Unless I'm assaulting a small formation with no save (a rare occurrence) I can't imagine a time when I wouldn't want more hits. Even 12 hits on something like a devestator formation will only statistically just kill it assuming it has rhinos.

Genestealers have a shitty save so rely on overkilling their target to survive and win engagements.

2x4+ is more 'swingy', but the significantly better average is enough to ensure it will very rarely be the case that rolling 1x2+ per unit is preferable (see previous post).

If however, swingy-ness/variability per se is the problem, then 1x2+ is the way to go. Personally I think swingy-ness is fluffier for Genestealers. Utter carnage if things go their way!

Your stats are fine and that is indeed true, but the "parity point" does depend how many units you have and I think your number of units is fairly low as an example. If you increase the number of units to 5, it is now 3 hits or more. That is, both CC stats have about the same chance of getting 3 or more hits (96% for 2+ vs 95% for 2x4+). For a formation of 8 units this "parity point" goes up again to between 5 and 6 hits.

Basically what this means is that, even though 2x4+ has a rather substantially lower average number of hits than 1x2+, 8 units actually have the same (slightly higher in fact) probability of getting at least 5 hits. Depending on your perspective, you might say that's good enough or you might not. Given first strike, your likelihood of winning the assault with 5 hits will depend most on your target's armour value, but remember you have a 97% chance of getting this result, and still a very high chance of getting more. This makes them a very predicable/dependable unit. If you think you need more like 8 hits then you're playing a risky strategy using 2x4+ stealers in this way, with only a 50% chance of getting this. You are likely to conclude they are poor performers as a result. If you want 12 hits, this is incredibly unlikely.

Working out below:

5 dice, 2+

chance of 0 = 1 * ( 1/6 * 1/6 * 1/6 * 1/6 * 1/6 ) = 1/7776
chance of 1 = 5 * ( 5/6 * 1/6 * 1/6 * 1/6 * 1/6 ) = 25/7776
chance of 2 = 10 * ( 5/6 * 5/6 * 1/6 * 1/6 * 1/6 ) = 250/7776
chance of 3 = 10 * ( 5/6 * 5/6 * 5/6 * 1/6 * 1/6 ) = 1250/7776
chance of 4 = 5 * ( 5/6 * 5/6 * 5/6 * 5/6 * 1/6 ) = 3125/7776
chance of 5 = 1 * ( 5/6 * 5/6 * 5/6 * 5/6 * 5/6 ) = 3125/7776

chance of 2 or more = 7750/7776 = 1
chance of 3 or more = 7500/7776 = 0.96
chance of 4 or more = 6250/7776 = 0.80
chance of 5 = 3125/7776 = 0.40

10 dice, 4+

chance of 0 = 1/2 ^ 10 = 1/1024
chance of 1 = 10 * 1/2 ^ 10 = 10/1024
chance of 2 = 45/1024
chance of 3 = 120/1024
chance of 4 = 210/1024
chance of 5 = 252/1024
chance of 6 = 210/1024

chance of 2 or more = 1013/1024 = 0.99
chance of 3 or more = 1 - (45+10+1/1024) = 0.95
chance of 4 or more = 1 - (120+45+10+1)/1024 = 0.83
chance of 5 or more = 1 - (210+120+45+10+1)/1024 = 638/1024 = 0.62

8 dice, 2+

chance of 0 = 1/6 ^ 8 = 1/1,679,616
chance of 1 = 8 * 1/6 ^ 7 * 5/6 = 40/1,679,616
chance of 2 = 28 * 1/6 ^ 6 * 5/6 ^ 2 = 700/1,679,616
chance of 3 = 56 * 1/6 ^ 5 * 5/6 ^ 3 = 7000/1,679,616
chance of 4 = 70 * 1/6 ^ 4 * 5/6 ^ 4 = 43,750/1,679,616
chance of 5 = 56 * 1/6 ^ 3 * 5/6 ^ 5 = 175,000/1,679,616

chance of 3+ = 1,678,875/1,679,616 = 1
chance of 4+ = 1,671,875/1,679,616 = 1
chance of 5+ = 1,628,125/1,679,616 = 0.97
chance of 6+ = 1,453,125/1,679,616 = 0.87

16 dice, 4+

chance of 0 = 1/65,536
chance of 1 = 16/65,536
chance of 2 = 120/65,536
chance of 3 = 560/65,536
chance of 4 = 1820/65,536
chance of 5 = 4368/65,536

chance of 3+ = 65,399/65,536 = 1
chance of 4+ = 64,839/65,536 = 0.99
chance of 5+ = 63,019/65,536 = 0.96
chance of 6+ = 58,651/65,536 = 0.89

_________________
Kyrt's Battle Result Tracker (forum post is here)
Kyrt's trade list

@Kyrt, i don't quite follow your argument.

We should probably take the Math-hammer to another thread... (apologies all)

Adam, the only thing I'll note about your math is that I'm more interested in the odds of rolling 2 hits, or 3 hits, etc. Not the odds of rolling 2 or more hits, 3 or more hits etc.

I'm fine letting you guys debate this further but the stats are standing for the reasons I stated, regardless of how people feel/what they think. It's for the playtests at this point.

