ShadowJak said:

Actually, it is supposition. Just because the formula seems to work, does not mean this is actually the formula that FFG came up with and uses.

dvang said:

This is true. But I'm fairly certain at least for the small size ships. I wouldn't be surprised if the big ships get a discount because having a bigger base is a disadvantage.

Dan said:

It seems you were correct about the discount for large ships, or they have an altogether different formula for them (I think probably the former, -10 looks about right).

Oddly, it looks like the skill one YT did not get any discount for being large, it conforms to your prediction of 26. Maybe for game balance reasons, at 16 points you could have a squad of six!

I apologize for the thread necromancy, but I can't figure out how to use this forum software to send a PM.

Shadowjak, would you be willing to share your XLS file? I'd like to try a different modeling technique (exploratory factor analysis) and see what kind of fit I can get.

I have no problem sharing the Excel file but there isn't anything there that hasn't been shown in the original post. The regression analysis table is just text and numbers because that is how Excel does it. If you want to change the numbers, you have to output a whole new report.

-----------------------------------

Looking back at what my projections were, it is interesting that my estimate of the YT-1300 named pilots was off by 10 but my estimate of the 1 skill pilot was almost exactly right (they apparently added a point to it the same way they did for the Academy Pilots and most named ships). It seems like they added a discount not based on being a large ship but based on the fact the squad point costs would have been more than half the team before upgrades.

I appreciate all the work you've done on the regression analysis, since I think having this info is important for those of us that want to come up with "official", and/or useful "unofficial" stats for fighters and other vessels that FFG may not ever produce, or are unlikely to produce in the near future.

I imagine we might also inquire directly with FFG to see if they are willing to share their points valuation formulae with us. A number of rules makers in the past have been willing to do that, so hopefully FFG will be willing to do so also.

If not, I guess we can just continue to proceed independently, as you've done already.

In looking at your formulae, the results sound very good, so I really appreciate your willingness to share your info with us.

I tried a different, less complex approach, assuming that the FFG folks might just go with something like that, and then make minor tweaks at the edges, in order to try to come up with an acceptable playbalance, after some playtesting.

The follow system, which is very similar to yours, seems to work well for the two Tie models, and the X-Wing. It also works for the Y-Wing, but only with a bit more substantial tweaking. Still, given that the values for even the Y-Wing are consistent, I think it might be valid, or close as well.

The formula is as follows:

2 Points for each fighter hull

Then, 1 Point each for each of the following: Attack Value; Defense Value (Agility); Hull Value

5 Points for each Shield (seemed a bit high to me at first, but works. I didn't try increasing the Hull values and reducing the Shield values)

1 Point each for the Pilots' Skill Rating Value, e.g. Skill of 2 = 2, Skill of 4 = 4, Skill of 9 = 9, etc.

Using the above values, we get to within one point each for the Tie, Tie Advanced, and X-Wing values given in the game, so I think that works fairly well. With the following minor modifiers, it works spot on:

+1 Cost Point for the Academy Tie Fighter Pilot (presumably, as already mentioned, a minor tweak to keep 9 x Tie fighters from being built in a 100 Point force, so the max is 8 x Tie Figther pilots, with a points cost of 12 each, instead of the 11 points calculated using the formula).

-1 Cost Point for Tie Advanced and X-Wing fighters, if their Pilot Skill rating is 4 or less. Presumably, since they have no special abilities.

For the Y-Wing, as mentioned, the above formula doesn't fit the standard. However, since it is the only Agility 1 (Defense Rating) spacefighter in the game thus far, I submit that a major tweak was made to it, that still fits with the original formula.

After running the numbers, I believe that is the case. Use the standard formula listed above, for a base value, but then make a major adjustment, due to its lack of Agility, and also due to its slower speed (unless using a Red maneuver), by subtracting 8 from the final points total. Depending upon the pilot chosen, this results in either a 25% reduction in capability for the best rated pilot, and a 33% reduction for the worst pilot. If this formula is followed, the point values for the Y-Wings match exactly with those in the game for the rated pilots (e.g. 8 points less than the calculated value, for all of the pilot variants of the Y-Wing).

I haven't tried using this system for any of the larger vessels, like the YT cargo vessel, or Millenium Falcon yet.

