Turn 5 Twister Strategies Compendium

[kukurumei]

There is a ridiculous amount of confirmation bias going on with this mechanic, and much of it from people who have never experienced the false positives that Sonya describes. For that matter, i'm not even sure that anyone has a "true" understanding of how the mechanic works (BG included), but instead some have an understanding of what methods work more consistently than others. Its hard to get a read on how a mechanic truly functions when even the BEST case scenario is (again, as given by BG) an 80-90% success rate over the course of a string of attempts.

The "oh, you should have KNOWN its a circle" stuff should probably just go ahead and die off. You can relate nearly anything to any other thing in some way, similar to the Forer effect used by psychics. Call out random associations and eventually one is bound to stick. If the solution was to run in a straight line to the northeast, SOMEONE would equate it to "Oh, tornadoes mostly move northeasterly, the solution was SO clear... how did you not get it?"

The bottom line is that twister as a whole is a questionably designed mechanic, at best, and for a very long time to come even the best farm guilds are going to have nights where they get streaky wipes simply because the method didn't work for an essential raid member at an essential time (read: not a scholar). Using BG's numbers of 80-90% success, assuming only 3/4 targetted people are essential, you are still looking at best a 15% potential kill rate, per attempt, assuming everything else goes flawlessly and you only see 4 twisters. Those odds aren't particularly exciting to me as a raider.

That's pot-kettle there.

How about asking BG doing what they did, instead of saying "freak coincidence" mouthing off.

A basic probability of 80% over 8 members 4 times. (not including all the previous trial and error runs that made it to a stream-line video win) is heavily against your argument. Anyone who studies basic dice roll theory will tell you that.

And theories about ping and lag...also heavily against your argument, because they were a mix group of nations, (half they admitted used battle ping, and half didn't). Thus that holds little water as well.

Face it someone took word-play leap of faith and payed off, now people say "Bah i could have done it too, if it weren't for your luck"

Phsaw, why can't some people take it honestly.

"Haha developers you got us" it's simple.

[Liander]

-

I'm not sure anything you said directly has anything to do with what Xil has stated, you just kind threw words together...

The basic design problem of twister in its current form is surviving when you shouldn't.
Example:
Attempt 1: Twister starts casting and I run in a straight line. I survive.
Attempt 2: I run in a straight line. I survive.
Attempt 3: I run in a straight line. I die.

All three are the exact same with different outcomes. Why am I surviving by running in a straight line when running in a circle is the correct way to do it? Now multiply that by hundred of combination of running out of twister and you can see the problem. There was never a clear picture of the mechanics intention. BG found circles was a bit higher % of surviving which we've observed as well. But 10-20% chance to die from a mechanic you're doing right shouldn't exist.

[Soukyuu]

The bottom line is that twister as a whole is a questionably designed mechanic, at best, and for a very long time to come even the best farm guilds are going to have nights where they get streaky wipes simply because the method didn't work for an essential raid member at an essential time (read: not a scholar).

I agreed with the rest of your post but: what? are you saying a scholar is not essential?

[Mjollnir]

That's a loser's argument. If this was in the main game, yes, it would be unfair, but this is a category "secret post game boss". You are entitled to nothing, but to pit against sly developers.

Maybe the age of RPG is gone, and people expect to be spoon fed missing links, but any RPGer will get the crude but effective developer's joke.

I wasn't stating that it being a very hard mechanic to fathom was a bad thing at all, but clearly English isn't your first language, so you've misunderstood. I was highlighting that to say there were obvious hints present is easy after the event (I don't even consider the word "Twister" a hint as to it's own evasion method, like you do).

It would have been a "developer joke" if the solution to evade it was to put your left hand on red, your right hand on blue and your left leg on yellow; or to deploy a DOROTHY in the middle of it.

[Xilrasis]

I agreed with the rest of your post but: what? are you saying a scholar is not essential?

In terms of meeting the DPS requirement of the last two phases of the encounter, a scholar is a non-essential raid member. The twister phase is solo healable (there is zero extraneous damage) and as long as he's back up to support healing in the final phase, you are fine. Meanwhile, to use their video as an example, if a bard, monk, or BLM had died to twister, they would have very unceremoniously met the enrage timer due to having a DPS with the weakness debuff.

That's pot-kettle there.

How about asking BG doing what they did, instead of saying "freak coincidence" mouthing off.

