Are the Eldar movement rules broken? Whats the alternative?

[Sigoroth]

Averages represents the references from which the game mechanics resolve
as a gamer as well, you know as i do that we roll dice expecting to beat the average
this doesn't mean if enough data point is collected, that the result is significantly different from the average

yes, i acknowledge there is a role for ranges and confidence interval in addition to averages, but averages should remain the reference point because with any random distributed bell curve, ranges and confidence interval are mirror over the average

Well I assume you mean variance, not CI. Also in some cases the distribution curve is skewed, such as with bomber attacks runs, etc. And yes, as gamers we tend to simply use averages as a measure, though I'm usually quite happy to just reach the average, rather than beat it. Usually when I beat the averages it's on something stupid, like massively overkilling an escort. However, the point is that you can't use averages to work out the probability of getting a kill (unless you know the shape of the distribution and the standard deviation). As gamers we're also interested in finding out the odds. So if you're shooting your Lunar at that last Eldar escort it's good to know you're total odds of destroying it is around 63%. However, when shooting at a squadron your average total kills (worked out using probability) is 0.78 escorts.

with the current mechanics of my turn your turn, everybody is indeed moving and turns represent snap shots. and that is why i don't have a problem with eldar MSM because it is multitasking in action rather than sequential move shoot. the snap shot comes a little later, that is all. would it been better if MSM isn't accompanied by eldar ability to turn on a dime?

 the LotR system of i move you move i shoot you shoot we fight is less abstract.

Yes, limiting the turn rate definitely would help. It doesn't however eliminate the problem where you can hide after shooting (so long as there's terrain in your path). Also, if the Eldar can't run and hide then they will simply lose every game given their current level of resilience. So they'd need that upped.

[fracas]

i would be more interested in CI than variance, as in how likely (95% for instance) an occurence would occur.
yes sometimes the distribution is skewed but this pertains to the interpretation of the results rather than the results of the dice themselves.

i think you should expect better than average rolls. positive attitude helps for me and my opponent.

i understand better what you said earlier now, that it would be hard to tease apart MSM from holofield as representative of eldar movement. in some ways both represent eldar choice in technological development to stress movement (and with vast distance with the time space continuum a form of uncertainty of placement/location) rather than the necrons, the other ancient race, which stressed resilience.
all the more reason i think it is characterful.

[horizon]

Yes the LotR system is awesome. But that means changing the whole system of BFG, a bit out of scope I guess.

With MSM, the msm holofield combination protection works in a way, yes. And in a battle with sparse terrain it can even work against fleets with only short ranged weaponry and slow ships. But versus fleets with long-medium range weapons (Chaos, Tau), fast fleets (Marines, Necrons) and ordnance fleets (Tau, Nids, even Orks!), the Eldar go down and down faster.

As for DE movement: well, it isn't bad but DE are currently designed into a one-trick pony for dull battles. Playing longer battles makes it death DE for sure. In character perhaps but it could be done better I think.

[lastspartacus]

MMS DE. Duh

[Plaxor]

Horizon,

What is this 'one trick' that I keep hearing about for DE?

I think that DE work close-to fine. They shouldn't win a long battle. They should be in and out real fast killing everything with high hitting power and then not stick around to be shot at.

[Sigoroth]

i would be more interested in CI than variance, as in how likely (95% for instance) an occurence would occur.
yes sometimes the distribution is skewed but this pertains to the interpretation of the results rather than the results of the dice themselves.

Er. The CI refers to the surety that your sample average is representative of the population average. When we're talking probabilities we already know what the population averages are, so no need for CI. Yes, while the results of the dice rolls may follow a standard distribution, what they mean does not, and that's the point of interest. When looking at bomber kills of escorts you will have a right skewed distribution.

If you had a standard distribution, the mean and the variance or standard deviation you could work out the probability of a kill. However, this is a convoluted method made for testing sample data where probabilities are unknown. This method is designed to draw inferences about the population from a sample, and so work out the probabilities of an individual conforming to some population parameter, such as getting heart disease, diabetes, etc. Since we know the probabilities (which is how we worked out the averages in the first place) then we already know the probability of getting a kill. No need to take sample data and make inferences about the population, no need for CIs and no need to try to work out the probability of a kill from the averages.

i think you should expect better than average rolls. positive attitude helps for me and my opponent.

i understand better what you said earlier now, that it would be hard to tease apart MSM from holofield as representative of eldar movement. in some ways both represent eldar choice in technological development to stress movement (and with vast distance with the time space continuum a form of uncertainty of placement/location) rather than the necrons, the other ancient race, which stressed resilience.
all the more reason i think it is characterful.