I've used this to come up with points for the Tie Defender too, since I'd love to see that fighter included in a later release. I rated it at 3 Attack, 4 Defense (Agility, since it is very fast 155 MGLT according to references, while the X-Wing is only 100. Matches up well with the A-Wing for speed, which I think is rated at 150 MGLT), 3 Hull (same rating as a standard Tie in one reference I saw), and either 4 or 8 shields, depending upon which reference you want to go with. That gives it a point value of either 32, or 52, NOT including the Pilot Skill Rating, or other mods, e.g. Ion Cannons, or Missiles - it can carry both.

For the Ion cannons, I was thinking a rating of 2 Points for those, since they are fixed in place, and can only fire forward (Y-Wing's Ion Cannons cost 5, but can be used in a 360 degree arc), and the standard points values for the missiles.

Not sure how the Agility rating is derived, so went with a more conservative 4 rating, instead of giving it a 5, even though it is greater than 50% faster than the X-Wing, and still very maneuverable. Presumably, due to its speed, it will be less maneuverable at the faster end of the scale, hence the more conservative rating.

For speed, I was thinking of giving it an 8, but perhaps only a 7 for playbalance, with 8 being a Red Maneuver, so it isn't too powerful. I was also thinking it would have less turning capability at the higher speeds, e.g. for speeds 6 and 7, perhaps only easy turns, and at 8 it can only fly straight. Not sure how to handle the 180 degree maneuvers, but perhaps only those at speeds of 4 and 5. Hard turns permitted at speeds of 5 or less, in addition to all the other maneuvers, since it is supposed to be similar to a standard Tie. It should be capable of barrel rolling too.

Anyway, let me know what you think of the above.

Also, I noticed the defense stats for the Tie's hull seem to be about 50% greater in X-Wing than in the ratings for them in some of the computer games. Not sure if that is due to a different, upgraded variant, or if they also tweaked that for playbalance too, since otherwise, it seems as if you'd need THREE Tie Fighters to take on ONE X-Wing in an even match, instead of just TWO. I imagine they didn't want to include 3 x Ties in the box, along with just 1 x X-Wing mini.

Of course, you could probably just as well chalk it up to their better maneuverability, not wanting to make the number of defense dice for the Ties seem like it was too overwhelming to the Rebel pilots.

Might make for an interesting game variant, especially if you are a Rebel player, and like to see those weak hulled Tie fighters destruct when hit, by giving them only 2 Hull Points, instead of the 3 they have now.

I really enjoyed reading through this entire thread, though it's taken me some time to wrap my head around it. After consuming everything and going over the numbers for my own amusement, the values that ShadowJak came up with fit the core set very nicely for all of the vanilla ships.

What happens next is what happens in the development of every game. Playtesting. The first sign of this is the cost of the Academy Pilot being bumped from 11 to 12. As ShadowJak mentioned, and is likely the case, in testing a squad of nine TIE Fighters was too overwhelming, or perhaps the games lasted too long, or the board was too cluttered. Whatever the case (go test it yourself if you really care) the designers agreed that eight was the max number of TIE Fighters you could use at 100 points.

Then there are all the advanced aspects of a ship, the actions and upgrade points, which have a value of 0 in the linear model. To me this looks like another way the Developers simply balanced out game play. Perhaps, X-Wings and Y-Wings were tested with Evade and/or Barrel Roll and it was found that they simply could not be taken out that way. Conversely, perhaps TIE Fighters had the Target Lock ability and it was deemed too powerful (I can assure you that it is too powerful. A friend of mine and I tested this theory). Maybe the TIE Advanced getting to keep all four actions is a trade-off for giving up the Droid upgrade slot.

We'll never really know, unless FFG comes out and tells us all the little tweaks they made in developing the game. I'm not surprised to learn that the Wave 2 ships don't fit the same linear model. The TIE Fighter is the only outlier in the core set, with a base cost of 10 points. The Y-Wing is 16 points and the X-Wing/TIE Advanced are 19 points. Relatively speaking these are very close numbers. On a linear regression, this increases the likelihood that the model fits. The Wave 2 models continue to deviate from that line and thus it shouldn't shock anyone that they don't fit the linear model.

There will never be a simple (or complex) formula for calculating all of the ship values, but linear regression gives us some very important information. Namely, we have learned that for these four ships, the developers felt that red dice and green dice were worth similar point costs on a ship. We also know what a ship gets for free (2 Weapon, 2 Agility, 3 Hull, 0 Shield) before it needs to pay for improvements. Another valuable piece of information is that Pilot Skill is a linear upgrade.

Named Characters also stray from the linear equation, which should surprise no one, since it's extremely difficult to put a linear point cost on the kind of complex abilities the characters have. Again, point totals are the result of copious playtesting. What we learn from the linear model is not what the designers envisioned the squad costs to be, but what did they feel was more/less important to gameplay and squad building at a high level.

