ETC 2019 UD Lists

This site uses cookies. By continuing to browse this site, you are agreeing to our Cookie Policy.

The latest issue of the 9th Scroll is here! You can read all about it in the news.

Our beta phase is finally over. Download The Ninth Age: Fantasy Battles, 2nd Edition now!

  • Getyrom wrote:

    I am fully aware that me insinuating that you think the Sha Guarding is OP and need nerfs and not buffs was very very far-fetched :P but honestly, I have no idea how to otherwise interpret what you wanted to say with your post without more information (which you added in your next post). I was trying to be funny while making an obvious bad point.
    However, I agree with you only a little. I think it is the other way - instead of players needing a lot of time to react to metas and know-hows in order to start using all the options available to them I think it is the leaders of T9A that are slow to react to obvious balance problems that exist throughout the game. They are miles better than GW (who did it once every 5-6 years or so), but still far from where I would want it to be.

    Exact balance is impossible to achieve in a complex game as this. When balancing, determining if unit X should increase +1ppm or +2ppm is almost impossible to decide, but using just a teeny bit of math knowing if unit X should benefit from +ppm or -ppm is not that difficult, ie telling you in what direction you should be headed. The balance is then perceived by continuous and constant change that shift the powerlevel between units, instead of the current situation where everybody have had the time to understand that list X is the best list and thus is seen 90% of the time. This is IMO bad balance.
    Having been one of the people responsible for balance changes... I can only tell you that I disagree with at least some of what you are saying. A full exploration of why does not belong here and would be a foolish exercise via text anyway.
    If you look carefully over the last 18months, I think you will see sufficient counter-examples to what you are saying to allow you to develop your own ideas further without any input from me.

    However, remember that stability is also a key thing: the project got a lot of criticism for how fast things were changed.
    Yearly points updates is a compromise that the project has settled on: the constraints, concerns and considerations leading to this are not all about balance.

    Indeed, isn't it funny how much abuse the project got for caring only/too much about balance, but now there are only yearly points updates, suddenly the project should be worrying more about balance...
    Being supportive & giving useful criticism aren't mutually exclusive.
    Are you supportive of the project? Do your posts reflect that?

    List repository and links HERE
    Basic beginners tactics HERE
  • DanT wrote:

    Indeed, isn't it funny how much abuse the project got for caring only/too much about balance, but now there are only yearly points updates, suddenly the project should be worrying more about balance...
    Haha yes, I have noticed this. When updates are frequent "you change too much", and when updates are scarce "you change too little". It's always a loosing battle :D

    I think concerning what the community wants the project to be, T9A project has taken good measures to follow the wishes of the community. But this is not the same as balance, as you say.

    We'll have to agree to disagree.
  • Yearly is still too often. xP

    Five years would be optimal. Two would seem to be the minimal interval. The few people who want constant change could just switch which army they're playing.
    Just because I'm on the Legal Team doesn't mean I know anything about rules or background in development, and if/when I do, I won't be posting about it. All opinions and speculation are my own - treat them as such.

    Legal

    Playtester

    Chariot Command HQ

  • Just_Flo wrote:

    And just last week I was told in HE Board, that collecting and using data on list usage was a waste of time ...

    On the topic of speed, I guess it is a little bit of both. Yes out precedurtes take more time than we would like them to take and yes, most players happen to not test seriously try out or test every part of their book during one circle of the book being the exactly the same.

    Yeah but. HE board....
  • @Fnarrr

    5 years isn't longer than it will take to do all the FABs. I certainly hope the lifespan of the books is planned for longer than the total development cycle of all books. (I estimate we'll realistically get up to ~2 books a year at maximum output, which means at least 8 years for all books to get FAB treatment).

    The fact is, the game is externally balanced to a level where army book differences << player skill differences. At which point, there's no need to rejigger points yearly - and it's counterproductive, because more often than not the lax level of statistical significance used to justify changes will turn up false positives way more often than actual balance problems, which means most of the time we'll be making the game less balanced rather than more balanced.

    We don't have the data required to address 1% or 5% differences in external balance, and to get that data we'll need long time horizons - assuming those are even large enough concerns to bother with.
    Just because I'm on the Legal Team doesn't mean I know anything about rules or background in development, and if/when I do, I won't be posting about it. All opinions and speculation are my own - treat them as such.

