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 was actually out of the loop during the SM release. It was right around a period of high stress at university and I just didn't have the time to get the codex. Thus, I missed the initial "OOH AHH" and "THOSE UNITS SUUUUCK" that is associated with every new toy's release.

But, it's a new year, and 40k has forgotten its SM problems for the greener pastures of, well, Tyranids.

Recently I ran a game with a 6-man centurion devastator squad deathtar, two 4-man centurion devastator squad deathstars, kantor, tigurius and 3 squads of scouts. I was amazed at the massive overkill that is associated with centurion devastators. Alas, I completely overlooked one key thing, when taking my Grav Cannon/ Grav Amp Missile Launcher centuriosn, with perfect timing and prescience. That's something I'm noticing could better my actual army, with a Pedro-kantor focus on sternguard.

I have one unit in that army, a 10 man devastator squad. I decided to do a comparison of the two.

My current devastator firebase is as follows:

10 Devs
4 ML
Drop pod (to make up the number of drop pods 3.)

devs hide behind ADL for 4+ cover save, boltgun guy mans icarus lascannon.

this costs 235 points + ADL

I could move the drop pod to another unit and take a unit of 3 Devastators for 280 points. this would give me:

3 TL lascannons
3 TL Missile Launchers

I could hide this in a bastion, or behind an ADL (I haven't read the rules regarding this just yet, so forgive me if my rules are wrong on the bastion bit. I will update when I've done that reading), have it sit with kantor and be my firebase.

What are the benefits, and what are the downsides, I suppose must be the second question

Devs
More bodies, ablative wounds.
BS5 on one guy
can overwatch
more attacks when I hide kantor in the unit
29%

Centurion Devs:
S/T5 with a 2+ save and 2 wounds. Never fear a helturkey again!
can hide kantor effectively, so that he can provide BS5 for the icarus.
More, more accurate firepower.
No fething around with wasted boltguns


In my mind, the price difference is negligible, and when hidden in a bastion the body lack wouldn't matter, especially with a 3+ cover save to protect against AP2. but the extra attacks are golden. If I can, I'll hide the devastators in the bastion and have a 5-man unit with kantor to man the ICL.

So, thoughts?

Edit: I brain farted the math.

I'm a fan of

Tigurius
6 Cent Devs w/ ML, Grav-cannon

Tac Squad x5 w/ Drop Pod, Locator Beacon
SPAM
...


The idea being you roll for gate of infinity and use the Pod beacons to never scatter as you warp around the field dealing an insane amount of damage. It has a lot of weaknesses like lacking an invuln which can be easily fixed by allying Dark Angels with a Librarian with Terminator Armor and a PFG.

Iron Hands chapter tactics,

5 Cents, with grav cannons and missiles if the points allow, 1 librarian with mastery lvl 2 if the points are available and a Land Raider Crusader.

endlesswaltz123 wrote:Iron Hands chapter tactics,

5 Cents, with grav cannons and missiles if the points allow, 1 librarian with mastery lvl 2 if the points are available and a Land Raider Crusader.

I've run them with IH chapter tactics when I ran 14, however in this instance I don't have that choice. I'm already taking Pedro Kantor, forcing me to take IF rules.

Also, Grav cannon/grav amp with ML doesn't work for me because I need range. That said, I've seen what Tigurius, 6 IH cents and an IH tank chapter master can do. It is both glorious and evil at the same time.

S.K.Ren wrote:I'm a fan of

Tigurius
6 Cent Devs w/ ML, Grav-cannon

Tac Squad x5 w/ Drop Pod, Locator Beacon
SPAM
...


The idea being you roll for gate of infinity and use the Pod beacons to never scatter as you warp around the field dealing an insane amount of damage. It has a lot of weaknesses like lacking an invuln which can be easily fixed by allying Dark Angels with a Librarian with Terminator Armor and a PFG.

Again, I've run this before (not using gate). Something for you to consider, instead of spamming Tac squad and drop pods is Farsight from Tau. He's a BB for marines and he gives no-scatter deepstrike if he's your warlord. He also give you a small bonus to combat and some shooting. Of course, you have to go marine allies, but that's not really a problem. tigurius with scouts and then centurion devastators, with a tau army to support would be very devastating.

Back in fifth ed codex marines, I used to run a libby with GOI and null zone in a farsight bomb. after an assault one time, I hit and run 13" and then GOI 24" to contest an objective and to claim linebreaker. It was awesome.

