Balance/Power Creep in Malifaux

[lusciousmccabe]

9 hours ago, Fetid Strumpet said:

As someone said above, how good is your low def model with the high armor, hard to kill, and regen going to be when Raspy, or Molly, or anyone else with a paralyze effect targeting Df just keeps you perma paralyzed for most or all of the game?

... Conversely, in general there isn't anything in the game that lets you just ignore high a high defensive stat. The closest being models which in an aura give negatives to the defensive stats, and all of those to one degree or another have issues (for good reasons).

I would add Leveticus and Victoria to the list of models who really don't care if you have Armor 57, Regeneration 9 and Hard to Kill twice. In Levi's case it makes absolutely no difference and with Mark of Shezuul the lower Df target is actually more vulnerable. That's what really hurts low def models, since your presenting vulnerability they can exploit. 

If you look at the things that do let you interact with defensive stats they either need to hit that stat (slop haulers) or penalise all defensive stats to the same extent (Arachnids, Daydreams, the various no cheat/negative flip effects). If there was an ability that said something like "This model treats Df scores greater than 4 as 4 when making attack actions" then you could compare it to the various work arounds for defensive abilities, but as it is there's this all or nothing effect for Hard to Kill (for example) wheras a higher Df score works against anything other than indirect attacks, which also neutralise a fair amount of defensive abilities as well. 

[Math Mathonwy]

Stuff that ignores Hard to Wound is actually pretty rare. Viks, Levi, and Joss spring to mind (though in Joss' case it doesn't matter much). And Moon Shinobis which actually benefit from it but they aren't very common. I'm sure there are more but I can't think of them off hand. Armor is ignored a whole lot more.

But it's a difficult subject to approach. HtW is great against certain damage tracks - I'd take HtW over +2 Df when fighting against something that does 2/5/7 damage. But if an attack causes Paralyze or even Slow after damaging that +2 Df would likely be way better.

But really, I'm not sure that this is about powercreep in the absolute sense - I don't think that many people would suggest that Book 3 has somehow a higher power level than Books one and two. But average stats going up does put pressure on future stats going up as well. Myyrä's analysis of 8SS with attack stat of 5 being useless is kinda chilling in a way. If the defensive stats were on a wider scale, then maybe that wouldn't be true?

[Math Mathonwy]

13 hours ago, katadder said:

No I dont hire arachnid swarms, just summon spiders for them, but they are still not poor models because of having an Ml5. 

I do like them but most my other crews tend to be fixed (and with arcanist generally play to theme) so squeezing in an 8SS beater instead of a 10SS one is difficult. They have still had some good kills for me and swung games

Arachnid swarms would fit the themes of Mei Feng and Ironsides just fine but no one hires them. You saying that they aren't bad goes kinda contrary to that. I mean, they certainly aren't bad when Summoned "for free" with Ramos!

But if no one hires them, that suggests that they are overcosted or, in other words, bad. I mean, I don't think that there exists a model that is so bad that no one would take it if it were cheaper.

If they had Ml 7 I think they would be more popular hires (they would also be unholy terrors but that's beside the point).

[solkan]

15 hours ago, Math Mathonwy said:

Do you hire Arachnid Swarms often? I never see anyone hire them (outside of Adran, naturally). They have built-in synergy with half the Masters of the Faction yet they aren't hired. Now, granted, their design is such that if they had a high Ml stat they would be absolutely horrible but the point remains.

Do I get bonus points because the option to hire Steam Arachnid Swarms for Mei Feng, or to hire enough Steam Arachnids to form up into Steam Arachnid Swarms once they cross the table, is why I have (through trade and happenstance) thirteen Steam Arachnids magnetized so they can combine into four swarms? 

It was fun when I got someone to say "You hired how many?" playing with Mei Feng and hired Steam Arachnids at a tournament.

Clearly, no one would hire a Steamborg Executioner with Df 5 because it would be too easy to kill. 

 

[Math Mathonwy]

16 minutes ago, solkan said:

Do I get bonus points because the option to hire Steam Arachnid Swarms for Mei Feng, or to hire enough Steam Arachnids to form up into Steam Arachnid Swarms once they cross the table, is why I have (through trade and happenstance) thirteen Steam Arachnids magnetized so they can combine into four swarms? 

It was fun when I got someone to say "You hired how many?" playing with Mei Feng and hired Steam Arachnids at a tournament.

But yeah, points, definitely.