Thanks for putting up all this info ShadowJak and thanks to everyone who contributed to the discussion. It has been enlightening.

Any chance of posting the spreadsheet files?

DoctorMike

Would love to know how the formula works for Wave 2 if you've done that. Curious if it is the same, or if they came up with something slightly different for this wave since it isn't quite as symmetrical.

I was playing around with his formula for Wave 2 just now.

Assuming my math is right:

Alpha Squadron, Avenger Squadron, and Sabre Squadron are all undercosted by 1 point

Fel's Wrath, Turr Phennir and Soontir Fel are all bang on.

The Firesprays are a different story all together as previously mentioned. A Basic firespray should come in at 43.75 points assuming this formula:

Squad Points’ = 2 + ‘Pilot Skill’ + (‘Weapon’ – 2) x 8 + (‘Agility’ – 2) x 8 + (‘Hull’ – 3) x 4.25 + ‘Shields’ x 4.5

Plugging in the Firespray numbers it looks like this:

Squad Points = 2 + 3 + (3 - 2) x 8 + (2 - 2) x 8 + (6 - 3) x 4.25 + 4 x 4.5

This should work out to 43.75 where as the cost of a bounty hunter is only 33. Moving up the ladder, the costing discount stays within 1 point of that standard. Meaning it seems like the formula is correct with a -11 modifier on it and then rounding up. Alternatively, it's a completely different one

The Rebels work out a bit differently.

All the A-Wings are overcosted by 1.25 according to the formula. Not sure why and again it assumes my math is correct. The best thing I can figure is that it's for the sake of balance.

The YT-1300 is again a bit of a freak. The Outer Rim Smuggler with it's vastly different statline is actually overcosted at 27 points, which was a surprise for me since the Firespray held true to the formula assuming we give it the -10/-11 modifier. The ORS comes in at a statline value of 25.75, or overcosted by the same 1.25 that the A-Wings are purportedly overcosted.

All the named YT Pilots, Chewie, Lando, Han, all come in undercosted at -7, with their value coming through the formulas at 49,51 and 53 respectively, versus their stated costs of 42,44 and 46.

I would say that things hold true by and large with the fighter sized models, however the larger ones seem to play differently, as demonstrated here.

From the rebels side, it seems as though a ORS YT-1300 is a bit of a bad deal compared to the other options points wise.

Who wants to do wave 3? C'mon... ya'll can't be tired of statistics that much.

Seriously. I would like to see if the #'s are holding true. I'm developing some campaign stuff so I need to know...

If you need to know, you can do the work.

Gladly, if that's what it comes to. This was such a hot topic though so I thought I'd save myself some work if someone else has already done the comparison. We're also on a community forum here... you know... that place you go for help, ideas, and discussions...

I plugged the numbers in for Wave 3.

Formula/Actual

B-Wing

Blue Squadron Pilot - 26.5/22

Dagger Squadron Pilot - 28.5/24

Ibtisam - 30.5/28

Ten Numb - 32.5/31

HWK-290

Rebel Operative - 5/16

Roark Garnet - 7/19

Kyle Katarn - 9/21

Jan Ors - 11/25

TIE Bomber

Scimitar Squadron Pilot - 16.75/16

Gamma Squadron Pilot - 18.75/18

Captain Jonus - 20.75/22

Major Rhymer - 21.75/26

Lambda Shuttle

Omicron Group Pilot - 35/21

Captain Yorr - 37/24

Colonel Jendon - 39/26

Captain Kagi - 41/27

B-Wings, slightly undercosted.

HWK-290, vastly overcosted.

TIE Bomber, dead on.

Lambda Class Shuttle, way undercosted.

This is all based on the formula provided, obviously.

... and thats the problem with performing a regression analysis on a subset of the total data points. My guess is they developed probably a year or two of releases using their formula, so using one wave to predict the next wont do. Once we have more data and rerun the regression the actual formula will be more accurate.

... and thats the problem with performing a regression analysis on a subset of the total data points. My guess is they developed probably a year or two of releases using their formula, so using one wave to predict the next wont do. Once we have more data and rerun the regression the actual formula will be more accurate.

I agree. I also think it is just as likely that they made arbitrary decisions. Can we really have a 5 point ship? Add 10 points and playtest it. Its possible big ships get a point break for being such huge targets. Do they factor in firing arcs? We don't know.