Take severin loth with red scorpions for cheaper guarenteed GoI.
His rules are free on forgeworld.

Well, there is a reason why most people run grav cannon Cents; but that doesn't mean LC/ML is terrible. I certainly don't think they're overcosted or anything.

Point being, your idea is fine. It may not be the best use of DevCents, but they'll still be effective.

 The Shrike wrote:
Well, there is a reason why most people run grav cannon Cents; but that doesn't mean LC/ML is terrible. I certainly don't think they're overcosted or anything.

Point being, your idea is fine. It may not be the best use of DevCents, but they'll still be effective.

If you run LC/ML, you should be using Imperial Fist CTs. LC/ML and Tank Hunter means that you have a decent chance at killing a Land Raider without melta.

Ok yeah, Tau with Tigurius/Cent Devs and Gate is insane. You don't have to waste points with Drop Pods and you get the benfit of Tau, y'know, being Tau.

Chance of Tigurius getting gate = 75%. Chamce of Loth getting gate = 100%... just saying...


Automatically Appended Next Post:
Chance of rolling gate if you go for prescience as one of your rolls = 5/9.

The Shrike wrote:Well, there is a reason why most people run grav cannon Cents; but that doesn't mean LC/ML is terrible. I certainly don't think they're overcosted or anything.

Point being, your idea is fine. It may not be the best use of DevCents, but they'll still be effective.

I've seen Dev Cents with Grav cannons in action, and I know they decimate. But three of them footslogging (and footslogging is the best I can do with them) down the field is going to get chewed up before it can do anything. The other part of it is, well, my army doesn't need Close Quarters shooting. It's a Pedro Kantor Drop Guard army. What I do need, is an effective way to deal with armour and monsters after I've used up my sternguard's combi-weapons by the third turn.

McNinja wrote:

 The Shrike wrote:
Well, there is a reason why most people run grav cannon Cents; but that doesn't mean LC/ML is terrible. I certainly don't think they're overcosted or anything.

Point being, your idea is fine. It may not be the best use of DevCents, but they'll still be effective.

If you run LC/ML, you should be using Imperial Fist CTs. LC/ML and Tank Hunter means that you have a decent chance at killing a Land Raider without melta.

As I've said like seven times by now, This is kantor so of course I have IF CT. Also the chance of getting a wrecked land raider from 3 ML and 3 LC with full re-rolls is about 26%, with 1.6 glancing and .8 penetrating hits per turn. (You should wreck it in two turns.)

Poly Ranger wrote:Chance of Tigurius getting gate = 75%. Chamce of Loth getting gate = 100%... just saying...

Automatically Appended Next Post:
Chance of rolling gate if you go for prescience as one of your rolls = 5/9.

Neither parts of your math make sense.

Tigurius has 3 chances to get gate, Two if you go Prescience.

This gives you a 16% chance with re-rolls, twice.

That's about a 51% chance, not a 75%. Are you just pulling numbers out of your arse? what's the story?

Also, does Loth get access to Centurions? Because if he doesn't he's not as useful as tigurius because he has to be your ally and thus, disallows you from taking farsight for the flawless deepstrike

Edit: spelling.
Edit Edit: Funnily enough, my math was off. (forgot it was only two rolls, not three.)

If rolling 3 dice on the table you have a 50% chance of not getting the power. If you don't get the power, you roll all three again, another 50% chance of not getting the power. Chance of not getting the power is 1/2 and 1/2. In probability and means multiply. Chance of not getting the power is 1/2 x 1/2.
When multiplying fractions you multiply the numerators together and you multiply the denominators together. This gives you 1/4.
Chance of all mutually exclusive events = 1 in probability. You either get the power or you don't, so 1 - 1/4 = 3/4.
Now we divide 3 by 4 to get a decimal. 4 does not go into 3 so we carry the 3 over onto a place holding 0. 4 goes into 30 7 times with 2 left over which we carry onto a place holding zero. 4 goes into 20 5 times. Giving us an answer of 0.75.
Now we convert this decimal into a percentage by multiplying by 100. To multiply by 100 we move the decmal place 2 places to the right (technically the decimal doesnt move - the digits do , but lets imagine we are moving the decimal for the sake of clarity.) This gives us 75% chance.
Another method of converting 3/4 to a percentage is to find the multiplier which makes 4 into 100 (as we want to get the denominator to be 100 since 'percent' means out of 100). This multiplier is 25. Now when finding equivalent fractions we must keep the numerator and denominator balanced. In other words 'what ever we do to the top we must do to the bottom and vice versa'. Therefore we must multiply 3 by 25. This gives us 75. So our fraction is 75/100. Again percent means out of 100 so we can just get rid of the denominator now it is 100. Giving us 75%.

