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.

Ailaros, Most of your post is spot-on, but I disagree with your point about the sample size. You're right that you don't roll nearly enough dice in a single game of 40k to be able to expect statistical averages, but there's a statistically significant difference between the number of dice you roll for your all-foot guard armies and the amount you would roll for a deathwing army. When you roll the to-hit rolls for, say, 100 lasguns, the variance is going to be much lower than the roll on, say, 30 storm bolters.

I wanted to make an unrelated point, too: horde armies tend not to care about elite equipment. To use the most extreme example, a grot dies just as easily to a bolter as he does to a plasma gun. Weapons upgrades are thus less efficient against horde armies than elite armies.

Well, I think part of the variance problem is that 40k players use statistics incorrectly. Because it's easy and obvious to calculate, most players use the "average" method. So if I roll 1d6 I'll roll an "average" of 3.5. You roll a 100 dice and sum up the result and you'll get an answer close to the average.

The problem is that the average method doesn't include variance. The real way to calculate would be to find the average +/- standard deviation. Dice creates a perfect bell curve and results are finite and known. That means if you choose to charge, you should be able to figure out that you'll kill 3 models +/- some deviation.

So if the answer is that you'll kill 3 +/- 2 models, you'll know that you have a 66% chance of killing between 1 and 5 models. Because the SD is high, you should expect extremely varied results.

In other cases, you'll find your decision has a very tiny SD.

It seems strange to me that people talk about "random", but the variances are finite and the possibilities are countable.

Edit:

I forgot to tie it back to the topic. Anyways, hordes have a statistical advantage in having more models because a higher model count necessarily leads to lower variance. You will note in the example above that the Standard deviation for 1d6 is incredibly high relatively compared to the sum of 100d6.

And boomwolf is right as well. Not every failure means a critical fail. Not every success is a critical success. In an objective mission, killing successfully may not be a critical success if there is even one model left on the objective.

Another effect one must take into consideration is the critical success/failure possibilities.

If you increase the number of dice in a way that the average remains the same you gain the fact that will critical failure remains as bad-IE 0 successes-the critical success is far greater, and while rarer with more dice, far more devastating.

That's why I love the CiB over then BC.
Having 5 S3 shots is superior to 3 S5 shots because while on average you get the name number of wounds on almost anything, while both "fail horribly" with 0 wounds, the BC "does great" with 3 wounds, while the CiB "does great" with 5 wounds. (the chance to get AP1 aside, and the odd cases of Sv 4+ targets)

Its not much, but it CAN tilt the result, more dice thrown with the same average is in a way superior.

And if you're talking about tournament play, there are also some outside factors to consider.

1. Horde armies are weak to strange variances of terrain. This is a serious problem in tournaments.

2. Horde armies take longer to deploy and play, and time is a critical factor during tournament play

But my major point in my first post still stands: The armies that find ways to completely mitigate dice do the best whether they are horde or not. Siren prince: I pass a leadership 10 test and you don't get to shoot at him. Your firepower simply becomes irrelevant and you don't get to play. THAT is what gives a particular army (or build, or list or whatever) a huge advantage.

Hesperus wrote:When you roll the to-hit rolls for, say, 100 lasguns, the variance is going to be much lower than the roll on, say, 30 storm bolters.

Sure, but there's still going to be a ton of variance for 100 lasguns. Remember, 100 is less than INFINITE by a pretty wide margin. We're so far over on the "small numbers" side of the spectrum, that anything you're going to actually see in 40k is going to be practically indistinguishable most of the time.

Also, I'd note that not all die rolls in the game are equally important. How any single die roll out of 100 lasgun shots rolls isn't likely to make a big difference. The one single roll for who goes first, or those last couple of armor saves you throw against a terminator contesting an objective, on the other hand, will.

Skim off, say, the top five most important luck-based events, the ones that will really make a difference, and you're looking at a truly tiny number of dice rolled.

Really horde armies have the greatest benefit from deviance, as any deviance only increases their points value because they normally they die anyways. So they can only really get better as their baseline is already about as bad as it can be.