16 minutes ago, solkan said:

Clearly, no one would hire a Steamborg Executioner with Df 5 because it would be too easy to kill. 

I think Howard would be a lot less used if his Ml was 5.

[Myyrä]

4 hours ago, Math Mathonwy said:

I mean, I don't think that there exists a model that is so bad that no one would take it if it were cheaper.

Governor's proxy comes close. With certain masters I wouldn't take it even if had a cost of -2ss.

[D_acolyte]

18 hours ago, Omenbringer said:

Dont assume that many of the more vocal posters in this thread haven't play tested for Wyrd before. Many of us have and some still do. In my experiences the play testing process wasn't super effective. Often times when valid overly potent interactions were discovered during play testing it was met with shall we say a very aggressive and vocal  opposition that drowned out the signal. More than a few of these wormed there way to publication despite being discovered during play testing. Perhaps it has changed...

The 0 point upgrades and power reduction of the Metal Gamins are interesting examples. Wyrd didn't just change them on their own, threads such as these prodded them to look at the models more acutely. I will offer that they are a bit quicker than last edition to change things.

A discussion about Strategies and Schemes would take its own thread and would likely result in the same posters digging in as they have before.

I am not assuming that people in this post are not did not participate in playtesting, I have participated in all the open play testing myself. As for the effectiveness of play testing the larger the group the more dissenting voices are drowned out, it is often an unfortunate aspect of open playtesting. The main purpose of this thread from what I can see is delving into worries about new models coming down the line not the old models power, the only way to try and change what is to come is by seeing if you can join a play testing group when they put out the call for it.

I used the various examples to show that Wyrd is still committed to trying to balance there games. Sure topics on the forum can have an effect the issue is separating what is just noise from what is not.

I am not saying the Strategies and Schemes are out of balance but rather that they change the balance of the models or how you look at them.

I fully believe there are vary few things that brake the current ed of Malifaux alone but rather they do it from taking multiple of them or having compounding effects.

[LunarSol]

4 hours ago, Math Mathonwy said:

If the defensive stats were on a wider scale, then maybe that wouldn't be true?

I'm not sure how you put them on a wider scale.  The game essentially operates on a 2D13 dice curve (with a dozen or so caveats) with the point where Df = Ml acts as the central peak with a success rate of about 54%.  Each point of variance shifts the probability along the curve and while there's little problem with expanding the range against average stats, it does create issues when models in the expanded range clash (Df 8 vs Sh 3 for example).

That's not really the issue with Ml 5 on 8 SS+ models though.  The issue is just that they generate far less AP for your crew than equal points in cheaper models, but that AP isn't notably more powerful or reliable.  You're looking to fail on half your attacks and you're looking at probably dealing 2 damage.  You can get that for 5 SS easily.  The tendency for min 2 to be the norm tends to shadow over other factors.  I'd wager Ml 5 would be far less of an issue on a model if it was Min 3.

[Adran]

18 hours ago, Omenbringer said:

What I dont agree with, is that compensating for lower Df through abilities such as Armor is a great method. The model is still going to get hit and the margin of success is also going to be higher. Higher Margins of Success translates to higher damage potentials (even with damage abilities considered) and unless the Black Joker shows up still means 1 Wd inflicted versus the always 0 Wd's of a successful Defense Resist.

I'd disagre, I think  that you can compensate for lower stats with other defensive abilities 

I'd also like to point out that not all damage comes from attacking df. 

Models with Armour like abilities and Hard to kill get those protections against blast damage and pulse damage and condition damage (which can sometimes be placed without any duels)

Quite a few models attack Wp, so a high Df doesn't help, where as the hard to wound  or armour will still help. 

If you want to cheat the damage flip (lets say you do blasts on severe or such) then focus and attack works better against a high Df model than a lower Df model with Hard to wound. 

 

And there are always corner cases, where a model with low Df and High Armour is a great model for you to use to target yourself and get blasts off from. 

