I am curious if anyone knows why it is that sometimes the pie chart shows a clear advantage to a player and then when you actually do the battle, it was way off. This cannot be blamed on bad luck or bad battle decisions either. You can clearly see the attack odds show something each attack that was not reflected in the pie chart, resulting in a completely different outcome no matter how many times you reload. Quite often the pie chart is all we have to go on to decided whether we want to fight or not. I wish it was more accurate.



Another related complaint is how if you are the attacker you cannot back out of the attack even when you find the pie chart showing you are going get wiped.



Both of these issues find me saving almost every turn. A quick save option would have been helpful here.



Lastly on a good note, since the big patch, the AI seems much better. Its more aggressive now and much more fun. Good job here.

I believe the pie chart is not intended to show "attack odds", but instead shows the ratio of total army HP. HP is an important stat, but obviously, the other stats and abilities still have some effect on the actual outcome.



But it's not "all you have to go on"--you can see the individual units participating in the battle and what all of their stats and abilities are. (This is more laborious, of course, but I'm not sure there's a good middle-ground between "look at the raw data" and "ignore your personal skill and trust the computer's judgement".)



You can also examine the units in the enemy army (and their stats) before you attack them, which is probably why they don't feel the attacker needs a "retreat" option. (Though it's possible to pull in enemy reinforcements you didn't previously see.) Also, retreating costs an action point, and you already used up your action point by launching the attack. And it would be kind of weird if both sides retreated (I guess you both lose half your HP, because of...what?)

Antistone wrote:
I believe the pie chart is not intended to show "attack odds", but instead shows the ratio of total army HP. HP is an important stat, but obviously, the other stats and abilities still have some effect on the actual outcome.

Close - it's an invisible unit attribute called "military power". A unit's military power is calculated based on all its other attributes (attack, defence, hitpoints, damage, etc) and is also influenced by its capacities (a unit with circular attack will have 1.5 times more military power, for example). An army's military power is the sum of its units' military power values, and the chart shows the relative military power of the two armies.



Needless to say since military power is only an approximation and since it doesn't take into account the mix of different units on the field, the terrain and so on, having a higher military power is no guarantee of success.

wilbefast wrote:
Close - it's an invisible unit attribute called "military power". A unit's military power is calculated based on all its other attributes (attack, defence, hitpoints, damage, etc) and is also influenced by its capacities (a unit with circular attack will have 1.5 times more military power, for example). An army's military power is the sum of its units' military power values, and the chart shows the relative military power of the two armies.



Needless to say since military power is only an approximation and since it doesn't take into account the mix of different units on the field, the terrain and so on, having a higher military power is no guarantee of success.

Doesn't this throw the AI off? Since I'm assuming they use this to calculate their chances at victory. Or am I missing something?

Slashman wrote:
Doesn't this throw the AI off? Since I'm assuming they use this to calculate their chances at victory. Or am I missing something?

The AI originally used just this balance of military power to decide whether to fight or to retreat, but it now also uses a bunch of other metrics like the size of its deployment zone, the presence of forests and high ground, etc. This does mean that the AI's behaviour contradicts the chart sometimes though.

wilbefast wrote:
The AI originally used just this balance of military power to decide whether to fight or to retreat, but it now also uses a bunch of other metrics like the size of its deployment zone, the presence of forests and high ground, etc. This does mean that the AI's behaviour contradicts the chart sometimes though.

Ah OK. I get it now. Thanks for the reply!

wilbefast wrote:
Close - it's an invisible unit attribute called "military power". A unit's military power is calculated based on all its other attributes (attack, defence, hitpoints, damage, etc) and is also influenced by its capacities (a unit with circular attack will have 1.5 times more military power, for example). An army's military power is the sum of its units' military power values, and the chart shows the relative military power of the two armies.



Needless to say since military power is only an approximation and since it doesn't take into account the mix of different units on the field, the terrain and so on, having a higher military power is no guarantee of success.

Ok, thanks for the info.

To me it seems that sometimes the being "off" comes from not correctly calculating in reinforcements. Normally with only 2 armies, one on each side, its correct and reliable and the autobattle is mostly according to the graph. However if there are reinforcements it seems to me that in some cases only the first row gets calculated for humans at least or on defense while the attacker has all power on his side, the defender doesnt get calculated correctly. But its just a feeling, since i had that often repeating...being so weak that autobattle would result in a heavy loss but since reinforcments were there and i just wanted to try it out...i always won, often without losses. So my guess, there is a bug in there somewhere.