Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you: No adverts like this in the forums anymore. Times and dates in your local timezone. Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance. Email notifications for threads you want to watch closely. Being a part of the oldest wargaming community on the net. If you are already a member then feel free to login now.

I did some simulations of Blood Talon results a while back here: http://www.dakkadakka.com/dakkaforum/posts/list/399462.page#3365139

Calculating the probabilities explicitly can be done individually, but it's quite annoying.

The chance of wiping out 30 Boyz is practically zero. I ran simulations for a 21 man IG Powerblob, and average number of dead was about 6.5, with 0 wipeouts in 100,000 simulations. A Guardsman is equivalent to a Boy for this purpose, as the Furioso wounds both on a 2 and hits both on a 3.

That said, I wasn't accounting for the possibility of the enemy breaking and getting run down. Against a 20 Boy mob, you've got a pretty good chance of that happening, as you only have to kill 10. Against 30, it isn't happening, at least in the first round. Against the blob, it isn't relevant until you manage to get a wound on the Comissar.

Cheers Mr E. Impressive stuff.

I have always found that it is more than possible for an ork mob to klaw a blood talon dread before dying playing orks.
I struggle to fit dreads into my or any BA lists, but that is a conversation for another day.

I like mathhammer in theory. It lets me get a rough estimate of what will happen. However, luck plays a huge place in the game. For example, at my shop im the guy who will get a 50% gets hot chance instead of the 16.66666% chance. So that means if i double tap my plasmavets ill lose at least one guy. Which happens alot. About 50% of the time. Not complaining tho. Mathhammer is fun to engage in. And i think we all do a instictual mathhammer when we play. However say what you will, luck is bigger then mathhammer imo.

Rolling 22
is a chance of 0.000 000 000 000 000 075 9%
Winning the lottery in the UK(6 out of 49)
is a chance of 0.000 007 155%

Still, I have seen more people rolling terrible rolls than winning the lottery

Jidmah wrote:Rolling 22
is a chance of 0.000 000 000 000 000 075 9%
Winning the lottery in the UK(6 out of 49)
is a chance of 0.000 007 155%

Still, I have seen more people rolling terrible rolls than winning the lottery

I'd say you'd see more people rolling than lotto draws though

MFletch wrote:Or draw a 3d representation of scatter, i.e. a series of rings the height of which shows how likely they'll hit that point.

Should be easy enough, 2D6 is simple compared to blood talons.

MFletch wrote:Finally can give the fun example of 30 orks charging SS/TH assault termies. Essentially it is quite possible that orks will win the fight but also quite possible the termies will, however if you work out the result using averages you do not see this and may be bitterly disappointed when the orks actually lose

Should have all three of these done for you by this evening, unless something slows me down.


Automatically Appended Next Post:

ObliviousBlueCaboose wrote:I like mathhammer in theory. It lets me get a rough estimate of what will happen. However, luck plays a huge place in the game. For example, at my shop im the guy who will get a 50% gets hot chance instead of the 16.66666% chance. So that means if i double tap my plasmavets ill lose at least one guy. Which happens alot. About 50% of the time. Not complaining tho. Mathhammer is fun to engage in. And i think we all do a instictual mathhammer when we play. However say what you will, luck is bigger then mathhammer imo.

Overall, you'll end up getting close to these results if you play enough times. You just need to bear in mind that a 16.666666% chance can happen quite often, however had it been 5% chance you would have been very unlucky. The new system will be far more helpful.

I'd just like to point out that 30 boyz shooting and charging 10 Assault terminators will almost definitely mulch the terminators within two combat phases. Simply talking from experience here.

Jidmah wrote:I'd just like to point out that 30 boyz shooting and charging 10 Assault terminators will almost definitely mulch the terminators within two combat phases. Simply talking from experience here.

Yeah, I haven't done the full workings yet, but looking at the averages (it's best to start with basic math hammer, it's easier) 30 charging orks will kill 5 termies straight off, then the attacks back will cause around 6 deaths, then one for fearless, which isn't enough by a long way.

