So on average and across the entire meta game, 2 A-wings do slightly more damage than one ship with 3 attack dice. This is before considering the effect of the white K-turn on the Defender's action economy.

I'm assuming your values take some kind of consideration for number of green dice defenders (not Defender) will roll? Because 2 attack ships x 2 also means double the number of agility dice rolled when comparing to a single 3 attack ship. So for example, a TIE Defender will have a much easier time putting damage through the a-wing's 3 green dice than they will have trying to put damage through its 3 green dice (for example).

Right - obviously a Dagger performs exactly the same as a Blue if all your opponent brings are Academies. But I thought the conversation here based on the OP was more about the Defender's cost, and not on specific matchups.

Straw man argument. The only place I tried to associate a value to the Defender are based on my Lanchester's thread math, and the jousting value there most certainly includes both the offense and durability. It's also equally valid to take two sets of ships ships in a specific matchup, and calculate the damage PDFs and their resulting relative jousting values.

So on average and across the entire meta game, 2 A-wings do slightly more damage than one ship with 3 attack dice. This is before considering the effect of the white K-turn on the Defender's action economy.

I'm assuming your values take some kind of consideration for number of green dice defenders (not Defender) will roll? Because 2 attack ships x 2 also means double the number of agility dice rolled when comparing to a single 3 attack ship. So for example, a TIE Defender will have a much easier time putting damage through the a-wing's 3 green dice than they will have trying to put damage through its 3 green dice (for example).

Yes. Attacker damage is an average across a variety of defenders, and also range 1-3, and focus on offense and defense. If I recall correctly, the math to derive about 1.75x assumed approximately an equal chance of shooting at agility 1, 2, and 3.

The high end of a curve never makes up for the low end. If I can dodge 99 out of 100 shots I'm still dead at some point. If that ratio is relevent against the number of games in a tournament I'm in serious trouble.

It's not flawed, it's Murphy's Law. And it's how you have to plan if you want to play at your best.

Yes, it is flawed. Deeply. This is like never playing your star quarterback because he may get hurt, especially if it's over the course of 20 games.

Risk management is about more than just the impact - it's also about probability and mitigation. You seem to that assume even slight probability is guaranteed, and mitigation is impossible. That makes you leave a lot of potential sitting on the sideline.

I mean, no one should be running Fel, right? I've seen him oneshot at Range 3 a few times. So, that obviously means he can't win tournaments....

Right - obviously a Dagger performs exactly the same as a Blue if all your opponent brings are Academies. But I thought the conversation here based on the OP was more about the Defender's cost, and not on specific matchups.

...

Citation? You're clearly putting words in my mouth. The only place I tried to associate a value to the Defender are based on my Lanchester's thread math, and the jousting value there most certainly includes both the offense and durability. It's also equally valid to take two sets of ships ships in a specific matchup, and calculate the damage PDFs and their resulting relative jousting values.

You said this in relation to the 2 A-wing vs. Defender comparison:

Actually, after the refit, they do.... So on average and across the entire meta game, 2 A-wings do slightly more damage than one ship with 3 attack dice.

That's purely cost:attack, completely ignores the reality of matchups, and compresses everything to a single value that you then compare to the Defender.

The point I was trying to make is that you cannot ignore matchups when you make comparisons like this, and you cannot distill down the entirety of the metagame to one number that says 3 attack is X better than two. How much better it is will depend heavily on the matchups in the specific game. 2 dice against 1 agility isn't bad; 2 dice against 3 agility is going to see a great many useless shots. 3 attack can hold its own against 3 agility, making it more comprehensive for a variety of opponents.

This makes your 1.75 multiplier meaningless, for two reasons. First, as I pointed out earlier, it does nothing to account for the frequency of scenarios - it can't. Second, half your data points fall away once the game starts. If I'm facing a TIE Swarm, it doesn't really matter what my 2 attack can do against 1 agility. And then you tack that onto cost, which becomes yet another meaningless element once the game starts.

Okay, clearly this is no longer an argument. The folks in the "Tie Defender Sucks" camp will not be moved from their positions, and the folks in the "Give it a shot" and "It's worth its points!" camp will not be moved from theirs.

I'm going to go actually do a comprehensive thread on the math of 2 Chaardan Prototypes vs the Delta. Debate me then.

