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.

In your personal experience has mathhammer worked for you??

or

Has bitter personal experience told you to consider other things beyond strict percentages??

Not usually. The dice is almost always waaay off what the math tells me it usually is. I tried math hammer and it never translates well to the actual game at hand. So I no longer bother and rely on real experiences to determine the effectiveness of a unit.

Yes it works, if you treat it like what it is: A estimate of how a unit will act on average over the course of lots and lots of games.

Of course it only tells you about probabilities, and doesn't account for stuff like survivability, or vector; You'll still have to do some thinking yourself.



Anybody that tells you that Mathhammer doesn't work, either is bad at math, is using weighted dice, or falls for the all to common "confirmation bias".

DrunkPhilisoph wrote:
Yes it works, if you treat it like what it is: A estimate of how a unit will act on average over the course of lots and lots of games.

Of course it only tells you about probabilities, and doesn't account for stuff like survivability, or vector; You'll still have to do some thinking yourself.



Anybody that tells you that Mathhammer doesn't work, either is bad at math, is using weighted dice, or falls for the all to common "confirmation bias".

^exactly this.

Also peoples common misconception is that the mean (what most people think of as the average) is more likely to come up than all the other results put together. This is just not the case most of the time. An obvious example is rolling a 7 on 2d6. It is the most likely result compared to any other individual result, but it still is only 1/6 of a chance! So then people say - 'well I rarely roll a seven', well yes, but it is still the most likely result, there just happens to be lots of potential results.

It's not like this is a matter of opinion

The guys above are right, and if you observe a different behavior: just repeat it more often and the results will get closer and closer to the distribution.

This: http://en.m.wikipedia.org/wiki/Law_of_large_numbers

Math hammer is important for two main things;

1) Designing an army - You can accurately estimate the typical values related to any given unit. The firepower a tau squad dishes out, the added benefit of an ethereal, how much firepower 5 scouts can handle etc.

2) Judging how well or how poorly your rolling has gone to the norm. I.e. if you throw a single bs3 lascannon shot and miss, you cannot judge how lucky you were; however if you throw 3 lascannon shots - hit with 3, strip 3 hull points, you've done well and you also know that is not typical (it was lucky) and not to expect that for the game - But you can plan accordingly. >> TL: DR Judge and react accordingly, based on fact.

If "mathhammer" didn't apply to games with an element of chance than I'm pretty sure casinos would be in a spot of trouble.

Mathhammer works as a good estimate of a units average damage output, toughness against specific circumstances, and general probabilities. Like everyone else has said.
Is it useful for list building and general planning for 40k? Sure.
Is it useful for guiding tactical decisions in-game? Nope!

Waaaghpower wrote:
Mathhammer works as a good estimate of a units average damage output, toughness against specific circumstances, and general probabilities. Like everyone else has said.
Is it useful for list building and general planning for 40k? Sure.
Is it useful for guiding tactical decisions in-game? Nope!

This is what i was meant to say on my earlier post. My earlier post came out really dumb now I re read it.

Waaaghpower wrote:
Is it useful for guiding tactical decisions in-game? Nope!

It absolutely is useful for in game tactical decisions. Whenever you're assessing what order to fire weapons, and at what targets, to maximise your chances of eliminating primary targets you're using Mathhammer. But again, as stated everywhere, it's a guideline not a guarantee.

Chrysis wrote:

Waaaghpower wrote:
Is it useful for guiding tactical decisions in-game? Nope!

It absolutely is useful for in game tactical decisions. Whenever you're assessing what order to fire weapons, and at what targets, to maximise your chances of eliminating primary targets you're using Mathhammer. But again, as stated everywhere, it's a guideline not a guarantee.

100% right. The math gives you information whether shooting this gun at that unit is a desperate move hoping for good luck or if you can expect half of the enemy unit to get wiped out if you shoot at something different. It's the base you need for solid tactical decisions. Depending on the situation it CAN be still the right thing to shoot with bolters on that jinking FMC since everything could be lost if it is not downed. But you can make better decisions if you know that the probability for that is 3% while the probability to kill those Genestealers which will be charging in the next round is 40%. Decisions, decisions

Btw, anyone not believing in the correctness of probabilities and math is invited to have a game with me. We'll throw 2d6 for 200 times and I will win 5 Euro on a roll of 7 and you will win 10 Euro on a roll of 12

Works for my orkses. The more dice you throw - the closer you are to the average.

Mathhammer has 2 aspects, percentages and averages. With the first usually being much harder to calculate. Build lists with averages and play them with percentages. The other way around brings only disasters.