 

So yes, I probably would rather a point of Df over a point of Armor in general, but that's not always true. And the ability to have a range of different choices as to how you want a model to survive is one of the reasons there can be such a huge variety in peoples crews. By model a having a lower df and armor, perhaps it has a larger ml range to make it equal points, or it gets some other ability. A model with Df 4 and Armor+2 will typically last longer in a fight against Howard than a model with Df 5 will, even though it will get hit more often.  (unless they are very low wounded models)

 

[Myyrä]

Found the code I used for comparing defenses of different models. What I had ready was an Austringer attacking 4 different models

1. Df 5, Wd 5
2. Df 5, Wd 5, Manipulative 11 (unactivated)
3. Df 5, Wd 4, Armor +2 (codename Metal Gamin)
4. Df 4, Wd 5, Armor +1 (codename Ice Gamin)

With focused attacks it takes on average approximately the following amounts of ap to kill the targets, (including the ap spent on focus):
5.1
9.7
7.0
5.9

Without using focus it takes approximately following amounts of ap:
3.9
7.4
5.1
4.4

The calculations are done using Monte Carlo simulation. The flips are produced by simulated decks. No cheating is included. Also, Black Joker isn't always chosen on attack flips when using focused attacks apparently, but I'm too lazy to fix it right now.

Moral of the story:

Don't use focus with Austringers if you don't need the extra range and don't intend to cheat. Even low manipulative can be surprisingly good if you haven't activated. In certain cases armor +1 can be more useful than an extra point in Df.

If someone has some specific models they would like to compare and the list is short enough, I might be willing to run some more calculations.

[Omenbringer]

20 hours ago, Redbeard said:

Really? It was pretty obvious that he was creating an extreme example to show that defense abilities other than defense do matter. Thus why he said "we can argue about where that line is."

So you can see the sarcasm in Justin's post but not in my response? Interesting.

As this thread has likely already reached the point where sides have dug in (honestly probably began that way) and nothing good is likely to really come out of it any more, I will be bidding it fair well before the bans start flying (as often happens in these threads).

One last thing; Being a "fan" of the game doesn't mean you can't criticize it's deficiencies, it means you are obligated to work at improving them.

[Math Mathonwy]

3 hours ago, LunarSol said:

I'm not sure how you put them on a wider scale.  The game essentially operates on a 2D13 dice curve (with a dozen or so caveats) with the point where Df = Ml acts as the central peak with a success rate of about 54%.  Each point of variance shifts the probability along the curve and while there's little problem with expanding the range against average stats, it does create issues when models in the expanded range clash (Df 8 vs Sh 3 for example).

I meant "wider scale" in the sense that there would be more models with lower stats. It was indeed a bad expression.

3 hours ago, LunarSol said:

That's not really the issue with Ml 5 on 8 SS+ models though.  The issue is just that they generate far less AP for your crew than equal points in cheaper models, but that AP isn't notably more powerful or reliable.  You're looking to fail on half your attacks and you're looking at probably dealing 2 damage.  You can get that for 5 SS easily.  The tendency for min 2 to be the norm tends to shadow over other factors.  I'd wager Ml 5 would be far less of an issue on a model if it was Min 3.

Out of the models that Myyrä listed previously, The Sow has 3/5/6 (as well as three-headed), Slate Ridge has 2/3/5 or 4/5/7 depending on his health, Rail Golem has 3/4/6, Rat King does 2/3/4, Arachnid Swarm does 2/4/5 (plus in-built Overwhelm Trigger and Strength in Numbers), Ama No Zako does 3/5/7, Candy does whatever she does, Cojo does 2/3/6, and Sue does 2/3/4 but has a built-in Critical Strike. So essentially four have min 3, two might have it, one has other stuff to make up for it, and two are just kinda bad (Rat King and Cojo). OTOH almost everyone has an extra AP available through something be it Smell Fear, Casting Expert, Overwhelm Trigger, or a (0)Action that can provide essentially extra AP.

[Myyrä]

38 minutes ago, Math Mathonwy said:

Out of the models that Myyrä listed previously, The Sow has 3/5/6 (as well as three-headed), Slate Ridge has 2/3/5 or 4/5/7 depending on his health, Rail Golem has 3/4/6, Rat King does 2/3/4, Arachnid Swarm does 2/4/5 (plus in-built Overwhelm Trigger and Strength in Numbers), Ama No Zako does 3/5/7, Candy does whatever she does, Cojo does 2/3/6, and Sue does 2/3/4 but has a built-in Critical Strike. So essentially four have min 3, two might have it, one has other stuff to make up for it, and two are just kinda bad (Rat King and Cojo). OTOH almost everyone has an extra AP available through something be it Smell Fear, Casting Expert, Overwhelm Trigger, or a (0)Action that can provide essentially extra AP.

Most other models in their price range also have similar things.

[LunarSol]

50 minutes ago, Math Mathonwy said:

I meant "wider scale" in the sense that there would be more models with lower stats. It was indeed a bad expression.

