Also MJ never said that a ship can't win or even do well, just that it's not as efficient as another ship.

Very first post in the thread, VDM:

The TIE Interceptor's stat line cost efficiency is actually a hair higher than the X-wing.

And the TIE Interceptor has boost and barrel roll, so if the PS1 Interceptor isn't viable, then the PS2 X-wing certainly is not either.

The PS1 interceptor is "not viable". PS2 X-Wing is "not viable".

We're obviously missing the full post and relevant context but what you just quoted and then "quoted" using quotations are 2 different things.

Just saying

I can appreciate the sheer volume of effort that's gone into it, but I find it more than a little irritating when it is used (not necessarily MJ, but certainly quite a lot of people on this forum) as a be-all, end-all argument to effectively write off large proportions of the game as competitively worthless, when the experience gained form using them, trying them and finding out yourself is certainly anything but.

And I'd agree. I've never been a blind follower of mathwing, I've said a number of times that mathwing is a good place to start, but is not the end all-be all of X-Wing. There's number of things it can't take into account.

Defenders

And that's my beef with it. No matter what the Math says about the Defender, it is a lot stronger than that efficiency. MJ even admits having only a medium degree of certainty about it.

I think what you really need to do is take a step back and reconsider why you are choosing to "have a beef" with Mathwing.

Mathwing doesn't have feelings. Mathwing doesn't have an agenda. Mathwing is data. Mathwing has never been billed as the end all be all of X-Wing.

People sometimes make points or arguments based upon that data. Sometimes they are good arguments and sometimes they are flawed.

Mathwing never said anything mean. Mathwing didn't hurt other people's feelings. Mathwing is data.

It's not the Mathwing itself, it is the effect it has on people. Because of mathwing, the Defender, by all rights a good ship if slightly overcosted, sees no use because everyone follows math as the end all be all. It isn't.

There's more to a ship then the mathwing value, and while that tells us a lot about the ship, it isn't everything.

EDIT: Read Vanor's post just above mine, that is exactly how I feel.

Defenders

And that's my beef with it. No matter what the Math says about the Defender, it is a lot stronger than that efficiency. MJ even admits having only a medium degree of certainty about it.

I think what you really need to do is take a step back and reconsider why you are choosing to "have a beef" with Mathwing.

Mathwing doesn't have feelings. Mathwing doesn't have an agenda. Mathwing is data. Mathwing has never been billed as the end all be all of X-Wing.

People sometimes make points or arguments based upon that data. Sometimes they are good arguments and sometimes they are flawed.

Mathwing never said anything mean. Mathwing didn't hurt other people's feelings. Mathwing is data.

It's not the Mathwing itself, it is the effect it has on people. Because of mathwing, the Defender, by all rights a good ship if slightly overcosted, sees no use because everyone follows math as the end all be all. It isn't.

There's more to a ship then the mathwing value, and while that tells us a lot about the ship, it isn't everything.

EDIT: Read Vanor's post just above mine, that is exactly how I feel.

eh...that's a bit too simplistic an outlook

the Defender's awesome, but part of that awesomeness is how weird and unique it is. Weird and unique does not necessarily equate to "easy to fly" regardless of mathwing. The defender's pretty skill intensive especially with it's horrible range of red and green manuevers and the need to make the most of the white 4k. It requires a very particular playstyle to make effective, and not every one is going to enjoy it or the effort required, mathwing or no

People are sheep and it's not specific to any field of human endeavour the majority prefer to follow rather than lead.

So when someone assumed to be an authority says something it's taken as gospel, it's an aspect of our psychology that's led us into some really dark times.

Using maths is fine as long as you keep your scepticism turned on.

Wow stop blaming the math and the people putting it out and maybe start blaming to dumb sheep.

