Pure generics were not the best combos so replacing them is not "power creep" as I (and you) define it. It is true that the game has crept upward, but the top end is entirely dominated by named Aces with the notable exception of the Stresshog. TLT did not exceed the "best ships/combos", it doesn't even exceed pure generics in a faceoff between the two (TLT vs. Pure Generic). It only plays better against the top end, allowing otherwise inadequate generics an opportunity to fare adequately against the best.

In other words, TLT allows the mediocre to have better game against the best, but is not itself pushing the top end upwards. By your own definition TLTs would need to be replacing Soontir, Whisper, Fat Han, Poe, Corran Horn and their like. They are not. The most they can do is reduce the dominance of those ships.

There are different categories of best kinds of ships. To name a few:

• cost effective filler
• arc dodgers
• turrets
• control
• alpha strike

You're right that some of the aces are both arc dodgers and cost effective, and that is a problem that has been pushing generics out. However it is also true that TLT is universally better than the previous best cost effective filler in the game, and essentially completely invalidated basic Z-95s, TIEs, and B-wings from the meta overnight. That's also power creep.

To reiterate:

Equivalently, power creep may also be observed anytime a new ship/upgrade renders an old ship/upgrade completely obsolete.

The 24 point TLT Y-wing has rendered the previous best vanilla ships obsolete. This was mathematically predicted and has been empirically observed starting at Worlds. Therefore it is power creep. This is not to say that it is necessarily the best ship in the game, nor that you would want to take 100 points of pure TLT Y-wings, but it is still power creep, and the impact on the meta game is significant.

And again, there were three PURE TLT spam lists in Worlds Top 32. There was only 1 pure generic non-TLT/Crackshot spam list in Top 32. (BBBBZ).

Look at Paul's list. He has:

• 36 points of cost effective filler
• 26 points of control
• 38 points of PS10 Ace killer

24 points of that 36 point cost effective was previously Z-95s, and was replaced by TLT Y. The only reason that the 12 points of bandit is still in there, is the quantization problem where you you can't get 36 points of cost effective TLT. That will change in wave 8 with Ezra + TLT/PtL/Chewie, which Paul is 5-0 in the vassal league.

TLT is definitely top dog so far in SC season, both before and after the cut.

Here's the Feb Top 10, all upgrades, before and after the cut (Feb set a record in terms of total # of lists submitted to LJ, btw):

Here's the list again, but tweaked so the 2nd column is the conditional effectiveness of each upgrade advancing to the cut.

TLT 15.5% | 1.15

HLC 6.0% | 0.83

Engine 5.9% | 0.97

Autothrusters 5.7% | 1.0

Palp 5.2% | 1.56

PTL 4.9% | 0.96

R2-D2 3.9% | 0.84

ATC 3.5% | 1.43

VI 2.6% | 0.92

Related, consider this a feature request for a new set of plots.

X-axis: ship/pilot/upgrade overall usage. (column 1)

Y-axis: ship/pilot/upgrade cut rate (column 2)

TLT is by far the most used upgrade (by points), and yet still has a positive cut rate coefficient. That tells me that the meta still has not fully reached steady state yet. Likewise, Palpatine crew and ATC (on just Vader, presumably) are even more effective, leading to the conclusion that Palp Aces is easily one of the strongest archetype builds right now.

Even a hardcore playtesting effort is going to have a hard time precisely pinning down the true cost of something; conversely, the pure mathematical model is probably also going to miss some of the subtleties of actual play.

A Card has a fixed printed cost that exists in a vacuum. But the worth of that printed cost can vary depending upon the card's use inside a 100 point squadron.

Yes, and each use case has to be analyzed separately. It's their system not mine, I can only work within their constraints.

I don't see why FFG would need to pay $50K to have Mathguy on a retainer, so he can tell them that the printed cost of a TLT should be 6.4 points +- 0.5 points ?

The product would certainly be better if it was balanced more accurately (more meta diversity = more fun = more sales), but quantifying that and deciding if it is worth the tradeoff of spending more money on development is fundamentally a business decision not a technical problem. Related to this:

1. FFG is well known for hiring new employees / designers with relatively little professional experience, and then paying them below industry norms for their position.
2. The full-time equivalent pay for a Technical Balance Director, aka "Mathguy", that could reliably provide accurate results would be in excess of $100k / year. (see discussion upthread).

