3.2 PvP: Feast Game Mode Discussion Thread.

Quote:

---

Originally Posted by Februs

Saying that you would rather have the system judge player rankings based on top percentage within specified jobs rather than top overall is fine and dandy, but what are the criteria for making it to the top ranks? How would we be evaluated? If the answer is simply "by our overall number of wins," then not a lot actually changes.

---

Not sure if there is actually a fair way to determine that in a team game. If it say judged an individual based on their damage or healing output, it wouldn't account for all the variables in a pvp match, like a whole team relentless focusing on you and using stuns and sleeps to reduce your performance. It's not like a training dummy exercise where a person can perform a certain way with consistency, so I think any quantifiable metric isn't telling the whole story. Things could only really be fair if it was a 1 vs 1 type of game and the jobs were balanced so well that they were basically all the same.

Quote:

---

Originally Posted by Februs

As for Ranking rewards being separated by jobs, how would you guys propose this be done?

---

Ideally I'd like to see something on the scale of the old Hamlet Defense system;
http://vignette1.wikia.nocookie.net/...20120426015428
Kills +
Assists +
Deaths -
Stuns + (actual Stuns, where you interrupt something, this can be measured in PvE for Limit Break building, no reason it can't be here)
Stuns - (just building Stun resist, stop it FFS, might be a bit iffy when it comes to non-casters though)
Sleep/Bind +
Breaking Sleep/Bind -
Collecting Medals +
Holding Medals for X time +
Losing X Medals -

Stuff like that, obviously I'm not going to work exact values in there, but you get the idea. All of that is tallied, all completely transparently so you can look at it and go "Oh, I get points for using Sleep? Better start using Sleep then", then the final result acts as either a bonus or multiplier. Did really well but just got outplayed in the end? That's less punishing than straight up sucking from the start. Your team won, but all you did was Hide in the corner with Medals or die a lot? Hopefully you'll lose rating. That's one reason why I'd want a multiplier, if you end up getting a negative score because you sucked, but your team still carries you through, that multiplier increases your negative score. Being carried actively harms your rating that way...

IMO, this is the kind of system that would be ideal for rating. Does it make it easier to advance? Probably, so what though? It makes it easier for everyone. The only issue is if you just slapped that on the current season. Paladins and Black Mages would have an obvious advantage since they can do a lot of Stun and Sleep/Bind respectively. That's why I'd prefer Job based rankings. You can have a system like this then, because it doesn't matter if Paladins earn more because they have Stun on GCD, Paladins are fighting each other for the #1 spot, not the entire data center.

Quote:

---

Originally Posted by Dimitrii

Ranking by roles I think would dampen this noticeably and overall make the rankings more meaningful than if you did it by jobs.

---

I really don't think they should do ranking by roles... All that will do is highlight how utterly imbalanced certain Jobs are... If they can balance everything then great, but part of the reason I'd suggest Job based rankings is that it lets you postpone that. Warrior versus Dark Knight is pretty much unfair, but Dark Knight versus Dark Knight is entirely balanced.

Quote:

---

Originally Posted by Dimitrii

Unless you are gonna slash it to something like ranking it as the top 5 for each job.

---

Kinda? Right now we have the top 100 getting gear, the top 10 getting a crown, and the top 1 getting a piece of worthless furniture.

Switch that to Job based rankings, and I'd say the top 100 get a worthless piece of furniture (meaning 1,300 per data center), because that wont be hard at all. The top 10 get the gear (130 per data center, close to the current 100), while the top 1 gets something else. Right now that would be the crown, in future I'd suggest something like a unique weapon. The most balanced way to work it I guess would be for the top 8 to get the gear? That would mean 104 sets of gear going out. Top 5 would only result in 65 sets of gear.

It depends on how many people actually play a given Job... Do we actually have those numbers? Obviously it can be adjusted, but even something like the top 10 Dark Knights doesn't seem to diminish the gear that much IMO... We'd go from 100 sets of that gear, to 130, and that isn't considering the possibility of tailoring the gear for specific Jobs/roles like they've done for solo/party. Getting a trophy would be incredibly easy, but is that a reward anyone particularly cares about?

