It depends. Speaking only from experience, but at a tournament level, the uber-competitive have a tendency to abandon reasons like that, in favor of the best tactical options. It's probably the weakest of the points, but it does have the potential to skew any purely math based calculation, that would need to be factored out.

Someone shows up to a Tournament with the First Company (5 Terminator formations and a couple of Land Raider formations) and gets smashed. Does that mean Terminators aren't an effective formation? How that's calculated into the formula is important. I was just trying to spell out some of the variables that would need to be factored in. I didn't even cover luck. Which, over the short term, or with a relatively small sample size, can be a significant factor.

Morgan Vening

Morgan Vening is of course right. There's no scientific way to use this data to get details of the effectiveness or otherwise of particular units.

Finally, a convincing argument for the howling girlie's!

I agree on sample size, bigger is better :-) But it's hard to draw the line beyond which we say 'insignificant'.



But what if someone else turns up with all Tacticals (their particular turn on) and gets smashed? If we then compare Terminators with Tacticals we see no advantage to either. Similar to my last point, what's to suggest that such fluffy choices are unevenly distributed amongst unit types? I expect they are to some degree but necessarily significantly so?

These variables do need to be considered but I don't think they should be used to dismiss the proposed measurement out of hand.

_________________
Uti possidetis, ita possideatis.
May your beer be laid under an enchantment of surpassing excellence for seven years!
An online epic force creator: Armyforge

What would you say if we had a trillion samples* and armies with Assault formations were highly correlated with losers?

What would you say if we had a trillion samples and armies with Assault formations were highly correlated with individuals doing worse than they usually do?


* (apart from being over the moon that the entire planet had started playing EA)

_________________
Uti possidetis, ita possideatis.
May your beer be laid under an enchantment of surpassing excellence for seven years!
An online epic force creator: Armyforge

But we don't. Discussion over as far as I see it.

_________________
Frogbear is responsible for...
Previous World Eaters
Previous Emperor's Children
Previous Death Guard
Previous Imperial Fists
Previous Chaos Squats

Is was a hypothetical question to establish whether the idea of stats is complete anathema or open to discussion.

I can see where you stand on this, thanks for your contribution.

p.s. i happen to think the best way to assess army lists is player feedback, but i also find the stats issue an interesting problem

_________________
Uti possidetis, ita possideatis.
May your beer be laid under an enchantment of surpassing excellence for seven years!
An online epic force creator: Armyforge

Agreed. But with so many factors, and so many options/combinations, even if everyone on the board was playing once a week, for a year, I doubt it'd make a decent statistical sample.

I'm not saying it'd all be one way. But like randomness (hey, my Hellblades got shot down by Marauders two games running, for no effect!), it's a factor that does influence the numbers, that is neither easy to quantify, or easy to analyse.

I don't want to see it done, and then see it used as "proof" one way or another, without a statistically meaningful sample. And with the variables involved, all the factors, from metagame, to the environmental (what I referred to), the strategic (the sheer number of combinations of units and formations that can be taken), and the tactical (Hena mentioned losing a game due to Sustaining and not Doubling. How's that factor into the effectiveness of that formation, all the formations in his army, all the formations in his opponent's army). The sheer number of combinations require a sample that even if it takes into account most of what I listed, make the "trillion" seem not so unreasonable. Given we see maybe hundreds of actual results annually on here, out of maybe thousands of games played by TacCommers, the sample pool is just too small, and the data collection and factoring too great.

I'll be happy to be proven wrong. But the calculation needs to be robust, or it's meaningless, or worse than meaningless.

Morgan Vening

I doubt we'll get anything useful out of it either, I'm just interested to see what people come up with :-)

Quote:

Given we see maybe hundreds of actual results annually on here, out of maybe thousands of games played by TacCommers, the sample pool is just too small, and the data collection and factoring too great.

For some measurements but not all. And some might be interesting by jove!

_________________
Uti possidetis, ita possideatis.
May your beer be laid under an enchantment of surpassing excellence for seven years!
An online epic force creator: Armyforge

I say make Frogbear do it.

_________________
author of Syncing Forward and other stories...

It's a dog-eat-dog world, and I've got my Milkbone underwear on.

+1

_________________
Uti possidetis, ita possideatis.
May your beer be laid under an enchantment of surpassing excellence for seven years!
An online epic force creator: Armyforge

It might give an indication, but it certainly wouldn't be scientific.

Do you mean conclusive rather than 'scientific'?

_________________
Uti possidetis, ita possideatis.
May your beer be laid under an enchantment of surpassing excellence for seven years!
An online epic force creator: Armyforge

Mosc. You are wasting time on this thread.

Get over to the Tunneller thread and help us make a difference by including those carriages that have been giving you headaches

_________________
Frogbear is responsible for...
Previous World Eaters
Previous Emperor's Children
Previous Death Guard
Previous Imperial Fists
Previous Chaos Squats

No, I mean scientific. There are too many variables at work for it to be a scientific conclusion. You can possibly draw a conclusion from such a set of data, but not one of any scientific worth.