I hope I explained this as well as I do to my pupils who I teach maths to every day. If not I apologise, it is 12:30 here after all.

So no not 'out of my ass'.


Automatically Appended Next Post:
I can go into the 5/9's if you like? You will have to wait until tomorrow though as I need sleep for an early start tomorrow (pupils 1st day back... they are going to be wild).

Haha, he done got told!

Poly Ranger wrote:

I can go into the 5/9's if you like? You will have to wait until tomorrow though as I need sleep for an early start tomorrow (pupils 1st day back... they are going to be wild).

Patronising someone who clearly also understands statistics is a very odd move. Especially since you didn't read the rules before pretending to have all the answers.

Page 418 of the big Rulebook gives you the rules on how to generate psychic powers (I don't know what page it is in the small one, so you'll have to find it yourself if you use that one.)

The rule reads...

Pg 418 of the BRB Generating Psychic Powers, third and fifth paragraphs wrote:
... to Randomly generate a psychic power, first choose one of the psychic disciplines known to the psyker. then, roll a D6 and consult the chosen psychic discipline...
... if the psyker needs to generate more than one psychic power, repeat the above process until a number of psychic powers have been generated equal to the psykers mastery level.

The rules very clearly do not give you the option of rolling all three powers at once. Therefore, you roll your first power, you have a 30.6% chance of succeeding. You roll your second power, it gives you another 30.6% chance of succeeding. Finally, if you want to, you roll your third power, with - you guessed it - a 30.6% chance of succeeding. This gives you an overall success rate of 67%*. (1-.306 = .694, .694*.306 = .212, .694-.212= .482, .482*.306=.147; .306+.212+.147 = .665) 30.6% being the chance of you getting the number you want on a D6, if you re-roll a failed attempt. (.1667+((1-.1667)*.1667)=.306


Next time you want to be a condescending prick, make sure you understand the whole context of your topic. You can't know everything about something if you don't, y'know, know the context of the statistics you're about to spurt out.


Automatically Appended Next Post:
Now let me be a condescending prick back.

Poly Ranger wrote:
If rolling 3 dice on the table you have a 50% chance of not getting the power. If you don't get the power, you roll all three again, another 50% chance of not getting the power. Chance of not getting the power is 1/2 and 1/2. In probability and means multiply. Chance of not getting the power is 1/2 x 1/2.

except where three rolls does not give you a 50% chance of getting any power, it just means you'll get half of them.

When multiplying fractions you multiply the numerators together and you multiply the denominators together. This gives you 1/4.

Really! I didn't realise we still needed to work in fractions and couldn't just work in decimals, y'know, like a normal, functioning adult.

Chance of all mutually exclusive events = 1 in probability. You either get the power or you don't, so 1 - 1/4 = 3/4.

Please explain to me why you didn't go 4/4 - 1/4? This is the correct way to write out that expression, isn't it? Or is it acceptable to take shortcuts when "enlightening" the rabble?

Now we divide 3 by 4 to get a decimal. 4 does not go into 3 so we carry the 3 over onto a place holding 0. 4 goes into 30 7 times with 2 left over which we carry onto a place holding zero. 4 goes into 20 5 times. Giving us an answer of 0.75.

So you just re-wrote your problem, to prove that the erroneously concieved math you used is in fact correct? I didn't need proof to find that .5*.5=.25. I needed proof to find that the numbers were, in fact, .5 and .5. Which they aren't. They're .27, .27 and .27.

Now we convert this decimal into a percentage by multiplying by 100. To multiply by 100 we move the decmal place 2 places to the right (technically the decimal doesnt move - the digits do , but lets imagine we are moving the decimal for the sake of clarity.) This gives us 75% chance.

Again, I clearly already understand the basic concepts of percentages. I've already given you several percentages, and my posts are riddled with statistics.

Another method of converting 3/4 to a percentage is to find the multiplier which makes 4 into 100 (as we want to get the denominator to be 100 since 'percent' means out of 100). This multiplier is 25. Now when finding equivalent fractions we must keep the numerator and denominator balanced. In other words 'what ever we do to the top we must do to the bottom and vice versa'. Therefore we must multiply 3 by 25. This gives us 75. So our fraction is 75/100. Again percent means out of 100 so we can just get rid of the denominator now it is 100. Giving us 75%.

