None, uhm.. sorry.. SOME shall pass!

As you’ve no doubt already noticed, the big topic for debate at the moment is Mythic’s decision to remove friendly collision detection in patch 1.3.6. It’s been tested a couple of times on the PTS (Public Test Server) and appears to have been generally well received, particularly as people are quoting a 15% improvement in WAR client performance. 15% may not sound like much, but it is. Seriously folks, that’s a dramatic improvement for software and should not be underestimated. However, it comes at a price, a price that is not yet known.
 
It’s important to note that enemy collision detection is NOT being removed; it’s only friendly targets you’ll be able to walk through. Which would at first seem fine, after all it is enemy collision detection that allows us to form up shield walls and do the whole “none shall pass!” thing. And we’ve all been annoyed trying to push through a doorway that’s blocked by a blob of idiot/AFK players from our own realm, so combined with the performance improvement you might think that this all sounds great and that it won’t change anything really.
 
Hmmmm. I’m not so sure, in fact I’m a little worried. Not just because it’s a step towards WoW PvP, which is utter turd, no sir, because this isn’t some small tweak. This is a serious, core gameplay changing, “OMG THEY DID WHAT?” type change.
 
First off I’m going to on a little tangent. Shield walls, or tank walls as we call them in WAR, haven’t actually worked for a looooooooong time. All a tank wall does at present is provide easy targets for burning down and applying AoE on. I hear on some servers they still do tank walls, but they wouldn’t last 20 seconds on mine. Karak Norn is all about using terrain to create a funnel that the attacking players get, er… funnelled into and AoE melted to death instantly. That’s no ego or server pride statement; I’m not saying our server is better, just that it does work differently here. A friend of mine who was involved recently in a long discussion on Mythic’s vent with people from other servers told me that it came out in discussion that they’re just not using the same amount of AoE as we do. They have much more varied warband set ups, with like only 3 Bright Wizards or Sorcs in a warband (interested to hear comments on this), whereas many warbands on Karak Norn will have 6-8, backed up with guarding tanks, healers and maybe a token other DPS or two... A tank wall can not stand up to that (the stacked PBAoE instant kills), plus even if you’re not running a bomb spec warband/zerg, you can still easily crush tank walls, there’s rift/magnet, etc for one and the easiest thing is you just do a coordinated offensive morale bomb on them. Since morale damage is not mitigated by anything that makes a tank, a... uhm... tank… BOOM! Instantly dead tank wall, job done.
 
The reason I went on that tangent is people are quoting the use of tank walls in the testing and how they will still work and give outnumbered defenders a chance of holding out. Unfortunately for me this just doesn’t mean anything at all, since tank walls simply do not work. The only defence that gives any protection against a vaguely coordinated zerg is a funnel defence. Funnelling defences work not so much by using enemy collision detection, but more by using a combination of terrain and your enemy’s friendly collision detection. You use a choke point to set up and the enemy can’t all rush through the funnel at once, the friendly detection slows them down and as a result the defenders have time to melt them all with monstrous AoE.
 
So with friendly collision detection removed, is there any way that an outnumbered defender can stand up to a larger enemy force? Tank walls will continue not to work and with nothing to slow the attackers down, funnels will not work either. It sounds like it will become purely about the numbers you bring and nothing else. You may say that the underdog shouldn’t be able to win, that they should not have any advantages, well sorry but that is utter bollocks. For a start look at history (or just watching 300 will do for history-lite) and secondly the whole point of things like keeps is to provide a defensive advantage when facing larger numbers. An outnumbered, but organised defence should be allowed to make use of terrain to their advantage, but with friendly collision detection removed, what does terrain actually do? That’s the problem, because right now friendly collision detection is just as, if not more, important than enemy.
 
Against all odds is meant to encourage us to fight when outnumbered, but if there’s zero chance of standing up to larger numbers, will people bother? What about in smaller scale situations and scenarios where enemy players can now at all times freely kite you into their allies? How is that going to work out?
 
Having said all of this, you may be surprised to hear that I do actually think it should go onto the live servers, albeit with a proviso.
 
It is a huge performance improvement and with the other changes in 1.3.6 combined with the tricky to predict live environment, maybe things like tank walls will work again. We can all theorycraft to our hearts content about how it will work out, but this is a major change to a core component of the gameplay and as such its hard to reliably predict the outcome. There is a chance that everything will be ok, better even. Alas testing it on the PTS just doesn’t prove anything about the real impact of this change. The only way to know is to test it on the live servers and see how it works out. So I think it should be put onto the live servers, but it is absolutely critical that Mythic are ready to hotfix it out if it does turn the game into an even bigger zerg fest.
 
You can say what you like, be it for or against, but it boils down to this being a major change to the core of WAR’s combat system. Mythic need to be extremely cautious and must react very, very quickly should the risk versus reward prove not worth it.