It's not the Mathwing itself, it is the effect it has on people. Because of mathwing, the Defender, by all rights a good ship if slightly overcosted, sees no use because everyone follows math as the end all be all. It isn't.There's more to a ship then the mathwing value, and while that tells us a lot about the ship, it isn't everything.EDIT: Read Vanor's post just above mine, that is exactly how I feel.

The numbers really are everything, you just aren't thinking about them in the right way. A YT-1300 with Gunner really does have a value of X (whatever it would be assigned). However, that X valuation is meaningful when that ship is up against twin Decimators. Everything can be assigned a value that, depending on why it was assigned and how it is used, can be valuable to players.

Also, mathwing had very little to do with the sinking of the Defender. I am sure that plenty of players, myself included, playing a games with the Defenders and constantly twisting the dial to look maneuvers that it does not contain is what keeps people from flying it. I did not see or know of mathwing prior to concluding that the Defender is a sub-par ship.

People are sheep and it's not specific to any field of human endeavour the majority prefer to follow rather than lead.

So when someone assumed to be an authority says something it's taken as gospel, it's an aspect of our psychology that's led us into some really dark times.

Using maths is fine as long as you keep your scepticism turned on.

Nailed it.

All this talk of 'netlisting' is exaggerated. Squadron lists in X-Wing are really not that complicated. People like to stroke themselves and pretend that they have unique ideas that no other player could conceive until the read about it on the internet. The reality of this game is that Han Solo with C-3PO, Predator, and Engine Upgrade is a very simple conclusion to arrive at.

I don't doubt that some people read about a list and then try it, but the idea that those same people could not eventually arrive at that list while independently optimizing their squadron is hard to believe.

What's also interesting, funny, and a little irritating sometimes is for certain swaths of the online community to feel the need to "name" those squads and then somehow claim ownership of that list "idea." Take Han and 3 talas to any tournament and you'll immediately be labeled as bringing "Paul's World's list." Oh that's "Panic Attack", ohhh that's "Terminator's" etc. Some, specifically in the podcasting community, are so quick to claim that that's this person's list or that person's list, even flat out saying that this online personality is "credited" with creating the dogfighting shuttle, or the doom shuttle, etc. etc. The poor common folk have no chance of getting any credit to their own list choices once someone out in wild talks about a list online.

credit really isn't terribly important,

we attribute these labels to certain lists in order to quickly distill their strengths and weaknesses so that we have some idea of what to do when we encounter them

panic attack is 4 rebel ships with stress generators (specifically r3-a2 ion y and 3 tactician bs), Heaver's han + 3 talas is specifically r2-d2 +c3po han made with the explicit purpose of out haning other hans (often confused with han + 3 talas, which is just han + insurance against generic spam), and BBBBZ is BBBBZ

sometimes they will be named after the person who popularized it, since that name belongs to the person who introduced it onto a wider stage even if he/she didn't invent it. You don't get any tangible benefit to patenting a list, and it's just easier for people to remember titles of more famous players.

It's not the Mathwing itself, it is the effect it has on people. Because of mathwing, the Defender, by all rights a good ship if slightly overcosted, sees no use because everyone follows math as the end all be all. It isn't.There's more to a ship then the mathwing value, and while that tells us a lot about the ship, it isn't everything.EDIT: Read Vanor's post just above mine, that is exactly how I feel.

Also, mathwing had very little to do with the sinking of the Defender. I am sure that plenty of players, myself included, playing a games with the Defenders and constantly twisting the dial to look maneuvers that it does not contain is what keeps people from flying it. I did not see or know of mathwing prior to concluding that the Defender is a sub-par ship.

The Defender maneuvers just fine.

credit really isn't terribly important,

we attribute these labels to certain lists in order to quickly distill their strengths and weaknesses so that we have some idea of what to do when we encounter them

panic attack is 4 rebel ships with stress generators (specifically r3-a2 ion y and 3 tactician bs), Heaver's han + 3 talas is specifically r2-d2 +c3po han made with the explicit purpose of out haning other hans (often confused with han + 3 talas, which is just han + insurance against generic spam), and BBBBZ is BBBBZ

sometimes they will be named after the person who popularized it, since that name belongs to the person who introduced it onto a wider stage even if he/she didn't invent it. You don't get any tangible benefit to patenting a list, and it's just easier for people to remember titles of more famous players.

