Power in Numbers
All other things equal, the team fielding the most players has the biggest advantage--this is why most professional sports teams field exactly the same number. Because there's so much emphasis on DPS there ends up being not a lot of time to react and even a very tanky character has difficulty dealing with two bursty opponents simultaneously. Add this to a relative lack of meaningful ability depth and you lose for the most part any competitive 1 vs X "skill overcomes numbers" play. This isn't entirely true but the exceptions are where the 1 is incredibly more skilled than the X and the X can barely mash the keys with a target. There is just plain too much DPS to deal with for any prolonged period of time in such an encounter and it doesn't take any real skill to do it.
The downed state causes other issues. If there were such a thing as a typical group vs. group fight it would not be uncommon for one or more players to be downed and revived. For a cohesive and well-coordinated group, it is exceedingly rare for the opposing group to get a stomp off especially since most classes have some sort of delaying tactic for would-be stompers. In small group encounters this is already bad enough but you look past that and pretend that it encourages group coordination. In larger encounters, especially those where one side outnumbers the other, it typically renders the larger team un-killable. If the size disparity is large enough, even dead-dead team members can be resurrected faster than they can be killed. Yet another advantage is run speed. A few professions have perma-swiftness, a paltry 33% movement bonus out of combat (also: no mounts) but many do not. Luckily a few professions can boost multiple people in a group periodically and since swiftness stacks, a large group can often get perma-swiftness before they even set off. So in a weird backward way, larger groups now move faster than smaller ones. This is a notable penalty in WvW where objectives aren't super far apart and smaller groups can't flip objectives anywhere near as fast as larger ones.
The speed differential becomes even more important when you consider how relatively hemmed in the maps are. They aren't especially large to begin with and with runspeed they're effectively 33% smaller, but they're also heavily channeled toward a very small number of chokepoints. This means that travel through the map tends to be along fairly well-defined routes. Fights then also tend to be along those well-defined routes which in turn means that you had better finish up fast before more folks come by to tip the tides. If enemies show, you're doubly screwed because even if you turn tail and run, you're probably still locked in combat so even if you've got swiftness you can't get away because your new opponents aren't in combat and can move at full speed. Zergs can almost always catch runners. Sorry.
Larger groups also carry more supply so while your 5 person group typically carries not quite enough to build two rams, a 10 person group carries enough for three. Multiple groups also trivialize the lord NPC part of the capture and note that the capture point flips faster with more people in it. Large enough groups can flip a tower before the swords show up on the map and within seconds of the map indicator for keeps. And before you say that there should be scouts there: a) scouting is completely unrewarded in GW2 so almost no one does so, and b) even one person per tower/keep is typically more than low tier servers can afford even if there were people to call in--10 is a zerg in T8 NA and typically two borderland maps are empty.
The last advantage we'll discuss is one of the worst--the AoE cap. For performance reasons, player abilities are capped at 5 targets (most siege enjoys a much higher cap of 50). So in a 5 vs. 5 encounter, every player from both teams can potentially hit every other member from the other team. In a 10 vs. 5 team, each of the larger team can hit every member of the smaller team with their AoEs but the smaller team can only effectively hit half of the larger team. This is even worse in actuality than it sounds since you aren't guaranteed that your attacks will hit the same half so in all likelihood your damage is spread out amongst the larger group meaning that your chances of even downing a target are now much slimmer.
The worst part about the AoE cap is what it does to player behavior. Since there's no real disadvantage to standing on the same spot, larger groups will typically stack to gain virtual immunity to AoE damage. The larger the group, the higher the virtual immunity--call it "zerg immunity" if you like. Now note that the vast majority of the attacks in GW2 are AoE. The game is almost completely devoid of penalties to standing on top of one another and with all the benefits it's no real wonder why the game is played this way.