Elite armies can go either way due to deviance, so they must be inherently worse. That said, the chances of them deviating positively is greater because of the higher chances on individual dice rolls. Yet the fact that they can both suffer greatly, or suffer great improvement, doesn't balance out the fact that horde armies can only really get better with dice rolls.

Going from Ork to Plague Marine armies there definetly is a difference in the skill level, knowledge of the game, and using your units to their maximum potential for one to succeed. Making a bad tactical decision or forgetting to shoot a unit or their ability with an elite army is basically wasting points you already paid for. Orks can mess up 1 shoota boy save and not really notice it, but Grey Knights can lose a 40 point terminator and lose a lot of their effectiveness. So there's that aspect of chance and human error to take into account. Maybe that's why they're called elite armies though.

I had always thought that Chaos Space Marine army was a horde army, because of the way they are portrayed in the books, pc games, and short videos. However, I am wrong, because, as people commonly say, "their troops choices are other armies' elite choices."

However, in 6th edition, the Cultist units were reintroduced and CSM players can take allies such as Daemons, IG, Cron, and Orkz. IG and Orkz can supply more bodies while Cron and Daemons can patch up the holes CSM cant deal with.

The most effective horde style army CSM could field from what I heard and seen is Typhus and hordes of Plague Zombies. I feel that the cultist lack power compared to IG where their platoons can take melta, plasma guns, flamers, grenade launcher, missile launchers, power fists, heavy bolters, etc... So the only use for a cultist is to be a tarpit. The non troops choices would deliver the heavy hitting punches that the zombies lack.

In additiion, a Chaos Space marine is 2 points less than the Vanilla marine, so there is a small potential to horde mobs of CSM. However, they lack the effectiveness of Combat squads, and they shall know no fear.

So what happens to the statistics when an army can field horde element and elite units? In my honest opinion, CSM does statistically worse unless its a plague army.

That's almost on a whole other discussion there Filch. CSM can do a very fun mix of Elite and Horde which I don't think any other codex can match in variety and play style. Mix in allies like you said and thats some very potent list potential, but the ally potential would have to be taken into account for every other army in 40k. IG with a GK detachment. Orks with Tau Crisis suits and Broadsides. Necron Airforce with ork blobs on the ground. Adding allys would be a solid way of filling in weaknesses in your army or balancing out bad rolls with redundancy or alternative units. I think that was the whole point of the Ally addition, to add variety and fill holes.

My point was that paying more for a unit and not using it to its utmost potential could be looked at as wasting points. A mob of 30 ork boyz against a 5 man Terminator squad in close combat can take more bad dice rolls than those 5 Terminators. It would only take 5 1s from the 60 odd attacks the orks throw at them. But what if those terminators had combi flamers and forgot to use them. 6-10 dead orks before that close combat would of drasticaly changed the odds of the terminators surviving.

 Ailaros wrote:

To the point, in fact, where knowing statistics is all but useless in an actual game of 40k.

Whoah. Disagree. Knowledge of statistics, probability, and general math is what wins games.

You could have a million lasguns. Any army built from
flying transports or whatever bullsh*t mechanic people come up with doesn't care because the "brokeness" simply bypasses it.

 Lord PoPo wrote:
Obviously, luck plays a large part in the outcome of the games we play, and though a truly random set of dice will, in the long run, probably even out statistically. But we've all seen the outliers, the upsets, the just plain (un)lucky.I guess my question is multi-tiered, but here it is: Are certain builds more susceptible to chance? If so, what are they? and, are there cases in which you have to consider chance when building an army list?

Any army that rolls less dice is more susceptible to chance, one way or the other. This is in accordance with the Law of Large Numbers.

This means that a deathwing army relying on a 2+ save has a greater chance to completely rock it, or completely flop than an ork army -- which is throwing out 90 dice a turn when shooting. Simply put, the ork army itself is more reliable and resistant to complete dice flops due to the volume of dice.
This does not mean that high volume armies are immune. Everyone has seen the ork lootas that shoot 30 times, hit 23, and wound 20. It simply happens less often.

