Cut the guy some slack.. Its not a definitive test but at least give the guy credit for attempting to test things out himself..

These forums seem to get more and more unfriendly as time goes on. Nobody here is perfect.

The best we can do as a community is create theories and test them as best as possible. They are not definitive but every little bit of information and testing provides a piece of a larger picture..

Not everyone is a math wiz that, through numbers, can provide flawless evidence. Some of us can only create a theory and do our best to present our findings to the community in hopes that it sparks new discussion.

God forbid if said theory goes against someone's opinion.. All hell breaks lose..

He can have an A for effort, but he still gets an F for failed methodology. I'm saying that in the friendliest way possible, of course.

So the percentage is for Wave 3, that is to say, before the antiswarm Wave. And thus not that useful for Wave 4 meta.

Except for the fact that the primary anti-Tie swarm ship is the Z-95. Which itself can be made into a swarm of its own.

I can't help but feel you're missing what I keep saying. You realise tournaments do not consist of solely TIE swarms, yes? By Juggler's numbers (the ones you quote fairly often) they've actually all but disappeared, with the dominant squadrons being Phantoms and supported YT-1300s.

Something, I am ashamed to admit. In my experiments, I mentioned that I've play the Zs and let AI run the Ties and that the Zs would win alot. I also flip it around and let the AI run the Zs and I play the Ties. Guess what - the Zs (AI) would win fairly regularly (2/3 of the time). I am an Empire player and Tie swarm is my favorite meta.

(I think you mean squad or list, not meta. The metagame is predicting what people will fly and counterbuild, it's not your squad.)

Why didn't you mention that earlier? That eliminates one of the main issues in your experiment, assuming you were competently blocking with the TIE fighters.

Anyway, I'm going to run the 3 Z-95 experiment myself and see what happens. Which AI is it? I'm guessing this one: http://xwing.runbam.com.s3-website-us-east-1.amazonaws.com/xwing_ai_1.7.0b2/index.htm

EDIT: Done. Well, your first problem is the AI compensates by getting a free F/TL on whatever ship it shoots, ie: EVERY Z-95 gets better than Howlrunnered every turn for free. This makes the AI much better at the Z-95 because it's cheating. I only just managed to win with a 1hp TIE left, and that was because of some lucky rolls and the way the AI flies (get in close and it can't cope.)

Personally, I'd say that's the final nail in the AI's coffin as a valid test for anything: it's firepower is massively amped on anything that can target lock (it F/TLs for free every round) yet it maneuvers like it's drunk. Were it not for the free F/TL (giving it a near certain two hits a shot) I'd have butchered those AI Zs.

We need new subforum "Mathwing aces" to save young innocent forum user from huge amount of mostly useless statistics (as someone who studied statistics in my trip to university I can say you are tangling situation with way too many variables to get conclusive results out of it).

Cut the guy some slack.. Its not a definitive test but at least give the guy credit for attempting to test things out himself..

No, sorry. Performing the test isn't anything special; he or she went on to make a silly claim based on misunderstanding his or her own test, was unable to defend that claim, started condescending to people who (politely!) pointed out the mistakes, and continues to fail to understand why there's anything wrong with any of it.

The best we can do as a community is create theories and test them as best as possible. They are not definitive but every little bit of information and testing provides a piece of a larger picture..

It's pedantic, but the distinction between a theory and a hypothesis is an important one.

Not everyone is a math wiz that, through numbers, can provide flawless evidence. Some of us can only create a theory...

Hypothesis. And it's not about being a math whiz, it's about critical thinking: "three naked Headhunters can beat three naked TIE Fighters, so the Z-95 swarm is going to take over the TIE swarm's role in the metagame" isn't defensible for reasons that have nothing to do with math.

...and do our best to present our findings to the community in hopes that it sparks new discussion.

God forbid if said theory goes against someone's opinion.. All hell breaks lose..

The OP could have said something like this: "I've been noodling around with the Headhunter, and was really surprised at how well it did against vanilla TIE Fighters. There seem to be some battle reports coming up with similar results, too. What do people think about the Z-95 swarm's chances against a TIE swarm? I think they're going to be a big deal in the metagame for the next few months."

That would have gotten some attention and started a conversation about the Z-95's strengths and limitations, how you can build squads with them, and how likely those squads are to be competitive against the heavy hitters of late Wave 3 and the other new toys from Wave 4. But that's not what happened, primarily because the OP (a) represented himself (or herself) as bringing some unique new knowledge to the table and (b) was clearly wrong.