Add these together and it should be clear that this will never happen at FFG.

defining "power creep" as rendering old options obsolete, then every miniatures game in existence has powercreep in some form or another. They gotta get you to buy the new stuff if they want to be successful, after all.

That's a fair assessment. If so, then every minis game has had power creep, and X-wing is fundamentally no different in that regard. Lets call it like it is.

There are a few solutions to the problem of generics being useless now. I have one in particular that I know would work quite well, but given that I'm not getting paid to do FFG's design work for them I'm just being a giant troll and keeping it to myself.

What you are defining as "power creep", I define as "balancing". There was so much complaining about fixing and balancing the weaker ships, and now that it has happened for a few of them, we are now complaining that it's "power creep"? Be careful what you wish for, or is there just no satisfying some people? TLTs have brought Ys and HWKs to a higher level and actually useful. They aren't breaking out above other strong lists in the meta nor do the numbers posted really statistically suggest that by taking TLTs you are much more likely to get into the top cut. The numbers Sozin has posted show that they are popular, a lot of them are being taken, and almost the same percentage is expected in the top cut, maybe slightly higher, but that could very well be in a margin of error.

They haven't made other generics "obsolete" either. I haven't seen any measurable data that shows that. There are plenty of instances where generics have made it into the top cut or have won store championships. "Obsolete" would mean we'd see none at all. Actually a turreted small based TLT is a much different ship than what I'm guessing what you are labeling as generics. TLTs are strong against arc dodging aces, which is probably why they are "strong" right now. Even if you are going against auto thrusters, at least you get a shot. However, they aren't good at blocking or up close like many other generics.

They haven't made other generics "obsolete" either. I haven't seen any measurable data that shows that.

https://community.fantasyflightgames.com/topic/192577-2015-worlds-results/#entry1874003

Store Championships are generally a poor data point because the skill level between the top players tends to be large, so you don't really get solid data on the relative value of squads / ships until regionals.

No, it isn't a good picture of the top tier. But, it does give a better view of the overall health of the game.

They haven't made other generics "obsolete" either. I haven't seen any measurable data that shows that.

https://community.fantasyflightgames.com/topic/192577-2015-worlds-results/#entry1874003

Store Championships are generally a poor data point because the skill level between the top players tends to be large, so you don't really get solid data on the relative value of squads / ships until regionals.

I'd have to say that those numbers don't look so unhealthy when compared to other games - aside from the Scum situation, where it's split even between Y-Wings and Brobots. Hmm...

Y-Wings total make up 28% of the ships taken TOTAL in the top 32? I take that back. 33%~ of the ships in the top 32 were TLT carriers? That's a huge amount, considering the variety.

And generics WITHOUT Palpatine or TLT make up 13% of the total ships (out of 46% generic ships on the field), and therefore 75%ish of generics used were TLT-carriers or Palpatine? The thing is, though, that it'd be the shuttle carrying Palp, which was only 3% of ships used, therefore the total percentage of generic ships that were TLT carriers was in the 65%+ range?

That is a touch unhealthy. It's becoming healthier, because at least it's not a 50% split between two ships like 2014 was (TIE Fighters and YT-1300), but it's not healthy. If the amount of available pilot cards doubled between 2014 and 2015, and the amount of different pilots used shrank by a third between 2014 and 2015, then that's bad.

It's an interesting note that Imperials have ALWAYS been an underused faction, and no doubt it's because reliable shields > fickle green dice any day of the week.

And yeah, it seems like a lot of the 2015 unhealthiness did stem from the TLT. Huh.

They haven't made other generics "obsolete" either. I haven't seen any measurable data that shows that.

https://community.fantasyflightgames.com/topic/192577-2015-worlds-results/#entry1874003

Store Championships are generally a poor data point because the skill level between the top players tends to be large, so you don't really get solid data on the relative value of squads / ships until regionals.

