3.0 Actual Games Conversation

[Kadeton]

2 hours ago, whispersofblood said:

It's simple math, bonuses to save influence the smallest pool of dice and therefore have the lowest influence on outcomes. 

That doesn't sound right, math-wise. Everything in the attack sequence should be commutative, unless a special rule breaks the sequence (by converting it to a mortal wound, or adding bonus hits, or whatever).

You could roll saves first, then to-wound on any failed saves, then to-hit on any successful wounds, and it would have no effect on the statistical outcomes.

[Beliman]

If your army doesn't have enough source of mw, I find that all units are a bit more resilient this edition than before. 

It's all about how many +1 saves can you stack without wasting CPs, and then reacting with AoD to an attack by the most buffed enemy unit (without mw).

At least, that's my experience playing with a Rend-Based army.

[Neil Arthur Hotep]

25 minutes ago, Kadeton said:

That doesn't sound right, math-wise. Everything in the attack sequence should be commutative, unless a special rule breaks the sequence (by converting it to a mortal wound, or adding bonus hits, or whatever).

You could roll saves first, then to-wound on any failed saves, then to-hit on any successful wounds, and it would have no effect on the statistical outcomes.

In the basic case that's correct, but if you have an ability that triggers on 6s on the wound or save roll, having a higher chance to hit in the first place makes those abilities trigger more often. So modifiers to the hit roll do have a consequence for rolls down stream. However, saying that this automatically makes hit modifiers the most impactful is a massive oversimplification.

[whispersofblood]

1 hour ago, Kadeton said:

That doesn't sound right, math-wise. Everything in the attack sequence should be commutative, unless a special rule breaks the sequence (by converting it to a mortal wound, or adding bonus hits, or whatever).

You could roll saves first, then to-wound on any failed saves, then to-hit on any successful wounds, and it would have no effect on the statistical outcomes.

It's not cummulative, while calculating averages we are actually calculating the chance of a single attack on a profile being successful, and then imply the multiplication required for the profile. 

But that isn't how it works in actuality. Each step is a new set of likelihoods, and the number of dice rolled creates a weighted value to any modifications to a "successful roll". 

So a 16% increase in successes on 30 dice. And a 16% decrease in success on 6 dice aren't symmetrical in influence to how much damage a unit takes generally speaking. And while we can find fringe cases where a combination of wound characteristic, rend value, save characteristic, and damage value create equilibrium. When deciding *influence* you want always get the most value by targetting upstream. Especially because the result is removing models you also want in increase the number of steps where you can benefit from outliers, particularly when you can positively influence the appearance of negative outliers. Getting more hits than you should, should result in more damage per die. A fractional gain on a small pool statistically has a high chance of resulting in *no* change in outcome.

Now that isn't how we feel about the save roll, but that is kind of hear nor there. That's before you figure out what the cost of each action is. This is functionally the same methodology people have used to intutuited that Rend -1 is probably has the lowest value to a player. 

Edited July 12, 2021 by whispersofblood

[yukishiro1]

4 hours ago, whispersofblood said:

It's simple math, bonuses to save influence the smallest pool of dice and therefore have the lowest influence on outcomes. 

Ironically, that's not how the simple math actually shakes out, though. Because of the way saves work, the relative values are important. For example, take a profile of 3+/3+/0/1 against a 4+ save - applying -1 to hit reduces the damage by less than applying +1 to save. The disparity becomes even larger vs a 3+ save. On the other hand, if you have a 5+/3+/0/1 profile vs a 4+ save, -1 to hit has more impact than +1 to save; if it was a 5+/3+/0/1 profile vs a 3+ save, -1 to hit is statistically the same as +1 to save. Etc etc. 

A blanket statement that the -1 to hit has more impact than the +1 to save is not only wrong as a blanket statement, it's also usually wrong in the specific circumstance we were discussing, when a hammer hits an anvil with a 3+ save. For a 3+/3+ profile, which is the most common hammer profile, you need rend 2 to make the two equivalent, and you need rend 3, which almost nothing in the game has, to make the -1 to hit better. 

 

Edited July 12, 2021 by yukishiro1

[whispersofblood]

12 minutes ago, yukishiro1 said:

Ironically, that's not how the simple math actually shakes out, though. Because of the way saves work, the relative values are important. For example, take a profile of 3+/3+/0/1 against a 4+ save - applying -1 to hit reduces the damage by less than applying +1 to save. The disparity becomes even larger vs a 3+ save. On the other hand, if you have a 5+/3+/0/1 profile vs a 4+ save, -1 to hit has more impact than +1 to save; if it was a 5+/3+/0/1 profile vs a 3+ save, -1 to hit is statistically the same as +1 to save. Etc etc. 