Great. So you've given me three different proofs to prove that .5*.5 does, indeed equal .25. I'm glad we're on the same page. However, this still doesn't help us, when none of the values are relevant.

I hope I explained this as well as I do to my pupils who I teach maths to every day. If not I apologise, it is 12:30 here after all.

Oh, I get it! you're a math teacher! What an excellent paradigm you have there, being condescending to your "Students."

So no not 'out of my ass'.

So, very "'out of my (your) ass'.", since you didn't bother to check the rules before making your claim.

Automatically Appended Next Post:
I can go into the 5/9's if you like? You will have to wait until tomorrow though as I need sleep for an early start tomorrow (pupils 1st day back... they are going to be wild).

No, I know that 5/9 is 55%. Just like I knew 3/4 was 75%.

Mate, it was very clear before you spoke that I understand the math we're using. You and I both know that. I asked you where you were pulling .75 from because your math didn't align with mine, not because I didn't understand the concept of 75%. I apologise if you were upset by me using that turn of phrase; maybe that was wrong of me, but that doesn't change the fact that you were wrong in one place - your rules. If you had simply said "3 rolls with re-rolls for a 1/6 chance is equal to 75%", or something equivalent, I could've pointed out your mistake, we could've both moved on without an issue without me now thinking you're a prick.

Also, if it wasn't clear, the bold is my comments.

Have a lovely day!

Various Edits:
Downtoning and spelling, also making things a little more clear where they were poorly written.

Lets see the odds of getting a specific power with 3 chances and rerolls and no Primaris is:

1/6=6/36; 1/6 chance 1st try (duh )
+5/6*1/6=5/36; 11/36 chance 1st try w/ re-roll
+5/6*5/6*1/5=5/36; 4/9 chance 2nd try
+5/6*5/6*4/5*1/5=4/36; 5/9 chance 2nd try w/ re-roll
+5/6*5/6*4/5*4/5*1/4=4/36; 2/3 chance 3rd try
+5/6*5/6*4/5*4/5*3/4*1/4=3/36; 3/4 chance 3rd try w/ re-roll.

Note this is the total probability of all rolls leading up to each result. Hopefully that's obvious.

Though if FW has a IC that gets GoI 100% of the time, I may be inclined to go with him though re-rolling failed Psychic tests can go a long ways.

 S.K.Ren wrote:
Lets see the odds of getting a specific power with 3 chances and rerolls and no Primaris is:

1/6=6/36; 1/6 chance 1st try (duh )
+5/6*1/6=5/36; 11/36 chance 1st try w/ re-roll
+5/6*5/6*1/5=5/36; 4/9 chance 2nd try
+5/6*5/6*4/5*1/5=4/36; 5/9 chance 2nd try w/ re-roll
+5/6*5/6*4/5*4/5*1/4=4/36; 2/3 chance 3rd try
+5/6*5/6*4/5*4/5*3/4*1/4=3/36; 3/4 chance 3rd try w/ re-roll.

Note this is the total probability of all rolls leading up to each result. Hopefully that's obvious.

Though if FW has a IC that gets GoI 100% of the time, I may be inclined to go with him though re-rolling failed Psychic tests can go a long ways.

I don't understand what this math accomplishes?

It isn't a "roll this one, assume it doesn't succeed so roll this one" scenario. It's a "calculate the chance of this one succeeding, then remove that amount when you calculate the next ones chance of succeeding." scenario.


It IS NOT .25+.25+.25. It is .3056+.2122+.1474.

The chance is linked to the last one's chance of succeeding. It has to be. Otherwise, they're three entirely different rolls.

You don't calculate BS4 with re-rolls as 12*.667+12*.667, why is this any different?

What are you talking about?
Initial chance is 1/6

1/6 times you get the power you want and move on
5/6 times you re-roll that 1/6 chance
1/6 times you get the power you wanted on the reroll
5/6 times you keep the power you didn't want and generate Power 2
1/5 times (because you have 1 of the powers already and did not take the Primaris) you will get the power and move on
4/5 times you re-roll that 1/5 chance
1/5 times you will get the power you wanted and move on
4/5 times you keep the second power you didn't want and proceed to generate power 3
1/4 times you get the power and move on
3/4 times you reroll that 1/4 chance
1/4 times you get the power you wanted
You are now out of powers to generate

You can even write it like this
1/6+5/6(1/6+5/6(1/5+4/5(1/5+4/5(1/4+3/4(1/4))))) =3/4
Its still the same chance overall.