Excellent, now you just gotta understand the logical absurdity of a pop-out attack on the scale of BFG and you're there.

[fracas]

Are you mixing power with ci?
Since we are using stats to estimate the probability of something will happen rather than evaluation is rather whether our data is representative of a measured population.
With dice and gaming the probability of certain rolls are known

Now more than ever I think we should keep both MSM and holofield!
Maybe change how much they can turn

[horizon]

Now Fracas, Tell me, how to fix this with msm:

Raider scenario 750 vs 1500:

Eldar attacker, turn 1:
move onto table:
destroy ship in shooting phase (750 pts of Eldar can do that easilty).
destroy or cripple another ship in ordnance phase (again -> they can do that
ordnance move = turn and fly from table.

So opponent gains 10% from "fleeing" Eldar = 75.
Eldar destroyed a ship and/or crippled one. = more then 75
Even if hulk was left it is an Eldar victory.

So game over. Eldar opponent has only rolled Brace for Impact.
Fun?


Also: one of the strongest aspects from the Corsair Eldar is that they excel in fleet engagements and escalating engaments. Pretty odd for a raider fleet isn't it?

[Sigoroth]

Are you mixing power with ci?

No, I'm not. Power refers to the ability of your experimental design to significantly detect differences and is related to type II errors. The confidence interval is the range in which we are 95% (or whatever value is being used) sure that the population mean falls based on our sample mean and the variance. So it represents how sure we are that our sample mean represents the population mean.

Since we are using stats to estimate the probability of something will happen rather than evaluation is rather whether our data is representative of a measured population.

What? English please.

Now more than ever I think we should keep both MSM and holofield!
Maybe change how much they can turn

How the hell do you come to that conclusion? MSM is ridiculously stupid. It should never have been introduced in this game.

[fracas]

in gaming statistical analysis, it isn't about detecting a difference between your sample and the studied population. gaming statistics is about predicting an outcome that hasn't happened yet. thus a 95% ci means that there is a 95% chance something will happen rather than a 95% chance the observed matches reality. predictive vs analytical.
whereas variance is a measure of the difference of the sample distribution from the true value. no?


early on i posted we have differing opinion on what constitute characterful. we are back to where we started from.

[horizon]

But you haven't adressed both problems I described.

[fracas]

horizon

the scenario you described to me is a problem of the scenario rather than the MSM rules
perhaps the player as well

with any fleet i can concentrate all my fires to get kills. as soon as i have killed more than i have lost + disengagement, i disengage and win. you can do this against necrons easily.
not a problem with MSM imo
redesign the scenario for minimum turns ... or one which each fleet is tasked with taking out a "flagship" as in hunt the bismark in the north sea or carrier hunt ala battle of midway

it is also a player problem as well because a game that is not fun for both players regardless of who wins or lose then it was a game poorly played.


also ... one off scenarios do not adequately reflect the challenges of a fleet commander that a campaign has
destroying enemy warship without losing yours (interesting enough, when you have killed your intended, you really should disengage, no?)
replenishing your losses when it take months and years to build a new ship
protecting your supply lines, when potentially the supply ship is more valuable than your warship
etc

again, scenario problems and victory point assignments rather than game mechanics

[Sigoroth]

in gaming statistical analysis, it isn't about detecting a difference between your sample and the studied population. gaming statistics is about predicting an outcome that hasn't happened yet. thus a 95% ci means that there is a 95% chance something will happen rather than a 95% chance the observed matches reality. predictive vs analytical.
whereas variance is a measure of the difference of the sample distribution from the true value. no?


early on i posted we have differing opinion on what constitute characterful. we are back to where we started from.

Seriously, does no one know anything about statistics on this forum?

[fracas]

Statistical analysis either look for differences within the studied population by looking at distribution of the observed data (analysis of variance) and/or difference between the observed population vs the supposed true population (power and significance)
With dice there is only one population
With gaming rolls there will be no difference between the observed rolls and the possible rolls if enough rolls are made. But since each set of roll is it's own true population once rolled the analysis resets. It is a predictive analysis applying to that roll alone. Variance doesn't really matter.

Your suggestion that variance is useful is off
We know what the distribution will be: a bell curve
We know what the average will be as well thus this value is useful
Knowing how tight the bell curve is gives you probability which is then the confidence interval but still not as useful as average but more useful than variance

The chance of a lunar killing a sword moving away is X and for killing a hemlock is Y
Doesn't change regardless of how many matches

[horizon]

So, you rather change a scenario which works for all other fleets then solve the origin of the problem?

The problem being the Eldar rules.



But as said: it is easier to kill a Hemlock then a Sword.*


* If no terrain is present ans ship is within range.