Yes it would be easier to get the gear by jumping on something like Dark Knight, but how many others would have that idea? We'd see the rankings for day one, perhaps a lot of Warriors and only 5 Dark Knights, so people switch over for that low hanging fruit, then suddenly we've got a lot more Dark Knights. The community in general would also be completely aware of this kind of meta, if you see someone in the Dark Knight specific coat for the top 10, and you know only 9 people even bothered with Dark Knight on your data center, you're not going to take that seriously. Meanwhile if you see a Warrior with the #1 Great Axe equipped, and you know there were over 100 Warriors fighting for that? Yeah...

Perhaps I am being biased when i say I want more people to have the outfit because of the previous rank removal on other pvp gear?

Anyway, its Day2 of Ranking and the flaws of the system are starting to show, if you notice a majority of people in solo ranking have 50-60% percentiles in the top 30 showing just how flawed raw wins are in a RNG composition (a good premade wouldn't be near this percentile) and to me that is disheartening.

As for Nalien, how would it compensate for a match being too short or too long? If you make a system based on hitting thresholds in different area, then time would be a huge factor

A duration modifier, perhaps? So a short match where you do the best possible ends up rewarding the same as a long match where you do the best possible.

Alternatively it can "wear off". So say you use Stun properly at the start, then never again, in a long match that wears off and you don't get the rating from doing it, since you didn't use it towards the end. That way a short match in which you used Stun properly would effectively be the same as a long match where you used it properly. If you keep using it, you maintain that rating bonus, if you stop using it, you lose that rating bonus. That way I guess it's effectively a one time thing? You'd still need to "play properly" throughout to get that reward, but longer matches wouldn't allow you to farm more of it.

Quote:

---

Originally Posted by Nalien

Stuff like that, obviously I'm not going to work exact values in there, but you get the idea. All of that is tallied, all completely transparently so you can look at it and go "Oh, I get points for using Sleep? Better start using Sleep then", then the final result acts as either a bonus or multiplier.

---

I can definitely agree with this, because added transparency in a competition is almost never a bad thing (unless it reveals an exploit or gives someone an unfair advantage). Personally, I'd like to see all this data anyway (including job rankings), just for the sake of comparing stats and player data. That said, l there are a few inherent problems in such a system, especially in an MMORPG like this.

The first problem is establishing weight of general actions. How do you determine the value of contributions? For example, is Dps rated higher on the multiplier than CC actions, or are they the same? If they're the same, a person could run around the entire match simply stunning people willy-nilly without any sense of tactics and score the same as a person who busted their butt to actually score a kill. That's an easy exploit for maximum ranking without any effort, which obviously is not a good thing. However, if they're not weighted the same, then you have the reversed situation in which players with high CC are scored lower just for using it, basically penalizing them for doing their jobs. You highlight Pld's and Blm's as jobs that would likely have an easier time climbing ranks, but in reality it could be the exact opposite. The fact that they have such a high CC and have to actively divide their attention across multiple things puts them at a disadvantage over players who's job it is to simply burst down a target (especially Pld's, who have to constantly juggle Dps support with Healer support with enemy Healer suppression). High CC players could score lower by default even when they are playing better.

Another problem is this:

Quote:

---

Originally Posted by Nalien

Your team won, but all you did was Hide in the corner with Medals or die a lot? Hopefully you'll lose rating.

---

Don't get me wrong, I hate the current state of ninja's hiding with a stack of medals to secure a win. However, not every player who does this is being intentionally dishonest by exploiting the move. Sometimes they don't have a choice. Say, for example, the tank on the enemy team is an idiot and gets 9 stacks of vulnerability and dies. You could let his medal drop expire and work up from scratch, but you'll probably run out of time before that point; so, there's only really one option. Someone has to pick them up. That leaves the Tank, who's likely going to pop some serious buffs and play ring around the rosie on the staircase until the match expires, or the Ninja, who is likely going to hide. That player is actively doing exactly what they have to do to secure the win for their team, bu with this system they'd be punished for it instead of rewarded, especially if they happened to have a somewhat lack luster start to the match (sometimes you just get an off game).

I do like the system, and I would definitely want to see it implemented in some form or another. (even if just a leaderboards thing so that I could view my comparative stats). It's just that I would want to make sure that the implementation is fair if we're going to have it effect our rankings, especially when there are certain things that could potentially be exploited by dishonest players. Though, in the end, it doesn't really make a difference. I can't imagine SE implementing a system like this, simply because of the amount of work it would require to do so. They've given us enough gripe about system limitations and storage space that I can almost guarantee that this will never be a thing.