And yes you do calculate To-hit the same way. BS4 TL weapon has a 2/3 chance to hit and a 1/3 chance to re-roll that 2/3. 2/3+1/3*2/3=8/9

What I'm trying to say is I think you are misunderstanding what the probability means. We are finding the odds of rolling 1 specific power on a table assuming that you kept rolling on that table until you either got the power or ran out of powers to generate. On average that will be 75%. On any given roll however the probability is 1/6, 1/5, 1/4, etc, based on how many powers you already have from the table.

 S.K.Ren wrote:
I'm a fan of

Tigurius
6 Cent Devs w/ ML, Grav-cannon

Tac Squad x5 w/ Drop Pod, Locator Beacon
SPAM
...


The idea being you roll for gate of infinity and use the Pod beacons to never scatter as you warp around the field dealing an insane amount of damage. It has a lot of weaknesses like lacking an invuln which can be easily fixed by allying Dark Angels with a Librarian with Terminator Armor and a PFG.

I've run this to devastating effect. I usually go over to Divination after picking up Gate, since I can get Perfect Timing or the one that gives an invulnerable save. Awesome stuff!

Using the Drop Pod spam version my plan was to ally in a Dark Angel Librarian in Terminator armor and a PFG. 4++ for everyone, I'm guaranteed Prescience and can devote all 3 of Tigurius' powers to getting GoI.

 S.K.Ren wrote:
What are you talking about?
Initial chance is 1/6

1/6 times you get the power you want and move on
5/6 times you re-roll that 1/6 chance
1/6 times you get the power you wanted on the reroll
5/6 times you keep the power you didn't want and generate Power 2
1/5 times (because you have 1 of the powers already and did not take the Primaris) you will get the power and move on
4/5 times you re-roll that 1/5 chance
1/5 times you will get the power you wanted and move on
4/5 times you keep the second power you didn't want and proceed to generate power 3
1/4 times you get the power and move on
3/4 times you reroll that 1/4 chance
1/4 times you get the power you wanted
You are now out of powers to generate

You can even write it like this
1/6+5/6(1/6+5/6(1/5+4/5(1/5+4/5(1/4+3/4(1/4))))) =3/4
Its still the same chance overall.

And yes you do calculate To-hit the same way. BS4 TL weapon has a 2/3 chance to hit and a 1/3 chance to re-roll that 2/3. 2/3+1/3*2/3=8/9

What I'm trying to say is I think you are misunderstanding what the probability means. We are finding the odds of rolling 1 specific power on a table assuming that you kept rolling on that table until you either got the power or ran out of powers to generate. On average that will be 75%. On any given roll however the probability is 1/6, 1/5, 1/4, etc, based on how many powers you already have from the table.

Please please please use decimals. I'm finding it really hard to follow the math behind all the fractions.

Anyway, you have a chance of rolling any number between one and six, so a probability of one. You have a 30.56% chance of successfully rolling this. Your math agrees. You don't then roll against a probability of one again, you roll against a probability of 69.44, because you're adjusting for the fact that you could've already gotten this value.

You roll your first time with a success rate of 31/100. You roll your second time with a success rate of 69/100*31/100, you roll your third time with a success rate of 48/100*31/100. You need to compensate for the previous rolls or the previous rolls are not linked to the current roll and we might as well call the current statistic 93/100.

Uh, question, too. Why are we assuming you don't take the primaris?

Because if you take the Primaris you are left with an unclaimed table of powers. If you roll a power you already have, you roll again until you get a power you don't have. This doesn't count against the re-roll. Its the reason why in my math I use smaller and smaller fractions to note successful rolls. With a full table its a 1/6, with one power claimed on the table its now a 1/5, with two powers its 1/4 etc.

And the reason I use fractions is because they are exact. But if you want percentages:

16.66..% of the time you get the power you want on the first roll.
83.33..% of the time you re-roll

So far we can express this as 0 .1666.. +0.8333..*(Outcome of the reroll). It is expressed this way because the goal of this probability is to find all successful solutions to obtaining a specific power. Since the goal is solely fixed on success, we look at the odds of any subsequent rolls as if they only exist when the prior rolls fail.