A basic probability of 80% over 8 members 4 times. (not including all the previous trial and error runs that made it to a stream-line video win) is heavily against your argument. Anyone who studies basic dice roll theory will tell you that.

And theories about ping and lag...also heavily against your argument, because they were a mix group of nations, (half they admitted used battle ping, and half didn't). Thus that holds little water as well.

Face it someone took word-play leap of faith and payed off, now people say "Bah i could have done it too, if it weren't for your luck"

Phsaw, why can't some people take it honestly.

"Haha developers you got us" it's simple.

How about you ask BG how many attempts they did (i'd put money on it being in the triple digits) where their main tank or DPS or both healers died to twisters inexplicably and it forced them to reset, or they couldn't meet the DPS check in the twister or final phase as a result?

Pretty much every group out there who has been doing twisters for any extended period (Blue Garter included, as they gave with their per-member success rates) has attempts where no matter what you do, someone just dies. This turns the encounter into less of an "execute perfectly and beat the encounter" to an "execute perfectly and hope the right people survive so we CAN beat the encounter." There is a ceiling to how successful you can make yourself on the mechanic, and once you meet it you simply throw yourself at it until something sticks.

As for the probability -- do you even understand basic dice roll theory yourself? Or is this a case of "lets throw out some random words to try and reinforce my point"? The basic probability calculation is extremely simple. 4 people are targetted per cast, at (lets give them the benefit of the doubt) a 90% success rate. Lets say they ALWAYS only see 4 twisters during the phase. The SIMPLE probability is that they have a 65.6% chance of all players surviving per twister cast, which across 4 casts gives a total of 18% chance of full raid survival. If you drop the success rate to 85% with 5 casts of twister, you are looking at a 3.8% chance of full raid survival. This is, of course, assuming perfect player play throughout all other phases... this is just an 18% (or 3.8%) chance of moving through ONE specific phase cleanly. Unfortunately, people are only able to see the ONE attempt where they have actually killed the boss as opposed to the hundred(s) before, and i'm sure tens on monday where the tank or a DPS or both healers flopped over dead.

As for ping and lag, they make a very clear and decided difference in many other mechanics in the game, i'm not sure why this would be any different (Especially with such tight timing requirements).

[KroLeXz]

they nerfed twister GG, you can now run in a full circle and dodge 100% of the time. Happened last night, and an official response from devs why they nerfed it.

[Xilrasis]

they nerfed twister GG, you can now run in a full circle and dodge 100% of the time. Happened last night, and an official response from devs why they nerfed it.

Honestly, it was a necessary nerf. The entire difficulty of the encounter was getting your group through enough twisters to meet the DPS check. Now that twisters can be dodged reliably, you can expect a lot of guilds who previously invested a substantial amount of time in the encounter to kill it today or tomorrow, depending on raid schedules.

[Teloc]

You can walk in a line to move from twister now. No circle or zumba needed.

[Rokuthy]

You never had to run in a circle to dodge twisters, it was simply a method of dodging twisters that was consistent due to the mechanic of twister itself. As long as you had an understanding of the mechanic, you could formulate numerous strategies for dealing with ways to dodge it.

The circle dodge became popular for a few days because that's what was used in Verily's POV, but it wasn't the only method that had a high success rate. Grats on your kill.

Edit: now that Carraway has posted the full strategy breakdown, I just wanted to touch on a few things I've read over the weeks on this thread:

So how did they work?

At a certain point during the twister (for the sake of our success rates, we'll call it 50%), the server chooses your character's location and places a twister plume marker at your feet. At this point (50% through the cast), you're tethered to that location based on distance. The key was to be standing completely still 50% through the cast (our initial strats), and then begin moving. At this point, the entire goal is to stay with a certain, completely unknown but specific distance from your original tethered location. If you move off of your plume, and not too far, you live. Simple as that. If you did move too far, and the server updated your location, then at that point, regardless of the cast bars current remaining duration, the plume is adjusted to your new location.

This means if you moved in too wide of a circle, or too far from your original location, your location was updated, and a new plume was placed at your feet. At this point, you probably wouldn't have time to move off of it, and you would die. The reason that you could live while moving before 50% (while still staying within the initial tethered distance), is that direction changes would force position updates. This is why circle movements, or tight boxes, and things of that nature were so successful: you stayed within distance to your original tethered location while also forcing constant server updates. (People feel that tight directional changes do not force positional updates. Let's assume that it's absolutely certain that do not - another simple explanation for being able to dodge effectively while moving before twister chooses your initial location, would be that you stayed within your safe zone and the twister wouldn't reset its location so long as you didn't force the plume to move).

