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.

So I'm just wondering, how do you "randomly select an enemy model"

I noticed this is said A LOT in the Necron book for various abilities... I'm always used to me selecting where wounds land on my troops when they're attacked, or I'm used to shooting at a vehicle and picking which weapon I want destroyed... but when it says "Randomly select an enemy model..." i've tried finding this in the BRB and had no luck.

Few things that do this - Mindshackle Scarabs, Transdimensional beamer (exile ray)

Terribly sorry if this is a ridiculous question, I'm sure the answer is right in front of me somewhere, I just can't find it :(

Thanks guys

It's not easy.

Total up the number of models in the enemy army.

Assign each model a number.

Then, roll a dX where X is the number of models. I suggest http://www.random.org/ to make this easier. Otherwise, get out a d100 (or two d10s) and prepare to reroll as needed.

The method used is unimportant. All that matters is that it is random, this having an equal chance of affecting all eligible models. Use whatever method works for you and your opponent.

Looks like i'll need to dig out my d30 die lol

It doesn't need to be dice of course. You could draw lots, etc.

Dice can be used in the majority of situations. Just number the models and roll the amount of dice necessary to incorporate the highest number (so for 14 models you need 3D6) and either make it so that any numbers higher (in this case, 15 through 18) count as rolling the highest possible result attributed to a model (in this case 14) or simply re-roll until you reach a valid number.

You would need to reroll. Simply letting the 15 to 18 count as 14 gives the model assigned the "14" a higher chance of being hit....thus it isn't truly random.

Using non-casino grade dice isn't truly random or generating the most random result possible either, so honestly it would depend on the game you're playing. A casual game could support either way, with the former being quicker, but a tournament would require the re-rolls.

Oppent writes down each model and assigns them a number on paper, so like

1 - Serg
2 - MS
3 - Marine
4 - Marine
5 - Marine

You pick a number with out seeing sheet and then look it up on what you got?

Best way I have found is to roll a die for each model, one at a time, and whomever gets the highest result is the random winner.

In the event of a tie, all those with the highest result re-roll.

It is fairly quick once you get the hang of it.

Or you could nominate half the unit and roll a single Die, odds left side evens right side, repeat til there is one guy left. (At most you will need like 5 rolls.)

Just use your army list. First model (usually HQ for me) would be '1' then the next unit (lets say a 6 model command squad) is 2-7 and so on. Then roll the appropriate dice.

zeshin wrote:Just use your army list. First model (usually HQ for me) would be '1' then the next unit (lets say a 6 model command squad) is 2-7 and so on. Then roll the appropriate dice.

That works for units with even numbers, but D7 are tough to come by

It really shouldn't be too hard to figure out how to randomize. If you have 6 or less choices, assign each of the targets a number and roll a dice, rerolling until you get one of the target numbers. If greater than 6, easiest way is to split the group into two or three sets of equal size, then roll to see which one is picked, then continue the process. For example

# of targets - solution
1 - you got it
2 - roll a D2
3 - roll a D3
4 - roll a D6, rerolling 5 and 6 OR roll two D2s
5 - roll a D6, rerolling 6
6 - roll a D6
7 - kind of complicated. Roll a D6. Subtract one from the roll, then multiply by 6. Roll a D6. If the result is (6,6), start over. Add the two numbers together, divide by 7. The result is the remainder of the division.
8 - roll a D2, then a D4
9 - roll a D3, then a D3
10 - roll a D2, then a D5
11 - complicated - do it like a D7, but roll a D6 for the second dice, start over on the results of (6,4), (6,5), and (6,6) and divide by 11.
12 - roll a D2, then a D6
13 - complicated - easiest would be draw a card from a deck of cards, J = 11, Q = 12, K = 13
14 - roll a D2, then a D7
15 - roll a D3, then a D5
16 - roll a D2, then a D2, then a D4.

etc. Prime numbers are hard, so I'd just use a random.org on someone's phone.

Avatar 720 wrote:Dice can be used in the majority of situations. Just number the models and roll the amount of dice necessary to incorporate the highest number (so for 14 models you need 3D6) and either make it so that any numbers higher (in this case, 15 through 18) count as rolling the highest possible result attributed to a model (in this case 14) or simply re-roll until you reach a valid number.

