I don't get the combat preview

[bob0809]

At first I thought the combat preview shows the odds for each result to happen, i.e. a big black part means an equally big chance for unit destruction. However, I notice that the combat preview changes slightly every time when you undo a move and redo it. Can you explain what this actually shows? I mean, the odds would have to be always the same, no?

[Pat]

I'm just replying so you know you're not being ignored.

The Studio is closed for Christmas and I can't say when you might get a useful reply. This is a question that's come up before and I remember us debating it when we were first developing Bulge but that seems so long ago now I can't for the life of me remember the reason. I'm sure someone will be on to give you a good answer but in the mean time have a great holiday.

[bob0809]

Thanks! Take your time and enjoy the holidays, I can wait.

I Found this by Miguel:

Just a quick thing I should point out is that the number in the combat preview is a count of units in the battle Not a ratio. Hit the "v" on the combat preview to see a combat outcome histogram.

(http://www.shenandoah-studio.com/Forum/ ... =119&p=707)

An interesting decision to display a histogram and not just the odds. The only reason I can think of is to preserve a certain degree of unpredictability. Or is there another reason?

[Pat]

Miguel's actually talking about the mini-combat preview that we have in Bulge. That's not present in Moscow which just sopens the full preview with the histogram.

[bob0809]

Yes, I shoild have just quoted the second sentence. The histogram still works the same right?

[Pat]

Yes the histogram should be as good as identical.

Pat

[daviddunham]

And it is not odds, it is a Monte Carlo simulation, which is why it can be a little different each time.

[s_nkarp]

And it is not odds, it is a Monte Carlo simulation, which is why it can be a little different each time.

Per David's comment, a combat is run some dozens of times. So there may be some variation between one batch and another, but it will be quite close.

[sa_gibson]

Do you guys have data which track the predicted result - that shown by the preview - vs the actual result of the combat? I understand the subjective nature of recollections, but it seems to me the preview is more often wrong than right. I'm sure that's not the case, but some data would be helpful in convincing my left brain...

[s_nkarp]

We do test the preview against the theoretical values - it matches quite well. The thing to remember is that even the most likely result is often unlikely: there are a lot of possible outcomes for any fight, so there's a good chance that one of many unusual events crops up in any given battle.