Guild Wars 2

So let me get this right that these are the highlights.

-Can now watch matches on sPvP servers you join!
-Yes, you have to actually join an sPvP server and can only watch matches from that one server.
-By "matches" we mean only what is currently going on, you can’t watch old matches or save matches
-sPvP team size increased from 5 to 6 thanks to live-spectate and VoiP

I also see that this seems to be modeled exactly on "spectator" mode used by FPS games. That mode in those games doesn’t really see much use, but more importantly it takes up a player slot on the server.

While it wasn’t perfect, the spectator mode from GW1 was superior to this in every way aside from the GW2 spectate mode’s ability to directly see builds.

JonathanSharp.7094:

If you don’t want people looking at your builds, just disable spec mode on your custom.

Welp! Nevermind.

Syeria.4812:

Epic statistics fail.

The OP conducted this over a miniscule sample size for the purpose of data acquisition. He’s testing for something that presumably has a rate of occurrence significantly below 1% but only has a sample size of about 400. A “massive sample size” would need to be in the tens of thousands at the very least.

I see “massive sample size” posted a lot in regards to statistics and usually it is incorrect. It ultimately depends upon what is attempting to be discerned as to how large the sample size must be. Increases in the sample size reduce error and for particularly low percent occurrences getting the accuracy within an acceptable range can be difficult; trying to determine tenths of a percent when your error is around a whole percent is not terribly useful.

However regardless of how “massive” or “tiny” the sample size, using the P value (confidence interval) we can make conclusions from the above test. The idea that it wasn’t a “massive sample size” so its useless is completely erroneous despite its pervasiveness on online forums.

In this case with a 440 trials with no drops we can say for a P value of .05 that the drop rate is less than .7%.

Now, we can’t really say much else about the drop rate from this test, but there are conclusions that can be drawn.

Dungeons are in a really weird place right now. Many of the enemies are going to hit very hard and at the same time take a lot of focusing to take down. There are many encounters where the first person through a door is going to get dropped and there really isn’t a way around it; the two dodges you get just don’t go far enough and the only other option is for some to spec themselves, well, tanky.

Control is in a really terrible spot right now. The stuns/dazes/etc last for only a few seconds and when it takes 10-20 seconds to focus down one enemy it is just insufficient. I greatly dislike the WoW styled 60 second recastable sheeps, but right now if two mobs target the same character that isn’t specced tanky then that character is going to go down.

Melee is also in a very bad spot. Melee mobs can almost without exception be kited by ranged to mitigate nearly all of their damage. At the same time the ranged mobs themselves can get off the lion’s share of their damage and their auto attacks alone are often enough to down a group. All characters are going to be exposed to ranged damage and melee characters are going to have to expose themselves to both ranged and melee in order to be effective. This usually ends with the melee characters getting melted in a very short period of time and further confounding this is that the melee characters are going to get downed away from the rest of the party.

Right now there are several things that are putting dungeons in a weird place. It really should be no surprise that folks are attempting to go back to tank/healer/damage. While there is so much vitriol towards the “trinity”, it evolved because it is a very effective form of managing things.