    Legal

    Playtester

    Chariot Command HQ

  • Squirrelloid wrote:

    @Fnarrr

    5 years isn't longer than it will take to do all the FABs. I certainly hope the lifespan of the books is planned for longer than the total development cycle of all books. (I estimate we'll realistically get up to ~2 books a year at maximum output, which means at least 8 years for all books to get FAB treatment).

    The fact is, the game is externally balanced to a level where army book differences << player skill differences. At which point, there's no need to rejigger points yearly - and it's counterproductive, because more often than not the lax level of statistical significance used to justify changes will turn up false positives way more often than actual balance problems, which means most of the time we'll be making the game less balanced rather than more balanced.

    We don't have the data required to address 1% or 5% differences in external balance, and to get that data we'll need long time horizons - assuming those are even large enough concerns to bother with.

    Three per year is plausible (which is a 5 year cycle, more or less, on all books). Not in terms of doing one book in four months, but in terms of having three books on staggered release every four months. Currently the BGT is one of the bigger logjams in the process and we're working on that. Art and layout is the other major logjam, and I don't know how much we can do about that - but looking internally at the process for rules? It's fine. The ID LAB team has consulted periodically with the RT but we don't absorb their full attention, and we're only using a percentage of the RT, BGT and ADT members for the LAB team.

    I'm not saying we will get there, but it's plausible.


    Onwards:

    Player skill mattering more than army book is true, but also not much of a boast. I've seen someone win six simultaneous chess games with pieces missing from their starting board in each game. (Relevance? Well, if our game is "as balanced" as chess would be if Black gets a full set of pieces, but White is missing a Knight and Purple is missing a Bishop, it's not exactly particularly well balanced now is it?)

    Also, your logic doesn't track. The difference between a false positive and a real positive is that once addressed the false positive is more likely to have a "actually, change it the other way" signal crop up and get nudged back to where it was.


    And, yes a third prong: changes matter for more than just balance. They matter for engagement. Starting a new army costs a LOT of money and time. Something you'd written off as not worth it getting a decent-sized cut can make a lot of models suddenly a lot more interesting. I know a lot of people love refining their army, or just getting used to it, but novelty matters.

    People *know* they ain't getting a new AB any time soon. They *might* get an adjustment though. And that's interesting to wonder about.

    Plus sure. You can start a new army. But other people have played that army, and worked out to a greater or lesser extent what the "good stuff" is. You can, however, be the first to work out the new possibilities opened up by the latest changes.

    Background Team

  • WhammeWhamme wrote:

    And, yes a third prong: changes matter for more than just balance. They matter for engagement. Starting a new army costs a LOT of money and time. Something you'd written off as not worth it getting a decent-sized cut can make a lot of models suddenly a lot more interesting. I know a lot of people love refining their army, or just getting used to it, but novelty matters.
    this goes the other way too, price cuts can invalidate investments. Squirrelloid can tell you all about that one. He backed the TMS UD kick starter to get a fun army and by the time it was delivered he could barely field a playable army, and definitely not the one he had bought initially due to "balance" fixes. And his list wasn't even using crazy insane combo pieces, it was just trying to build a fun fluffy list that became a casualty in the constant search for better balance.


    WhammeWhamme wrote:

    Plus sure. You can start a new army. But other people have played that army, and worked out to a greater or lesser extent what the "good stuff" is. You can, however, be the first to work out the new possibilities opened up by the latest changes.
    As DanT has repeatedly shown this is a false belief in the community, there are tons of unexplored builds and avenues, people just assume the best build has been found because people use it and then ignore everything else. There is just as much exploration to do now as there would be after a price adjustment because most of this community doesn't really look that long or delve that deep into the books.
    “You can never know everything, and part of what you know is always wrong. Perhaps even the most important part. A portion of wisdom lies in knowing that. A portion of courage lies in going on anyways.” -Lan Mandragoran, EotW


    Dovie’andi se tovya sagain.
  • WhammeWhamme wrote:

    DanT can beat most people on gameplay skill with a list handicap. That is not most people (by definition).