My issue isn't with the naming, it's the arrogance, at least in some, relating to claiming ownership or crediting a person with the list idea as if no one else could possibly come up with that list.

well, it's one thing to come up with a list

it's another to bring it to a high profile event with the expectation of winning, and then actually winning with it

the names given (apart from generic titles like panic attack or bbbbz) probably has less to do with "inventing" it and more to do with "flying it incredibly well"

so really, less the invention part and more the proving part

I haven't bothered to post in this thread yet since I have little interest in "defending" the work, or correcting people's misconceptions on the more technical aspects if they have clearly already made up their mind. I intentionally made the method totally public so it can be reviewed, so if you disagree with the conclusion then you can demonstrate where the math went wrong.

But, the conclusions from the original quote on the first page sounds to me rather like grating nails on a chalkboard, so since the OP brought it up again I feel somewhat compelled to point out what should be obvious...

Also MJ never said that a ship can't win or even do well, just that it's not as efficient as another ship.

Very first post in the thread, VDM:

The TIE Interceptor's stat line cost efficiency is actually a hair higher than the X-wing.

And the TIE Interceptor has boost and barrel roll, so if the PS1 Interceptor isn't viable, then the PS2 X-wing certainly is not either.

The PS1 interceptor is "not viable". PS2 X-Wing is "not viable".

I did not say the PS1 Interceptor is not viable, and I did not say that the PS2 X-wing is not viable. Take note of the (emphasized) logical conditional wording.

What the math lets you do, is take a known baseline and compare another ship to it. So for example, if "Ship A" is known to be largely ineffective in the meta, and "Ship B" is essentially the same or inferior to "Ship A", then by extension "Ship B" must also be largely ineffective as well. Replace "Ship A" with "PS1 TIE Interceptor" and "Ship B" with "PS2 X-wing".

Taken in context, the conclusion from my original statement is that if even the PS2 X-wing had boost and barrel roll for free, it still wouldn't be particularly attractive in the meta.

And of course, the classic "The Delta Squadron Pilot is less efficient than two TIE Academy Pilots and costs more".

The numerical analysis fundamentally dictates how much arc-dodging a ship needs to do to earn back its cost. The numbers have a range based on attack range probabilities, meta dependent matchups, and action economy, but the spread is still narrow enough to draw significant conclusions. In the case of the generic TIE Defenders, that number (adjusted for pilot skill) works out to be around the same as the better named turreted ships (particularly the Rear Admiral). Therefore, for the generic TIE Defender to earn back its points, its dial has to be roughly equivalent in combat effectiveness as having a turret. A dial can fundamentally never be better than having a turret (unless we live in a fictitious world where 100% of ships have autothrusters and the turret ship has the dial of a stalled out dump truck), therefore the TIE Defender costs too much for its capabilities.

When I first brought up the TIE Defender's cost problems before the dial was even revealed, it was met with mixed reactions. Several people who playtested wave 4 contacted me (publicly and/or privately) and told me that I was wrong, the ship is balanced fine, wait for the dial reveal. (Disclaimer: I'm not a playtester. I did look into it but had to decline due to an overly broad non-compete/IP ownership clause, despite Alex Davy's best effort at convincing FFG legal to reach an agreeable arrangement.)

Fast forward today, and we have nearly irrefutable evidence that the generic Defenders are not tournament viable. The math really is that powerful - it provided more forward looking insight than the playtesters and the game designers had. That is obviously a very strong statement and could be construed as confrontational, but ultimately it's a case of "I'm just calculating the numbers, don't shoot the messenger".

