Combat Effectiveness

[Hugmenot]

Shad,

Why is CE measured as ( W / (W+ L + D)) instead of ((W + 0.5D) / (W + L + D))?

I am very proud of some of my draws but they count for nothing towards the nation's CE

[plloyd1010]

Because

[vince hughes]

Daniel,

I agree. In many sports (though not often in the US, hockey excepted in part), there is often credit in a 'draw'. Cricket being the most highlighted of all. A team can be being trounced, but if they don't take all the wickets by the end of play, then it is a draw. This often leads to a situation where the last batsman is hanging on in there in the last few minutes, not worrying about runs scored but instead preserving his 'wicket' to salvage the draw.

See Michael Atherton and Jack Russell's heroic saving of a game against South Africa in 1995/96 at Johannesburg: Atherton had to bat 10 hours 45 minutes to do it. Memorable stuff. If anything gives credence to a 'draw', this does.

http://www.independent.co.uk/sport/crick...24253.html

[Shad]

Why is CE measured as ( W / (W+ L + D)) instead of ((W + 0.5D) / (W + L + D))?

Because

Peter nails it in a word.

CE was something I sketched out during an idle moment long long ago. Not a lot of thought went into it. I can vaguely recall having argued over the formula once before - perhaps with the moderator team... I think Vince wanted credit for draws?

Anyway, my rudimentary thinking now is the same as it was then:

• how often in war is a draw the desired result?
  

I don't think sports examples are relevant due to the massive difference in "stakes".

Counterpoints?

[waynebaumber]

I consider a draw against Herr Hughes (particularly if he is playing with early war Germans v my brave but out gunned Soviet's) a moral victory.

[Hugmenot]

My counterpoint is this is a competitive game, not war.

[Matt W]

To appropriately reflect the commander's performances the CE would have to include gradations for major and minor victories etc. While draws are useful stats with respect to a particular scenario it should be noted that many of the more recent scenarios have moved to an objective rather than a point based victory schedule that precludes draws but tends towards minor victories.

The real problem is that the statistic is based on data which, by scenario design, is not consistent so what are you measuring anyhow? The fact that Germans have a higher CE than the Soviets is more a result of scenario design. Either that or we all suck at playing the Soviets (always a possibility!).

[Shad]

The real problem is that the statistic is based on data which, by scenario design, is not consistent so what are you measuring anyhow? The fact that Germans have a higher CE than the Soviets is more a result of scenario design. Either that or we all suck at playing the Soviets (always a possibility!).

If you could compare your personal CE to the overall CE you'd have a very reasonable measure of your personal competence.

Revealing such stats might lead to hurt feelings and embarrassment, though.

[vince hughes]

To get really technical, shouldn't a program that were able to read such measures be able to give you a 'weighted' performance in respect of the scenarios you have played and their previous results.

Imagine your first 10 scenarios ftf you take the side that has lost the most in all 10 of them. Now, losing a game as the Soviets where the results in previous plays were 4 German wins & 3 Soviet wins would be far worse a loss than in a scenario that had 15 German wins and 1 Soviet win.

So there you have it, a program needs to be developed that can measure your weighted performance LOL !!! Make sure it is on my desk by Friday afternoon please Shad, I look forward to it.

[Shad]

That is eminently easy to do, but we're missing one thing: comparison data. There are too few repeat plays to support that analysis at the scenario level. Only broad strokes at the nation level would be doable at this point.