16.66..% of the time you get the power you want on the first re-roll.
83.3% of the time you are stuck with a power you might not want and proceed to roll the second power.

The expression is now 0 .1666.. +0.8333..*(0 .1666.. +0.8333..*(Outcome of Second Power Roll)). Notice how the probabilities of the second roll are multiplied by the probability of failure from the previous roll? At this point we also have a power so the probabilities of the next roll succeeding will be different.

20% of the time you get the Power you wanted on the second try.
80% of the time you re-roll

The expression is now 0 .1666.. +0.8333..*(0 .1666.. +0.8333..*(0 .2 +0.8*(Outcome of Second Re-roll)))

20% of the time you get the Power you wanted on the second re-roll
80% of the time you are stuck with another power you might not want and proceed to roll the third power.

The expression is now 0 .1666.. +0.8333..*(0 .1666.. +0.8333..*(0 .2 +0.8*(0 .2 +0.8*(Outcome of Third Power Roll)))). We now have two powers so the odds change again.

25% of the time you get the Power you wanted on the third try.
75% of the time you re-roll

The expression is now 0 .1666.. +0.8333..*(0 .1666.. +0.8333..*(0 .2 +0.8*(0 .2 +0.8*(0 .25 +0.75*(Outcome of Third Re-roll)))))

20% of the time you get the Power you wanted on the second re-roll
80% of the time you are stuck with another power you might not want and you are now out of powers to generate.

The final expression is 0 .1666.. +0.8333..*(0 .1666.. +0.8333..*(0 .2 +0.8*(0 .2 +0.8*(0 .25 +0.75*(0 .25 +0.75*(0))))))

0 is used because there are no more rolls left and thus a 0% chance of success.
The answer is still same.

Edit: Goin' to sleep. I'm willing to resume the discussion for a little bit once I wake up around 4AM PST

 S.K.Ren wrote:
Because if you take the Primaris you are left with an unclaimed table of powers. If you roll a power you already have, you roll again until you get a power you don't have. This doesn't count against the re-roll. Its the reason why in my math I use smaller and smaller fractions to note successful rolls. With a full table its a 1/6, with one power claimed on the table its now a 1/5, with two powers its 1/4 etc.

And the reason I use fractions is because they are exact. But if you want percentages:

Fractions are exact until you have to put them into practice. saying 33.333% chance means a lot more to anyone than a 2/6 chance. Also, they're harder to read.

16.66..% of the time you get the power you want on the first roll.
83.33..% of the time you re-roll

I don't give a feth about the reccurrer, it doesn't matter. if you want to have a sulk about the fact that you need to write extra numbers, go elsewhere. Or round up to 17

So far we can express this as 0 .1666.. +0.8333..*(Outcome of the reroll). It is expressed this way because the goal of this probability is to find all successful solutions to obtaining a specific power. Since the goal is solely fixed on success, we look at the odds of any subsequent rolls as if they only exist when the prior rolls fail.

We're on the same page here. it's about .3056.

16.66..% of the time you get the power you want on the first re-roll.
83.3% of the time you are stuck with a power you might not want and proceed to roll the second power.

We agree.

The expression is now 0 .1666.. +0.8333..*(0 .1666.. +0.8333..*(Outcome of Second Power Roll)). Notice how the probabilities of the second roll are multiplied by the probability of failure from the previous roll? At this point we also have a power so the probabilities of the next roll succeeding will be different.

are you really underlining every decimal recurrer? That's so adult of you!

Now hold up a minute. That bit I've italicised... You agree with me!

The probability of failing the previous roll is .6944, though. It's 1-chance of success=chance of failure! Your chance of success is .3056, so your chance of failure is .6944.

Now, there's something I didn't account for in my math - If you get the same result as the last time, you do re-roll differently (I hadn't thought about this, but you're quite correct.) unless you take the primaris (which, I guess you have assumed we won't.)

So that's why 1/5 suddenly comes in - this is where I was getting confused, because I didn't see how you jumped from 1/6 to 1/5. I think I understand the math you're using now!

Okay, so here's the equation in decimals (without the sulky attitude towards recurring numbers.)

Chance for first try:
1*(.1667+((1-.1667)*.1667)) = .3056
Chance for second try (Y)
x(.2+((x-.2)*.2)=y
Where x is (1-.3056), so .6944
.6944(.2+((.6944-.2)*.2))= .2075
Chance for third try (Y)
x(.25+((x-.25)*.25)=y
Where x is (1-(.3056+.2075))
=.4869(.25+((.4869-.25)*.25))=.1505

So then our answer is .3056+.2075+.1505 which is 66.36, right?

