Quick Retreat Change

MithicYort

Here is the defense:

We managed to kill : 743 Knight 1498 Infantry 1214 Pikeman 862 Heavy Infantry 1414 Archer 1054 Mounted Archers

But we lost : 1015 Knight 688 Infantry 609 Pikeman 1435 Heavy Infantry 664 Archer 1701 Mounted Archers

These losses were taken with quick retreat strategy in place. Even with this strategy set, I sustained massive casualties. I didn't see what my troops were exactly after hit, but I estimate it between 60-75%

The whole point of this strategy is to give up land to protect troops but it doesn't feel like its functioning as described. I've listed out some thoughts below on how this could be improved.

Probably a quicker fix:
1) put a cap limit on losses with this strategy. Whatever percent would make sense. I have a couple thoughts.
a) A max losses based on total castles for example, if I have 5 castles, maximum troop losses would be 20% with this strategy. I realize once enough castles are in a players possession this would probable be too much protection. Perhaps once a kingdom is beyond 10 castles, the losses would be a flat 10% from then on.
b) A flat percentage across the board: 25% of stationed for example.

A more robust change, requiring much more time and resources to produce:
1) Add a completely separate strategy set for defense
a) Hold the line! (fight until the last man stands)
b) (Very complicated coding I imagine) Generals Discretion: Have some sort of mechanic that the defending army can "make an assessment" of the attacking force. If the algorith predicts a decisive victory, then they fight with everything available. If a decisive victory call can not be assessed then a retreat call is made, with a troop casualty rate greater than "Give up ground!" (the next strat) Say its 40-50% losses taken?
c) Give up ground! Basically a better "quick retreat" massively decrease odds of winning the battle, but cap losses less than "generals discretion" referencing a similar cap on casualties to the quicker fix mentioned above.

Lord Thane

New strats could be interesting. Hold the line sounds funnnnn.

I'll have to run some 1:1 battle tests to see how this is functioning. The guard towers and medical centers were working completely opposite of how they were intended to work, so very may well have some strat bugs to check out lol

DougEfresh

Lord Thane I saw a br earlier today where the defender was on QR and got decimated as well

trupheus

I can’t say this enough

2 biggest problems I see with attacking mechanics right now.
1) you defend with your entire army and losses are devastating defenders are often zeroed out. Not fun experience for most.
2) fodder isn't being fodder. The game today looks at damage and says oh you should lose x% then you lose that % with some entropy thrown in so it’s not a crystal clear 69.420% of your military lost. You’d see some 65% you’d see some 74% but the losses don’t make sense. You would lose 74% of a unit that made up maybe 5% of your army.

Lord Thane

trupheus I think #2 is going to be solved with @Leo suggestion for target weighted damage allocation.

And I'm going to look into #1 because we have the game guide source material telling us in only semi-vague terms that you defend with "around 70%" if you have multiple castles. I'm not sure if we have that mechanic built in or if it's just not working - need to dig into it to confirm.

Obviously getting the battle calcs dialed in is important, that's why I've been dogging on it since the first day of protection ended. Problem in getting a fix in is sifting through piles of reports for evidence of issues, then finding time to run proper tests to confirm and troubleshoot the code.

We're gonna release our admin battle sim to everyone to help with battle tests. This will be just like your send attack button, but instead of sending a real attack it will send a mock attack with real mock results. Should be helpful in further testing.

@TheKeeper is in agreement of further refinement of the battle calc process. So I think the next step is going to be we actually implement the target weighted damage allocation. Depending on how that goes, then we may asses the idea of the ranged nerf.

The next tests I have in my log is:
Test strategies in 1:1 test to confirm they're working.
Test reduced defenders with +2 castles for defender (probably going to be some 70-80% formula based on number of castles if it's not already implemented/workkng).
Test deployment of target weighted damage allocation.

TheKeeper is working on some new features and known bugs. I think by the end of next week we'll be in a lot better shape for most important issues. If I have time I'll also try whipping up a rudimentary battle calc people can use with their truesights to better gauge if an attack will be successful or not.

trupheus

Lord Thane

it's the same thing. he just put a nice spreadsheet together.

not sure what testing scripts you guys have but I play wreck it ralph with new features all the time. happy to help test.

bierlym

I concur with the original poster. QR is worthless against players 2x your size...I have lost 100% of troops home many times in the first week or two of the round.

Since the change where only a percentage of your troops defend, I still lost 70-80% of my total troops defending. So the attacker still killed 100% of my troops that were defending. only troops surviving are the ones that didn't defend. While not 100% troops loss, it is still brutal losses. need 30+ hours to retrain the archers I lost with 1:1 training centers. I would gladly give up the 1 castle and 16 land and still have the other 70-80% of my army.

Maybe a white flag strategy since QR is worthless for sparing your troops? When your troops are outnumbers 2:1 or more, you surrender and they let the troops live?

Lord Thane

I'll have to look into this again. I know the attack is halved, maybe the damages aren't being halved? Maybe it needs a more complex formula like the castle defense thing - do something where only half your troops participate in the battle so you can only lose half. So with the castle defense and QR, you'd only be able to lose 35-40% of your garrison in a defense. We may need to vote on this since it goes against the holy documents in game guide.

Lord Thane

Something else I thought of while in the shower just now, is to not forget about med centers. If you max them out 2:1 you can save 20% casualties. So you can turn a "lose 4/5ths" into a "lose 2/3" really easy just by having the buildings. Say you have 100 men and defending with 20%, your 80 guys get wiped but with 20% MC you get 16 of those guys back, end result is you get to have 36 units after the attack VS only 20 without MC. Could be an interesting strategy for having troops available for retals.

bierlym

The MC do spare some casualties, but at 2:1 means your economy and military only have 80% of the buildings. So both will be much weaker. guess it is a trade off. Not tested the balance between them. Out of curiosity, in a regular round with finite resources, could anyone run that kind of MC or guard towers and still muster a decent army?

My recollection from 20 years ago is almost noone ran very many MS or Guard towers. but that was a long time ago, and my memory isn't quite as good as back then. lol

Lord Thane

bierlym yeah I think it's rare, but it depends on play style a lot too. Maybe someone who only logs in once or twice per day might run high MC, so they can have some units left alive to possible run an attack. GT are always a huge FU surprise too, if the attacker doesn't have high enough level magic for a detailed true sight, they might undersend an attack not realizing so much of their ranged units are getting nullified, end up getting spanked.

Some nerds say it's better to have stronger economy to rebuild armies vs wasting the building space. Idk if their math checks out, but I will prolly be running some GT in the early part of a real game round, and then maybe switch to MC later in the round when it can save mucher bigger troop counts.

trupheus

Lord Thane

In a real round I don’t ever see the opportunity cost of 2:1 MCs being worth it unless I’m farming at the top waiting for people to try to break me.