A blanket statement that the -1 to hit has more impact than the +1 to save is not only wrong as a blanket statement, it's also usually wrong in the specific circumstance we were discussing, when a hammer hits an anvil with a 3+ save. For a 3+/3+ profile, which is the most common hammer profile, you need rend 2 to make the two equivalent, and you need rend 3, which almost nothing in the game has, to make the -1 to hit better. 

 

Ah I get the disconnect, perhaps "simple" was an understatement. It's more akin to economics, than arithmetic. 

The conversation was about influence. And why All out Attack feels more impactful than All out Defence. My explanation kinda meandered, probably because I'm doing multiple things at once.

The difference is the value each command ability beings to a player. All out Attack allows a player to create situations where they can impact the chance to benefit from variance, particularly because they are upstream. 

All out Defence allows for a player to decrease the odds of failure down stream but because the influence is on only a few pools of dice it has consequently less influence on the outcome.

Because they have same cost the same for differing amounts of influence on any given situation the resulting value is different. This effect becomes more pronounced the more rules you add in, or the more you improve the underlying impact (short hand for the relationship between chance for success+amount of dmg) for an individual attack. 

[yukishiro1]

No, that's not what the conversation was about.

Responding to my post pointing out that All Out Defense has an outsized impact on boosting the survivability of 3+ save models vs All Out Attack does on boosting damage in that situation, you disagreed, saying:

Quote

 Realistically All out Defence should be -1 to hit rolls or - 1 A to a minimum of 1 to be as beneficial, or relevant

When someone called you on this by pointing out that no, -1 to hit would not be as beneficial as +1 save most of the time in this situation, you then said it was "simple math" that you were right. 

Which is just wrong, mathematically. In the situation we're talking about, a model with a 3+ save, it's very rare among the common profiles in the game that a -1 to hit would protect better than a +1 to save (except against MW obviously, as we all noted above). 

If you were trying to make something other than a simple mathematical claim that's fine, but why couch it in those terms if so?

 

 

[Boar]

31 minutes ago, whispersofblood said:

All out Defence allows for a player to decrease the odds of failure down stream but because the influence is on only a few pools of dice it has consequently less influence on the outcome.

Downstream/upstream you are constantly refering to this. If we would take it seriously than f.ex. in attack profile where hit and wound values are the same, we would expect better return so to speak, from boosting to hit value, than to wound. In reality it does not, average values stays the same, what could be expected is different distribution.
Downstream/upstream is just human psychology at work.

[Rachmani]

7 minutes ago, Boar said:

Downstream/upstream you are constantly refering to this. If we would take it seriously than f.ex. in attack profile where hit and wound values are the same, we would expect better return so to speak, from boosting to hit value, than to wound. In reality it does not, average values stays the same, what could be expected is different distribution.
Downstream/upstream is just human psychology at work.

Exactly, upstream/downstream only becomes a thing when certain outcomes allow you to bypass of the normal order. 

MW on a 6+ to hit are better than MW on a 6+ to wound, but that's about it.

[Christopher Rowe]

How long are people finding 1,000 and 2,000 point games take? I've got a game day scheduled next month and I've got two tables. I was planning on running two-hour blocks (six games) and taking the suggestion from my local Warhammer shop manager to cap AoS games at 1,000 points for that amount of time. But I'm getting some pushback from players who want to play 2,000 point games.

[Andrew G]

Recounting my 3.0 games from the perspective of a fairly competitive Big Waaagh! player (I win ~75% of the 15-25 local events, score in the top 25% at least at GTs).  Writing this for fun, rather than any serious, reasoned criticism of the new edition as I'm sure some salt will seep through.

Caveats:

1.  All my experience in 2.0 was Mawcrusherless, board/objective control style, Big Waaggh! I wanted to try out a Mawcrusher Build in IJ for my first 3.0 games
2. I'm rusty from the pandemic + playing other table top games
3. I know the warclans book is coming out soon.

