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.

.

Of course they matter. Now should we answer some other obvious questions? Yes, the sky is blue, water is wet, etc.

 Peregrine wrote:
Of course they matter. Now should we answer some other obvious questions? Yes, the sky is blue, water is wet, etc.

Citation needed.
1. 1. 1. 1. 1. 6. 3. 5.

I feel that a heavily Mathhammered list adds a different style and can make mirror matches much more interesting, and of course is nearly mandatory for competitive play.

 Verviedi wrote:
Some people I know say "This is a game of luck, ...

Sure, but it's a game of luck because it involves dice. Which makes it a game of probabilities.

The fact that a model with BS2 and a model with BS5 can both miss due to an unlucky To Hit roll doesn't change the fact that the model with BS 5 has a higher probability of hitting.

Whoever claims statistics do not matter in a game which using probability at its core is likely an idiot.

The skill is simply playing to make the probability lean more in your favor.

They matter, unfortunately most people don't actually understand statistics. You'll see people calculate the odds of passing such and such test, or how many hits you expect to get from a squad, and so on. But there are so many more things you need to know: how likely are you to be within 10% of the average, for example? Sure you expect to get one hit if you fire six snap shots, but how likely are you to get none? One-in-six is only a small part of the whole picture.

For one thing, it makes larger armies more reliable; the more dice you roll, the closer you'll be to the average.

However, tactics are more important than statistics. Your tau broadside team with all its missiles may be able to pop a rhino nine times out of ten, but if there's a building between you it drops to zero out of ten. Or you might pop the rhino that was just there to draw fire while the assault marines close in. This is where statistics can fail; you might have a unit that can, statistically, devastate an enemy army, but that in practice just can't catch them, or can't make it into range, or whatever.

Maths - it's only a theory. Teach the controversy!


Btw, do these people also think there's a 50/50 chance of winning the lottery? Either you win or you don't.

Statistics are extremely relevant. The problem comes when people are bad at statistics (frequently) or rely too heavily on a limited perspective (also frequently). If you want to judge a unit you have to keep in mind the situation and the environment you're expecting to use the unit in. (Flyrants are your most reliable AA in the 'Nid book, they're almost an auto-include for that alone)

office_waaagh wrote:
They matter, unfortunately most people don't actually understand statistics. You'll see people calculate the odds of passing such and such test, or how many hits you expect to get from a squad, and so on. But there are so many more things you need to know: how likely are you to be within 10% of the average, for example? Sure you expect to get one hit if you fire six snap shots, but how likely are you to get none? One-in-six is only a small part of the whole picture.

For one thing, it makes larger armies more reliable; the more dice you roll, the closer you'll be to the average.

However, tactics are more important than statistics. Your tau broadside team with all its missiles may be able to pop a rhino nine times out of ten, but if there's a building between you it drops to zero out of ten. Or you might pop the rhino that was just there to draw fire while the assault marines close in. This is where statistics can fail; you might have a unit that can, statistically, devastate an enemy army, but that in practice just can't catch them, or can't make it into range, or whatever.

Law of large numbers man. Reversion to the mean. Standard deviation.

Statistics do matter. But they don't much matter for any particular roll. You can throw dice away at snap shots and models with 2+ re-rollables and you'll eventually get lucky. But a smart player doesn't put his faith in poor odds. 40k is like poker, you play lots and lots of hands to get to where you need to be to win, and while you're gonna lose a 60% hand 40% of the time, at the end of the day you finish the day on top.

Stats are very helpful for understanding how powerful a given rule is. For example, say I'm running a priest in my guard blob:
- at leadership 7 he's got about a 60% to get his war hymns off. Do I use the hymn to re-roll armor saves or re-roll to wound?
- a re-roll of a weak save is less effective, my 5+ armor goes from a 33% save to a 55% save, which is good, but on my bullgryns that'd take a 3+ save from 66% to 88%. Note that the 5+ re-roll does less than half the amount of saves that I'll fail, while the 3+ re-roll cuts the amount of saves that I fail in third.
- a re-roll of to wound against T3 opponents makes my 4+ to wound go from 50% to 75%. But of course, since I'm likely striking last, I'm going to maximize the number of attacks I make by re-rolling armor saves.

office_waaagh wrote:

However, tactics are more important than statistics.