I would much rather draw from a large number of events across the world with various skills of players. You get a much better picture of how effective a ship is, and can normalize for popularity and "meta countering" selections (less specific meta dependent data in general).

well our 2 last local SC had a final of 2 TLT-K-wings with bombs galore beat a dual named K-wing TLT galore list and the other one a dual named TLT K-wing list won going 3-0(only 7 ppl showed) TLT onK-wing is broken because it's turretwing again, it has no donut, not arc to dodge, if you avoid the TLT you get 3-4 dice primary(Miranda)

Nerf please, i have 3 useless Tie Phantoms now , all these Rebel players should now feel the nerf bat.

Store Championships are generally a poor data point because the skill level between the top players tends to be large, so you don't really get solid data on the relative value of squads / ships until regionals.

How do you measure skill level at Store Champs Tournaments?

How do you define a large skill level difference at Store Champs Tournaments?

How do you measure skill level at Regionals Tournaments?

You need to provide some kind of measurable data for player skill level to support your statement of generality.

well our 2 last local SC had a final of 2 TLT-K-wings with bombs galore beat a dual named K-wing TLT galore list and the other one a dual named TLT K-wing list won going 3-0(only 7 ppl showed) TLT onK-wing is broken because it's turretwing again, it has no donut, not arc to dodge, if you avoid the TLT you get 3-4 dice primary(Miranda)

Nerf please, i have 3 useless Tie Phantoms now , all these Rebel players should now feel the nerf bat.

not really turretwing when those 2 dice primaries fall through + how expensive they are (31) to be wasting time throwing out a tie's worth of damage

miranda especially, her ability makes her seem like Han but then you realize: no gunner, no predator, no han's ability; has to both forgo regen and hurt herself. Plus, if she has no shields, she can't use that part of her ability

Also, Miranda can be outplayed, as I was shown at a recent SC. If your range control is good, you can deny her the ability to regen unless your exchanging fire, which makes her ability considerably worse.

Also, Miranda can be outplayed, as I was shown at a recent SC. If your range control is good, you can deny her the ability to regen unless your exchanging fire, which makes her ability considerably worse.

I don't think you would let me get away with that tactic again Ben.

Also, Miranda can be outplayed, as I was shown at a recent SC. If your range control is good, you can deny her the ability to regen unless your exchanging fire, which makes her ability considerably worse.

I don't think you would let me get away with that tactic again Ben.

I just wish I'd remembered to give the tactic a shot on Saturday.

Store Championships are generally a poor data point because the skill level between the top players tends to be large, so you don't really get solid data on the relative value of squads / ships until regionals.

How do you measure skill level at Store Champs Tournaments?

How do you define a large skill level difference at Store Champs Tournaments?

How do you measure skill level at Regionals Tournaments?

You need to provide some kind of measurable data for player skill level to support your statement of generality.

Why does he have to do that?`It's common sense that regionals draw more 'top' players than your average store championship.

Also, as far as i can see MJ has provided far more data in this thread than you or anyone else, so please stop making up bs requests like he has to prove something to you.

Why does he have to do that?`It's common sense that regionals draw more 'top' players than your average store championship.

Also, as far as i can see MJ has provided far more data in this thread than you or anyone else, so please stop making up bs requests like he has to prove something to you.

No, it is not common sense that Regionals draw more 'top' players than Store Champs.

FFG has an open policy for Store Champs and Regionals, where players of all skill levels are invited to participate and indeed they do.

ListJuggler provides a wealth of data about all kinds of squadrons.

It a useful database to crunch numbers about squadrons.

ListJuggler does not provide a single data point about player skill.

When I have submitted my squadron lists, I have never been asked to fill in my 'player skill'. So how can my player skill be measured and compared to other players in a quantitative manner?

Without a way to measure player skill, how do you propose to determine if there's a correlation between the 'top' players and the types of squadrons that they fly?

This is dragging the thread off topic from the central issue of TLTs.

There's an old saying about lies, damned lies and statistics.

It never hurts to ask someone for evidence to prove their claims and I am sure that Major Juggler can appreciate that.

I mean, it is kind of basic logic. Winning a 20 person tournament is a bit easier than a 80 person tournament. More rounds to test you and the squad.

