8 minutes ago, Dengar5 said:

I use a different simpler approach. Plus I disagree with some of the results of your ancient outdated Mathwing numbers. They have been a useful point of reference to confirm that my model is providing useful results. Here are some factors where I suspect our models differ:

• I discount 2-attack below 3-attack by a wider margin. The ratio between these two is highly subjective & open to interpretation. 
• I use 3-attack ships as the baseline because it is the most common attack profile.
• With TLT gone, 3-agility is more valuable than ever. 1-agility is overvalued in original Mathwing.
• I apply an 'early death factor' to account for 1-shotted TIE fighters. This was a helpful breakthrough. It's a curve to devalue ships that die faster.

It would be interesting to know a cost estimate you have for one of the new mobile arc turrets. Your model is likely better at costing those.

I have 3/2/x/x as doing about 70% more damage than 2/3/x/x. That said, there are different philosophies on how to even calculate average expected damage across a blender of the entire meta. Basically there is the series approach and the parallel approach. Using the series approach, you have to kill each target in it's proportion to it's appearance rate in the meta, and you get scored on your overall time to get through everything. In this approach ATT2 can get "stuck" on hard to hit targets. In the parallel approach, you attack everything in the meta, and score points based on damage done and how hard that target was to hit. So if you do a bunch of damage to X and nothing to Y, at least you score some points for X, whereas you would get "stuck" on Y in the series approach. So, I'm using the parallel approach for expected damage output.

The "early death factor" is quite an interesting rabbit hole. It works both ways; it decreases low PS (now IN) ship jousting values, and it increases the jousting value of high IN ships. It's also a function of how much of a tank vs glass cannon the ship is; if a ship is very tanky then it has less of an impact, but glass cannons really want to shoot first. Interestingly I haven't worked out an exact analytical equation for this, but I do have curve fits based on empirical (simulated) data.

The new mobile arcs are interesting. I assume you mean actual mobile arc, like the shadowcaster's ATT2 arc. I haven't gotten to it yet, but I'll likely treat the expected damage as a weighted average of the front ATT3 arc, and the side ATT2 arc. In general, for effects that can have a variable rate of triggering, I just run the numbers for 0%, 10%, 20%, etc etc, up to 100%. Then you can back into it with analytical playtesting to decide what % rate corresponds to reality, and you can pick that value off as representative of real world performance. Of course if someone flies better to get a higher trigger rate then it increases the value and cost efficiency.

12 minutes ago, MajorJuggler said:

I have 3/2/x/x as doing about 70% more damage than 2/3/x/x. That said, there are different philosophies on how to even calculate average expected damage across a blender of the entire meta. Basically there is the series approach and the parallel approach.

The "early death factor" is quite an interesting rabbit hole. It works both ways; it decreases low PS (now IN) ship jousting values, and it increases the jousting value of high IN ships. It's also a function of how much of a tank vs glass cannon the ship is; if a ship is very tanky then it has less of an impact, but glass cannons really want to shoot first.

The new mobile arcs are interesting. I assume you mean actual mobile arc, like the shadowcaster's ATT2 arc. I just run the numbers for 0%, 10%, 20%, etc etc, up to 100%. Then you can back into it with analytical playtesting to decide what % rate corresponds to reality

Interesting, I have 3-attack doing 84.9% more damage than 2-attack. I would describe the method I use as parallel. Weighting to the different agility levels is pushed towards higher agility levels. Also, weighting for range-1 attacks significantly affects the 2-attack profile. My range-1 weighting is conservative.

You very much understand 'early death factor'. I don't use IN in my model, but I'm mostly interested in generic ship efficiency. 'Early death factor' is dependent upon the ships defensive profile entirely. I think it greatly improves real world ship efficiency prediction.

Some turrets have front arcs, some don't. Some are double sided, some are single arc. It's a mess to estimate well. For the Shadow Caster I've currently settled upon 60% 2-attack & 40% 3-attack for the offense profile as a rough estimate. This heavily nerfs the ship & it will likely be way overcosted on release. Without boost & only a single K-turn, it will be difficult to utilize its 3-attack front arc. I have a hunch that the Shadow Caster title may boost the mobile arc to 3-attack. FFG may be doing this to eliminate effective double Shadow Caster builds.

1 hour ago, Dengar5 said:

Interesting, I have 3-attack doing 84.9% more damage than 2-attack. I would describe the method I use as parallel. Weighting to the different agility levels is pushed towards higher agility levels. Also, weighting for range-1 attacks significantly affects the 2-attack profile. My range-1 weighting is conservative.

You very much understand 'early death factor'. I don't use IN in my model, but I'm mostly interested in generic ship efficiency. 'Early death factor' is dependent upon the ships defensive profile entirely. I think it greatly improves real world ship efficiency prediction.

Some turrets have front arcs, some don't. Some are double sided, some are single arc. It's a mess to estimate well. For the Shadow Caster I've currently settled upon 60% 2-attack & 40% 3-attack for the offense profile as a rough estimate. This heavily nerfs the ship & it will likely be way overcosted on release. Without boost & only a single K-turn, it will be difficult to utilize its 3-attack front arc. I have a hunch that the Shadow Caster title may boost the mobile arc to 3-attack. FFG may be doing this to eliminate effective double Shadow Caster builds.