There are of course many improvements that can be made on the model, but I don't want to tip my hand on what I have up my sleeve next.

Also, when I publish an analysis of Missiles and Torpedoes then you can blame me for those not getting used too.

it's like you can see into my dreams...

Fast forward today, and we have nearly irrefutable evidence that the generic Defenders are not tournament viable. The math really is that powerful - it provided more forward looking insight than the playtesters and the game designers had. That is obviously a very strong statement and could be construed as confrontational, but ultimately it's a case of "I'm just calculating the numbers, don't shoot the messenger".

This is the part where you lose people. It comes off as extremely egotistical. Your math is not the only factor into things. Different people have different playstyles. And your math cannot predict such things. That's why playtesting is far more valuable than your math.

And I've seen your houserules. I'm not entirely confident in your math as the sole basis of game balance.

it's not his math, it's the math

his argument is that the raw value of the defender's stats per points was low, and that it would not be overcome by any means through the strength of its dial. In this particular case, the math surround the game's mechanics would have predicted the difficulty in effectively using such a ship even when playtesting and developing found no reason not to let it exist in its current state.

and make no mistake, the defender is a very difficult ship to use for this reason. Given the current horridness of m.o.v skews and the continuing prevalence of idiotic PWTs, I would personally not bring defenders to a competitive event that I planned on doing well in simply because they are too fragile to be able to ignore dice. If there's any chance of being able to actually outplay the opponent (as in the case of coming up against arc-locked aggressors), then by all means they could do well. Otherwise, it's an uphill battle against dice and no amount of skill is going to beat them (unless you're really "good" at rolling them; maybe picked up a trick or two ).

still, it's in a far better place than the x-wing despite lower value. At least the generic defender can actually transcend jousts with its 4k, roll, and cannon. The generic x-wing remains a jouster that can barely joust

it's not his math, it's the math

Perhaps you've missed how he doesn't want to reveal his updated formula? Math is not a singular entitity. You can use it in a variety of ways to judge the same thing. Math doesn't lie. But how the math is used can be used to lie.

And again, it isn't worthless. But it does not deserve to be the bible truth many hold it up to be.

People are sheep and it's not specific to any field of human endeavour the majority prefer to follow rather than lead.

So when someone assumed to be an authority says something it's taken as gospel, it's an aspect of our psychology that's led us into some really dark times.

Using maths is fine as long as you keep your scepticism turned on.

Nailed it.

The Cult of Pythagoras does not approve...

People are sheep and it's not specific to any field of human endeavour the majority prefer to follow rather than lead.

So when someone assumed to be an authority says something it's taken as gospel, it's an aspect of our psychology that's led us into some really dark times.

Using maths is fine as long as you keep your scepticism turned on.

Nailed it.

The Cult of Pythagoras does not approve...

Yeah well he can contact me via necroscope if he wants to argue about it, i'll be chillin with my homie August Mobius.

And your math cannot predict such things. That's why playtesting is far more valuable than your math.

Lets go down the list. We will assume that design and playtesting intends to make each ship viable, and compare the public model's predictions with empirical results.