I wouldn't say that. Part of tactics is that you take advantage of whatever you're good at and minimize what you're weak at. "Employ your unit in accordance with its capabilities". Without a good understanding of what your units are capable of, you can't employ good tactics. Statistics are pretty important in this regard.

YES they do matter. They show what units will perform better and in what circumstances. For example, Warp Talons do nothing better than Raptors except charging MEQ units outside of cover.

Statistics absolutely matter.

What they won't do is tell you exactly how the dice will turn out, but they will tell you what the dice should *typically* do and what kind of variation to expect, which should then inform your actions as to how to best direct your forces to their best potential, as well as how units perform in different roles relative to each other.

Keep in mind that statistics are only as useful as the model they're being fed through. I remember having a 'discussion' with someone back in the day about how ogryns weren't tough enough to survive in their role as a counter-charge unit (back when they were s4/t4). He would trot out his mathhammer about how unit x charges the ogryns and decimates them, which while being technically accurate missed the point of a 'counter charge' unit and not being a 'tank/anvil' unit.

Statistics are good when they're used properly. But there's also lies, damn lies, and statistics.

Averages certainly matter! As do numbers of dice rolled. For instance, I really like to avoid playing terminators of any kind, because too often do I see people roll 1s for their armor saves, losing 1 or 2 termies to just 4 or 5 rapid fire bolters!

If stats didn't matter I would charge Abaddon with a gretchen and expect to take a wound or two off. (I don't)

If you're running a fluff list, sure, stats are pointless. If want to win a few more games, yeah, look at the numbers.

However, you could always do things the ork way. If I think hard enough that my terminators are actually awesome, maybe paint them red and shout a bunch, they'll become much more deadly.

 Verviedi wrote:
I like calculating, and base my unit choices on statistical damage output.
Some people I know say "This is a game of luck, stats don't matter. Walkrants are WAY better."

This brings up a point. Do statistics matter in game? Discuss.

Yes they matter. This doesnt mean luck cant still win or lose you a game despite all the odds.
You'll find different sorts of gamers.
Some enjoy doing the math and conside it part of the game and calculate down to the point what points tospend on what models based on math in order to maximize the odds of winning a game.
other build their armies in different ways.
Personally, I take what I think looks cool at the moment and what kind of mood I'm in. This is why I'm not a national tourney winner. Only win or place in local and regional ones and then usually because I always max out on the modeling/sportsmanship aspects. I still have fun just as the math hammerers enjoy themselves. if your dont like the idea of mathhammer or maximizing stats to win, dont fool with it and enjoy yourself anyway.

No, i don't even look at my dice when play. I just throw them and do random stuff.

 koooaei wrote:
No, i don't even look at my dice when play. I just throw them and do random stuff.

That's narrative as hell, that is.

There are two different types of stats that people use with mathhammer. Averages (usually the mean) and probability.
Lets not confuse the two.
The average (mean, not mode or median), tells you the amount of dice you would expect to hit, say, or wound/pen. This, as mentioned above, does not take variance into account but gives you a rough figure of what to expect.
Then there is probability. Such as the percentage chance of a bs4 bolter wounding an ork boy, so 2/3 X 1/2. This gives a 33.3 (rec)% chance of wounding. That is FACT. It IS a 33.3% (to 1.d.p) chance. This does not mean from a small sample that you will wound an ork boy 33.3% of the time, but it is still the exact chance of doing so. This shows especially true the larger the sample (amount of times you try).

You can use these to help you decide what to shoot/charge. Do you really want to charge that 3+ SM captain with poweraxe into that large unit of sanguinary guard by himself? No? Why not?

Statistics also includes things like variance - the understanding that a unit has an average damage output, and that it is likely to vary within a certain range for any given encounter. 40k is a probabilistic game in the same vein as something like Poker - skilled players understand the basic statistcal laws governing it, and so know when to push their luck, or when to play conservatively.

Math matters.
Generally, if my statistical count is more than 50%, I'll go for it.