Out of the models that Myyrä listed previously, The Sow has 3/5/6 (as well as three-headed), Slate Ridge has 2/3/5 or 4/5/7 depending on his health, Rail Golem has 3/4/6, Rat King does 2/3/4, Arachnid Swarm does 2/4/5 (plus in-built Overwhelm Trigger and Strength in Numbers), Ama No Zako does 3/5/7, Candy does whatever she does, Cojo does 2/3/6, and Sue does 2/3/4 but has a built-in Critical Strike. So essentially four have min 3, two might have it, one has other stuff to make up for it, and two are just kinda bad (Rat King and Cojo). OTOH almost everyone has an extra AP available through something be it Smell Fear, Casting Expert, Overwhelm Trigger, or a (0)Action that can provide essentially extra AP.

The issue with there being more models with lower stats is that it becomes irrelevant if there are enough models with higher stats to fill a list.  That's the trick; no matter how "balanced" everything gets, you still have to leave most of it in the bag when you play an actual game.

I do wonder how much of that is an issue for some of these models.  I originally wrote that up on the back of the SA Swarm discussion, but I really should have dug into the list further.  I do suspect a lot of it is just opportunity cost.  The Rail Golem is a lot harder to justify over Howard after all, and its a point level where there's rarely design space for both.  I'm a bit unconcerned when it commons to summonable models as they're cost tends to exist to mechanically facilitate other model abilities more than to make them a competitive hiring choice (Rat King is the big one, though Cojo is an example of a summonable model in need of compelling reasons to summon).

What I really failed to mention is that for most everything here, there's a roughly equal cost model with the same damage track or better at Ml 6 (or better).  I mean, Rail Golem is min 3, sure, but Howard is 4.  My point is that if a model has a lower Ml it needs to have some considerable compensation elsewhere and generally they don't.  Howie vs the Rail is an easy target, we're talking about one model that hits Df 6 for at least 4 damage 3:5 times while the other hits it only about 2.25:5 for 3.  The Rail Golem needs to make nearly twice as many attacks to equal Howard's damage output.  (Flip that min damage around and they're actually close to even surprisingly...)

That's what I was really trying to say.  Ml 5 is a drawback, but one that can be accounted for.  I think there's a decent number of models out there that aren't properly compensated, but I can also see an argument to the effect that there isn't enough attack volume in the game to make up the difference.  That gets messing with some of the core concepts of the combat engine though and would require some considerable thought.

[Durza]

Rail golem is also min damage 5 if it has burning and the attack was from locomotion.  

[Myyrä]

13 minutes ago, LunarSol said:

What I really failed to mention is that for most everything here, there's a roughly equal cost model with the same damage track or better at Ml 6 (or better).  I mean, Rail Golem is min 3, sure, but Howard is 4.  My point is that if a model has a lower Ml it needs to have some considerable compensation elsewhere and generally they don't.  Howie vs the Rail is an easy target, we're talking about one model that hits Df 6 for at least 4 damage 3:5 times while the other hits it only about 2.25:5 for 3.  The Rail Golem needs to make nearly twice as many attacks to equal Howard's damage output.  (Flip that min damage around and they're actually close to even surprisingly...)

You didn't take into account the Rail Golem has. Their odds of hitting without cheating are actually very close to being equal.

[LunarSol]

These are the things that happen when I try to math from memory without cards.  

[bertmac]

+flips to attack and the bonus damage when locomoting plus triggering extra attacks can  compensate for the lower ml stat.

[SpiralngCadavr]

...if you're going for reliability rather than being able to play something with certainty from your hand.

I consider a positive twist valuable for spamming attacks, but it's still not the same as an unbeatable attack with your hand to help that your opponent will want to be stopping.

[Myyrä]

2 hours ago, SpiralngCadavr said:

...if you're going for reliability rather than being able to play something with certainty from your hand.

I consider a positive twist valuable for spamming attacks, but it's still not the same as an unbeatable attack with your hand to help that your opponent will want to be stopping.

Yeah, probability of hitting without cheating is not nearly as relevant for a model whose attacks will almost always be cheated if they happen to miss.

[solkan]

20 hours ago, Math Mathonwy said:

I think Howard would be a lot less used if his Ml was 5.

Howard Langston wouldn't be a 12 SS Enforcer if his primary attack was MI 5.  Or that attack would be MI 5 for Howard to still be 12SS. 