The high end of a curve never makes up for the low end. If I can dodge 99 out of 100 shots I'm still dead at some point. If that ratio is relevent against the number of games in a tournament I'm in serious trouble.

It's not flawed, it's Murphy's Law. And it's how you have to plan if you want to play at your best.

Yes, it is flawed. Deeply. This is like never playing your star quarterback because he may get hurt, especially if it's over the course of 20 games.

Risk management is about more than just the impact - it's also about probability and mitigation. You seem to that assume even slight probability is guaranteed, and mitigation is impossible. That makes you leave a lot of potential sitting on the sideline.

It isn't so durable as to be unkillable.

It doesn't have amazing firepower. If anything it's a little lackluster for the cost.

What sets it apart?

A White K-Turn. It isn't worth the cost. 30 points is a whole lot of points to spend on any single ship. I want it to do something amazing. The Firespray and YT both do this, for their cost. The Defender though. It's eh. Like non Dutch Y-Wings. It's not Arvel. It's not the tie Advanced. But it isn't as cost efficient as an X-Wing, despite its ability to take them one on one.

3 of them still probably lose to 4 X-Wings.

1 of them and Ties probably loses to a Firespray and Ties.(Because Firesprays rear arcs are the single best counter to a Defender.)

Etc.

Find me a list that includes a Defender and can reliably stand up to a player just as skilled as you fielding a competitive squad. They're probably rare.

It's weak to Stress. It's weak to Turrets. It's weak to Firesprays. It's Weak to Buzzsaws. That's a lot of weaknesses to things we'll be seeing running the field.

It is strong against Tie Swarms. But probably not strong enough to put up with a full Salvo better than a Firespray or YT.

I can understand this discussion for the tried and tested Advanced but the TIE defender hasn't even been released yet.

You said:

You cannot simply remove half of the math and declare victory because of that one time at band camp you saw Vader get violated. A-wings do NOT have better offense for the cost, because it's not linear.

That was the context. Nothing else. I then said:

A-wings do NOT have better offense for the cost, because it's not linear.

Actually, after the refit, they do. See earlier in this thread: the average damage of 3 dice is about 1.75x that of 2 dice, taking into account an average of the meta for various agility defenders, and a variety of action economy scenarios.

2 x 1 = 2 > 1.75

So on average and across the entire meta game, 2 A-wings do slightly more damage than one ship with 3 attack dice. This is before considering the effect of the white K-turn on the Defender's action economy.

That was all I said. I was specifically talking about cost:attack because it was in direct reply to your comment, which was about cost:attack.

As far as statistical averages being meaningless - you are entitled to your opinion. The fact is, that applying those statistical averages to the Lanchester's model results in ship value predictions that are VERY correlated competitive tournament usage, for the ships that the method should be able to accurately predict value for (ships with non-unique traits). In this case the method won't work well on the TIE Defender because of its dial anyway, so it can only make general statements about its jousting value.

I'll be happy to run the numbers again limiting it to agility 3 targets. It'll only take a couple seconds to load up, change 1 variable, and rerun.

Edit: Also, we are in total agreement that particular matchups can and will vary from the statistical averages across the entire metagame. I posted as much in the Lanchester's thread. I pulled out that 1.75x multiplier to compare the Defender and A-wing in a general sense, not for their specific matchup against each other.

Find me a list that includes a Defender and can reliably stand up to a player just as skilled as you fielding a competitive squad. They're probably rare.

It's weak to Stress. It's weak to Turrets. It's weak to Firesprays. It's Weak to Buzzsaws. That's a lot of weaknesses to things we'll be seeing running the field.

It is strong against Tie Swarms. But probably not strong enough to put up with a full Salvo better than a Firespray or YT.

How is the defender weak vs buzz saws? pretty sure that white K-turn makes it easier to get behind them

Find me a list that includes a Defender and can reliably stand up to a player just as skilled as you fielding a competitive squad. They're probably rare.

It's weak to Stress. It's weak to Turrets. It's weak to Firesprays. It's Weak to Buzzsaws. That's a lot of weaknesses to things we'll be seeing running the field.

It is strong against Tie Swarms. But probably not strong enough to put up with a full Salvo better than a Firespray or YT.