Game 1 vs. Lumineth

Played against a 1-drop Teclis+30 Archers+Cathallar+stuff to stand in front of the before mentioned things. 
Lumineth T1: Shoots all my warchanters off the board and buffs his army (very hard to hide with shortened board + no-LOS restrictions).
IJ T1: Sacrifice a unit 5'ardboyz to unleashed hell. Mawcrusher and a unit of goregruntas get engaged. Net result, 10 spears dead and 0 archers (Archers took 0 damage from 3 goregruntas).
Lumineth T2: I ignored a few of the spell directed at Mawcrusher, but "reroll shooting against target" spell attacks goes through. Archers delete mawcrusher in one turn (this is the defensive build - Ironclad, 5+ Ward, ignore spell on 4+), teclis and spears deal with engaged gore gruntas.
IJ T2: Sacrifice another 5'ardboyz to unleash hell, goregruntas+ a different unit of 'ardboyz sneak around screens and hit archers. I actually kill 6 this time.
IJ T3: All I had left was a unit of Brutes, a larger unit of 'ardboyz engaged with his archers, and a couple 5 man units and wizards. I was ahead on objective points, so figured I'd play it out. I end up not doing much. 

Lumineth T3: Pretty much tables me. 

Take-aways:  I have no idea how archers slipped through cracks this edition. They were already good, and just got better with the introduction of unleashed hell, shortened board, more levers to pull to increase defensiveness. Scenario played a big role, as he was able to castle and remain in protection of Teclis with his entire army while contesting half the objectives. 

Game 2 vs. Slaves to Darkness: coming once I get some more spare time at work. 
 

[Overread]

2 hours is a tight fit for a 2K game especially considering its a brand new edition so people are going to be playing slower than normal. I'd agree with your store manager, if you want 2 hour games right now then 1K is a more sensible value to pitch the game at than 2K. 

You "might" get away with 1.5K, but it might be a rush. 2K would certainly feel like a rush in 2 hours for people new to a new edition 

[whispersofblood]

2 hours ago, Boar said:

Downstream/upstream you are constantly refering to this. If we would take it seriously than f.ex. in attack profile where hit and wound values are the same, we would expect better return so to speak, from boosting to hit value, than to wound. In reality it does not, average values stays the same, what could be expected is different distribution.
Downstream/upstream is just human psychology at work.

 

2 hours ago, Rachmani said:

Exactly, upstream/downstream only becomes a thing when certain outcomes allow you to bypass of the normal order. 

MW on a 6+ to hit are better than MW on a 6+ to wound, but that's about it.

Which is why my expanded explaination of the position included the increased impact of additional rules. 

But fundamentally the underlying issue of cost becomes relevant. 1 CP to create situation which decreases the detriments of variance in large pools, while retaining the benefit of variance vs. 1 CP for improved success rate + variance across a small pool(something like 45% of the original pool of a 3+/3+ profile).

This also explains why boosting a 3+ to a 2+ is "better value" than a 4+ to a 3+ while the straight probable gains for spend remains the same, the room for variance in the actual performance undermines the particular gains from the 1 CP spend. Think of it this way; you roll 12 dice on a 4+ and roll 6 success, but when you check your '3' you would have only gained 1 additional success because the majority of the failures (of which there were a statistically average amount) were '1's, and '2's. The benefit you would have gained is small for your spend, despite appearing average in the instant. 

The value isn't in the particular gains (+1 to hit, -1 SV) it's in the relative difference in the effect of the spend. This relative difference matters in game, because dice rolls are instantaneous and not statistical averages. This generally isn't relevant because we haven't run into magnitude of failure dice rolls, so we haven't needed to consider them. As such averages have been a suffeciently deep consideration. Now we have an on demand ability to turn marginal failure into success, which means we have to consider how the variation of specific values before we can acertain the value of the ability/spend.

There will be situations where even attacking second you will net a larger advantage by going All out Attack yourself rather than All out Defence based on the deeper probabilities. 

Wait until we start having conversations about the fluxuation on the value of a on CP across phases and factions. 

 

[stratigo]

9 hours ago, whispersofblood said:

I'm not an anything player, or should we get into a pissing contest counting our event appearances and factions appearances? I can assure you my nether regions are suffeciently swoll. 

It's simple math, bonuses to save influence the smallest pool of dice and therefore have the lowest influence on outcomes. 

Flatly, saves are extremely influential unless you are playing an army that regularly ignores them.

 

So LRL and Tzeentch don't care that much about saves, no. All out defense is weak against them.