    Hope =/= reality. Changes to books can give hope.
    and yet other people have success with his concepts after he proves they aren't complete garbage. Which suggests it's not just his skill, but also an unexplored avenue that the community couldn't see past their "this is the only thing that works" blinders.
    “You can never know everything, and part of what you know is always wrong. Perhaps even the most important part. A portion of wisdom lies in knowing that. A portion of courage lies in going on anyways.” -Lan Mandragoran, EotW


    Dovie’andi se tovya sagain.
  • Having tinketered around intinsively with SA, having played as succesfull or not as I am I still havent tried around half of the builds that I can envision.

    Oh that does not count the builds I could not envision but have seen elsewhere.
    Like Shinys build which went from why does he do that doesnt it hurti to it is the let us copy it list in a few months.

    Advisary Board Member

    Workfields: Tournament Analysis, Army Community Support, Playtesting, Community Engagement, Translation/ United Nations DE Blog: Inside TA. The biggest german Tabletop Board: tabletopwelt.de
  • Squirrelloid wrote:

    The fact is, the game is externally balanced to a level where army book differences << player skill differences. At which point, there's no need to rejigger points yearly - and it's counterproductive, because more often than not the lax level of statistical significance used to justify changes will turn up false positives way more often than actual balance problems, which means most of the time we'll be making the game less balanced rather than more balanced.
    That's mostly true, but not really. Especially considering the context of posting this on the UD lists thread, who are without a doubt an outlier.
    Even in the context of a relatively balanced system, when left untouched small outliers will slowly grow as people converge onto them. So you need to crop the tops and weed them out when they emerge strongly, or they stagnate the entire game.

    I'm aware of your lack of faith in fine-tuning, and don't agree at all with your approach - you are very strict on avoiding type I errors (wrongly demonstrating there are differences between armies) but are happy to accept Type II (wrongly concluding armies are balanced) which makes absolutely no sense in this context and the risk-reward ratios of making each decision.
    Hristo Nikolov
  • DanT wrote:

    Midgaardsormen wrote:

    DanT wrote:

    @Kermit
    Nice! 5 wins with a sha guardian, and then one loss which it seems is primarily due to a single mistake.
    Kudos.



    Remind me how many times this model has been declared useless or terrible in the last 12-18months? :P :D
    You dont her to pick your matchups in a 1v1 ;) So i would say it is Niche at best.
    Haha here we go again...

    People think teams is this mystical place where players only get good matches.

    How many players on this forum talk about the fact that they would take their list to a singles event?
    How many top team players talk about how small they think the differences are between teams and singles?
    Heck, even randomus' tacticas talk at one point about why one should bring balanced lists to teams.

    For the record, I have seen the sha guardian played at UK singles events too, and it didn't do particularly badly.


    Putting the result down to "teams" is kinda exactly the sociology my post was implying.


    Let me try putting this differently: niche? I dno, this seems ill defined to me. Meta specific? yes... but I think the point is that maybe we are in a good meta for it, and this is more important than teams vs singles.
    but there usually are armies that are made to being paired good (just not as many as people claim).

    Also team events allow armies not to be fielded against 1-2 worst pairings, but then again due to the very same reason not against 1-2 best pairings

    this allows for more skewed armylists to be fielded, but not to the extent people claim, so can in the longrun be disregarded
  • DanT how come you totaly disregad team event paring influence? In single event your next oponen is totaly ranndom. Only thing you can assume the better you did, the better oponent you will get next round. In team event that is not certain as there always is a skill difference among enemys team. And whats more important is up to your team if you can at least choose between two oponents. One can argue to what extent team is trying to utilise it with same player to protect him, but there are deffinietly tools to do that. As Shino said it is safe to say that you wont get 1-2 best or worst parings. If your top priority is to build all around list it may not be important to you (which will help other team members) but you cannot underestimate this effect at all.
  • (All of the below is my own opinions stated without justification. You buy it or you don't, no problem :) )

    @Nicreap @WhammeWhamme
    You are both very kind...

    I think you are both right to an extent. Some of my more esoteric builds that are designed very specifically for me or with very specific play in mind, have not really been replicated.

    However, a substantial number of my builds have been picked up (sometimes tweaked) and used with reasonable success. Including at least one exact copy used at ETC.
    What I found more surprising was that a beginner-ish level player picked up the same list and found it beginner friendly.

