Soldier Desertions

[Clampy]

To confirm, it is possible to have elites desert due to insufficent soldiers, but then have say thieves not desert due to sufficient soldiers.

For example:
If you would expect to have the following amount of each troop desert:
500 elites
50 thieves
300 defspecs
200 offspecs
50 soldiers

and you had 280 soldiers, you would lose:
500 elites
50 soldiers (instead of thieves)
300 defspecs
200 soldiers (instead of offspecs)
30 soldiers

It checks fresh each time if you have enough soldiers.

[Leonidas]

To confirm, it is possible to have elites desert due to insufficent soldiers, but then have say thieves not desert due to sufficient soldiers.

For example:
If you would expect to have the following amount of each troop desert:
500 elites
50 thieves
300 defspecs
200 offspecs
50 soldiers

and you had 280 soldiers, you would lose:
500 elites
50 soldiers (instead of thieves)
300 defspecs
200 soldiers (instead of offspecs)
30 soldiers

It checks fresh each time if you have enough soldiers.

If this is confirmed then the formula needs updating to ignore prior desertions

[Persain]

If this is confirmed then the formula needs updating to ignore prior desertions

i'll do that then "soon"

[Andurilas]

I still don't understand Pathetic Sheep's prior point that if you have enough soldiers to "cover" the deserting other-troop-type, then you will lose only 10% of the soldiers on hand instead of the soldiers actually substituting for the otherwise-would-have-been-lost units. That is, I thought the logic was:

1. Calculate metric = m.
2. Calculate number to desert: # of [elites] * m = x.
3. Check for x soldiers.
4. If # of soldiers ≥ x, lose x soldiers.
5. Repeat for next troop type.

But he seems to suggest that step four is actually:
4. If # of soldiers ≥ x, lose # of soldiers * m.

Is this really true?

[pathetic sheep]

I still don't understand Pathetic Sheep's prior point that if you have enough soldiers to "cover" the deserting other-troop-type, then you will lose only 10% of the soldiers on hand instead of the soldiers actually substituting for the otherwise-would-have-been-lost units. That is, I thought the logic was:

1. Calculate metric = m.
2. Calculate number to desert: # of [elites] * m = x.
3. Check for x soldiers.
4. If # of soldiers ≥ x, lose x soldiers.
5. Repeat for next troop type.

But he seems to suggest that step four is actually:
4. If # of soldiers ≥ x, lose # of soldiers * m.

Is this really true?

http://wiki.utopia-game.com/index.ph...Overpopulation

Leet Desertions = IF [(Leets home * Metric) > Solds home] Then (Leets home * Metric) Else 0

Either elites desert or 0 desert.
Either thieves desert or none desert. Same with specs and specs.

The last calculation is soldiers. The calculation that effects soldiers is not a "if/then" statement. It is a "max" statement. The number of soldiers that desert is determined after the number of elites have been determined.

When calculating the elite losses:

But he seems to suggest that step four is actually:
4. If # of soldiers ≥ x, lose # of soldiers * m.

4. If # of soldiers ≥ x, lose 0.

If you have elite losses it will lower soldier losses. When they say "get enough soldiers so you can lose them instead of elites" they mean "get your elite losses down to 0 by having enough soldiers". Only when you look at the whole tic event can you say "I had soldiers desert in place of my elites". But the number of soldiers that are lost is determined at the end. The numbers are not equal. It is not a substitution.

[TheOne]

Nsing a chained prov who is using soldiers to combat desertion a few seconds before tick is a top tier strategy used in top wars.

Sadly, not many such wars exists anymore.

[Andurilas]

Oh!

[2Utopian]

Wow, clampy sighting. I had no idea he was still around. Your contributions to understanding mechanics are missed by me at least.

[Andurilas]

Forgive my continued ignorance, but I'm having a little trouble figuring out what's going on with subtracting previous desertions from each troop type and then adding the number deserted back in at the soldier desertion.

[pathetic sheep]

...I'm having a little trouble figuring out what's going on with subtracting previous desertions from each troop type

You can word that as "If your elites desert then every trained type also deserts". All trained troops times metric.

Leet Desertions = IF [(Leets home * Metric) > Solds home] Then (Leets home * Metric) Else 0

Thieves Desertions = IF [(Thieves home * Metric) > (Solds home - Leet desertions)] Then (Thieves home* Metric) Else 0

Do the algebra. If a > n then (n - a) < 0.
Your (Thieves home * Metric) has to be a positive number.
For positive real numbers a and b if a > n then b > (n - a)

then adding the number deserted back in at the soldier desertion.

Look at the brackets:

Solds Desertions = Max[(Solds home-{Leet Desertions + Thief Desertions + Defspecs Desertions + OffSpecs Desertions}) * Metric,0]

(Solds home-{Leet Desertions + Thief Desertions + Defspecs Desertions + OffSpecs Desertions})
(Solds home-{stuff})
"stuff" has to be 0 or it has to be larger than "soldiers home".

So that could be stated "if anything trained deserts then your soldiers do not desert".

[Andurilas]

Thanks again. Those mechanisms seem a little random (like, why don't soldiers desert anyway?) but I understand them now.

[pathetic sheep]

Thanks again. Those mechanisms seem a little random (like, why don't soldiers desert anyway?) but I understand them now.

Look at at 4 hour period. Hour #1 elites+ desert, #2 elites+ desert, # 3 specs desert, #4 soldiers desert

If your soldiers had been deserting in hour #1 then you would have had elite desertions in hours #3 and #4. You would never have enough "soldiers home" to cause zero elite desertions.

[Andurilas]

Well, why not allow soldiers to cover each troop type independent of the previous? So in the first hour, soldiers can fail to cover elites (that is, Soldiers < Elites* Metric) but then maybe there's enough to cover your thieves (Thieves * Metric < Soldiers) so they don't desert.

[Leonidas]

Well, why not allow soldiers to cover each troop type independent of the previous? So in the first hour, soldiers can fail to cover elites (that is, Soldiers < Elites* Metric) but then maybe there's enough to cover your thieves (Thieves * Metric < Soldiers) so they don't desert.

Clampy said this is how it works, Persain said he didn't have the chance to confirm the formula. I'm not hoping I get a chance to check this during war of course, but if someone does, do let us know!

[Andurilas]

Ha I'm glad I'm catching up to the adults in the conversation, ever so slowly. Thanks for humoring my questions.