Like I said, the Store Champs isn't where you get the idea of what the tier 1 stuff is, but you get a good sense of the tier 1.5 and tier 2 stuff, which is should hopefully be a larger selection than the tier 1.

The assumption appears to be that the World Championship should reflect the theoretical success rate of ships most clearly. This is not an unassailable fact, but a reasonable starting point to test the model. It is possible that a LBW variation wins a store championship, but it is unlikely that such a list would come very far at Worlds. That event just has the biggest potential of 'filtering out' statistical outliers.

So while I can see that the decision to corroborate mathwing at Worlds can be questioned, it is hardly a weakness. I think MajorJuggler presents some good positive evidence, much better than the typical 'I know this game, this is what I see, trust me' that seems to be the norm here.

MJ has provided a model for determining which ships can trade fire effectively. Period. That is one part of what makes ships work, and a very significant one, but not the entire thing and certainly not as effective an indicator as many/most want to assume it is.

This entire thread is insane. Look just a couple threads up at the Store Championship results thread. The last page shows a huge variety of lists and upgrades. Hawks, blaster turrets, M3As, aces, naked B-Wings, even A-Wings. Ships that people on this thread are dismissing as useless are not just seeing play, but having success. MJ has provided his own predictive data, and crafted a theory of effectiveness based on his model. But real world results are NOT following the predictions spawned by that model.

Note that there is a difference between a model (the math he presents) and predictions made based on that model. A model can be 100% accurate within its scope (the variables accounted for) and still have poor predictive accuracy. This is because models can only incorporate known variables and there are always unknowns in real-world applications. More, the exact weight of the different variables (ex, barrel roll and boost are both helpful for maneuvering, but which is more helpful, by how much and by what measure) are not easy/impossible to determine within a model complex enough to be useful.

That predicted "jousting efficiency" correlates directly with popularity is as much (IMO more) due to people picking ships based on the model than it is a reflection of the accuracy of that model. Ships such as the TIE Defender or the K-Wing are perfect examples of the shortcomings in the "math". Both of them are completely inferior to alternative ships in terms of "efficiency" and yet the other less quantifiable advantages they have (white K and SLAM foremost) mean that on the table they can perform every bit as well or better than their jousting "superiors".

Major Juggler is presenting "I created a model, trust me, it says these ships should win". Others present "I know this game, this is what I see, trust me". But then there are folks who are saying "Look at the RESULTS, not the predictions. The evidence is that the game is incredibly diverse, there are a wide variety of viable ships and builds, TLT spam is NOT even remotely dominating, pure generic ships are seeing success both in swarms and as filler".

You don't have to trust me, just look at the evidence! The real world, it actually happened, recorded not predicted, evidence. The game is in the best place it has ever been.

And there is the problem - if you think there is better evidence, then by all means present it. I don't know what state the game is in, nor do I take anyone's word here as gospel truth. I just see that one view is argued for in a better manner.

While it is true that I will just have to trust MJ that he built a rational model based on proven knowledge, I do think that his position is much better than one of simple personal trust. His source may be a black box to me, but it does look like a structured method. In philosophy, there is a concept called 'heuristics', which represents the methods, examples, theoretical framework and even habits of a 'research programme', a way of gaining more knowledge. Even the semblance of heuristics is preferable to none at all. And that's what I see here. A choice between either a way of handling these points costs and going forward in drawing conclusions about them, or a rather passive way of just letting things happen and deciding if it's about right by drawing on personal experience.

While the TLT is strong, it's nowhere near the pre-adjustment Phantoms (and whatever shenanigans the Scum players will come up with soon).

I'm not saying the phantom was balanced but the phantom also didn't win any worlds either actually I dong remember it in the top 2 at worlds. TLT was in nearly every rebel and scum list in the top 8.

It was in every scum list because most scum ship suck. The only good ones are 95s, IG, and Y wing...oh and Firesprray. When you miss on the other ships sure you are going to see a similar meta as the god player want to attempt to win with their shiny new toys. It is just too bad the M the K the SV and HT suck so bad. Fun and fine for causual play but can't hang in regionals. 4 bad ships and 4 good ships...too bad 3/4 good ships where already old ships repurposed. That was the worst mistake FF made besides the phantoms was making the scum only ships come up short. The got tons of people to bit on spending money on a new faction and they ended up with just a few competitive builds. You see some many TLT y-wing because the only other answers is the more boring bro bots.