Just to explain my desire would be to represent scatter without BS change I would want a diagram of:

1/12'' height circle of radius 12''
2/11'' height circle of radius 11''
3/10'' height circle of radius 10''
4/9'' height circle of radius 9''
5/8'' height circle of radius 8''
6/7'' height circle of radius 7''
5/6'' height circle of radius 6''
4/5'' height circle of radius 5''
3/4'' height circle of radius 4''
2/3'' height circle of radius 3''
1/2'' height circle of radius 2''
half the total of above at the centre

So you want as much height not over terrain as possible when deepstriking.

The other boys against ss/th would require simulation, that is a lot of effort. The point being that a high number of attacks doesn't mean the deviation away from the average is small, expecting 2 out of 90 attacks to kill, which should be the average, is close to result 0 or 1 attack out of 90 being a kill.

MFletch wrote:The point being that a high number of attacks doesn't mean the deviation away from the average is small, expecting 2 out of 90 attacks to kill, which should be the average, is close to result 0 or 1 attack out of 90 being a kill.

Number of attacks reduces the standard deviation relative to the mean, but the variance can still be 'large'. The other key factor is the probability of a single success, the closer to 50% your chance of achieving X the greater the variance will be.

mean = number * probability
variance (NB: not Standard Deviation) = mean * (1 - probability)

20 shots at 0.1 chance = 1.8 variance (mean 2)
20 shots at 0.5 chance = 5 variance (mean 10)


Going back to the previous example I used,

4 missiles = mean 20/9 and variance 80/81.
20 bolter shots = mean 20/9 variance 160/81.

So both provide the same mean, but the larger number of shots results in a variance twice as large as the smaller number of shots.

Using these calculations is what I'd be doing to find a ballpark figure for probable range mid game rather than the 'standard devation' method that was linked to early in the thread. It avoids lots of square roots/squares/adding mid game, mean and variance is just multiply 2 or 3 numbers.

mean 20/9 (2) +/- 80/81 (1) hints at a likely result of 1-3
mean 20/9 (2) +/- 160/81 (2) hints at a likely result of 0-4

It may not be accurate in a strict sense , but it is good enough for mid game maths in 40k.

puree wrote:It may not be accurate in a strict sense , but it is good enough for mid game maths in 40k.

Sorry to sound like a broken record, but my new system is far quicker and far more precise, even if you only get the same answers out of it.
Just trying to help !

the quickest way is doing it instinctively in your head, I think most people who have learned the game well and can think critically will not need this, as they already have internalized what their units and the units they fight are capable of.

This would be pretty helpful for someone with a new army or just starting out though.

TBH, any mathhammer other than the most basic is a complete waste of time.

Any information important enough that you actually need to know it, can be done by basic multiplication. How many orks can a fire warrior kill while shooting?

BS3= 0.5 hits, x 0.666 chance to wound x 1 for no armour save=0.333 dead orks per shot per turn.

im2randomghgh wrote:TBH, any mathhammer other than the most basic is a complete waste of time.

Any information important enough that you actually need to know it, can be done by basic multiplication. How many orks can a fire warrior kill while shooting?

BS3= 0.5 hits, x 0.666 chance to wound x 1 for no armour save=0.333 dead orks per shot per turn.

Yes, but say you're in the middle of a battle and the orks are closing fast. If you send the kroot to help the 6 fire warriors against those 6 orks then the opponent could roll well for his "WAAAGH!" run move and possibly assault the broadsides. However, if the orks get into combat (even one of them) then they'll almost certainly be able to destroy the broadsides next turn.
The question: Is it better to risk the shooting with the fire warriors and keep the kroot where they are, or move the kroot away and hope that the enemy gets under 3 on his run move?

Normal Mathhammer: 1/3 x 12 (rapid fire) = 4 dead, you'd have to be lucky. But is it better with the 1/3 chance or the shooting?

My system: ok, it's not done yet, I needed to use my calculator a bit, but it's aproximately 0.369 chance of managing with the shooting, compared to 0.333 chance of them failing at running. So shoot, and don't send the kroot.

I'm not saying you have to use it. I'm just saying that when I'm done, it'll be better than normal mathhammer

There is math... and then there is luck... and then there is crappy dice.

If you haven't checked out Game Science dice before, it's well worth the time.

http://www.youtube.com/watch?v=bR2fxoNHIuU

p_gray99 wrote:

im2randomghgh wrote:TBH, any mathhammer other than the most basic is a complete waste of time.