But the attack on the steam arachnid swarm probably has MI 5 because it has a built in statistically declining recursion trigger--the sort of thing which produces unreliable yet occasionally spectacular results.  But it's also runs into the decapitate scenario--the attack is more likely to drain the opponent's resources rather than kill something--because preventing the recursion either through cheating to win the defensive duel or preventing the damage is going to pretty attractive to the defender.  And there are people who are going to despise the attack because it produces unreliable yet occasionally spectacular results.

If you don't see a lot of 8SS arachnid swarms running around, that could be because people creating their crews go "8SS?  That seems a bit expensive for a minion.  Maybe I should go with two 4SS models instead for more activations."  But I'd think that something like Hunting Party or Public Demonstration would make people pause to reconsider the list of available minions.

 

[Tawg]

Without quoting specific messages (Since I'm late to the party), I just wanted to say a little 2¢.

I know I disagree with a handful of things that @Fetid Strumpet believes (Namely Master influx), but I do think their notion of lower Df being exclusively worse is at least somewhat relevant.  For the same reason they mentioned when replying to @Justin, which is, when you look at the game as "How quickly can I kill this" some of the defensive Abilities do swing things in favor of those models.  But the volume of Abilities which forego the bottom line of damage in favor of fringe benefits (Paralyze and Slow, among a plethora), needs to be comparatively put next to such things, because they strongly effect the validity of models when faced with having lower Df.

As a counter point, the number of models which can invalidate conditions such as that are far less frequent (As in to say, "Cannot be paralyze" or such), and something such as Save Face Upgrade from the Ten Thunders is an invaluable tool when utilizing a model like Ototo against those with Terror, knowing that he would struggle against Paralyze.  But against similar "Issues" when not complimented by Misaki, Ototo could quickly be invalidated by his low Wp (almost directly a parallel for Df in all cases) despite his defensive Abilities.

So while yes, I can value Ototo by way of being able to laugh off (No pun) damage and often stick in an ugly fight with his self-heal, likewise I can count on him to be reliably poor against those which ignore his "strength", when I'm not actively trying to supplement his deficiencies.

So while the question of Df (Or Wp) vs Abilities which supplement them is relevant, and I agree with @Justin that at a certain point they break cover the deficiency, however that coverage of deficiency are often not all inclusive, which leaves them open to more than just plain Df // Wp would, if they could avoid being hit at all.

Thus, the issue that needs to be examined is not simply stat lines (Which realistically seems to be @Justin's core point), but a more holistic look at what could combat the various key "strengths" of models.  Which is often where models come out on top, for with reliable Slow or Paralyzes or, when dealing with damage the Ability to ignore typical defense Abilities while hitting reliably.  Conversely, models which deal with such status effects are often valued highly as well (Which high Df // Wp do by default), due to a unique niche they can fulfill in dealing with those issues, which would be one of the reasons a Witchling Stalker is hard to invalidate as a 4ss model (They are 4ss, maybe 5?  Cheap with a useful tool, either way).

And there are obviously other metrics the game measures as "useful as hell" such as Activation control (The recent topic of Rat-Engine), or awkward interactions that are unintended but not game breaking enough to demand errata (Like Papa-loco-in-a-box-buffing, or the likes).

Granted I don't play competitively, nor am I a terribly great player, so take what I say with a grain of salt.

Thinking about it, a unique way to punish higher Df (or Wp) models that may be useful, while not being terribly more powerful against the "lower Stat" models would be effects which apply on missing, such as an attack doing 2/3/4, but on missing applying Slow or such.  Basically creating a dichotomy where missing is not strictly the best outcome for the defending model.

[trikk]

22 hours ago, Myyrä said:

Governor's proxy comes close. With certain masters I wouldn't take it even if had a cost of -2ss.

Would totally get him for Lady Justice! We all know that Brutal Effigy should be her totem

[solkan]

5 hours ago, Tawg said:

Thinking about it, a unique way to punish higher Df (or Wp) models that may be useful, while not being terribly more powerful against the "lower Stat" models would be effects which apply on missing, such as an attack doing 2/3/4, but on missing applying Slow or such.  Basically creating a dichotomy where missing is not strictly the best outcome for the defending model.

An attack which applies an effect if it fails is essentially an action which automatically succeeds.  I don't think that's creating a useful dichotomy, especially is a game where the attacker can just cheat aces to produce the "fail" result.

Otherwise, by my count there are four models that have "After failing" triggers on their attack actions.  And both of the after failing triggers that cause damage require a non-automatic suit.

 

[LunarSol]

Another possible option is to change some of the on hit effects to "after dealing more than 1 damage".