I have been playing around with various range distributions. Right now I assume the same range distribution for all ships, but I might generalize it going forward. My current distribution is:

range 1: 40%

range 2: 35%

range 3: 20%

range 3 + obstruction: 5%

Since so many attacks are at range 1, it helps ATT2 get some damage through, which might partially explain the difference. The flip side is that AGI3 durability takes a hit when you have that many attacks at range 1. Some of the rest of the difference will come down to how often you assume a 3/2/x/x statline has focus for attack, vs how often a 2/3/x/x statline will have focus for attack. Since I'm doing the hyperspace calculation explained above I don't have to "guess", it's always a pretty accurate number, the exact value just depends on some of the other assumptions like shots-per-round, and what kind of attacks are incoming. But everyone gets treated equally, so it's at least consistent.

5 hours ago, Hobojebus said:

Wat u mean I can only have 2 defenders?????!!!!!!!

BioPhysical list 2.0

3 hours ago, MajorJuggler said:

If his numbers are that far off then his approach is probably fundamentally wrong. I don't know what his approach is though, so I can only see his outputs.

Edit / P.S.: This isn't to say that his numbers won't be what FFG initially prices them at. I honestly have no idea how FFG determines their pricing.

Dartboard.

9 hours ago, Dengar5 said:

I believe they intentionally gave the kihraxz another hull so it would be costed 1 point below the X-wing at 42. It sets it apart from the mangler scyk, which very much should be 39-40 points for 5x in a list. I find it likely that mangler canons will only change to crit in the bullseye arc. Canons now getting range bonuses really helps the mangler scyk.

I could see Manglers working like that, although it's worth pointing out that with cannons and turrets now getting range bonuses, that at the range most likely for you to get a bullseye is going to also be the range that you'll be rolling 4 dice.

It's hard to say if Mangler will or wont be in, but since I don't feel it was intended as a hard counter to something specific (and Tem-Numb is packing an entirely different ability now) then my gut says it probably will make it in in some form. For reasons outlined above, it would either have to see a price increase, or get an Atk decrease.

On a similar line of thinking, I'd imagine Autoblasters and Turrets are probably gone now. Or if they aren't they will be heavily reworked, perhaps with the ability to turn blanks or focuses into hits, something that Concussions Missiles and Proton Torps both lost, for example.

Btw the Lancer doesn't need to get to 100% efficiency. A true 1.0 turret only needs to get to around 80% efficiency, assuming 100% efficient jousters exist. A mobile arc is pretty close to a 360 turret when played right. Calculating the straight line efficiency and comparing to the reference point is one way to see if it's costed right. (And that's including the PS effect btw)

The other option, which is more useful from a player's perspective, is to treat the lancers attack as 3/2/x/x, PLUS some additional 2/2/x/x damage. This represents the increased firing duty cycle that you get from a turret / mobile arc. Keep increasing the proportion of "free" extra shots until the ship hits 100% efficiency, and that's your answer for how many more shots it needs to get to break even.

Oops, hyperspace discussion was in the other thread here:

https://community.fantasyflightgames.com/topic/276239-iso-confirmation-can-you-spend-multiple-calculate-tokens-at-once/?do=findComment&comment=3340916

26 minutes ago, MajorJuggler said:

Btw the Lancer doesn't need to get to 100% efficiency. A true 1.0 turret only needs to get to around 80% efficiency, assuming 100% efficient jousters exist. A mobile arc is pretty close to a 360 turret when played right.

I have a turret ratio variable to match turret efficiency to jouster efficiency. I want the final % efficiency list should be comparable across all ships. Readability is important. You are correct about 80% turret efficiency. I would say 1.0 turrets only needed 75% even.

I have bypassed using fire duty cycle as a variable. For the Lancer I'm guessing it will have to hit 87% before applying the turret ratio. I think its mobile arc is vastly different from a true 360 arc. For the double arc turrets, their ratio is a much more conservative 80% equivalency to 100%.

12 minutes ago, Dengar5 said:

I have a turret ratio variable to match turret efficiency to jouster efficiency. I want the final % efficiency list should be comparable across all ships. Readability is important. You are correct about 80% turret efficiency. I would say 1.0 turrets only needed 75% even. 

I have bypassed using fire duty cycle as a variable. For the Lancer I'm guessing it will have to hit 87% before applying the turret ratio. I think its mobile arc is vastly different from a true 360 arc. For the double arc turrets, their ratio is a much more conservative 80% equivalency to 100%.

The efficiency depends on the PS-adjusted value. Chewie / Leebo was a pretty golden standard for what should be a "good" efficiency for a turret, and it was around 80% after considering PS. Going the last step to figure out how much shooting first helps is a pretty big deal to getting the model fine-tuned. I'm reporting both the PS-agnostic numbers, as well as the PS-adjusted numbers. The latter are almost always higher, since the PS-adjusted values are normalized to the lowly PS1 TIE. Chewie / Leebo were probably around 75% in PS-agnostic efficiency, but I'm too lazy to look up the numbers right now. :-)

The new "bowtie" turret arcs are unique because we have never had that mechanic before, so it'll be interesting to see what their target efficiency should be. Without a gunner crew to turn it into a legit 360 turret like 1e, the target efficiency will have to be >80%, since by definition it's inferior to a 360 turret.

By the way, I think you might be the only other person I know who is actually calculating jousting values! Which is pretty cool. To get all the levels of details I think you'll need something more complicated than a spreadsheet, but it's a good start!