First of all i'd like to apologise for beimg a 'condescending prick'. I was insulted by you claiming I was 'pulling it out of my arse' which was disrespectful but no reason for me to be condescending.
Now even rolling one at once with a reroll makes no difference if rolling for one power on one table 3 times. It is still a 75% chance. It does however make a difference to the chance of wanting to get a power from a different table afterwards, because nobody will carry on rolling on that table if they already ave the power, so the 5/9 is not correct.
I do not have the time now but I will go into it when I get back from work tonight and explain why it makes no difference to the 75%. And will go into the probability of trying for prescience as well.
I do want to point out though, and I bring this up with everyone who says they don't like fractions, not just you. Everything we use is a fraction in disguise. Percentages are fractions with a denom of 100, decimals are fractions with a denom over a power of 10, whole numbers are fractions with a denom of one. Once you start thinking of all numbers as fractions it makes everyrhing in maths a whole lot easier.

Poly Ranger wrote:
First of all i'd like to apologise for beimg a 'condescending prick'. I was insulted by you claiming I was 'pulling it out of my arse' which was disrespectful but no reason for me to be condescending.
Now even rolling one at once with a reroll makes no difference if rolling for one power on one table 3 times. It is still a 75% chance. It does however make a difference to the chance of wanting to get a power from a different table afterwards, because nobody will carry on rolling on that table if they already ave the power, so the 5/9 is not correct.
I do not have the time now but I will go into it when I get back from work tonight and explain why it makes no difference to the 75%. And will go into the probability of trying for prescience as well.
I do want to point out though, and I bring this up with everyone who says they don't like fractions, not just you. Everything we use is a fraction in disguise. Percentages are fractions with a denom of 100, decimals are fractions with a denom over a power of 10, whole numbers are fractions with a denom of one. Once you start thinking of all numbers as fractions it makes everyrhing in maths a whole lot easier.

It's not that I don't like fractions, it's that reading them makes it harder to follow the math. (For me.) However, Decimals can all be expressed as either /100, or /1000 (and so on) and so are much easier to follow, even in fractional form.

2/6 is not quite the same as 33/100, but 33/100 is adequate for this kind of math (the difference is negligible when it comes to decision making, I think we can all agree on that.) 33/100 is also a lot easier to read, or 20/100, than reading math that involves a large number of varying denominators. That's why I ask for decimals.

Really, I don't think we can *accurately guess* at what the percentage chance of us getting, say, perfect timing from divination, without going into multiple trees of thought. You have to consider that the primaris power in div is pretty useful in most situations, while roughly 50% of all the powers are next to useless, even between varying armies.

So, 50% of the time with each roll, you'll continue having a 1/6 chance of getting a power, and alternatively you'll have a 1/5 chance the other 50%.

Finally, I don't consider the phrase "pulling from your arse" offensive, so I didn't expect it to offend. Since that clearly has, I apologise for being offensive.

The probability of failing the previous roll is .6944, though. It's 1-chance of success=chance of failure! Your chance of success is .3056, so your chance of failure is .6944.

And this is your mistake. You're taking the chance of failure for the whole system and applying as if it were the chance of failure for the actual dice roll.

Chance for first try:
1*(.1667+((1-.1667)*.1667)) = .3056

(0.1667+0.8333*(0.1667))=0.3056 aka
1/6
+5/6*1/6

Chance for second try (Y)
x(.2+((x-.2)*.2)=y

Where x is (1-.3056), so .6944
.6944(.2+((.6944-.2)*.2))= .2075

Ok I see where your mistake is. The probability of the failure for the second reroll is not (0.6944-0.2), it is (1-0.2) The first 0.6944 only exists as the product of 0.8333^2. The sum of the odds of each individual die roll should always equal 1. Given Chance of Success+Chance of Failure(...), Chance of Success+Chance of Failure =1

0.1667+(0.8333*(0.1667))+(.6944*(.2+(0.8*0.2)))=0.3056+0.2410=0.5466. Notice that rounding is showing its ugly head. This should be closer to 0.5555..

If we use the distributive property on this we can write the expression this way:
0.1667
+0.8333*0.1667
+0.8333*0.8333*0.2
+0.8333*0.8333*0.8*0.2
This starting to look familiar?