While we're on about math, i'd just like to take this opportunity to note that I'm bummed that the rebel prime list is 119 points, and the imperial prime list is at 101 points .... need to figure out a way to shave a single point...

(Imperial prime list)
Academy Pilot

Scimitar Squadron

Obsidian Squadron

Night Beast
Mauler Mithel

Soontir fel + VI

That's a bit harsh, Vorpal. The thread title is a question and, at least on the first page (I haven't read every page since I'm growing weary of these same debates everywhere), the OP is pretty clear that his or her belief that the Z will take over the meta is a question.

We don't have to follow rigorous, peer reviewed methodology on this board and personal experience is still valuable data. Yes, a single game can vary widely from the norm and is open to all sorts of bias and data--which should be observed--but it does show us something.

In this case, three named Zs beating three naked TIEs supports the Z's higher joust rating. OP then explicitly asked for others to run experiments with fuller swarms. That's very cordial.

You're right to point out that OP made some serious errors in his or her post, but that inquiry was not one of them. I should hope that we can retain that courtesy on both sides of the debate--whether we're graciously accepting corrections or humbly offering them.

Vorpal...

Fair enough. I understand what you are saying. He was wrong, we get it. He did not handle this gracefully in the least bit..

But does that give us credence to help devolve these forums into the shitfest it has become over the last few months? How many threads have you witnessed turn into comeplete **** during this week alone?

There is nothing wrong with debate and friendly critism. But this place is becoming just another internet forum.. And I'm pretty sure none of us want to see that happen.

If we aren't part of the solution to that, we are part of the problem.

Translation for the math wiggers:

63x + 5838(39) / 46y * 73x / .875 = Fly Casual, Post Casual!

(That was a friendly dig and by no means was meant to disrespect math as I understand its value)

So clearly what's working against the swarm isn't the Z-95 chassis, but rather the missile.

The ship and the weapon are not entirely separate things. The missile provides the damage, the ship enables missile strategies that you couldn't use before wave 4. Take the z-95 away and assault missiles might as well not exist for rebels.

Maybe it will help clarify my point to translate it into another context: say I'm a biologist* testing antibiotics. I have a new chemical compound (B) I want to compare to the current standard (A), so I set things up this way. I have 20 agar plates covered in bacteria (created from the same single initial plate, grown with the same ingredients at the same temperature in the same space). I treat four plates with A, and four plates with B. I also treat four plates with both A and B, because I'm thorough. Finally, I also have four controls ©, which get no treatment, and four comparisons (C') which I treat with the same purified water I use to administer the A, B, and AB treatments.

[*I'm not actually a biologist, so I apologize to real biologists if I got any details wrong here.]

So I do all of this work, measure the growth or contraction of the bacteria on my plates, and I end up with the following:

A alone causes 30% more bacteria to die off than in C and C'.

B alone causes no change in comparison to C and C'.

A and B together cause a 50% kill in comparison to C and C'.

If you go off and report excitedly to your colleagues that you've discovered a potent new antibiotic, they're going to be confused: your results clearly say that B has no effect at all on the bacteria. What's actually going on is that B is a "mediator" of A's effect--there's an interaction between A and B that kicks bacterial butt, but B alone (clearly) doesn't do anything.

***

Of course bacteria are the TIE swarm, A is Assault Missiles, and B is the Z-95. And this picture is simpler than the real one because Assault Missiles never happen alone; they always co-occur with some ship, and the Z-95 is the game's most cost-effective delivery platform. But since Z-95s without Assault Missiles are not particularly effective against the swarm, and other ships are also effective against the swarm when they take Assault Missiles, it's clearly Assault Missiles that are doing the job.

(To return to the analogy: the standard of treatment is that A is always administered with X or Y. Even if you find that AB is significantly more effective than AX or AY, the conclusion isn't that B has strong antibacterial properties--it's that B is a more effective mediator of A's effect than X or Y is. And even if there seems to be one preparation of B that's more effective than other Bs, but that still doesn't make B an antibiotic.)

Vorpal...

Fair enough. I understand what you are saying. He was wrong, we get it. He did not handle this gracefully in the least bit..

But does that give us credence to help devolve these forums into the shitfest it has become over the last few months? How many threads have you witnessed turn into comeplete **** during this week alone?

There is nothing wrong with debate and friendly critism. But this place is becoming just another internet forum.. And I'm pretty sure none of us want to see that happen.