Mathhammer is always right.

Your interpretations might be wrong, but Mathhammer cannot be wrong by its very definition.

/e: Dear lord, people seriously voted No? O_o

Math hammer is a very useful tool, provided that you use it as a guide, rather then a be all, end all. There will be outliers, definitely, but we all use math hammer, even if we don't sit there going 3/6... x4/6... x1/6. If you have a bolter and melta, and an ork boy and a trukk both within 12", you aren't going to fire the bolter at the trukk and the melta at the boy. That in itself is math hammer.

Averages. One bad experience is not 'disproof', it is simply one of those random times when the universe decides to 'fetch you, mortal'.

As others have mentioned, the point about statistical averages is that they, by definition, include outliers and randomly high or low rolls (some refer to these as 'hot' or 'cold' streaks). The problem with Mathhammer is not whether it's true or false (hint: it's true), but how people implement it in their list building and tactics.

Mathhammer is designed to give you an average interpretation of what a unit is likely to accomplish over a large number of rolls and a large number of games. It's not going to be 'proven wrong' by a single game where your Lootas may have rolled mostly hits or your Guardsmen killed a Daemon prince in CC. That's actually assumed to occur on a rare basis and, given that it does occur rarely (enough that people are excited to talk about it), only reinforces the concepts of statistical probability and combinatorics. You can use the averages that you get from Mathhammer to understand how the damage output of one unit will, on average, will be higher than another (due to ballistic skill, or higher strength weapons, etc.) and then make a determination on their respective value for points:

10 Guardsmen with S3 Lasguns at BS3 are not very likely to cause a lot of damage to a Dreadknight. 10 Space Marines with S4 Bolters at BS4 are more likely to do so. 10 Grey Knight Paladins with S7 Psycannons and BS5 (for whatever reason) have an even greater chance. That doesn't mean it's not possible for the Guardsman to kill the Dreadknight in a single round of shooting and the Paladins to completely miss with every shot, but it's extremely unlikely and anyone who would build a list around assuming that would happen is probably in for a rude awakening 99% of the games they play.

Of course this doesn't take into account the Toughness of the Marines vs. the Guardsmen, the Armour saves, the number of wounds, the special abilities, and wargear. Mathhammer in these sorts of situations gets very complicated, very quickly, and that's not even considering the synergy and interactions between the rest of your army, let alone your enemies army.

If you really want to see Mathhammer in action, play a game of 40k using statistical dice rolls (e.g. if you're BS4, just assume 2/3 of your shots hit, alternating rounding up and down by unit) and you'll quickly see how the game can actually be quite well balanced from a tactical perspective. Then again, some people really enjoy the wild, 'dice gods' potential to roll all or all because it makes the game more dynamic. There's nothing wrong with either way, although I'll admit it's always fun to roll dice.

DoW

The only problem with mathhammer is that while the theorycraft and math is solid the website quicktool hasn't been updated to include some special rules and the new 7th edition damage table. Running the numbers yourself is a pain, but I do recommend doing it. Also, whenever checking a riptide remember to add in a 4/6 factor if you want to check the reliability of the nova weapons. In fact, I've found that the nova Ion cannon is horribly unreliable having a 5/9 (.5 repeating) multiplier before you even try to hit anything, and a 44.44% chance to inflict a (easily saveable) wound on the riptide! Wether you save the wound or not you would lose the weapons functionality AND potentially the functionality of your sidearm if you take it from the gets hot test. The ion cannon tide is still good, but I wouldn't novabuff the ion cannon unless you have an emergency of THAT UNIT NEEDS TO DIE NOW!!!

Working in fractions makes it easy to calculate basic averages, it's when you go beyond 3 dice (hit, wound, save) that things tend to get tricky, or when you add in rerolls. Guardsman shooting at SM for example is 3/6 x 2/6 x 2/6 =12/216 =3/54. Assuming 20 shots you get an average of 1 kill. Takes all of 40 seconds, easy enough to do in an opponents turn. But, you shouldn't depend on those values, it's just a rough 'guesstimate' as to whether you fire at the terminators or the bikers.

Also, make sure you run different kinds of calculations for weapons. I still run Railguns because despite the lower odds of inflicting damage, when I do inflict damage the odds of the target getting critically impaired or destroyed are VERY high.

Mali wrote:
In your personal experience has mathhammer worked for you??

or

Has bitter personal experience told you to consider other things beyond strict percentages??