Think of it being akin to counting cards when you play blackjack (after all, your just doing the math to figure statistical probabilities). There is nothing written in the rules to stop you. So of course, the casinos in Vegas wont mind if you do it.

In my eyes, it goes against the spirit of the game. Of course, thats my opinion. Its one thing to eyeball it and figure "yeah, my guy doesnt stand a chance if he assaults this unit but he might do some damage oer there" and another to whip out the calculator. But if my opponent chooses to do so, thats up to them and I wont tell them they cant.

Statistics are tactics. A successful player is one who knows how to leverage his units so that the statistical outcomes are as heavily in his favor as possible. That doesn't protect them from ending up in a situation where they can't roll anything but 1s because luck is against them, but it does mean whenever they have to roll dice, things are already in their favor.

On a related note, I had a conversation last week with one of the local players. Nice guy, but the conversation turned to codex balance which turned to eldar which of course turned to him arguing that wave serpents are perfectly reasonable for their points. As an eldar player, I felt the urge to debate this point. His argument essentially came down to, "Well, averages don't matter. Sometimes you'll get unlucky with your rolls, and then the unit is balanced."

O_O Please do not be that guy. I've had 3 wyches survive 6 flamer/heavy flamer wounds thanks to 6+ FNP, but this does not mean my wyches are a reliable damage sponge.

 EVIL INC wrote:
Think of it being akin to counting cards when you play blackjack (after all, your just doing the math to figure statistical probabilities). There is nothing written in the rules to stop you. So of course, the casinos in Vegas wont mind if you do it.

In my eyes, it goes against the spirit of the game. Of course, thats my opinion. Its one thing to eyeball it and figure "yeah, my guy doesnt stand a chance if he assaults this unit but he might do some damage oer there" and another to whip out the calculator. But if my opponent chooses to do so, thats up to them and I wont tell them they cant.

There is no spirit to this game other than the spirit of GW getting mad stacks of cash.

Martel732 wrote:
There is no spirit to this game other than the spirit of GW getting mad stacks of cash.

Now that's something I can agree with.

However, it is also usually true that critical die rolls don't have have enough dice in them to obey the law of large numbers of dice.

statistics matter sometimes.

Statistics, means, averages, etc only really come out[are reliable as a method of planning] when there is a very large number of dice being used.

This means it is more important if you see someone rolling 60 shots to hit from an unit, than if you see someone rolling 2 shots to hit from an unit.

The chance for something to deviate from statistical average becomes lower the more times it is repeated, this is how casinos win and you lose. The more you play the more likely the casino gets your money because you are more likely to keep within the bounds of standard deviation at large numbers.

So for armies that use large pools of dice for things, this is important. For other armies, you cannot count on it as much.

This is compounded by the fact that your statistical math hammer unit is not going to be doing what you are calculating every turn of every game. Chances are you wont be firing your unit every turn, or assaulting every turn, etc. This further moves the dice pool from rolling a lot of dice to a few dice.

for example. 30 shoota boyz, shooting things. Fire 60 dice a round they shoot. If a Game lasts 6 turns [total of 360 shots] and they fire every turn you are more likely to generate a number close to the statistical average of BS 2 (33% chance to hit) than a squad of tankbustas that has 5 models firing two times in a game which although they have BS2 since rolling is random there is a much more likely chance to roll above or below the statistical average due to the small dice pool, ie it is reasonably possible to roll 3/5, or 4/5 hits instead of 2/5.

The basic reasoning behind this is it is very possible to roll a 6 on 1 die. It is less possible to roll 2 6's on 2 die, and on and on. The chance of rolling high or well for what you need is the same, but the chance to have a large number of high rolls is lower the more dice you roll.

In effect the results of rolling dice is a curve with outlier possibilities above average and below average. The variances becomes smaller the larger the pool becomes because the chance to have a statistical % change in a random set of numbers becomes lower with more dice, ie its harder to roll a yahtzee than a single 6 on a d6.

The result is you are going to see more variance in dice rolls with small dice pools than large dice pools.

It maters, but most people (players) have no real idea what probability is or what it actually means. Take a stats class and mathammer becomes useful.