Could we get away from all of the mathhammer and get back the Centurions

Sorry lol. Cents are great but expensive. Lacking an Invuln you either need a DA Libby with PFG or to roll the 4++ Power under Divination.

 S.K.Ren wrote:
Sorry lol. Cents are great but expensive. Lacking an Invuln you either need a DA Libby with PFG or to roll the 4++ Power under Divination.

Or, a Skyshield Landing Pad, or I could put them in a bastion for a 3+ cover save.

I can name a handful of units that can get AP2 with Ignores Cover. Most of them are Tau. And Tau rely on Pathfinders, possibly drones to do this. My list can deal with both Pathfinders and drones.

As I said, they're a firebase, not a roving attack dog, like Grav cents are.

To me the Grav Centurians would be a good choice for a Gun Line army for dealing with "Agresive Armies"

I think with the Nid's dropping, the Gravcannon Centurions could be a great anchor for gunlines against all dem big gribblies. I actually know a salamander player who always runs at least two squads in his gunline and it decimates everyone it touches.

I have been toying around with a GravCent loadout and a few spring to mind for a WS fast army support. One is tossing them in a WS land raider (obviously) for quicker deployment via scout and getting all the power of a land raider to boot! you could run an allied DA biker librarian to stay in a nearby squad and jump to the Cents when they disembark for a 4++ and next turn divination.

Another idea is pretty much the same but with allied Ultra marines and tigurius in the land raider. maybe overkill points wise but still a scary unit.

Depending on your local ruling on shrouding (i know, some places allow it and some places don't) is to get Cypher and attach him to the unit for infiltrate and shrouding. also, maybe an inquisitor for cheap divination. cheaper than the land raider ideas!

also, attaching khan to the unit allows scout for 6"

Sorry to get back to it but I said I would explain after work how it is 75% chance whether you roll all 3 then reroll or roll 1 and reroll followed by 1 and reroll followed by 1 and reroll.
The reason is that the rolls plus their reroll are mutually exclusive events from the other rolls. In other words if you are going to be rolling all 3 rolls on the same chart, the rolls do not effect each others probability. This is because with multiplication, order makes no difference.
I will work with fractions here but you can easily calculate the decimal by dividing the numerator by the denominator to see what I am saying if you prefer.
Now to calculate the chances of getting the power we must calculate the inverse (chance of not getting it) and subtract from one.
Now this only applies if rolling all on the same chart. So if it is rolled first or 6th the other rolls will not effect it. 1/2 x 1/2 = 1/4. 1 - 1/4 = 3/4. Now I've already shown this.
However if we roll one at once with its reroll we get 5/6 x 5/6 for the first roll plus reroll, 4/5 x 4/5 for the second roll plus reroll and 3/4 x 3/4 for the third roll plus reroll. Respectively that is 25/36 and 16/25 and 9/16 of NOT getting the power in any of these rolls. To work out the probability of all these happening we must multiply them together (and means multiply). 25/36 x 16/25 x 9/16 = 3600/14400. Simplified this gives us a 1/4 chance of NOT getting a specific power if rolled for three times with rerolls each time. Therefore again giving us a 3/4 chance of getting the power (75%).
It doesn't make one jot of difference what order we do it in in this case as they are mutually exclusive events since you are going to be rolling on the same table as one event does not effect the others... and the order of multiplication doesnt matter in maths.
However if wanting to make a roll on another table based on the success/lack of success of previous rolls it DOES matter as these are now no longer mutually exclusive events as getting the roll early on effects the decision of later rolls.
But this still means Tigurius has a 75% chance of getting gate if you want it and are prepared to roll 3 times on that chart.


Automatically Appended Next Post:
The chance of getting gate and prescience is less as you can only roll twice on the required chart if you definitely want prescience (in other words you need to save a roll if going for prescience not matter what). Again we need to calculate the inverse and subtract the result from one.
Chanve of not getting gate:
5/6 x 5/6 and 4/5 x 4/5
25/36 x 16/25 = 400/900
Simplified this gives us 4/9.
1 - 4/9 = 5/9 to get gate with two rolls with reroll whilst definitely saving one for prescience.
HOWEVER
There is a chance that you may get it in the first roll and give you the chance of two rolls on the divination table. This does not change the probability of getting gate but it also gives you a 1 -(5/6 x 5/6) chance of getting two rolls on divination
1 - 25/36 = 11/36 chance of getting gate and still having two rolls (with reroll) on divination.