    More excitingly, was when one of my lists seemed to change someone's opinion of a whole book.
    From boring and bad to fun and competitive.
    That felt good :)

    So yes, I do think that communities are slow to realise all the possibilities in their book, and get trapped in "conventional wisdom".
    A lot of the top team lists at the ETC for example don't quite fit with the conventional wisdom, and I think this is the tip of the iceberg.
    However, I have a hypothesis for this conventional wisdom issue...


    List writing
    I think that there is a vast amount of sociology and herd mentality involved in communities declaring things that are good or bad.
    But I also think that this is partly because t9a list building should not be done like it was in legacy.

    The lower power level/game impact level, and improved balance, means that lists where the parts work well together beat lists that have individually more points efficient units.
    (This is also why so many units are perceived to be meaningfully too expensive, because on some level every unit is overpriced until it is put into the right list).

    This is potentially a problem because it means that (A) more skill is required to write lists and (B) people are not used to writing lists in this way.
    I am currently putting my thoughts together on this (I have already started to do it in my WOtDG thread), and I plan to write a short guide from this perspective in future, to help players who want to start approaching things from this perspective.

    For someone like me, it is a big selling point for t9a: I can play with all sorts of unorthodox things without feeling handicapped, and it is one reason I am getting more out of my hobby than ever before.


    Balance tolerance
    My version of what squirrelloid said is this:
    "T9a games are by and large won by the more skilled player, mostly regardless of lists (where I am referring to the power of lists; clearly some matches will be skewed by a rock-paper-scissor effect induced by the exact lists chosen). This was not true in legacy."
    (I think this also applies to relatively small skill gaps, not just ""ETC player vs club level" type mismatches.)

    I also think that players' view of internal and external balance are shaped by a combination of:
    local meta, their playstyles, their regular opponents' playstyles, their skill, the typical skill of their opponents, the typical local skill differential, local gaming culture, local approach to terrain, the number of local players and armies (and which ones) in regular circulation, etc etc.
    I claim that this combination of things has a greater effect than the majority of the claimed imbalances in the game. Which is why every unit is OP/UP according to someone, somewhere on this forum.
    There are exceptions, but I am talking in generalities.


    UD
    However, since this is a UD thread, let me state that I think UD are an exception to this.
    I actually don't believe that UD's biggest issues are at the top tables, I think the "value added" by playing UD at top levels is less than the value added for playing UD at mid-levels. They are the army that adds the most to a weaker player. This is in exact opposition to e.g. HBE, which are a negative effect at mid-level play.
    Since the majority of games are not played at the top tables, I think this is why the "UD are OP" calls are so loud. Their effect on the skill curve is heavily distorting, and they are one of the few instances where an average game is more likely to be decided by list than skill.

    This is compounded by not being fun to lose to in the majority of instances, because the lists are predominantly blocky, with powerful magic and healing, and don't lend themselves to intrinsically interesting or cinematic games.
    "Ah, you've killed my unit and yours is now close to full health again... cool..."

    All of this is why the majority of my personal suggestions around UD are about (A) really focussing on the internal balance between constructs and non-constructs, (B) removing some of the brain-dead play and giving the opponent more counterplay options, and (C) trying to do these things with a relatively light touch.
    Of course, I gave up the right for my opinions on these things to taken any more seriously than anyone else's when I left RT.


    Meta cycling
    I personally don't get why 2-4 new books a year are not enough meta churn/freshness. In legacy that was all we ever got, and I personally never got bored.
    I don't see why annual points adjustments are suddenly needed for people to enjoy their hobby.
    It also means that those with a more hobby/narrative/Stillmanian approach don't get the rug pulled out from under them. We do want those players, right? And not just those chasing the latest tournaments?

    (Aside: I do find it interesting how many people only come to the forum to moan about balance issues, given the project has taken so much flak for being balance focussed at all costs, and has made an effort to move away from this for 18+months.)

    I don't believe that the meta is stale, and I write at least one new list most weeks that I am excited to use (although note this is not one list per book; I typically only look at one book each week).

    I don't have enough time to use all of those lists, which is why I post some of them on the forum.
    Given the fun that some others have got from my lists, I plan to start a list repository for them. If I can get even 5 people enjoying their hobby more, it will be worth every minute (and hour) I have put into writing these lists and putting them online ^^
    Being supportive & giving useful criticism aren't mutually exclusive.
    Are you supportive of the project? Do your posts reflect that?

