss wvw without fear of infractions
That’s a good idea. Basically, we won’t see anymore top5 players with 4 loses and 0 negative points. Assuming they should win all their games, loses will start to hurt, and everything will be better
Just a simple little question you may have overlooked: Aren’t high MMR players favored to win more often? (In fact that’s the only way to get high MMR is to win often…)
So, you’ve simply encouraged the same practice that currently gets people to the top of the farmboards.
Step 1: Play constantly.
Step 2: Win around 45-50% of your games.
Step 3: Glorious Hero’s Armor???
yes, the only way to get to high mmr is to win more than you lose… but to maintain high mmr you only need to win 50%. unless youre constantly getting poor matches against much worse opponents.
uhm no, if all games have the same weight and you keep winning half of them your mmr will get closer and closer to 50% the more games you play. The only way to keep an mmr of 70% (for example) winning 50% of the games is if the games you win are 70% more important toward your score than those you lose.
As someone who does not understands very well how ladder systems work, can anyone enlighten me on what is the difference between the current testing ladder and, say, for example, League of legend’s ladders?
