View Issue Details

IDProjectCategoryLast Update
0014688The Last FederationGameplay IssueApr 28, 2014 1:42 pm
ReporterHistidine Assigned Tokeith.lamothe  
Status resolvedResolutionfixed 
Product Version1.005 Beta (Holy Smokes It's Big) 
Fixed in Version1.006 Beta (Besieging And Federating) 
Summary0014688: Defender armadas instagib attackers
DescriptionIn the attached save the Skylaxians are launching an attack on the Boarines. They have a large qualitative and quantitative advantage, but it doesn't matter because any Skylaxian armada that touches the Boarine planet gets killed almost instantly while inflicting only trivial losses, even if the fleet has near-parity in strength rating just prior to the combat tick. Something seems to be way off with the combat calculations.

As far as I can see, this has been an issue since the alpha.
TagsNo tags attached.
Internal WeightNew

Relationships

related to 0014615 resolvedkeith.lamothe Superior fleet (5 times power) crushed by boarine defenses 

Activities

Histidine

Apr 26, 2014 6:32 am

reporter  

DefendersOP.save (498,493 bytes)

Histidine

Apr 26, 2014 7:30 am

reporter   ~0036840

Last edited: Apr 26, 2014 7:33 am

Sabotaging all the ion cannons on the planet doesn't help.

Some time later however, the Boarine armadas disappeared (didn't quite catch it, but I think the Boarines decided to go on the offensive and took their turn at losing all their armadas to this bug), and the Skylaxians sent a fresh fleet and pulverized their planet instantly.

wyvern83

Apr 26, 2014 7:05 pm

reporter   ~0036854

I've seen this as well, same races and roles too.

keith.lamothe

Apr 28, 2014 1:42 pm

administrator   ~0036932

For 1.006:

=== NPC Fleet vs NPC Fleet Autoresolve Combat Overhaul ===

* Basically this was using logic that was mostly from way before alpha, and that logic was much better suited to the strategic realities obtaining at that time than now. Now most significant NPC vs NPC fleet combat involves dozens if not hundreds of armadas, and (at least later in the game, with tech bonuses) very large power values.
** This may result in NPC fleet battles happening substantially faster in some cases (but waaaay slower in others like those we tested, since they were insta-gib situations). If it's out of control please let us know and we can tweak some dials.

* Previously individual armadas would pair off and shoot at each other, with the weaker armada only having a very small chance of damaging the stronger one (and even then only a little bit, relatively), while the stronger one got to basically unload on the weaker one. This made some sense with one-vs-one combats but not in dozens-vs-dozens where there would often be a large mix of very-small and somewhat-larger armadas, so the overall result was being unduly impacted by which armadas happened to "draw" which opponent armadas for each round.
** Now in a round of such combat each armada will do its damage to the other armada, simultaneously. So the less-powerful one will still get mauled (relatively, it still takes months for similarly-sized forces to really smash each other) but the more-powerful one will now get hurt in due proportion.

* Previously the various bonuses which multiplied a fleet's autoresolve strength (tech bonuses, boarine rage multiplier, burlust turmoil multiplier, etc) were only being applied to armada attack power, not to its ability to absorb damage. This led to insta-gib situations later in the game. A much-more-rapid conclusion is still possible if one race has a MUCH higher effective bonus than its opponent, but that is desirable.

* Previously if hundreds of attacker armadas pounced on a planet with only a dozen or so defender armadas, each day all of the attacker armadas would do a full round of combat with a defender armada. Interestingly, this often tended to massively favor the _defender_, as the defender generally has larger individual armadas and was getting full shots at every single one of the attacking armadas despite being greivously outnumbered. So it was possible for an attacker which (in terms of the numeric power displayed on the overlays) out-strength'd the defender by over 5x to still actually _lose_.
** Now each attacker armada still attacks a defender armada each day, but:
*** It always picks a defender armada that has not yet fought that day, if there is one.
*** If there isn't such a defender armada, it picks one that has fought the least number of times that day (so if they've all fought once, it will pick them all once again before any one is picked a third time).
**** In the example where all the defenders have fought once, and one is picked again, it divides the attacker's effective attack power by 1.15 and divides the defender's effective attack power by 1.4.
**** If all the defenders have fought twice, and one is picked a third time, the divisors are 1.3225 and 1.96, respectively. And so on, basically 1.15^x and 1.4^x.
**** This represents the growing difficulty to efficiently fight in crowded conditions, and that the defender is going to get the worse end of such a combat since all the attackers are coming in will full weapon charges vs their outnumbered weapons, etc.

* Fixed a bug where the space power numbers displayed in the overlays (for both attacker and defender) were not actually accounting for all the various things which impact autoresolve combat.


Thanks :)

Issue History

Date Modified Username Field Change
Apr 26, 2014 6:32 am Histidine New Issue
Apr 26, 2014 6:32 am Histidine File Added: DefendersOP.save
Apr 26, 2014 6:32 am Histidine Description Updated
Apr 26, 2014 7:03 am L4m3ness Relationship added related to 0014615
Apr 26, 2014 7:30 am Histidine Note Added: 0036840
Apr 26, 2014 7:33 am Histidine Note Edited: 0036840
Apr 26, 2014 7:05 pm wyvern83 Note Added: 0036854
Apr 28, 2014 1:42 pm keith.lamothe Internal Weight => New
Apr 28, 2014 1:42 pm keith.lamothe Note Added: 0036932
Apr 28, 2014 1:42 pm keith.lamothe Status new => resolved
Apr 28, 2014 1:42 pm keith.lamothe Fixed in Version => 1.006 Beta (Besieging And Federating)
Apr 28, 2014 1:42 pm keith.lamothe Resolution open => fixed
Apr 28, 2014 1:42 pm keith.lamothe Assigned To => keith.lamothe