Biggest Rewards for the Least Risk
GW2 has some interesting design philosophies, many of which are good. One of these is that no one should ever be sad that a friendly has shown up. This leads to all kinds of fun things like never having to rush to a resource node before someone ninjas it, all XP being awarded in its entirety amongst contributors, and generally making group play very convenient. In PvP modes this has many adverse consequences.
WvW's rewards are largely all personal and include XP, coin, karma, and WXP, the WvW version of normal XP. At max reward minus any bonuses it takes 83.33 player kills to gain one level. WvW levels tend to be worthless outside of titles in batches less than 10. So you're looking at killing a veritable army of enemy players if this is your primary mode of play and want to progress. WXP is also awarded for objectives on the map like any other world quest in the game. These include any flip/capture that yields match points and to a much lesser degree, defending a land objective from enemy players. It is notable that these objectives on the whole deliver much more WXP than killing the average player who is most likely worth a very small number of points (double digits if you're lucky).
So, as a gamer wanting to progress as fast as possible to get some nice WvW level bonuses, how would you go about maximizing your WXP? That's right, you'd go flip a bunch of objectives as quickly as possible. Enemy defenders? Go somewhere else! There's usually tons of things to choose from and it's not like a heavily defended objective gives any more points--the defenders are probably going to take off before you can kill them regardless. The faster you can zerg them down, colloquially known as "PvDooring", the more points you get. Forget skilled play, forget balanced builds, just group up with as many of your close personal friends as you can and steamroll everything. As an added bonus, the larger your zerg, the more likely you are to completely dominate your opponents which typically leaves them completely demoralized and less likely to show up tomorrow. This is how much of WvW is played across most servers and it is scarily effective.
There are lots of valid reasons to break up a larger zerg to fight a thinking enemy and even more valid reasons why this never happens. The long and the short of it, though, is that matchup scores are almost solely determined by population and coverage. No amount of skill and organization can overcome these two aspects for any statistically significant variation.
Scoring and Matchmaking
WvW(vW) scoring and matchmaking is especially ridiculous. It has not only changed a lot, but it often hasn't made any sense whatsoever and is occasionally very, very broken. Look away to escape the madness!
In the early days, servers were all ranked based on who they'd beaten in the past using a modified Glicko system. This meant that if your server blew out your opponents, you'd get a pile of points and might move up into the next tier. It also meant that if you got totally curb-stomped you might lose enough points to drop down a tier. This seems to be reasonable and good. Minus any confounding effects or botched maths, you'd generally be matched up with servers in the same ballpark as your home server. The problem is that there is a confounding effect, and it's a doozy.
Anyone at any time can transfer to any server they like so long as the server isn't full. Your server of choice full? Wait until like 4am when it's less likely to be and voila you're on a "winning" team. Again I have to note that we should expect gamers of all people to choose the shortest path to the cheese and in fact, this is exactly what happens. A huge number of WvW(vW)ers stack on the top three servers until drama shatters them at which point they bandwagon to the next new server. This causes the Glicko system to go into open revolt because the formerly top rated server had to spend weeks plummeting down to its new rightful place (the "Glicko Ghetto" in many cases) now that the "hardcore" guilds had all left. The new server upon gaining a giant influx of guilds now goes through the process of completely blowing out their opponents for weeks on end until they find competitive matches in the upper tiers. The really tragic part of this is that most of the folks who don't transfer end up with the short end of the stick on both trips. If you're a native, then you're being "propped up" by the newcomers who "know better" from "higher tiers" and once the bandwagoners leave you got curbstomped for weeks until the server finally settles likely worse for the wear. I would say "this happened frequently during the first N months of the game" but in fact, it's still going on.