1. Generic X-wings (and the vast majority of X-wing pilots overall) will fall out of favor, this was predicted well before X-wings became "extinct" in the competitive meta.
2. Generic Y-wings without a turret are not competitive. Generic Y-wings + Ion without BTL-A4 can't pull their weight.
3. TIE Fighters are good.
4. TIE Advanced are terrible.
5. A-wings are overcosted by about 2 points (later fixed with Refit). All named A-wings other than Jake aren't really cost effective.
6. The YT-1300 is the first turreted ship and essentially serves as a baseline, but it can easily be concluded that the generic is very overcosted relative to the named pilots. If you need the higher PS then Han is a bargain price compared to Chewbacca.
7. Generic TIE Interceptors are marginally better than generic X-wings, but still aren't quite cost effective enough to be competitive. Soontir Fel is easily the best pilot even after Imperial Aces.
8. All Imperial Firesprays have poor cost effectiveness.
9. Generic B-wings are essentially universally better than X-wings, and more versatile to boot. Named pilots are generally very lacking except for Keyan.
10. The Rebel HWK is extremely lackluster, with possible exception of Roark for when you must absolutely go nuclear with Pilot Skill.
11. Missiles and Torpedoes are overcosted, so the TIE Bomber is essentially dead on arrival, having worst straight up cost efficiency (as it should) compared to the standard TIE Fighter.
12. The Lambda Shuttle is very cost effective, although once you look at durability based on shots-to-kill not necessarily by as much as you might think. Valuing the dial requires some analytics and still isn't done as well as I would like. The model also incorrectly weights the named pilots as being "reasonably" costed because the PS curve I use doesn't recognize that a high PS Shuttle still has almost no options to take advantage of its higher PS. Technically Kagi is still good in Epic, but overall I count this as a model failure, although it is no worse than the design + playtesting process.
13. The generic E-wings are very overcosted, as is E'tahn by a lesser margin. Corran is tricky to model directly, but his ability is so disruptive (and his PS price progression is so cheap) that he is clearly the most cost efficient pilot of all the E-wings.
14. Generic Z-95s are on par with TIE Fighter for cost efficiency and will forever fundamentally change rebel list building.
15. Advanced Cloaking Device TIE Phantoms murder everything in cost efficiency when they out-PS bid their opponent, even before considering the declock movement.
16. Generic Defenders are dead on arrival despite having a very unique dial. Named pilots aren't costed as bad and could see occasional use, especially Vessery in the ideal TL-happy squad (not truly feasible until post-Raider).
17. Outrider with Mangler is a pretty reasonable ship for balance. 58 Dash is absolutely insane, but only if you disregard the range 1 donut hole. YT-2400 without Outrider can work with HLC, but should generally be a corner case addition to an existing list.
18. Rear Admiral Chiraneua has absurdly good cost efficiency for a turret, and the flexibility of 3 crew slots favors him (as the most expensive VT-49) the most. Kenkirk is essentially obsolete compared to the Admiral. Oicunn is hard to model, but should generally be favored much less than the Admiral. Generic PS3 is basically useless except as a 43 point Vader Doom Boat, but even that only works in a very specific meta vs Soontir Fel.
19. Scum Z-95's are basically the same as Rebel Z-95's. N'Dru is great when he works, but is such a glass cannon that a good opponent will nuke him early.
20. Generic Scum HWK is similar to Rebel Scum HWK: not very good. PS3 Mux is very versatile. Palob has arguably the best ability in the game, and murders every other turret's cost efficiency, but R1-2 blaster turret is the weakness. Dace's double damage on Ion doesn't make him cost effective.
21. Scum PS5 Firespray is priced better than PS3 Imperial Bounty Hunter but still has a slight uphill climb. Scum Kath and Boba are both solid, although I still haven't re-done the raw numbers properly on Boba.
22. Generic StarVipers are overcosted by about 2 points, essentially making them dead on arrival. Named StarVipers can be OK.
23. M3-A are not really competitive with any loadout.
24. IG-88 I still haven't done the due diligence on analyzing (ironic since that's what I have been playing), but the raw cost efficiency is competitive even before upgrades, adjusting for PS6 cost progression. Cannon upgrade is mandatory. Pilot Skill 6, large base and forward arc is a liability. Unique pilot abilities still not considered.

Those are the predictions so far. More will come as wave 7 ships are spoiled. Feel free to tally up the results to see which has been better at predicting game balance. If the model was used during the design phase, then the ships would clearly have been balanced much better than by playtesting alone. We would still be flying X-wings and taking Proton Torpedoes on Y-wings without Turrets, and they would be making a consistent proportion of Top 8 cuts. Again, don't shoot the messenger.