Does it make a big difference? IMHO, not enough to change your playstyle for.

As a terminator player through and through, I have talked about this issue before in my group at length, with the results being that elite armies are worse than hordes when it comes to dice.

As an aside, the counter argument is that hordes take longer to play, because of more models to move and more dice to roll. This was expressed earlier in a way I quite agree with in that a good player benefits from the harder to play but less luck dependent army, while the worse player benefits from the easier to play but more luck dependent elite army. We see this crop up in video games a whole lot with certain "noob" weapons that a vet doesn't need but that a "noob" can still earn kills with. It's an important part of balance in games that focus on fun instead of ensuring the better player always wins such as in chess.

Anyway, terminators take a wound and fail an armor save 1 in 6 times. An Ork costs 1/6 the price and fails an armor save 5 in 6 times. Without any additional modifiers and with equal points, 6 ap5 bolter wounds on the termies kill 1 on average and kill 6 orks all the time. So it would seem both are balanced. However, 1 bolter wound can never kill 6 orks, but it can kill a terminator. On the other hand, a terminator has about a 33% chance to survive 6 bolter wounds, and the orks have 0 chance. Why this is important is because you know with certainty orks will never die to 1 wound and always die to 6 wounds. However, you can only make predictions about what the terminators will do. If you need that terminator to pass a save to win the game, you count on luck to win. With orks, you can play with certainty that less than 5 wounds won't wipe the squad.

In addition, a huge wrench to this equation is cover. So while up to this point, the terminators paying for their save meant the terminator and the Ork were balanced in point cost versus bolters, cover is free and give a save to orks all the time, while never giving a save to terminators. Thus, a huge part of the game only benefits the orks. This directly makes terminators worse.

After that, we get to snap fire. Elite armies tend to have higher bs scores than horde style ones. However, both fire snap shots at the same bs. A bs4 model loses 3 points of bs when making snap shots, while orks for example only lose 1 point of bs.

Finally, we get to special weapon shots. A plasma gun has no additional effect on orks beyond better wounding. However, a plasma gun takes away the 2+ save of a terminator. A flamer, that takes away cover saves, may appear to only hurt orks, however the orks paid nothing for that cover save, thus they lose nothing--without cover they are still just as efficient as the terminators as we covered earlier on a point for point basis.

So in conclusion, terminators and other elite style forces suffer from unexpected results much more often than cheaper redundant forces. What's more, many rules of the game have a much bigger effect on cheaper models than elite models, such as cover and the snap shot penalty. Good players can overcome the difficulty of quickly moving and fighting with a horde, but can do nothing to overcome dice variance in an elite army as dice roll outcomes on a fair toss can not be manipulated by player skill.

my bad, I was trying to answer this question by relating this to the only army I play which is CSM.

If we eliminate allies for CSM, then how do we define the CSM army? Is it still elite even with the small addition of cultist and marginally cheaper standard marines?

Do I lose too often because I do not know how to play an elite army or is it because elite armies get hurt more severely when they lose units?

I believe that a dice roll can be manipulated by player skill. I just hate it when my opponent finds some excuse to use single saving throws for their terminators with TH/SS cuz there is a IC with different stats, or a mad doc granting the squad feel no pain since he is a different toughness. I noticed that when my opponents make single dice rolls, they spin it sideways with the 6 on top so that it would be more likely to land on a 6. That in my opinion is manipulating the dice roll, but it is accepted by everyone else at my FLGS. Hell I have seen people use big dice and just drop it straight down and get 6s.

 Filch wrote:
I noticed that when my opponents make single dice rolls, they spin it sideways with the 6 on top so that it would be more likely to land on a 6. That in my opinion is manipulating the dice roll, but it is accepted by everyone else at my FLGS. Hell I have seen people use big dice and just drop it straight down and get 6s.

This is cheating, period. It's deliberately "rolling" the die in such a way that it isn't random, and you should tell them to stop being TFG. If they refuse, take your business elsewhere and be sure to inform any of your non-TFG friends about the cheating and suggest that they do the same.


