The Complete EDU Guide

[Aradan]

No idea what's the maximum, theoretically you can give a very high value, but it's probably capping internally at some point, which is a bit complex to estimate. The lowest value is 0 obviously.

[Apostle Zodd]

thank you

[Athenogoras]

Thanks Aradan and all who have contributed to this guide.
An observation from my own edu-modding on RTW/BI.

Stat_heat penalty: Penalty ranges from 0 to 10. Only positive and even numbers count.
That is 0,2,4,6,8 and 10.

I used 3 units with same stats and skeleton.(spearmen). Also tested cavalry( two exact units).

[atraps]

Stat_heat penalty: Penalty ranges from 0 to 10. Only positive and even numbers count.
That is 0,2,4,6,8 and 10.

I do not think that bit is true as the vanilla EDU uses odd, and negative numbers for heat penalties.

[Athenogoras]

I do not think that bit is true as the vanilla EDU uses odd, and negative numbers for heat penalties.

That is what I thought also. But after testing I noticed no difference between 0 and -2(-4,-6 etc) in fatigue.

[Aradan]

We ran quite a few tests on stat_heat about... 18 months ago, that suggest odd and negative values do count. I must find the tables with the data or at least the results, and post them here.

[atraps]

Well, negative heat penalties either mean that the unit gains a fatigue benefit in a desert climate, and/or, gains a fatigue penalty in winter climates. Have you tested both possibilities?

Also remember that the max could be 2 for the negatives, as that's the highest I've seen in the vanilla EDU.

As for the odd numbers, I really think they count, as if they don't lots of the heat penalties in the vanilla EDU are meaningless.

[Aradan]

Okies, HERE is what we posted at the Org back then. IT doesn't mention anything about odd values, but I'm pretty sure it does make a difference. I just can't find the data atm. As you can see negative values do make a difference (the amount of difference depends greatly on the climate_heat of the map)

[Athenogoras]

Well, negative heat penalties either mean that the unit gains a fatigue benefit in a desert climate, and/or, gains a fatigue penalty in winter climates.

Fatigue is tied to the climate_heat of the battlemap. The results are the same although the difference is much smaller in winterterrain than in desertterrain.

Also. Training affects fatigue. A highly_trained unit tires less fast than a trained unit, all else equal. Although the difference is much smaller than the stat_heat numbers.

Edit: All test done on desertmap custom-battles.

Edit2: Atraps: Do a test. Take Nile spearmen and Pharao guards. Make sure they both have the "hardy" attribute and that they both are highly trained. Give Nile spears -2 in stat_heat(they have -1 I think) and Pharaos Guard 0. Use syrian flats-map. Set them of phalanx. Run them across the map.

[Aradan]

Well, it' actually always better when you do the tests with the exact same unit or with clone units (identical units, just different internal name), just to be perfectly safe.

Training affects fatigue loss rate? Can you post some data?

[Athenogoras]

I did all infantrytest with fs_medium_spearmen. All edu-stats similar.

Did a time-test.
With highly_trained. From exhausted to fresh: 10:38
With trained. From exhausted to fresh:13:51
Same unit.
There is only difference between highly and normal/untrained. No difference between normal-untrained.

[Aradan]

That's a huge difference there, 30 percent or so! Thx for the info, very useful.

[Athenogoras]

Yes. Tests were done on desert map with stat_heat 10 only to amplify the difference. The difference is not so much(although there is a difference) when are getting exhausted as to how fast they recover.

PS I spent 30 minutes with a clock in my hand watching a screen. I actually checked my window so noone was watching. Its kind of tragic.

[Aradan]

Haha, that's unit stat testing for ya!

[florin80]

Do you think the general`s bodyguards should get the same moral as the corresponding regular troops(companion cavalry for example) or should they get lower? I`m saying lower because I`m thinking it might help them stay alive longer by not being too cunning, but I`m not sure if this has any real effect on their behavior. Any idea?

[Aradan]

I'd actually say higher; from a gameplay point of view, you can't have the general's bodyguard break off and run before regular troops do, i will cause massive routs etc.

[florin80]

Regarding stat_cost. Are the different costs(from recruitment to upgrade) completely independed of each other or is it necessary/recommended to have a certain balance between the sums?

[Aradan]

Depends on the balance system you prefer.

I prefer having the upkeep cost to be training cost/modifier, where modifier is a value that depends on the unit class. Elite/superior/veteran units that would form a factions standing army get a higher modifier, which means a lower upkeep cost/recruitment cost ratio, while peasants/levies/militias and the various irregular forces would get a low modifier, which means that they cost more to upkeep comparatively (as a ratio, not an actual value). So, low-class units are cheaper to train, but cost much to maintain and are better disbanded after fulfilling their purpose, while better units are expensive to train, but are ideal to form the standing army.

Armour/weapon upgrade costs depend on a value that we name "overall campaign cost" (depends on unit stats+size+difficulty-to-acquire+other stuff) which in turn also affects recruitment cost, but recruitment cost and upgrade costs are not directly linked.

Makes sense?

[florin80]

How could it not? I`ve never seen it better expressed. Thank you! One question though. How`s the ai handling this system? The most widely available units will probably be the basic ones and because they are also the cheapest, won`t the ai be tempted to buy many of them instead of buying more of the professionals? Because in the long run it might affect it economically since the ai is not smart enough to disband afaik.

[Aradan]

Yes, the AI does struggle a bit, so you have to find a balance. An addiitonal 'trick' used is that as the barracks get larger, you start removing the lower level units, so (for example) the AI no longer has the ability to train militias when it is in a position to train superior units, etc. The player also loses that ability, but at least he/she is smart enough to take his militias to a lower-tier settlement for retraining. It's not perfect, yes, but it's a solution...