If we aren't part of the solution to that, we are part of the problem.

There was the thread where someone came in and said he expected to hate the forum, and by the way we were all wrong about the Z-95. People said "no, we're not, and there's evidence" and he said "Oh yeah? What does your evidence say about MY WIFE'S TRAGIC DEATH?! I was right about you!" I think that says less about the forum than the poster in question.

There was the thread where someone--actually, I think it was jpltanis, but I can't check now--was promoting a YouTube battle report, and a few people decided that their discomfort with gender issues was relevant. That is about the forum, but it's not about people shouting down a friendly debate.

Then there's the Nth go-around of "X is underpowered and overpriced" and the oh-my-God-just-kill-me-already iteration of "PAHNTOM IS BROKEN GUYZ!!1!11 OK NO WAIT ACTUALLY ITS TEH OUTRIDDER TOTALLY BREAKIN MAH GAME!!!", sometimes in the same thread. Honestly I treat these mostly as noise, at this point, although sometimes there's someone who seems reachable and sometimes I just like to poke the bear. That is about the forum, but I think for the most part it's a symptom of the same problem jpltanis seems to have here: people aren't thinking critically.

...there's this idea that's popular in education right now of "metacognition", which just means thinking about thinking. Particularly in science ed, the idea is that metacognition is an important skill for students to acquire. Once you learn to set a thoughtful little sentinel between your input and your output, it becomes much easier to revise your own misconceptions and mistakes rather than relying completely on someone else (i.e., the teacher) to fix them for you.

(shrug) I can't fix the world, and I can't even fix the forum. I can fix my own behavior and level of frustration in response to people being thoughtless, and I know for sure I've flown off the handle in the past--notably but by no means solely at MajorJuggler (to whom I owe an apology), and I'm trying to keep a better eye on that tendency.

But I've read back through this thread, and I'm comfortable with my conduct. I'm not being nice (although I think I started out fairly politely), but my criticism is consistently directed at the OP's conclusions and behavior and not at his or her person.

There was the thread where someone--actually, I think it was jpltanis, but I can't check now--was promoting a YouTube battle report, and a few people decided that their discomfort with gender issues was relevant. That is about the forum, but it's not about people shouting down a friendly debate.

Is that what happened? I only noticed that thread was missing because I woke up one morning with likes, and when I clicked to see which posts they were the thread had disappeared. There was nothing particularly contentious about it when I'd gone to bed the night before, so I was left scratching my head at that one.

-Nataris

It's very hard to have an honest discussion with someone that wants you to buy their assertion that 2+2=5.

We've had one poster in this thread put forward that he think little of the Z-95 because it's two Agility dice stack up unfavorably to the 3 Agility dice of a TiE. Another posters points out that at their respective Hull and Shield values that actually the Z-95 ends up being slightly more survivable despite the 2 Agility. The first posters then dismisses this because he doesn't want to muck up his games with "Maths". That's well in good but it doesn't change the fact that 2+2 does not equal 5.

Then we have the OP in this thread. He wants us to believe that because of his testing that Z-95s may be the superior swarm ship and will overtake TiEs as such. His backing for this is 3v3 games using an "AI" simulator that not only can't fly a low PS TiE at all but cheats. He wants us to believe that something is there because not only does he win as the Z-95, but so does the AI. Well that must prove it right? Oh yeah I mentioned the AI cheats right? Yeah the AI gets a free Target Lock every round for every ship, and doesn't take stress. So yeah if you make every Bandit Squadron Pilot into Vader, then 2+2=5. but otherwise you are just arguing nonsense.

Also, I'd just like to point this out (because someone alluded to it)

A theory is an idea that you'd like to know the validity of. A hypothesis is a clearly defined and testable situation with recorded and repeatable parameters. And generally, if you have a theory, you're just thinking. If you have a hypothesis, you're testing it. And if you test it with a direly flawed parameter, then it's not valid. That's what seems to be causing all the hot water in this thread, not people being impolite. If OP wants to test his hypothesis, then he needs to play a Headhunter swarm against a variety of skilled and unskilled opponents, and in turn have a variety of opponents play against him with one. Probably wouldn't hurt to have some Z on Z swarm games too for comparison as well…

Of course once you consider Howlrunner you need to consider Blount. For example:Blount, assault missileTala x6vs.HowlrunnerTIE fighter x610 points of upgradesNow the imperial player has a difficult choice: make good use of Howlrunner's buff at the cost of having only 2 HP on all of their ships, or spread out to avoid splash damage and have trouble focusing on a single target while the ball of z-95s is free to blob up and smash its targets one at a time.