Automatically Appended Next Post:

DevianID wrote:
This was expressed earlier in a way I quite agree with in that a good player benefits from the harder to play but less luck dependent army, while the worse player benefits from the easier to play but more luck dependent elite army.

You have that backwards.

The weaker player benefits from the easy to play horde army because it's predictable. It will be hard for them to luck their way into "undeserved" wins, but it's also much less likely to have the random game-changing bad luck that a weaker player won't be able to handle.

The good player benefits from either army. They have the skill to make the elite army work effectively (using its advantages in force concentration, fewer kill points, etc), and recover from occasional bad luck. On the other hand they also benefit from the horde army because it's predictable and minimizes the chance of a single incident of bad luck costing them the game, allowing their superior skill to decide the game.

The only players who care about model count are extreme newbies who are still busy learning the game and benefit from having things be as simple as possible, but the skill threshold for ignoring this factor is pretty low.

I disagree peregrine. The weak player only gets 3 turns tops with the hordes army. With the elite army, they can play to completion, and when dice are hot can overcome lacking tactics with statistically anomalous results. The horde army, even when played to completion, performs roughly average thus requires a steady plan, lest it be destroyed by a better player with a force multiplier that turns the average result of the horde into a destroyed unit.

DevianID wrote:
As a terminator player through and through, I have talked about this issue before in my group at length, with the results being that elite armies are worse than hordes when it comes to dice.

As an aside, the counter argument is that hordes take longer to play, because of more models to move and more dice to roll. This was expressed earlier in a way I quite agree with in that a good player benefits from the harder to play but less luck dependent army, while the worse player benefits from the easier to play but more luck dependent elite army. We see this crop up in video games a whole lot with certain "noob" weapons that a vet doesn't need but that a "noob" can still earn kills with. It's an important part of balance in games that focus on fun instead of ensuring the better player always wins such as in chess.

Anyway, terminators take a wound and fail an armor save 1 in 6 times. An Ork costs 1/6 the price and fails an armor save 5 in 6 times. Without any additional modifiers and with equal points, 6 ap5 bolter wounds on the termies kill 1 on average and kill 6 orks all the time. So it would seem both are balanced. However, 1 bolter wound can never kill 6 orks, but it can kill a terminator. On the other hand, a terminator has about a 33% chance to survive 6 bolter wounds, and the orks have 0 chance. Why this is important is because you know with certainty orks will never die to 1 wound and always die to 6 wounds. However, you can only make predictions about what the terminators will do. If you need that terminator to pass a save to win the game, you count on luck to win. With orks, you can play with certainty that less than 5 wounds won't wipe the squad.

In addition, a huge wrench to this equation is cover. So while up to this point, the terminators paying for their save meant the terminator and the Ork were balanced in point cost versus bolters, cover is free and give a save to orks all the time, while never giving a save to terminators. Thus, a huge part of the game only benefits the orks. This directly makes terminators worse.

After that, we get to snap fire. Elite armies tend to have higher bs scores than horde style ones. However, both fire snap shots at the same bs. A bs4 model loses 3 points of bs when making snap shots, while orks for example only lose 1 point of bs.

Finally, we get to special weapon shots. A plasma gun has no additional effect on orks beyond better wounding. However, a plasma gun takes away the 2+ save of a terminator. A flamer, that takes away cover saves, may appear to only hurt orks, however the orks paid nothing for that cover save, thus they lose nothing--without cover they are still just as efficient as the terminators as we covered earlier on a point for point basis.

So in conclusion, terminators and other elite style forces suffer from unexpected results much more often than cheaper redundant forces. What's more, many rules of the game have a much bigger effect on cheaper models than elite models, such as cover and the snap shot penalty. Good players can overcome the difficulty of quickly moving and fighting with a horde, but can do nothing to overcome dice variance in an elite army as dice roll outcomes on a fair toss can not be manipulated by player skill.

Thanks for this post. Not only did you say everything I was going to say, you added quite a lot and said it better than I could!