    List repository and links HERE
    Basic beginners tactics HERE
  • I am so bored of these teams vs singles arguments.
    My position for 2 years now has been
    "analyse both, together and separately, and look what differences there are".
    I don't see how any other approach has any advantages over this one.

    All I do on the forum, is point out to the extreme camps, what each other's arguments are.
    Because often they occur on different army forums (cross ref "(sometimes) army forums are echo chambers and pick the data analysis that gives the answer they want"), and then each attack the project, whilst acting like their opinion is the only reasonable one.

    What would you think/do if you were the project in this situation?
    Different communities vehemently arguing for different things, claiming theirs is the only justified approach, and all saying you must listen to them because "community driven".

    Shino wrote:

    but there usually are armies that are made to being paired good (just not as many as people claim).
    Also team events allow armies not to be fielded against 1-2 worst pairings, but then again due to the very same reason not against 1-2 best pairings

    this allows for more skewed armylists to be fielded, but not to the extent people claim, so can in the longrun be disregarded
    Right, but how is the correct take away from this anything other than what I have said above?
    "analyse both, together and separately, and look what differences there are".


    hypnotic wrote:

    DanT how come you totaly disregad team event paring influence?
    I dont, I merely pointed out that some people claim this (see above).

    ...
    As Shino said it is safe to say that you wont get 1-2 best or worst parings. If your top priority is to build all around list it may not be important to you (which will help other team members) but you cannot underestimate this effect at all.
    And yet he also said that in the long run these things can be disregarded, which is exactly the position you attacked at the start of your post.

    My view has always been "analyse both, together and separately, and look what differences there are" (see above).
    Being supportive & giving useful criticism aren't mutually exclusive.
    Are you supportive of the project? Do your posts reflect that?

    List repository and links HERE
    Basic beginners tactics HERE

    The post was edited 1 time, last by DanT ().

  • Right, but how is the correct take away from this anything other than what I have said above?
    "analyse both, together and separately, and look what differences there are".
    there are a few issues.
    1) samples are too small and prone to outside factors

    2) strength of army varies on opponents view of it. In singles you might want to play safe when paired against negative mu, and in teams armies viewed as strongest and ones that can more easily score big will be paired so more often. Great example would be ud. look at etc this year and previous year. They have been nerfed more than other armies, yet they have higher average than last year. It might mean few things:

    a) many people didn't see the strong builds
    b) they weren't viewed as so op army, so not as often used as hammer armies
    c) different people played these armies, as often better players, tending to maximise their scores change armies to better ones
    d) luck involvment

    I believe that a can bemostly discarded, as builds didn't differ that much from last years, but still it's just that there were more uds that were meant to scor big. Still their number across the teams shows that they are believed to be the strongest army, and it seems to be so now. It's just the percent of outlier that might be exaggerated, or they other way round - diminished, by different people both playing this army and against it
  • DanT wrote:



    <snip good stuff>

    Meta cycling
    I personally don't get why 2-4 new books a year are not enough meta churn/freshness. In legacy that was all we ever got, and I personally never got bored.
    I don't see why annual points adjustments are suddenly needed for people to enjoy their hobby.
    It also means that those with a more hobby/narrative/Stillmanian approach don't get the rug pulled out from under them. We do want those players, right? And not just those chasing the latest tournaments?

    (Aside: I do find it interesting how many people only come to the forum to moan about balance issues, given the project has taken so much flak for being balance focussed at all costs, and has made an effort to move away from this for 18+months.)

    I don't believe that the meta is stale, and I write at least one new list most weeks that I am excited to use (although note this is not one list per book; I typically only look at one book each week).

    I don't have enough time to use all of those lists, which is why I post some of them on the forum.
    Given the fun that some others have got from my lists, I plan to start a list repository for them. If I can get even 5 people enjoying their hobby more, it will be worth every minute (and hour) I have put into writing these lists and putting them online ^^

    Some people like more frequent changes than legacy ever gave them. This shouldn't be surprising; legacy was never perfect.

    If you *do* like constant changes, T9A prior to gold was kinda your jam - it was changing all the time. Exciting times!

    The new status quo might be *enough*, but that doesn't mean you're not going to miss the "good old days". :)


    (Also, comp. Stuff like Swedish Comp shook up things more often than new books released. In a way, the WHFB community "churned" the metagame artificially on an irregular basis with "sorta" points changes that made some units more or less desirable without changing how they worked. In NZ it seemed like half the tournaments had different composition requirements, and I personally always tried to build to the event at hand. 'unno about elsewhere.)