_________________
Dave

Blog

NetEA Tournament Pack Website

Squats 2019-10-17

I think this is a straw man argument. Totally wiping out a devastator formation with rhinos using kills needs about 11 hits, which is actually very unlikely at 2x4+ too. In fact very few formations (if any) can put that number of hits out with any degree of certainty. Remember that although more dice means more variance in the extremes, the probability of reaching those extremes is substantially smaller.

The fact is you don't need to wipe it out this way, getting 5 or 6 hits (translating to 3 kills or so) is going to give you a good chance of winning the assault, and there is a very high chance of getting this number of hits. And this is for marines, who with their high armour are tough targets for genestealers.

That's not to say I don't think genestealers have a problem, but I think their problem is with their armour, which makes them very dependent on the number of hits they cause and highly variable in success against targets with different armour values. I agree also though that they need all the hits they can get.

_________________
Kyrt's Battle Result Tracker (forum post is here)
Kyrt's trade list

Using 'X or more' because in most situations getting more than X is at least as good as getting X, if I consider X to be sufficient.

So, let's say I think that at least 5 hits are required to make an attack worthwhile.

If I'm to compare 1x2+ vs 2x4+ then I'll look at their probabilities of getting 5 or more, rather than exactly 5 hits.

My previous post indicates that it will never be the case that 1x2+ will be significantly more likely to get you those X or more required hits (even at low X, which is the surprising bit).

Err yeah, numbers getting a bit extreme, but maths is fun!

The basic gist is that although there's no denying that 1x2+ is likely to produce fewer hits than 2x4+, there is a question over whether it makes you less likely to win the assault.

A large enough formation can be equally likely to win an assault even with a poorer CC ability, because any hits beyond what is "needed" are inconsequential. It's the same as what your stats are illustrating, but rather than thinking of it as being "only 4% more likely to cause at least X hits", if we want to know if a downgrade to 1x2+ actually affects results it's more relevant to look at where the new statline starts to be poorer than the old. Hypothetically, if one stat line was just as good at getting 50+ hits as another stat line, it's isn't really a nerf because 50 is always enough. If on the other hand the new statline has a lower chance of getting just 2 hits, one would conclude it probably is a nerf because it is going to affect the ability to win an assault.

That point where the new stat line is worse than the old is the lowest number of hits for which the new stat line has a poorer chance of achieving them. In your example of 3 genestealers, this was 3 hits - below this, both statlines are as good as each other (they both have about the same chance of getting 0 hits, 1 hit and 2 hits). Again, I don't care if there in a 4% increase in the chance of getting 2 hits, it's such a high chance as to be inconsequential. I'm interested in knowing in what kinds of assaults does the stat change make it a worse formation.

So, my numbers were exploring what this number of hits is for larger formations (i.e. more like their starting size). As it turns out, in a formation of 8 stealers, 1x2+ has just as good a chance of getting between 5 and 6 hits as 2x4+ - this is the crossover point. So, to conclude:

1. In a formation of 3 stealers, the new statline is poorer in situations where you need 3 or more hits, and is no poorer where you only need 2 hits.
2. In a formation of 8 stealers, the new statline is poorer in situations where you need 6 or more hits, and is no poorer where you only need 5 hits.
3. The smaller the stealer formation, the more difference the stat line change makes.
4. The larger the number of hits you need to win (e.g. target formation power or armour), the more difference the stat line makes.

Does this make sense?

_________________
Kyrt's Battle Result Tracker (forum post is here)
Kyrt's trade list

It sounds like you're agreeing with me , i.e. 1x2+ and 2x4+ are more or less equivalent for lower numbers of required hits. 2x4+ gets significantly better in situations where you need more

It might be that our intuition is to mistakenly treat 1x2+ as 'similar' to 1x1+.
See second table in previous post to see how they compare.

Or maybe it's a case of incorrectly extrapolating from the 1 unit case (in which 1x2+ is more 'reliable'):

As I've always said:
- 2+ is significantly worse than the current stats
- Dave is pushing through the downgrade even though most everyone else thinks the current stats are fine, or even underpowered.

There's nothing more to be said, really. We're gonna have to show genestealers being even more rubbish than they are now in playtests.

_________________
Currently doing a plastic scenery kickstarter

2+ is always worse than 2x4+. But it's situationally a 10% downgrade and not always a 17% downgrade.

Dunno about the horde of teeth and claws, I've always considered genestealers more a precision ambush killer. Probably because they were so far above anything else in CC ability in 40k 2nd.

Situationally, as in under certain situations. Situationally 2x 4+ is also 27% better than 2+.

But on average, the most common results will fall on a balanced bell curve that shows a 17% drop in average result.

_________________
Currently doing a plastic scenery kickstarter

And my argument is that the situation where the difference is smaller is the more common one. Especially as genestealers will never use their CC value in defense.

It's not. You might as well say that the situation where the different is larger is equally common to the situation where the difference is smaller.

_________________
Currently doing a plastic scenery kickstarter

I think there are too many possible situations for us to attempt to quantify how much better 2x4+ is over 2+ in general. However we do know:

1) it is significantly better in many situations and slightly worse in a few
2) it is not, in general, somehow less 'reliable'