I've been running this z-95 swarm for a few weeks now and I have been having enormous success. A good friend of mine has recently started making battle reports and here is one that features my z-95 swarm vs a pure defender list.

http://community.fantasyflightgames.com/index.php?/topic/110421-7-z-95-headhunters-vs-3-tie-defenders-176mb-of-images/

Oh yeah I mentioned the AI cheats right? Yeah the AI gets a free Target Lock every round for every ship, and doesn't take stress.

Not sure what AI the OP was using, but the one I use doesn't do either of those things... Not saying the OP's testing is valid, but the AI I use doesn't cheat.

The AI he uses which he linked in this thread a page or so ago tells you to take a free action TL with any ship able to TL prior to taking your actual action.

That is why the Z-95s are winning regardless of whether he or the AI plays them. The AI can't fly the TiEs and cheats when flying the Z-95s.

In reality if his tests mean anything it is the opposite of his theory as the Z-95s are being given substantial outside advantages in order to win that match. Good thing for the OPs theory that the testing he is doing means nothing.

The AI he uses which he linked in this thread a page or so ago tells you to take a free action TL with any ship able to TL prior to taking your actual action.

That is why the Z-95s are winning regardless of whether he or the AI plays them. The AI can't fly the TiEs and cheats when flying the Z-95s.

In reality if his tests mean anything it is the opposite of his theory as the Z-95s are being given substantial outside advantages in order to win that match. Good thing for the OPs theory that the testing he is doing means nothing.

No, that is not how one decide what action to take.

The application's wording is kind of weird, but I believe the intent is to select the target lock action, if you can spend it in combat or else take the focus action.

Like the actual game, you can only take one action. So you guys are doing it wrong.

There is only one difference. When the AI makes a red movement, it does not perform an action, but no stress token. (See app's instructions on page.)

Choose first action that matches criteria, based on list on the main page for the selected ship

NOTE: Target Lock is a free action, choose this plus first other action that is available in the list.

That's taken directly from the instructions on the website you linked. All ships with the TL Action available have it listed as

Obtain on targeted ship as a free action.

Clear Target Lock at end of turn.

Like the actual game, you can only take one action. So you guys are doing it wrong.

Are we?

(See app's instructions on page.)

You might want to go take your own look first.

I didn't use it the way you had described in my experiment. I tried to play it as close to a human person and with the actual game rules as much as possible. It was primarily to have it choose the movements. I want it to emulate a person as much as I could.

Just tried the 8 Ties versus 8 Z match friday night (human players at both ends) and I no longer thinks that the Z have a big advantage over Ties.

I found that the Ties have a much larger firepower advantage than I thought over Zs.

This is confirmed by my mathwing calculus:

Discounting crits, the only range at which Ties have less than a 4/3 firepower advantage is at range 2 if the defender is wiithout focus and even then it's still close to a 4/3 advantage. If the defender has a focus, then the Ties have more than a 2 to 1 firepower advantage at range 1 and 3, and only close to 5/3 at range 2.

Crits can obiously be a factor, but even in the best case scenario, a Z has less than a 1 in 6 chance of dealing a crit to a Tie (range 1 focused shot on a focusless tie) so an important crit will be rare.

Thanks for doing this test.

May I ask, how competitive was this battle? Was it one-sided? You hinted on the dice effect. Did one side just have extremely good or bad rolls?

This is confirmed by my mathwing calculus:

Discounting crits, the only range at which Ties have less than a 4/3 firepower advantage is at range 2 if the defender is wiithout focus and even then it's still close to a 4/3 advantage. If the defender has a focus, then the Ties have more than a 2 to 1 firepower advantage at range 1 and 3, and only close to 5/3 at range 2.

Could you explain this in more detail? How exactly does the TIE have a firepower advantage when both ships have two attack dice and you don't have Howlrunner? If anything the z-95s should have the advantage since they can take target locks to boost their initial long-range shot and then any ship that didn't spend its target lock will have a focus + TL stack on the following turn (probably at range 1).

If you're referring to the fact that the TIEs are shooting at 2 agility with 2 attack while the z-95s are shooting at 3 agility with 2 attack then you also need to account for the fact that the z-95 has an extra HP to make up for that lower agility.

Crits can obiously be a factor, but even in the best case scenario, a Z has less than a 1 in 6 chance of dealing a crit to a Tie (range 1 focused shot on a focusless tie) so an important crit will be rare.