Quote:

---

Originally Posted by xTysonx

Anyway, its Day2 of Ranking and the flaws of the system are starting to show, if you notice a majority of people in solo ranking have 50-60% percentiles in the top 30 showing just how flawed raw wins are in a RNG composition (a good premade wouldn't be near this percentile) and to me that is disheartening.

---

The fact the top solo-queue players are having 50-60% is healthy. Matchmaking is doing its job if it can pit players against each other to the point where they can only win half the time, which is supposed to allude to equal skill and chance of winning.

In fact, if you look at a competitive game such as LoL, you'll see even the best players have a win rate that's close to that. Fact is, if you have and maintain a consistent win rate that's at the very least above 50%, even if it's just 51%, you'll make it to the top just due to the law of really big numbers in this case.

Also, the reason you don't see good premades with this chance is because your average premade is not composed of avid, experiennced, coordinated PvPers. Now, if FFXIV were a game with an actual PvP scene, you wouldn't see these premades run as rampant as you do with their sky-high +90% winrate.

I don't know. The fact is that LoL has a huge playerbase, meaning that top players plays versus equally skilled player no? (Never played it, so I really have no idea how matchmaking and ranking work). So yeah it's seems fair to say that you have something close of a 50% win rate.
ffxiv doesn't have this huge player base meaning that you kinda play with what is avalable. It's not really pitting you against equal value players, it's just that the queue is a flip a coin factor so you have a 50% win. Then if you re unlucky enough you get increadible losing strike where you can't do nothing against even if you re good.
Don't get it wrong, I know that it is the same issue everywhere. But I don't feel at all that a 50% win rate here show that you have foes with equal strenght but more that you have half of the time the weakest link due to the not that high number of people playing.
And consider too that they changed the way you lose point. Before a 51% win rate allowed you to progress. Now, at least for what I saw in bronze you have a -15/+25 meaning that you need a 41% win rate to progress. Which probably show that it was way too hard and frustrating to progress because you needed this 51%.

If this game could just stop putting me with amateurs that can't even follow simple mechanics, that would be great. Here's what usually happens for me:

A. I get stuck with a healer that can't handle pressure and our DPS (with the attack buff) dies within the first 20 seconds with our healer dying soon thereafter
B. One of the DPS asks what the battle plan is, so I inform them and instruct them to focus on the marker, but they do something else anyway
C. Even after switching said marker, one of the DPS stick to the target they were originally attacking and have tunnel vision for the rest of the match, causing us to lose since we can't kill the new target with just a tank and one DPS
D. Everyone focuses on a different target altogether and splits our DPS throughout the entire fight, then wonder why nobody is dying while the rest of the team runs around with 3 members that have 50% HP or a little less at all times
E. Someone on the team doesn't give a shit about rank, says ranks are meaningless, then we end up losing because the other team actually does care about their rank(s)

I'm sure I'm forgetting something here, but while it does seem fair that we're all queuing up with randoms under the same circumstances, it sure doesn't feel that way at times. I can't tell you how many times I got one good team yesterday followed by three or four bad teams (usually the same guys) before finding a good team again. Hell, there was even one guy trolling in there by queuing up as a CNJ. He couldn't heal for shit and would always die pretty quickly, lol. Fortunately I didn't get stuck on his party, but I couldn't help but feel bad for the opposing team that kept getting stuck with his sorry ass.

At the end of the day, if you don't care about ranked, don't play ranked. Plain and simple. I don't care what you do in your free time, but once you start playing with other people, you better be giving it your best... Otherwise you're just wasting my time (and yours too). I'm trying to maintain my top 10 rank at the moment, so excuse me if I don't feel like laughing when you come to troll PvP for shits and giggles.

You forget what you don't see and is probably the situation I hate the most:
F. Tank who take 10 stacks then die, breaking the whole progress did before or throwing the game because he died 30sec before the end.

Considering its the weekend and the season started, queues are pretty crap today. Figures that the first day I can legitimately sit down and pvp for a few hrs in the past 3 days I get these 15+ min queues.