The problem with this method is it can be manipulated. For example for a 10 man squad:
If you roll 2d6 - 1 (so 1 is possible) the sixth guy is six times more likely to get picked than the first one. Total sum of seven on two dice is a 1 in 6 chance. Two ones is 1 in 36.

Macok wrote:

Avatar 720 wrote:Dice can be used in the majority of situations. Just number the models and roll the amount of dice necessary to incorporate the highest number (so for 14 models you need 3D6) and either make it so that any numbers higher (in this case, 15 through 18) count as rolling the highest possible result attributed to a model (in this case 14) or simply re-roll until you reach a valid number.

The problem with this method is it can be manipulated. For example for a 10 man squad:
If you roll 2d6 - 1 (so 1 is possible) the sixth guy is six times more likely to get picked than the first one. Total sum of seven on two dice is a 1 in 6 chance. Two ones is 1 in 36.

Exactly, that's why you do it like I posted above.

MasterSlowPoke wrote:

# of targets - solution

7 - kind of complicated. Roll a D6. Subtract one from the roll, then multiply by 6. Roll a D6. If the result is (6,6), start over. Add the two numbers together, divide by 7. The result is the remainder of the division.

11 - complicated - do it like a D7, but roll a D6 for the second dice, start over on the results of (6,4), (6,5), and (6,6) and divide by 11.

13 - complicated - easiest would be draw a card from a deck of cards, J = 11, Q = 12, K = 13

etc. Prime numbers are hard, so I'd just use a random.org on someone's phone.

A better way to do the prime numbers is just split them into two groups, roll a D2 to slit the group in two and then roll the appropriate die.

Something like this:

7 - D2, then a D3 or D4 depending on the D2 roll.

11 - D2 then D5 or D6 depending on the D2 roll.

13 - D2 so you will have a group of 7 and a group of 6, so either D2, then a D3 or D4 depending on the D2 roll like for a 7, or a D6.

Whatever way you do it, make sure your opponent knows what is going on.

DeathReaper wrote:

zeshin wrote:Just use your army list. First model (usually HQ for me) would be '1' then the next unit (lets say a 6 model command squad) is 2-7 and so on. Then roll the appropriate dice.

That works for units with even numbers, but D7 are tough to come by

I meant cumulative for the whole list, and do it in one roll. Say you have 74 models in your army total; you would simply roll percentile and ignore (re-roll) a result of 75-100.

The problem with rolling multiple dice and adding them together is that not all results are equally likely. 2d6 rolls 7 far more often than it rolls 12, so it's inappropriate to roll 2d6 if you want to randomly choose a model in an 11 man unit.

With the prime numbers you can't split them into 2 groups and then roll a different dice depending on the group. With the group of 7 given as an example above the models in the group with only 3 models are more likely to be selected than the models in the group of 4. You can do it by splitting into groups, selecting a group randomly and then rolling the largest dice for each group, so in this example a d4 for both. If you get an inappropriate number for the group (a 4 in the group of 3) you'll have to start over from scratch, you can't just reroll within the group or you bias the selection towards the models in the smaller group.

You know what would be simpler then that? Pick a number 1-however many choices (right it down), then your opponent picks a unit/ model and you work round clockwise (or what ever) till you get to that number ... that's the model/ unit effected.

Nether person can know the others choice so it will always be random ... and damn quicker then rolling.

http://www.gamestation.net/GameScience-White-d7

They make tons of wierd Dx's... Not that getting a ton of weird dice is in any way economic.

We decided on using cards since there are enough of them in one suit for most of our games.

DeathReaper wrote:

MasterSlowPoke wrote:

# of targets - solution

7 - kind of complicated. Roll a D6. Subtract one from the roll, then multiply by 6. Roll a D6. If the result is (6,6), start over. Add the two numbers together, divide by 7. The result is the remainder of the division.

11 - complicated - do it like a D7, but roll a D6 for the second dice, start over on the results of (6,4), (6,5), and (6,6) and divide by 11.

13 - complicated - easiest would be draw a card from a deck of cards, J = 11, Q = 12, K = 13

etc. Prime numbers are hard, so I'd just use a random.org on someone's phone.

A better way to do the prime numbers is just split them into two groups, roll a D2 to slit the group in two and then roll the appropriate die.

Something like this:

7 - D2, then a D3 or D4 depending on the D2 roll.

11 - D2 then D5 or D6 depending on the D2 roll.

13 - D2 so you will have a group of 7 and a group of 6, so either D2, then a D3 or D4 depending on the D2 roll like for a 7, or a D6.