Thanks! It's odd that the Tie's are so spot on. They have them down at least. Lambda Shuttle clearly got a discount (have to assume it's dial was seen as that bad) but not too far off the -10 speculation for large ships, except Kagi. BWings are really close too (because they are the most Tie esque of the rebels?).

One of the things that this formula does not do is account for the movement dial. The Shuttle may be way under costed, but the terrible movement dial makes up for it. Same with the special abilities of the HWK making up for their higher than expected price.

If there was some way to quantify the dial (and their may be, but with the large number of possible maneuvers and relatively small sample size, it may be difficult), we could come up with a more accurate number.

In the end, my guess is that these points are a good starting point, but play testing fine tunes the numbers.

... and thats the problem with performing a regression analysis on a subset of the total data points. My guess is they developed probably a year or two of releases using their formula, so using one wave to predict the next wont do. Once we have more data and rerun the regression the actual formula will be more accurate.

I agree. I also think it is just as likely that they made arbitrary decisions. Can we really have a 5 point ship? Add 10 points and playtest it. Its possible big ships get a point break for being such huge targets. Do they factor in firing arcs? We don't know.

I do anticipate a 10-12pt ship at some point but it will likely have very little def or offense and just do something else/support. Tie Swarm proved that even the lowest of the low can overwhelm in enough numbers if the agility or att is just enough that statistics are on it's side.

Actually - I think we're starting to discover that the maneuverability (dials) of the ships have costs, as well as the pilot abilities likely have varying costs, base size has an effect, and probably different costings for different kinds of firing arc.

I'm willing to accept 1 point per pilot skill, 3 point hull and 4 point shield at this point. I also suspect there is indeed a base amount of hull, attack and defense given to all ships to start with...

A good baseline might be the A-Wing and Tie Advanced as the only difference in maneuver dials is the green 5 forward (never straight!!!) of the A-wing. Maybe giving credit for an identical move wheel you can then compare the action line and try to extrapolate any changes to the formula.

Anyone think to ask a developer if they have a formula for ship balance and, if so, what is it?

I have some game design background, and here are my thoughts...

I suspect there are some things that have more ambiguous values assigned as well in wave 2. If they are following a strict formula, it would be unwise to assign a cost value to an upgrade option by itself, the smart way is to assign a value to a collected set of upgrade slots. IE:

1 elite talent =1 points

1 missile =1 points

1 elite talent AND 1 missile = 3 points.

Taking this a step further these could actually be taken as baselines equal to zero, with negative values for their absence.IE:

1 elite talent =0 points

1 missile/torpedo =0 points

1 elite talent AND 1 missile = 2 points.

No upgrades at all= -2 points

This is the sort of stuff that frequently gets missed in regression driven analysis since it frequently represents a change to a given formula after supposed constants have been defined, and are assumed to be correct.

Dials could work on a similar system, and since these are all based on values objectively perceived by the designer, we would never know.

It could also be that there is a second, conditional table involved, and that designs meeting certain criteria have points added or subtracted from their costs. This could explain the tie fighter academy pilot cost, or the 10 point gap on the firespray.

ALSO as if that weren't enough, they could be balancing ship stats completely independently from pilot cards with a set guideline for divorcing the two design functions.

After that I can almost guarantee that there is a review process wherein values are fudged according to synergies and meta-strategy.

This is fun and all, but I wouldn't expect a complete working solution unless they just outright tell us the process. This game wasn't engineered for "building your own ships". It gives the game designers more freedom to offer a wider variety of interesting combinations while keeping the learning curve relatively shallow using rule segmentation via cards, and a tight leash on the balance side of things.

No ship designed outside the process is going to "Fit" quite right, and Its a safe bet that it doesn't run on a simple point buy. The ultimate wrinkle is that as they continue to release more waves, the process will become MORE complex, with additional steps and an ever growing array of tables which will rapidly outstrip any attempt to categorize or reverse engineer their properties.

Of course, I could always be wrong. Maybe they just eyeball it and take their best shot.

Actually - I think we're starting to discover that the maneuverability (dials) of the ships have costs, as well as the pilot abilities likely have varying costs, base size has an effect, and probably different costings for different kinds of firing arc.

I'm willing to accept 1 point per pilot skill, 3 point hull and 4 point shield at this point. I also suspect there is indeed a base amount of hull, attack and defense given to all ships to start with...

Right-Titan's break down at the top of pg 4 is pretty spot on. For now I'll be using that and working from there. It would be interesting to see if the dials or some # of options makes a big difference... even if it is +/- 1 here and there.

I would like to see the ships separated from the pilots so I can work on campaign rules.