This may be true for a single ship, but a 1/6 chance is still significant when looking at an entire list. Assuming your 1/6 number is correct a whole squad of eight z-95s has about a 75% chance of getting a crit through each turn. And given the number of crits that inflict extra damage or cripple a ship there's a non-trivial chance that you'll be losing a ship that you wouldn't have lost if you had shields.

It was actually pretty one sided in favor of the Ties result wise with better rolls on the tie side. We'll probably do more tests because of the rolls even if we tried to correct by removing a damaged tie from play.

Range 3 saw 4 ties against 4 Z: the Zs did nothing while the ties stripped 2 shields. This is slightly better than expected for the Ties, but nothing to write home about.

Range 1 saw 2 damage on a Tie (average should be around 2 or 3) with 4 Z at range 1 and 4 at range 2 most with no action due to academy blocking. The ties destroyed the damaged Z and another one. WIth 4 Ties at range 1 and 4 ties at range 2 with focus, this is expected. We removed the damaged Tie to continue the test.

We stopped the test a couple of turns later, when the Z were reduced to 5 (2 or 3 damaged) and the Ties still had 7 ties (2 damaged).

During the engagement: at range 3 the ties have a firepower advantage. At range 1, a good proportion of the Zs can usually be blocked, meaning that the firepower advantage the Ties have is actually increased. The Ties are also more agile in a furball, meaning that they can concentrate firepower more while getting actions.

I appreciate the report, your honest observations, and most of all your willingness to do more tests. I can't thank you enough. I lack the Z95s to do this myself. I know with 16 ships on the table, the battle tended to be long. I hope you enjoyed playing it.

Keep us posted.

This may be true for a single ship, but a 1/6 chance is still significant when looking at an entire list. Assuming your 1/6 number is correct a whole squad of eight z-95s has about a 75% chance of getting a crit through each turn. And given the number of crits that inflict extra damage or cripple a ship there's a non-trivial chance that you'll be losing a ship that you wouldn't have lost if you had shields.

Of all crits, the only really crippling ones for a Tie in a swarm are:

direct hit, blinded pilot, weapon malfunction and minor explosion (only count 1 as it has a less than 50% chance of being a problem) this means that 12 in 33 crits are a big problem. Consider that a tie only has 3HP, this means that more than 1/3 of the crits will target your last hull point (could also be the 2nd hit on a single HP Tie (or third on a 2HP TIE )). Combine those 2 numbers and you will see that only 2/9 of all crits you receive will matter. Now also consider the 1/6 number I gave you and you will see why I said that meaningful crits will be rare.

For the firepower comparison : if you take a TL, then the Ties have an even greater advantage in firepower as you cannot use it in defense. Also consider that if you can lock, then you can be blocked on the following turn meaning no lock+focus stack.

Now also consider the 1/6 number I gave you and you will see why I said that meaningful crits will be rare.

No, they really aren't, because you're still just considering one ship. Assuming your 1/6 and 2/9 numbers are correct the swarm of eight z-95s has about a 26% chance each turn of inflicting a "meaningful" crit. IOW, each turn roll an attack die, on a blank remove one of your ships that would otherwise still be alive.

For the firepower comparison : if you take a TL, then the Ties have an even greater advantage in firepower as you cannot use it in defense.

Not really, because focus only works once and therefore the majority of defense rolls you make will be without focus (or will have no eyes to use your focus) either way. So in exchange for having one z-95 not have a defensive focus token I get all eight of my z-95s to have a target lock, and all of them get to shoot before your TIEs.

And of course you don't have to use the TL option if you don't want to, it's just silly to claim that the TIEs have a firepower advantage when the z-95s have the extra offense-boosting action available.

Also consider that if you can lock, then you can be blocked on the following turn meaning no lock+focus stack.

IOW, "if you're in range to shoot I get to block you next turn". Remember that the z-95s move after the TIEs, so they can move just barely into range 3 and get their target locks. And even if you do block them a leftover TL is better than no action at all.

And you're also ignoring anything that happens after the initial pass. Let's say we have a z-95 and a TIE with no shot this turn because everything is outside their arc. The z-95 can take a target lock for future use, the TIE can, at best, try to barrel roll into a better position. So, assuming the z-95 can maneuver back into the fight next turn it will either have a focus + TL stack or k-turn with a TL, while the TIE fighter just has its basic focus or actionless k-turn.