You obviously don't want to replace playtesting with math. That would be absurd -- although the math has a significantly better track record than playtesting for ships that don't have significantly unique capabilities.

Proper numerical analysis is a necessary but not sufficient component of game balance. Without it you get a system that has significant room for meaningful improvement and a competitive meta that is far more narrow than it could have been.

Perhaps you've missed how he doesn't want to reveal his updated formula?

Trade secret. If, in theory, I could help contribute to making a game that makes X-wing balance look like 40K by comparison, then there could be a significant financial value in such an approach. There are obvious reasons for not going public with it. I have posted neither methods or results from next-gen approach, nor do I have any obligation to do so.

The public model is a voluntary service to the community that I am under no compulsion to provide, but predictions from the current model are public and will be maintained going forward for new waves.

Again, I don't feel the need to "defend" the model (although specific feedback is always welcome), and going through the list of ships should be sufficient to demonstrate robustness. Already been dragged too far into the debate, likely the last post I'll bother with on the thread.

// resume regularly scheduled discussion

Too many people "disagree" with MJ's conclusions/numbers without bothering to dig into them and understand them.

I disagreed with his analysis of the Outer Rim Smuggler, because I find great value in an ability his math does not incorporate. Namely, the fact that it goes first and is really big, making it by far the most effective blocker in the game. I found that ability to be worth the loss of value (by MJ's metrics) in particular lists, but that didn't invalidate MJ's work. I just needed to know both the value calculated by MJ and HOW he calculated that value in order to accurately evaluate the ship. I find PS 1 Starvipers to be much more valuable than MJ's math suggests for the same reason. With PS 1, boost and barrel roll they are outstanding blockers. However, and this is a big caveat, if I were to try to plug either of these ships into a list that does not exploit their particular skills at blocking they would be every bit as sub-par as MJ suggests. So, I don't do that.

I also disagree with the conclusion that Defenders are not good value for their points, finding that the uniqueness of their 4k is of greater utility than what MJ values it at. This again does not invalidate his math. By making his assumptions open, MJ allows me to take what he has and apply it to my own lists and use it as part of the information I have available to make decisions.

"Mathwing" does not tell you what ships are best. It tells you what ships are best under the circumstances MJ has run the numbers within. If you plan to have a list that does not play in a way to support those circumstances, you can adjust your expected values. As he said, don't shoot the messenger. The reason his numbers don't always apply is because his circumstances (which he explains) don't always apply.

Fast forward today, and we have nearly irrefutable evidence that the generic Defenders are not tournament viable. The math really is that powerful - it provided more forward looking insight than the playtesters and the game designers had. That is obviously a very strong statement and could be construed as confrontational, but ultimately it's a case of "I'm just calculating the numbers, don't shoot the messenger".

This is the part where you lose people. It comes off as extremely egotistical. Your math is not the only factor into things. Different people have different playstyles. And your math cannot predict such things. That's why playtesting is far more valuable than your math.

And I've seen your houserules. I'm not entirely confident in your math as the sole basis of game balance.

Actually, this is probably about right. See, he is very clearly assuming that correlation implies causation. Since he is not seeing tournament results, he's assuming that it's proof that his equations and coefficients are correct. This is a common mistake actually.

He missed the point that the OP mentions, that it is a self fulfilling prophecy. When people look at his work, they use it to build lists, and therefore are not taking the Defender to tournaments. When more than 50% of competitors bring a variant of three lists, it's hard to derive any causation from the tournament results.

Also, the most important thing to remember is these games are designed to maximize fun. For some players, theme trumps all, it doesn't have to be competitive to be fun.

All models are accurate; some are useful. The devil's in the assumptions.

I have no problem with MathWing given its assumptions and trust the math for the particular model. It's not the sole possible model, and it won't model well those things outside its assumptions.

At best, it's an indicator, but one that shouldn't be used in isolation without understanding.