Uh. What? Talk about a false dichotomy. Mathhammer works for EVERYONE who doesn't have weighted dice. People who think it doesn't are just remembering the bad rolls and forgetting all the other ones. With that said, of course in this game there are other things to consider beyond strict percentages. So it isn't one or the other, it is yes to both questions.

The main problem is people who never had at least a bit of probability theory.

The number of rolls in a single game of 40k is low enough that the outliers may not get evened out. And if you lose a game or even a single loved unit due to this you'll remember that. The human mind is kinda biased with this, remembering the exceptional events better than the average ones.

I've had games where I couldn't kill a razorback with my railgun. The next game I rolled all sixes on the same railgun against a landraider. Odds are a good thing to plan around, but sometimes gak just happens.

Or rather, understand the odds but don't let them cow you into weak play for the sake of appeasing the dice gods.

I took a Shokk Attakk Gun to a tournament last year, 1st game I rolled boxcars on Ovesa Star, 2nd game I rolled boxcars on a squad of Thunderwolves.

Point being, sometimes you need to take a long shot to really pull out a win.

Ventiscogreen wrote:
I've had games where I couldn't kill a razorback with my railgun. The next game I rolled all sixes on the same railgun against a landraider. Odds are a good thing to plan around, but sometimes gak just happens.

And still when you play 1000 games and you count how often that railgun hit or didn't hit you'll see that the results will match the theory. Only the single sample taken at random will deviate, the mass will inevitably show the mathematical distribution.

It doesn't make me less rage filled when I couldn't kill a razorback with a railgun over three turns.

The people with math-hammer is the in game variables.
Should I take haywire blasters or heatlances?

A haywire blaster as VERY good odds of doing a single hullpoint of damage, and very poor odds of doing anything else.
On the otherhand, getting a weapon destroyed, or immobilized is much more likely with the heat-lances.

Sometimes, you need to kill, and sometimes you need to maim. From a list building side, it's tough to assign values to those, as it is dependent on the opponents list (maiming a razorback is fine, maiming a landraider full of assault units isn't good enough).

IMO, the problem with math hammer is two fold. People are trying to get the wrong information out of it, and people are really crappy at generating a sample population.

-Matt

I don't care what the math is, I always roll terrible, except when I expect to. I have a knack for getting off lucky shots and blowing things up I shouldn't, and having massed shooting fail entirely. Go home Nuffle, you're drunk.

Technically, just so those who didn't realize it can know, all mathhammer calculations have an inherent error in them unless you are using the kind of dice made for casinos. Because of the weight distribution of "pipped" die, that is, die that have divots carved out of them that are painted, you are more likely to encounter (if I recall correctly) ones than sixes. The rouned edges of many die also affect your distribution.

I believe math hammer has its place, but applied too liberally you're just theorizing in a vacuum. I still believe the absolute best lists in general scenarios (particularly, outside of tourneys) are well-balanced all-comers lists designed to be adaptable to numerous threats, not just be 100% efficient against A or B.

 Gratlugg wrote:
Technically, just so those who didn't realize it can know, all mathhammer calculations have an inherent error in them unless you are using the kind of dice made for casinos. Because of the weight distribution of "pipped" die, that is, die that have divots carved out of them that are painted, you are more likely to encounter (if I recall correctly) ones than sixes. The rouned edges of many die also affect your distribution.

I believe math hammer has its place, but applied too liberally you're just theorizing in a vacuum. I still believe the absolute best lists in general scenarios (particularly, outside of tourneys) are well-balanced all-comers lists designed to be adaptable to numerous threats, not just be 100% efficient against A or B.

While technically that is true, the difference from pips is completely negligible. The rounded corners are a bigger problem, but in actuality the biggest problem is casting imperfections. (Rounded corners might not be a problem at all as long as they are all equally rounded.) I'm sure you've heard of microwaving dice, where you ever-so-slightly melt them to make two rolls (Typically 6's and 1s,) more likely. (The sides that are facing up and down.) This happens naturally, though, too: Leave your dice in the heat, have a slight mistake casting them, a slight imperfection in the mold... That's going to be your biggest problem unless you are using casino grade dice.

Thanks all. =) some great replies with good thoughts and a few laughs.

I may be misunderstanding the definition of "Mathhammer" wrong.

Best example i can thing of and this is over the course of about 30 games.

Warwalkers. i know statistaclly duel lances or duel scatter lasers is supposed to do better.

How ever, in my experience.
when i run duel weapons i face armies that do not require the weapon i brought. ex if i run duel lances i face all horde and light tank armies.

When i run scatter and lance combo. i have a shot every turn and do more over all damage.

So is this "mathhammer" or something else??