 

But the math for saves is, in fact, more influential to the outcome of an attack them the match for hit or wound rolls for units with already good saves. Particularly because save stacking is an easy way to negate rend and incredibly common

[Rachmani]

34 minutes ago, whispersofblood said:

 

Which is why my expanded explaination of the position included the increased impact of additional rules. 

But fundamentally the underlying issue of cost becomes relevant. 1 CP to create situation which decreases the detriments of variance in large pools, while retaining the benefit of variance vs. 1 CP for improved success rate + variance across a small pool(something like 45% of the original pool of a 3+/3+ profile).

This also explains why boosting a 3+ to a 2+ is "better value" than a 4+ to a 3+ while the straight probable gains for spend remains the same, the room for variance in the actual performance undermines the particular gains from the 1 CP spend. Think of it this way; you roll 12 dice on a 4+ and roll 6 success, but when you check your '3' you would have only gained 1 additional success because the majority of the failures (of which there were a statistically average amount) were '1's, and '2's. The benefit you would have gained is small for your spend, despite appearing average in the instant. 

The value isn't in the particular gains (+1 to hit, -1 SV) it's in the relative difference in the effect of the spend. This relative difference matters in game, because dice rolls are instantaneous and not statistical averages. This generally isn't relevant because we haven't run into magnitude of failure dice rolls, so we haven't needed to consider them. As such averages have been a suffeciently deep consideration. Now we have an on demand ability to turn marginal failure into success, which means we have to consider how the variation of specific values before we can acertain the value of the ability/spend.

There will be situations where even attacking second you will net a larger advantage by going All out Attack yourself rather than All out Defence based on the deeper probabilities. 

Wait until we start having conversations about the fluxuation on the value of a on CP across phases and factions. 

 

From what I read here we should probably agree on "common semantics" here. Because I think I get what you're trying to say, but stuff like "better value" is very confusing because it's statistically untrue. You're however safer from anomalies the well, lower the number. 
Basically it's like this (I'll just use saves but that's true for every dice roll in the game): 

Going from 6+ down to 5+ to 4+ etc. your additional success value goes down with every step. 6+ to 5+ is a 50% gain in saves. 5+ to 4+ only 33% etc. So the first step is the most valuable in that regard (statistically speaking).

However, if you you come from the perspective of failure it's the other way round. The lower you get the less likely (percentage wise) it becomes to fail. 3+ to 2+ means 50% less failures. 4+ to 3+ you go from a 50% fail chance to a 33% chance to fail. 3+ to 2+ it's 33% to 17,5%. 

So, while that looks like it's the same it's actually not. If you take variance into account and accept that you don't necessarily roll according to statistics you can stretch the odds quite a bit.

Rule of thumb is: The higher the number of dice the better it is to go for bigger successes. The lower the number of dice the more important it is to avoid failure. In combat & particularly with saves you also have to account for wounds but the general idea stays the same, it just gets another layer.

[Dankboss]

I had my first game of 3.0 yesterday. It took about 4 hours to get to grips with it.

Troggoths (not Gloggs Megamob) vs Legion of the First Prince on the new Shifting Objectives.

Straight off the bat the game was back and forth thanks to Battle Tactics, which is a great addition. The smaller table meant things happened faster, and I had no issue maneuvering my Troggs across the battlefield. All Out Defense helped keep everyone's models on the table for longer which is just a positive, although my Troggs have no issue breaking armor in general. Loads of CP on both sides and reactionary abilities made the game so much more engaging for both players. The game went surprisingly well for a first game and I won 25-24, which is remarkably close all things considered.

When I first read about them I thought battle tactics were a bit meh, but having played them they're both easy enough to achieve and not intrusive to your playstyle that they're a welcome addition, and definitely lets an army that can't keep up on objectives to catch up, which in this case was Legion of the First Prince because Troggoths hitting on 2+ tear through everything.

Grand Strategies are probably too easy to score and so cancel themselves out, although I nearly managed to stop my opponent's Prized Sorceries after assassinating a Gaunt Summoner with a teleporting Hag vomit and nearly killing Be'lakor, who escaped thanks to Heroic Recovery (which is another great addition to keep your Heroes in the thick of it, since everyone wants their toys staying on the table).

Be'lakor can really put in work this edition and nearly took out my Hag immediately with +1 to hit and wound, although she just survived.

For any Gloomspite players reading, I feel like Troggoths are the way to go this edition, if you don't already.