How is the defender weak vs buzz saws? pretty sure that white K-turn makes it easier to get behind them

He probably means that the high agility is somewhat mitigated by getting a reroll from Gunner. But yeah, white K-turns, or even white turns for that matter, should be sufficient to get behind the Buzzsaw or at least dance behind it.

I can understand this discussion for the tried and tested Advanced but the TIE defender hasn't even been released yet.

Some of us like me are math nerds and point out that on paper it can't joust well, but the white K-turn throws a lot of standard calculations out the window.

Some of us like KO have actually played with the ship extensively on vassal, and have quite a bit of useful experience with it.

In either case, what it really means is that we're excited about the ship but are trying to figure out where exactly it will land.

I find it difficult to think of things as either overrated or underrated before they're even released.

Sometimes the only exercise I get all day is jumping to conclusions.

Aminar:

I understand your point, and you are correct that a tournament build must be above all things consistent. A list that relies on green dice is not consistent, regardless of averages, which is your point. On that, you and I agree.

However.

You are neglecting to consider that a squad that relies on red dice (B-Wings for example) because they cannot evade damage and simply must destroy the enemy quicker than they run out of hull points is no less vulnerable to outliers. I have lost games when my X-Wing firing 4 dice at R1 has missed entirely against a 1 hull YT-1300. I have lost games where an entire focused swarm nets 0 damage against a point blank range X-Wing. These things happen no matter what ships you are running.

The only real defense, as you noted, is to avoid getting shot at altogether. This is why I don't understand your unreasonable hatred for the Defender. It is VERY good at avoiding fire. It is easily the equal of A-Wings, Interceptors, and Adv Sensor B-Wings. The fact that a move is predictable doesn't automatically make it a bad move. My opponent can know I am about to K-Turn behind them all they want, but if they are stressed from K-Turning last turn there isn't a thing they can do about it.

The drawback for green dice in this game, as noted by others, is that up until now every 3 green ship in the game has relied pretty much entirely on them for defense. A single shot at range 3 has killed many, many interceptors and more than a few A-Wings. That 1/20 shot can end entire games if it happens to your 32 point Soontir on the first volley. But if your 1/20 event happens against a Defender, it loses shields. No big deal. You need TWO major outliers, happening to the same ship and early in the game, for the Defender to be lost. Those odds are very low, low enough that you can indeed plan around it.

Six health, 3 from shields, doesn't just mean it lasts longer it also ensures that an early negative outlier does not finish the ship off. The more shots it takes, the more the green dice will tend to "even themselves out". The Defender is one of the most reliable ships in the game because of this, you cannot just ignore the moderating effect of +3 shields on a 3 Agility ship if your argument is to be a valid one.

As far as statistical averages being meaningless - you are entitled to your opinion.

Let's be careful not to overstate my position here. I don't think statistical averages are meaningless - I think building statistical averages off of incomplete models which disregard things they can't handle is meaningless. "Average number of evade rolled on 3 dice without focus" is a statistical average, but that does not mean it can encompass the performance of a TIE Fighter.

If you actually have some of the data I think you're missing, I'd love to see it. For instance, how you modeled the distribution for number of shots at given ranges, with what sets of tokens on either side, and the distribution of target agility values. Because lacking that, any claim to a meaningful statistical average is just fiction.

If you actually have some of the data I think you're missing, I'd love to see it. For instance, how you modeled the distribution for number of shots at given ranges, with what sets of tokens on either side, and the distribution of target agility values. Because lacking that, any claim to a meaningful statistical average is just fiction not verifiable.

Fixed that last part for you.

The assumptions are in the Lanchester's thread. You can recreate the results by hand in Excel, it will just take a while to copy down the average damage for every single possible scenario. Its actually on my list to incorporate everything into Excel rather than in Matlab, so people can poke around with it, change the ratios for everything to see how the results play out. For any sort of reasonable spread of agility on the defending ships, the biggest factor in the results by FAR, is actually not target agility or the range bins: it's how often the defender has focus. I assumed the attacker had focus 2/3 the time, and the defender has focus 1/2 the time. If you revise the 50% focus on defense down further then 2 attack dice become more attractive.

If you actually have some of the data I think you're missing, I'd love to see it. For instance, how you modeled the distribution for number of shots at given ranges, with what sets of tokens on either side, and the distribution of target agility values. Because lacking that, any claim to a meaningful statistical average is just fiction not verifiable.