"OK!", says ArenaNet. "We know how to fix this!", they said. "We'll reset all ratings back to baseline, let each tier of servers fight it out for a week, then start the next set of tiers based on the outcome. Eventually it'll all even out, right?" Players in the know rightfully cried foul since all it meant was that the servers doing the most curbstomping would then be pushed to the top tiers and the most competitive of tiers (typically the top) would be pushed to the bottom because the scoring gaps are much smaller. This also coincided with the end of free transfers meaning that it would be marginally less convenient to transfer to a "winning" server (now it costs in-game gold to convert to cashmoney gems or just spend $10 for the gems directly). Many top tier servers tanked their matches only for ArenaNet to realize How Terrible an Idea This Was After All(TM) and not reset things. I honestly don't know why this wasn't obvious.
Months pass and while the bandwagoning might have been slowed, it was only slowed slightly given that the cost wasn't really prohibitive and destination servers were often more than willing to subsidize the cost of moving entire guilds in the name of "more competitive matches". Over the first year of the game it ended up polarizing the WvW populations toward higher tiers where things like player abilities stop working as the servers themselves start buckling under the load. Luckily (?) on many of the top servers the WvW queues were so long that you were often spared the misery.
During this time people started to note that a lot of the matchups were the same from week to week. While many of them were competitive, occasionally there were tiers where the Glicko maths didn't really work out. One of the more egregious examples was T8 in North America where one large server was completely and utterly destroying the other two servers week after week but since the smaller servers didn't have a huge number of points to bleed, the larger server never moved up. "Aha!", the ANet devs say. "We can fix this too!" Their solution was to make matchups pseudo-random. The servers would roll some dice and mostly stay in the same tier but would occasionally bounce one or two tiers higher or one or two tiers lower. So from the bottom of T8 you could be facing servers from T6 which would result in a complete blowout for you and very likely a boring week for them. On the other side, a T1 server could be matched up in T3 against a T4 and T5 server and lose rating because they didn't blow them out badly enough. Many players argued that this would be terrible for the player base and guess what? They were right again!
While this may have fixed some of the bad maths around ANet's implementation of Glicko, it also meant that every week there was a good chance that your tier would have a lousy matchup. The only real way to ensure that your WvW wouldn't be curb-stomped was to (dun-Dun-DUN!!!) move to one of the top tier servers with their giant queues and horrifical ability lag. At least then when the RNG struck you weren't going to lose--it'd just be a boring week. Those that remained got pretty used to this and the more optimistic ones settled with "well, because we're two tiers higher than we should be, we'll have an off week and just not WvW(vW) until the match resets."
Now fast forward to October of this year. To make WvW(vW) more interesting and to add more "meaning" to ranking and scores and the like, the devs decided that what they really needed to do was break the servers into leagues based on the current tiers then randomly match them up within the leagues to find out who was "best". Forget for a moment that we already had a pretty good idea who the top servers were (hint: they were at the top). Either way, winners in a league would be given some extra super special things which probably include a special finisher and probably more things (
WvW could have been so much more and it's so close to the magic that DAoC had that it's maddening they didn't pay just a little more attention. ArenaNet put all the right sounding pieces into play and then hoped that it would work out. What they actually did was put some of the right pieces onto the wrong map and put the wrong mechanics around them. Yes you can take and claim keeps in RvR but that wasn't what made it great. Yes, you could run around in very large groups (and a lot of skill-less people did so) but large scale combat just for the sake of large scale combat is not enough. Yes, there were three factions which make somethings interesting but it also makes some things really, really hard. It's like a dance troupe trying to mimic their favorite video but they can't hear the music. Did they have any seriously competitive PvPers on their design staff at all?
It is all the more disappointing because so much of the game is so well executed. They have a lot of the right kinds of things but it's seemingly bolted together without an instruction manual. They actively reward people who have the least skill who take the fewest risks which actively punishing those who take the largest risks. That is a death's knell for any game with competitive aspirations and one that will not hold a hardcore audience for very long. The turning point is where fiercely competitive people realize that their contribution doesn't really matter and that every competitive piece is determined by things out of their control namely population and coverage. Most of them, I suspect, are already gone.
The worst thing about the game is that it could have been great but they settled for good. Guild Wars 2 snatches defeat from the jaws of victory which is a goddamn shame.