[Boar]

1 hour ago, whispersofblood said:

This also explains why boosting a 3+ to a 2+ is "better value" than a 4+ to a 3+ while the straight probable gains for spend remains the same

You constantly get math wrong. Not too mention that your use of words like influence/value is lax that I am not really sure what you are referring too. Please Look at @Rachmanipost.

1 hour ago, whispersofblood said:

Think of it this way; you roll 12 dice on a 4+ and roll 6 success, but when you check your '3' you would have only gained 1 additional success because the majority of the failures (of which there were a statistically average amount) were '1's, and '2's. The benefit you would have gained is small for your spend, despite appearing average in the instant. 

You would gain 2 successes on average there. So again bad math. Unless you wanted to show something else in which case there is nothing supporting that line of tought there.

1 hour ago, whispersofblood said:

The value isn't in the particular gains (+1 to hit, -1 SV) it's in the relative difference in the effect of the spend. This relative difference matters in game, because dice rolls are instantaneous and not statistical averages.

If f.ex. one side boosts their elite attacks 3+ to hit, and -1 rend, and defender boosts their saves from 4+ to 3+ than damage inflicted on defender will be smaller.

Relatively in this - I would say common situation - all out defense gives you slightly more. And it helps you vs other attacks that may be directed against that unit in that phase too! It is absolutely in similar "power level" or "utility" as all out attack

And yes dice rolls are not statistical averages, we know. But rolls, their success rate is absolutely influenced by modifiers, and statistics gives us estimate of chances of succeding, and hence values of buffs. That some buffs will affect rolls with lower variance, and other with higher variance makes no difference in itself to outcomes across many games.

 

[Boar]

Double post,

however something was bugging me.

We are so far taking almost for granted that f.ex. 4+/4+profile will have less variance if to hit roll is buffed, than if to wound roll is buffed. Stemming mostly from smaller number of wound rolls. Even I granted it as such.

This is distribution of attacks with either wounds or hits buffed by one. 5000 sequences of 20 attacks each.

And one more:

Third's the charm:

It is another psychology thing, it sorta makes sense initially, untrue in the end.

[Rachmani]

We never took that for granted! It’s exactly like your graph says. The order makes no difference.

Only when the order gets changed or steps left out it does. Like with mortal wounds on 6+. Here hit is better than wound.

Edited July 12, 2021 by Rachmani

[whispersofblood]

4 hours ago, Andrew G said:

Recounting my 3.0 games from the perspective of a fairly competitive Big Waaagh! player (I win ~75% of the 15-25 local events, score in the top 25% at least at GTs).  Writing this for fun, rather than any serious, reasoned criticism of the new edition as I'm sure some salt will seep through.

Caveats:

1.  All my experience in 2.0 was Mawcrusherless, board/objective control style, Big Waaggh! I wanted to try out a Mawcrusher Build in IJ for my first 3.0 games
2. I'm rusty from the pandemic + playing other table top games
3. I know the warclans book is coming out soon.

Game 1 vs. Lumineth

Played against a 1-drop Teclis+30 Archers+Cathallar+stuff to stand in front of the before mentioned things. 
Lumineth T1: Shoots all my warchanters off the board and buffs his army (very hard to hide with shortened board + no-LOS restrictions).
IJ T1: Sacrifice a unit 5'ardboyz to unleashed hell. Mawcrusher and a unit of goregruntas get engaged. Net result, 10 spears dead and 0 archers (Archers took 0 damage from 3 goregruntas).
Lumineth T2: I ignored a few of the spell directed at Mawcrusher, but "reroll shooting against target" spell attacks goes through. Archers delete mawcrusher in one turn (this is the defensive build - Ironclad, 5+ Ward, ignore spell on 4+), teclis and spears deal with engaged gore gruntas.
IJ T2: Sacrifice another 5'ardboyz to unleash hell, goregruntas+ a different unit of 'ardboyz sneak around screens and hit archers. I actually kill 6 this time.
IJ T3: All I had left was a unit of Brutes, a larger unit of 'ardboyz engaged with his archers, and a couple 5 man units and wizards. I was ahead on objective points, so figured I'd play it out. I end up not doing much. 

Lumineth T3: Pretty much tables me. 

Take-aways:  I have no idea how archers slipped through cracks this edition. They were already good, and just got better with the introduction of unleashed hell, shortened board, more levers to pull to increase defensiveness. Scenario played a big role, as he was able to castle and remain in protection of Teclis with his entire army while contesting half the objectives. 