Fixed that last part for you.

Without changing it, as far as I'm concerned.

The drawback for green dice in this game, as noted by others, is that up until now every 3 green ship in the game has relied pretty much entirely on them for defense. A single shot at range 3 has killed many, many interceptors and more than a few A-Wings. That 1/20 shot can end entire games if it happens to your 32 point Soontir on the first volley. But if your 1/20 event happens against a Defender, it loses shields. No big deal. You need TWO major outliers, happening to the same ship and early in the game, for the Defender to be lost. Those odds are very low, low enough that you can indeed plan around it.

KO, great post. Again this is why I want to add a subroutine to calculate and plot the PDF for "shots required to kill a target". I already know that the Firespray and Defender's PDF must by definition look more like a bell curve than any of the other ships (Central Limit Theorem, everything converges into a Gaussian Distribution with enough averages), but I would love to actually SEE the curves...

If you actually have some of the data I think you're missing, I'd love to see it. For instance, how you modeled the distribution for number of shots at given ranges, with what sets of tokens on either side, and the distribution of target agility values. Because lacking that, any claim to a meaningful statistical average is just fiction not verifiable.

Fixed that last part for you.

Without changing it, as far as I'm concerned.

Well thankfully in this case it is perfectly verifiable and the assumptions were all clearly stated many months ago, it just requires the willpower and/or ability to recreate the work. That's the scientific approach.

Generally, if someone is unwilling or incapable of recreating an experiment which has been well documented and perfectly repeatable, but insists that the outputs of the experiment are wrong, then their argument holds very little weight.

I don't think that's exactly the case here. You seem to be more curious about the underlying assumptions, rather than the denying the mathematical accuracy of the singular numerical output. And obviously we both agree that particular matchups can and will differ from the averages.

You cannot simply remove half of the math and declare victory because of that one time at band camp you saw Vader get violated.

Oh God why!? I'm trying to forget it ever happened!!! ... Sometimes at night, I can still hear him cry.....

Bottom line is the Tie Defender will be a commercial success due two the upgrades cards coming with it. We will all own 4 of them.

Bottom line is the Tie Defender will be a commercial success due two the upgrades cards coming with it. We will all own 4 of them.

Ha ha ha, I don't know about everybody... but a true Scotsman would certainly own 4 of them!

Disclaimer: I am not Scottish, but I do have 4 on pre-order.

Edit: FFG making this game is like a license to print money.

You seem to be more curious about the underlying assumptions, rather than the denying the mathematical accuracy of the singular numerical output.

I'm more pointing out that they are assumptions.

Even with the amount of game state you choose to ignore for your model, there's a massive amount of information that you just invent numbers for. Anything which isn't printed on the cards, you basically have to make up. It's a Garbage In, Garbage Out thing. Without any way to validate either the inputs or outputs, the numbers aren't anything but fiction no matter how accurately the math is completed.

All this has been covered before. Those who like your numbers use them. Those who see weakness have pointed it out, and generally ignore them. But please, refusing to recreate a broken experiment with no meaningful validation on either input or output is not a sign of lacking either willpower or ability.

Six health, 3 from shields, doesn't just mean it lasts longer it also ensures that an early negative outlier does not finish the ship off. The more shots it takes, the more the green dice will tend to "even themselves out". The Defender is one of the most reliable ships in the game because of this, you cannot just ignore the moderating effect of +3 shields on a 3 Agility ship if your argument is to be a valid one.

This concept is known as "Regression to the Mean":

Among other things, it means the more times you run specific statistical likelihoods, the likelier that outliers will cancel each other out.

What are the minimum number of red dice to kill a Defender, and what are the odds that they will do so?

3 dice hit shields, and 3 dice hit Hull, or 2 dice hit hull if one of the two was a Direct Hit! or a successful Minor Explosion.

The former is 7/33 of the damage deck, and the latter is a 3/8 odds of a 2/33 pull.

This is theoretically possible in a number of ways. The most red dice one can throw at a time in this game is 7 (Advanced Proton's 5, + Opportunist, + Jan Ors. Darth Vader crew would also work, as instead of throwing a die, he simply adds a damage at the end).

Those 7 can theoretically 1-shot a Firespray-31, I might add.