Any information important enough that you actually need to know it, can be done by basic multiplication. How many orks can a fire warrior kill while shooting?

BS3= 0.5 hits, x 0.666 chance to wound x 1 for no armour save=0.333 dead orks per shot per turn.

Yes, but say you're in the middle of a battle and the orks are closing fast. If you send the kroot to help the 6 fire warriors against those 6 orks then the opponent could roll well for his "WAAAGH!" run move and possibly assault the broadsides. However, if the orks get into combat (even one of them) then they'll almost certainly be able to destroy the broadsides next turn.
The question: Is it better to risk the shooting with the fire warriors and keep the kroot where they are, or move the kroot away and hope that the enemy gets under 3 on his run move?

Normal Mathhammer: 1/3 x 12 (rapid fire) = 4 dead, you'd have to be lucky. But is it better with the 1/3 chance or the shooting?

My system: ok, it's not done yet, I needed to use my calculator a bit, but it's aproximately 0.369 chance of managing with the shooting, compared to 0.333 chance of them failing at running. So shoot, and don't send the kroot.

I'm not saying you have to use it. I'm just saying that when I'm done, it'll be better than normal mathhammer

The funny thing is math says the orks wouldn't destroy all the broadsides first turn irony.

But anyways, it is irrelevant, because it depends on the tactical importance you place on any given unit. And either way, the best choice is probably still to move the kroot towards the broadsides because they are (in most cases) a unit of great importance, and risking them at all just isn't a good idea. Move the kroot and fire with them and you can guard the broadsides will helping the FW.

If anyone has any mathhammer problems, there's an app for that on the appstore. It can calculate vehicle hits, everything, wounds caused against 50 Terminators with Storm shields against 50 railguns for example. It takes 5 seconds. No point in stupid graphs or constant halfing

Orkaswampa wrote:If anyone has any mathhammer problems, there's an app for that on the appstore. It can calculate vehicle hits, everything, wounds caused against 50 Terminators with Storm shields against 50 railguns for example. It takes 5 seconds. No point in stupid graphs or constant halfing

o.0

What's it called?

to me the most important thing is to remember the law of large number / probability convergence and central limit theorm.

they all apply here because if you have 25 or 30+ dice roll, your sample outcome convert closely to the expected value. think BA on assault +1 attack from sanguinor or eldar blade strom (you know shoot this turn next turn back into the minibus).

when the game goes on and the unit size start to reduce as your opponent start to kill of some of your guys, or if you are like me keep rolling 1s when landing into terrains; the probability get more and more random.

the games built this way so that you are in advantagous position if you have a large number of dices to roll, either throught loads of shot, unit with loads of attack on assult or simply large unit; having large number on your side you have a bit more control over randomness; and therefore give more more confident on taking down those armour 14; termis or monster creatures; that said more often then not when your number of rolls are that big you simply wipe out your enemry unit and get shot at next turn.

I like it. But for math hammer is at its best when you can use it at he table side. Estimated guesses by statistics are best at this

forgot to say as well, on probability convergence, if the probablilty of successful event is small, say, tactical marine trying to shoot and kill BA terminators whom have a priest attached, dice rolls sequence will have to roll 1 on armour save and have the chance halved again for FNP; it will take much larger number of dice before the probability to converge.

as a result, these kind of sequence, with the successful event quite far down the tail; like shooting enermy with good armour, or high toughness, larger unit will not give you that much control over the variation of the outcome; i.e. your result is still largely down to chance.

Im just curious here, did you intend for this to be used during games on not? because during game mathhammer with a tool like this seems to be taking things a little too far. its really not that important to win. its just a game and the use of a tool like a spreadsheet etc during a game seems a little like cheating. it does not seem to be in the spirit o things. I figure you can do what you want and if your opponent is ok with you using something like this mid game, then more power to them, but to me its not in the spirit of game play.

I think this is awesome p_grey.

Thats a great comparison superm. If you sat down at any legit gambling table and tried to start counting cards or doing any mumbo jumbo with a spread sheet you would get tossed out.

Anpu-adom wrote:There is math... and then there is luck... and then there is crappy dice.