Game 2 vs. Slaves to Darkness: coming once I get some more spare time at work. 
 

Curious what was your actual list. I've been noodling with my Warclans and keep pondering Boarboyz. 

@Boar @Rachmani

Rachmani is closer to what I'm saying, I'm clearly not explaining it very well. I'm not arguing the mathmatical significance of +1 to an averaged D6 roll, and it's commutative effect, but I understand how you might think that I am. 

I'm trying to describe the difference in value (the unit of measure for the total benefit of using 1 CP) of using All out Attack vs the value of All Out Defence. These abilities create two pools of dice marginal failures( failures close enough to become success after +1) and true failures (failures too large to become successes. The worse your stat, the more true failures you have, thankfully a '1' is almost always a true failure.

Next submission, historically our math has ignored the distribution of the specific numbers the die roll, as insignificant. A success was a success and a fail was a fail. If you had a 4+ save it didn't matter if your fails were 3-2-2-1 because they all failed. That isn't true any longer. The distribution of fails matters because of the previous distinction. Think of it this way, the Average of 3d6 is approx 10, which can be be rolled in a number of combinations. But, what if suddenly some of those combinations were special, regardless of the total value. That is where we are, where actually what is more useful is improving the total the dice roll rather than the success/failure of individual die rolls. Because a higher total means more natural success or more marginal failures once you look at the rolls. 

Because we have the ability to buy +1, we now must consider the distribution (the exact facings that make up the combinations of success and failures) and the potential variance of those exact facings that constitute failures generally. This is because you can spend CP to turn a marginal failure into a success. 

 Turning as many failures into success as possible, which you can do by spending your CP on the largest dice pool. For the single spend you have the opportunity to touch many more individual dice. Bigger pool also means less volatile outcomes, and more predictable amounts of marginal failures to turn into successes, but also more room for positive variance because you are rolling more dice, and less room for negative variance because your floor includes smaller numbers than before. 

@Rachmani explains the difference quite well better than I would have. All out Attack is increasing the rate of success of the largest pool. All out Defence decreases the rate of failure of the smallest pool, and because saves generally in AoS are around a 4+ the distribution of failures can be anywhere between a 3 which is a marginal fail or a 2 and a 1 which are true fails, and the smaller the dice pool the more vulnerable you are to irratic distribution due to small sample sizes. 

Because you have to do most of this calculation blind to the actual facings you are going to roll it makes the choice of investing much less straight forward than simply improving how many facings result in a success. 

If I was looking to invest 1 CP into a defencive action I think it would go: redeploy, Inspiring Presence, All Out Defence generally speaking. 

 

Edited July 13, 2021 by whispersofblood

[Kadeton]

11 hours ago, whispersofblood said:

It's not cummulative, while calculating averages we are actually calculating the chance of a single attack on a profile being successful, and then imply the multiplication required for the profile.

I said commutative, not cumulative. Learn some maths before you spout off about maths. I couldn't follow much of that word salad, but your concepts of "influence" and the "stack" are just flat wrong when talking about probability.

[Mutton]

2 hours ago, whispersofblood said:

I have no idea how archers slipped through cracks this edition.

If they were willing to put something like Unleash Hell into the game in the first place, it's safe to say they're still clueless on just how powerful shooting in the game is right now. But Sentinels specifically, yeah, it's unbelievable that they and Skinks are still so bloody cheap.

[stratigo]

1 hour ago, Mutton said:

If they were willing to put something like Unleash Hell into the game in the first place, it's safe to say they're still clueless on just how powerful shooting in the game is right now. But Sentinels specifically, yeah, it's unbelievable that they and Skinks are still so bloody cheap.

ironically, as a KO player, shooting feels much weaker because of how saves work now and the loss of reroll triumphs.

 

It really is just mortal wound shooting best exemplified by sentinels that shine through.

[Zappgrot]

11 hours ago, Andrew G said:

Recounting my 3.0 games from the perspective of a fairly competitive Big Waaagh! player (I win ~75% of the 15-25 local events, score in the top 25% at least at GTs).  Writing this for fun, rather than any serious, reasoned criticism of the new edition as I'm sure some salt will seep through.

Caveats:

1.  All my experience in 2.0 was Mawcrusherless, board/objective control style, Big Waaggh! I wanted to try out a Mawcrusher Build in IJ for my first 3.0 games
2. I'm rusty from the pandemic + playing other table top games
3. I know the warclans book is coming out soon.