Whatever way you do it, make sure your opponent knows what is going on.

The problem with that is each model won't have the same chance of being picked.

A group of 7 split into 2 groups, A of size 3 and B of size 4. Both of these groups have a 50% chance of being picked, so good. You then roll to pick a model out of each group - each model in group A has a 33% chance, and each model in group B has a 25% chance.

Model 1: 16.7% chance to be picked
Model 2: 16.7% chance to be picked
Model 3: 16.7% chance to be picked
Model 4: 12.5% chance to be picked
Model 5: 12.5% chance to be picked
Model 6: 12.5% chance to be picked
Model 7: 12.5% chance to be picked

The probability function is biased towards group A. Each model should have a 14.3% chance to be picked, so that method is flawed.

DeathReaper wrote:

MasterSlowPoke wrote:

# of targets - solution

7 - kind of complicated. Roll a D6. Subtract one from the roll, then multiply by 6. Roll a D6. If the result is (6,6), start over. Add the two numbers together, divide by 7. The result is the remainder of the division.

11 - complicated - do it like a D7, but roll a D6 for the second dice, start over on the results of (6,4), (6,5), and (6,6) and divide by 11.

13 - complicated - easiest would be draw a card from a deck of cards, J = 11, Q = 12, K = 13

etc. Prime numbers are hard, so I'd just use a random.org on someone's phone.

A better way to do the prime numbers is just split them into two groups, roll a D2 to slit the group in two and then roll the appropriate die.

Something like this:

7 - D2, then a D3 or D4 depending on the D2 roll.

11 - D2 then D5 or D6 depending on the D2 roll.

13 - D2 so you will have a group of 7 and a group of 6, so either D2, then a D3 or D4 depending on the D2 roll like for a 7, or a D6.

Whatever way you do it, make sure your opponent knows what is going on.

The problem with this method is that not every model has an equal chance of being selected, taking 7 for example, models 1-4 would have a 1/8 chance of being selected while models 5-7 would have a 1/6 chance of being selected.

What I've found that works for me is to carry larger dice around, I've got a d4, a few d8s, some d10s, 2 d12s, and a d20 in my dice box, with a pair of d10s, you can roll d100 which gives you 100 possible outcomes, which should be enough. If not, you can easily turn that into d200 or larger by rolling a d2, d3, d4, or d6 first.

You can also go smaller such as a d40 (40 possible results with d4, d10), or a d66 (36 possible results with d6, d6).

1/8 vs 1/6 I do not know anyone who would have a problem with this difference, it is random enough for government work.

Though if they do just roll a die for each model and the one with the highest score wins. as I have noted before.

That is totally random.

Run into the street, grab any randomer, ask them to pick a model from the squad and then let them go.

There, random.

Steelmage99 wrote:You would need to reroll. Simply letting the 15 to 18 count as 14 gives the model assigned the "14" a higher chance of being hit....thus it isn't truly random.

of course its still random,theres just a higher chance for model #14.

Tri wrote:You know what would be simpler then that? Pick a number 1-however many choices (right it down), then your opponent picks a unit/ model and you work round clockwise (or what ever) till you get to that number ... that's the model/ unit effected.

I really like this idea. Simple and elegant.

Ghaz wrote:The method used is unimportant. All that matters is that it is random, this having an equal chance of affecting all eligible models. Use whatever method works for you and your opponent.

Actually the rules don't say that. Randomness does not inherently mean each occurrence has an equal chance of happening. While that is definitely the best way to play it, effectively it is whatever you and your opponent agree upon.




Automatically Appended Next Post:

HamHamLunchbox wrote:

Steelmage99 wrote:You would need to reroll. Simply letting the 15 to 18 count as 14 gives the model assigned the "14" a higher chance of being hit....thus it isn't truly random.

of course its still random,theres just a higher chance for model #14.

Weighted chance does not preclude randomness...

But it does preclude fairness.

The best solution is to avoid including ridiculous "select a random x" mechanics in any codex ever again.

For the record, if it's not the same chance for each item, it's not random. I'd just get a d20 and roll until I got a valid result; that should be the simplest. Unfortunately, multiple d6s wouldn't work very well because then you can never get 1...

If its not equal, then its not random because one (or more) models have a greater probability of being chosen. Stacking the odds in one model's favor of being chosen is not random.