    Background Team

  • Fnarrr wrote:

    Squirrelloid wrote:

    The fact is, the game is externally balanced to a level where army book differences << player skill differences. At which point, there's no need to rejigger points yearly - and it's counterproductive, because more often than not the lax level of statistical significance used to justify changes will turn up false positives way more often than actual balance problems, which means most of the time we'll be making the game less balanced rather than more balanced.
    That's mostly true, but not really. Especially considering the context of posting this on the UD lists thread, who are without a doubt an outlier.Even in the context of a relatively balanced system, when left untouched small outliers will slowly grow as people converge onto them. So you need to crop the tops and weed them out when they emerge strongly, or they stagnate the entire game.

    I'm aware of your lack of faith in fine-tuning, and don't agree at all with your approach - you are very strict on avoiding type I errors (wrongly demonstrating there are differences between armies) but are happy to accept Type II (wrongly concluding armies are balanced) which makes absolutely no sense in this context and the risk-reward ratios of making each decision.
    When you can't even measure the change you're making, you can't justify it as 'balance-driven', because it's unmeasurable.

    Every change has a cost. That cost is measured in people's time and effort in building a force that can no longer be used as is. You can't just ignore that cost. Which is one reason why i'm more happy to accept Type II errors and more risk averse on Type I errors. Because people's collections are expensive and consumed a lot of time to put together. Every change invalidates existing collections in some sense (sometimes more obviously than others).

    Another reason to favor avoiding Type I errors over Type II errors is because type II errors should eventually get spotted as data improves. If you don't change things, your data stays comparable. On the flip side, if you knee-jerk react to every type I error, you make changes which make the data incomparable.

    Type I errors are also especially pernicious when you're making multiple comparisons (and every T9A analysis makes a minimum of 16 comparisons) and have confounding factors (e.g., player skill) which you don't measure and which likely have a distorting effect on your statistical assumptions. The multiple comparisons angle alone is enough to make T9A data practices very worrying. on 16 comparisons with a 95% 'loose' statistical standard, you expect almost one Type I error per analysis. That's ridiculously common, and we're not even using that loose standard. 66.7% standard means 1 Type I error per 3 comparisons. You'd expect 5 every time you do an analysis based on random chance alone. That it appears we don't even see that many strongly suggests that any apparent 'OP/UP' results we're getting are most likely Type I errors.

    And the ease of getting strong p-values directly correlates to the size of the effect. If it's so hard to tease a real p-value out to show something is 'OP', then the effect size (how OP it is) cannot be very large. In short, it's close enough to balanced for any practical purpose.

    Finally, I disagree UD are without a doubt an outlier. If that was the case, you could prove it statistically with real statistical confidence. (Like, p < 0.01 levels after correcting for multiple comparisons). We have hundreds of data points (games) on UD alone. That's more data than many scientific studies which do report those sorts of p-values.

    At the end of the day, there's a fundamental epistemological question here, which is how close to 'perfectly balanced' is close enough, since we'll never be able to measure perfectly balanced. Answer that question, and you can calculate how much data you need to answer it with real precision. What you can't do is degrade your statistical standards and set arbitrary short time horizons, and assume you're getting any sort of useful or consistent answer on how balanced things are.

    Basically, it feels like the project wants to manufacture imbalance just so it can change things. That's not a goal of balancing things, that's a goal of finding a way to justify changes, and it will necessarily sacrifice real balance along the way.

    Edit:
    Aside - the people who argue for annual point updates to 'keep things fresh' are not arguing for a balanced game. They're arguing for change solely for the sake of change, and that works counter to balance. If balance is the real goal, changes should become rarer and rarer as the game asymptotically approaches 'balanced', and ultimately reach a stasis point where no further changes are needed. Because while we can't measure it perfectly, any given 'balanced' state is an objective state that doesn't change.
    Just because I'm on the Legal Team doesn't mean I know anything about rules or background in development, and if/when I do, I won't be posting about it. All opinions and speculation are my own - treat them as such.

    Legal

    Playtester

    Chariot Command HQ

    The post was edited 2 times, last by Squirrelloid ().