That's just it, I did and you went right past it again.

The evidence is the RESULTS. Not the model, not the predictions, not the arguments. The actual real-world results, that you can easily see on the Store Championships thread. Take a look, especially the most recent results that reflect how the meta has continued to adapt, and you will see enormous variety.

And just in case...

https://community.fantasyflightgames.com/topic/198561-2016-store-championship-results/page-19

Honestly, I don't think they will ever get it right until they print an errata pack (which they won't because they need people to buy ships for cards like auto thrusters...and that would give people a way out) and adjust both the ship cards and upgrade cards ever so slightly to reflect the new reality of now years of play testing. At least they could print all the reprints of existing ships with the new changes. That would be I fine would be a fine compromise. I would actual spend money swapping out my old stuff with newer players who would not mind have the old stuff for money and just memorize all the errata. That is the great thing about 2nd edition board games, they have the opportunity to fix their mistakes. This game needs some second edition ships. I would pay $200 for all the fixes across the board. Getting it right is important to me IMHO ...whatever the costs. Essentially, I would pay $200 for the pseudo "upgrade pack" busting the meta wide open and making so many lists viable.

Sorry if you felt if I ignored your remark about the results, but I did adress that indirectly - I don't think you have some heuristics with which you interpret them. Results in themselves don't show much, they're just that. I'd like to see what case is built around them.

The assumption appears to be that the World Championship should reflect the theoretical success rate of ships most clearly. This is not an unassailable fact, but a reasonable starting point to test the model. It is possible that a LBW variation wins a store championship, but it is unlikely that such a list would come very far at Worlds. That event just has the biggest potential of 'filtering out' statistical outliers.

So while I can see that the decision to corroborate mathwing at Worlds can be questioned, it is hardly a weakness. I think MajorJuggler presents some good positive evidence, much better than the typical 'I know this game, this is what I see, trust me' that seems to be the norm here.

If I'm looking at the overall effectiveness of ships, Worlds is an extremely skewed set of data. The thought process of list building a good list does partially include ship effectiveness, but more about "does this list counter the other lists that are popular at this moment of time"? Paul Heaver even discussed the reasoning for his list. Basically, what's a good counter to both TLTs and Aces like Soontir or Vader? - Poe w/ VI, Auto Thrusters and regen. Stresshog for hard countering Aces, Brobots, etc. He read the meta very well, and created a list that countered it.

Since then Poe lists are popular, and so the meta question is, "What do I bring to counter Poe, and all the other popular lists"? Does this kind of metabating really tell you much about the overall effectiveness of specific ships? Not really, maybe some, but it tells you more about what ships counter what ships.

Sorry if you felt if I ignored your remark about the results, but I did adress that indirectly - I don't think you have some heuristics with which you interpret them. Results in themselves don't show much, they're just that. I'd like to see what case is built around them.

What kind of framework do you need? There is high diversity in the SC winning lists, which is consistent with the hypothesis that we don't have one element dominating the meta.

I see this is long thread but I wanted to add my two cents. I tend to stay away from the meta builds and do my own thing but during my tournament this weekend I did notice some things. First of all I didn't do as well as I normally do 2-3-0. But I did meet a lot of new nice people and they all had the same thing to say. "Oh I just started playing like a month ago" and all 3 of my loses where to similar builds... TLT builds, as a matter of fact only one of my matches didn't include them. I also noted that ALL of them would bring the smallest asteroids to the table and few of them even commented that they hate asteroids. Turrets where annoying before but not great vs jousters that flew well. I had some of the hardest matches ever this weekend and it was a lot of fun, but I do feel I outplayed each and every one of my opponents but I was handicapped from the start wile they just drove around in circles not having to get me in their arcs as they slow rolled me to death with high hp ships and TLT

When they ride circles don't engage them, let them build distance and await them when they return. Vorpal Sword made a really good thread on how to deal with TLTs, you might want to check that one out (I don't remember what it was called, though).