How do I explain the false positives? Simple - you run in a straight line, no forced server updates, so based on luck and latency, you could get a plume location that you were lucky enough to move far enough away from that would still allow you to live. In the case of false positives, more often than not latency was actually working in your favor and not against it.

So that's it, those were the basics of the original twister mechanic. You were tethered to a location that you couldn't move too far away from while still getting a position update to fire off before you were blown up by the twister. If you moved a step too far and the server updated your location, you died - hence all of the original "rng at the end of the cast" complaints that I read early on.

RIP in peace, twister.

Edit: There are a few ways of visualizing this tether theory. First is the simple bullseye image. You have Deadzone -> Safezone -> Deadzone. Pretty simple - move into the deadzone that's beyond the safezone and the twister plume resets on your current postion. Another way is to imagine a ball with a short string attached to it. The ball stays in place, and you attempt to stretch the string as far as possible without making the ball move. If you get to a safe enough distance, you live. If you move too far, you pull the ball with you. The last way I can think of to explain it is to imagine the way an inch worm moves:

http://www.youtube.com/watch?v=9Fhz5ZQK-14

Awful video, ridiculous way of explaining this but it works I suppose. You're the head, the twister plume is the ass. You want to get away from the plume, but if you go too far it's going to reset on your current position. Any of these visualizations can explain the majority of false positives or whatever that we were seeing. You could easily run in a straight line and just get lucky with the timing of when the plume reset on your position (hence people claiming 50/50). The trick was to just stay in the safe zone distance from where you were standing 50% through the cast. My personal dodge method was to strafe right about 70% through the cast for about 2 steps, then immediately do a hard arc back around where my original position was. 9 times out of 10, my plume appeared where my arc began after the 2 step strafe (hence my reasoning behind sharp direction changes forcing updates).

My other bit of anecdotal evidence for direction changes come from divebomb dodges. I play from Japan, and even with battleping I get around 300 ms when I play. Occasionally I'll forget to turn on battleping and I'll have to deal with 400ms. There's a sweet spot just after 300ms where I flat out have a hard time dodging the divebombs using our strat. What I found that would work with 99% success, was to dodge out like I always do, and just before twintania dives, I'd stop moving forward and simply strafe/turn around. If I continued to just run forward, I would get hit into the middle of the arena (again, 99% of the time).

With all that being said, even if direction changes have absolutely nothing to do with forcing server updates, the theory behind twisters still stands. You had a minimal distance you needed to move to live, and a max distance before that position reset, starting the process over. Sometimes you got lucky doing it improperly, sometimes you did not. Sometimes latency played in your favor, sometimes it did not. In the end, our individual methods of dodging twisters coupled with our best theory as to how they worked mechanically lead to us having a fairly high dodge rate across the board.

[Xilrasis]

*snip*

Several of us thought at one point that this was the case, but further execution showed that this was unlikely to be the case.

I'll dig up some video clips later to show why, but by far the most effective method of dodging that I ended up using (with similar success rate to what you guys saw, 85-90%) is a very light "Arc" beginning from the start of twister cast. As soon as I see it cast, I move, in perhaps a 1 radian arc from my original point. I'm not sure if this was simply allowing it to reset twice, or if the movement past the halfway point wasn't enough since it was in a VERY slight arc, but it did not indicate the mechanic as you describe at all. This, coupled with straight line false positives, single hop false positives (not moving until the cast is almost complete and doing a single "jump" at the end of the cast), single steps, right angles, and any other number of insane strategies attempted by the masses and ZERO visual feedback on the mechanic caused twister to be a question mark at best.

In all likelihood, we will never know what the original mechanic ACTUALLY was, but the bottom line is that achieving the kill simply required slamming your head against phase 4 until everything worked out. Even with the highest success rates reported by you guys and other FCs, its still a statistical dart throw to succeed. I'm pretty sure we saw Phase 5 a grand total of 4 times when we got our kill. If twisters were in their current state 5 weeks ago, we would have seen a week 2 kill, which would (in my opinion) have been almost as big of a travesty. It was clear early on that the kill was simply going to go to whoever spent the most time attempting the encounter, and many companies (my own included) backed off of progression a lot after week 4 or so. Through sheer time investment and perseverance alone, your kill is well deserved.