Game 1 vs. Lumineth

Played against a 1-drop Teclis+30 Archers+Cathallar+stuff to stand in front of the before mentioned things. 
Lumineth T1: Shoots all my warchanters off the board and buffs his army (very hard to hide with shortened board + no-LOS restrictions).
IJ T1: Sacrifice a unit 5'ardboyz to unleashed hell. Mawcrusher and a unit of goregruntas get engaged. Net result, 10 spears dead and 0 archers (Archers took 0 damage from 3 goregruntas).
Lumineth T2: I ignored a few of the spell directed at Mawcrusher, but "reroll shooting against target" spell attacks goes through. Archers delete mawcrusher in one turn (this is the defensive build - Ironclad, 5+ Ward, ignore spell on 4+), teclis and spears deal with engaged gore gruntas.
IJ T2: Sacrifice another 5'ardboyz to unleash hell, goregruntas+ a different unit of 'ardboyz sneak around screens and hit archers. I actually kill 6 this time.
IJ T3: All I had left was a unit of Brutes, a larger unit of 'ardboyz engaged with his archers, and a couple 5 man units and wizards. I was ahead on objective points, so figured I'd play it out. I end up not doing much. 

Lumineth T3: Pretty much tables me. 

Take-aways:  I have no idea how archers slipped through cracks this edition. They were already good, and just got better with the introduction of unleashed hell, shortened board, more levers to pull to increase defensiveness. Scenario played a big role, as he was able to castle and remain in protection of Teclis with his entire army while contesting half the objectives. 

Game 2 vs. Slaves to Darkness: coming once I get some more spare time at work. 
 

I have no idea what gw was thinking whit sentinals either. This 30 inch range ingore los ignor opponets armour to mortal wounds bs Is jus the mother of all anti fun designs. It's just impossible to make a fun and competative unit out of them now. Since there is no counterplay.  Gw could make them to cheap (as in as they are now) and it's no fun for the opponent. Or they could make them to expensive and then it would be no fun for the Lumineth player. It's just the mother of all bad ideas to make something strong ignor los and long range at the same time. 

[Zappgrot]

11 hours ago, whispersofblood said:

 

Which is why my expanded explaination of the position included the increased impact of additional rules. 

But fundamentally the underlying issue of cost becomes relevant. 1 CP to create situation which decreases the detriments of variance in large pools, while retaining the benefit of variance vs. 1 CP for improved success rate + variance across a small pool(something like 45% of the original pool of a 3+/3+ profile).

This also explains why boosting a 3+ to a 2+ is "better value" than a 4+ to a 3+ while the straight probable gains for spend remains the same, the room for variance in the actual performance undermines the particular gains from the 1 CP spend. Think of it this way; you roll 12 dice on a 4+ and roll 6 success, but when you check your '3' you would have only gained 1 additional success because the majority of the failures (of which there were a statistically average amount) were '1's, and '2's. The benefit you would have gained is small for your spend, despite appearing average in the instant. 

The value isn't in the particular gains (+1 to hit, -1 SV) it's in the relative difference in the effect of the spend. This relative difference matters in game, because dice rolls are instantaneous and not statistical averages. This generally isn't relevant because we haven't run into magnitude of failure dice rolls, so we haven't needed to consider them. As such averages have been a suffeciently deep consideration. Now we have an on demand ability to turn marginal failure into success, which means we have to consider how the variation of specific values before we can acertain the value of the ability/spend.

There will be situations where even attacking second you will net a larger advantage by going All out Attack yourself rather than All out Defence based on the deeper probabilities. 

Wait until we start having conversations about the fluxuation on the value of a on CP across phases and factions. 

 

What? Who the hell looks at  absolute probability gains when calculating something to begin whit?  Doubling your money from 1 dollar to to 2 is not the same  impact wise as from 1.000.000 to 2.000.000 ?  For all the impact certain buffs can have on the spread of possible results  ( is this what you mean by relative impact? ).  Most buffs have such high impact on the expected outcome that the spread becomes less relevant. That being said the larger the poll of dice the less the spread of the outcome. So if you want a big chance for a result that difference greatly from the statistical average less dice is where it is at.  So your whole point makes absolutly no ****** sense to me. But maybe that's just me 

Edited July 13, 2021 by Zappgrot