If you haven't checked out Game Science dice before, it's well worth the time.

http://www.youtube.com/watch?v=bR2fxoNHIuU

Nothing I can do about that unless you want me to do 999/6000 chances. Nice to know, though.

im2randomghgh wrote:it is irrelevant, because it depends on the tactical importance you place on any given unit. And either way, the best choice is probably still to move the kroot towards the broadsides because they are (in most cases) a unit of great importance, and risking them at all just isn't a good idea. Move the kroot and fire with them and you can guard the broadsides will helping the FW.

This helps decide what is of most tactical importance, the broadsides are threatened either way, and what if the kroot shooting is needed to help with the first lot of orks? This really would help.

Orkaswampa wrote:If anyone has any mathhammer problems, there's an app for that on the appstore. It can calculate vehicle hits, everything, wounds caused against 50 Terminators with Storm shields against 50 railguns for example. It takes 5 seconds. No point in stupid graphs or constant halfing

Graphs? what's wrong with graphs? It helps.

komosunder wrote:I like it. But for math hammer is at its best when you can use it at he table side. Estimated guesses by statistics are best at this

This can be done at the table side!

sennacherib wrote:Thats a great comparison superm. If you sat down at any legit gambling table and tried to start counting cards or doing any mumbo jumbo with a spread sheet you would get tossed out.

Well, WH40k isn't gambling. This is why it's ok to bring the spreadsheet, mostly because you don't really need it during the game. If your game is going to break on needing to kill exactly that single unit, you made an error during list building, deployment and/or movment, all of which are not random. A gaming situation doesn't drop out of the sky and forces you to handle it.

p_gray99 wrote:

im2randomghgh wrote:it is irrelevant, because it depends on the tactical importance you place on any given unit. And either way, the best choice is probably still to move the kroot towards the broadsides because they are (in most cases) a unit of great importance, and risking them at all just isn't a good idea. Move the kroot and fire with them and you can guard the broadsides will helping the FW.

This helps decide what is of most tactical importance, the broadsides are threatened either way, and what if the kroot shooting is needed to help with the first lot of orks? This really would help.

That does not help you decide on tactical importance, at all. Tactical importance is the value you place on the continued existence of a unit. It is how important it is to your overall strategy.

There was only one lot of orks in your scenario, and now they are attacking the FW while getting RF'd by both kroot and FW, and are blocked from attacking the broadsides by the bubblewrap carnivores.

im2randomghgh wrote:

p_gray99 wrote:

im2randomghgh wrote:it is irrelevant, because it depends on the tactical importance you place on any given unit. And either way, the best choice is probably still to move the kroot towards the broadsides because they are (in most cases) a unit of great importance, and risking them at all just isn't a good idea. Move the kroot and fire with them and you can guard the broadsides will helping the FW.

This helps decide what is of most tactical importance, the broadsides are threatened either way, and what if the kroot shooting is needed to help with the first lot of orks? This really would help.

That does not help you decide on tactical importance, at all. Tactical importance is the value you place on the continued existence of a unit. It is how important it is to your overall strategy.

There was only one lot of orks in your scenario, and now they are attacking the FW while getting RF'd by both kroot and FW, and are blocked from attacking the broadsides by the bubblewrap carnivores.

Perhaps you didn't understand. One set of orks was in front of the kroot, and would be able to attack the broadsides. Another set of orks was about to destroy the FW and then attack the broadsides.
Also, this helps decide what is most tactically important by working out what it could do, and thus helps decide to use everything to attack the smaller set of orks, whereas otherwise the wrong decision may be made.

p_gray99 wrote:

puree wrote:It may not be accurate in a strict sense , but it is good enough for mid game maths in 40k.

Sorry to sound like a broken record, but my new system is far quicker and far more precise, even if you only get the same answers out of it.
Just trying to help !

You don't get the same answer out of it, I wasn't talking about probability of a specific result, but the likely range referencing the earlier link to using standard deviation to work out the likely range.

More importantly I wouldn't really be interested in playing someone who is using spreadsheest and calculators mid game. I woudln't do it myself and don; expect my opponent to. As someone lese said it is going a bit far for a minis wargame.

Maybe I'm just a math boob, but I don't even understand your charts when I look at them.