Ship loss ratio


#1

There still seems to be no rhyme or reason to the ratio of ship loss when attacking/getting attacked. Sometimes you lose ten ships per the attacker’s one, sometimes the opposite. Please explain.


#2

Battle results depend most heavily on how well the attacker does – 0% destroyed will always be an abject failure, and above 70% some additional factors start to matter as well (e.g., Primarch stats). Relative army size also plays a small part.


#3

…that still sounds fairly vague. :expressionless:


#4

True. To be honest, I think we’re more communicative than any computer-based game I’ve ever played and I get a bit nervous that we’re actually sharing too much :stuck_out_tongue:. Personally, one thing I (weirdly) enjoy about games it running experiments and trying to figure out how they work. Kind of like taking stuff apart and putting it back together again. I think it’s been very helpful to be more open with beta, though I to be completely candid I do second-guess the level of detail I provide sometimes … we’ll see how it goes.

Posting a lot more detail on http://forums.wardragons.com/t/sailors-hats-regeneration/2721/15 shortly.


#5

So was there changes to both the rescued sailor function and the repair function?


Sailors/hats regeneration
#6

The rescued sailor function was completely removed and repair rates were slashed.

Attackers now get about 10% of troops back for repair with successful attacks and defenders get variable amounts for repair-anywhere from 20-50% or so in my experience.

The more successful an attack or defense is, the lower your repair percentage (you do get more Glory, however, as they are inversely related).


#7

I mean, I like reverse engineering formulas and stuff too but I don’t have the luxury to do that here. I got no troops to test with :joy:


#8

I think the beta makes less and less sense. Islands are essentially meaningless for 90% of the current users and the one thing people want to do is being constricted further. I really don’t get how this is going to work…


#9

Islands seem valuable to me because of these benefits:

  1. Huge boost to crafting material rewards --> more & better gear
  2. Daily token payout (varies with how much you own, but even medium teams can get nearly 10k extra tokens each month).
  3. More Bonus XP
  4. More gold payout & reduced troop training time --> more troops --> more event prizes

#10

You get a slight boost IF you can hold your island. Most teams can’t and the islands don’t give enough benefit to be willing to. Honestly in terms of value ships are worth way more than the crafting items or gold or even tokens. The islands don’t really provide that much value for most teams and even if they wanted to. They’re simply unable to.

As far as the sailors go, I’m not sure sure why that got changed and definitely repairs…those seemed really important for helping teams that just got steamrolled…

I think there is a clear divide between what PG thinks is value vs the players side. For example, if you gave out tons of dragon boosts but then made it so we could only have 5 heal potions. Sure, it’s great you can boost your dragon. Except you can’t use it. So people can get more gold to do what? Build up an island they can’t hope to defend? More shards, okay so your rider has better gear. Except you have no island or boats so you can’t actually do anything.

The value the islands give is ONLY if you can retain them. But since boats are purchaseable, these things can all be bypassed. One thing that I would be interesting. If you could only have so many boats or level your primarch so far based on your island infrastructure? Then you have to give the islands meaning. But you would also have to make them much more defendable.


#11

I agree with the sentiment here that changes need to be made. But I wanted to first provide some background for the discussion. Each troop is currently worth up to 1 glory (ignoring revives) (an arbitrary value, but it needs to be something … 1 is simple). This makes it possible to give out event rewards for glory, because troops have a predictable value. It also informs the glory costs for upgrading Primarchs and Dragon Riders.

Players who attack or defend poorly get less glory (potentially much, much less). To make sure they’re not hurt too badly, we let them revive a significant fraction of the glory “potential” they lost. For the sake of example, say a player loses 100 troops and earns 20 glory because they only killed 20 of the attackers’ ships. Then they’ve lost 80 glory potential, and so we might let the revive enough ships to net 80 glory … net is important because it means we account for the gold cost to revive. In this example, under the current params, we would allow the defender to revive 94.7% of their lost ships, or 94. At a glance, it seems like the defender got a really good deal … once they revive their ships, they’ll have more ships that the attacker if they both started with the same number of ships. However, that intuition is incorrect because revived ships actually cost resources to revive (gold, sailors and time). If the attacker spent the same resources building new ships that the defender spent reviving, then the attacker would have a net of ~20 more ships than the defender (a 20% difference!).

A detail I omitted in the already lengthy example in the previous paragraph is that there is a minimum revive percentage. This means that the winner of the battle is entitled to revive at least 10.9% of their losses no matter what. In the event that the winner won by a large margin, they would normally have nothing to revive because they fully realized the glory potential of their dead troops. However, thanks to the minimum revive % they get to revive ~11% of dead troops anyway … this actually increases the maximum glory potential of each troop geometrically.

The attacker also has the advantage of having decided the time and place of the battle and gained a big tactical advantage (ships now are better than ships sometime down the road). So the attacker is the clear winner.

Note that this situation can also be reversed. If we reversed the outcomes for the attacker and defender, then the defender would have enjoyed this exact same outcome. For the sake of the example and relative brevity, I’ll leave it at that.

As I said above, I think the current problem is that attackers feel like they lose too much of their army. If this is true, then defenders must feel this even more acutely – because, as I described above, the attacker is much better off than the defender (20% better off in our example!).

Based on the feedback in this thread, I want to try to increase that maximum reparable value because this will make attacking less punishing. Attacking should be fun and something players get to routinely participate in (say, every major glory event in Atlas). Unfortunately, it’s not as easy as it sounds. If I raise it any at all from where it is now, then losers who lose badly will be punished harder that now. Specifically, if an attacker kills someone 10:1, then defender currently gets a 99.9% revive rate … it can’t go higher (this is actually part of how the minimum revive percent was calculated, and why it isn’t a nice round number).

That is probably okay to some extent because losing should be a little painful. But if the spread between top attacker and worst defender gets too ridiculous then griefing could be exacerbated. So there’s a fine line which we need to tread!

Another factor which needs to be considered is the glory value of a troop. When the minimum revive percent is higher, then the potential glory earnings from one troop goes up (1 troop = 1 + 1minimumReviveRate^1 + 1minimumReviveRate^2 + …) because reviving a troop over and over is relatively more effective when this rate is higher. To avoid changing the glory economy, this means we need to implement a counterweight – e.g., reduce the base glory value of 1 troop (from it’s current value of 1.0) or increase the glory cost of upgrades and prizes. The former is preferable because it has zero impact (good or bad) on glory people have saved up. The latter is not so good for players because it deflates the value of unspent glory. So we’ll just consider the former option since it is strictly better.

All that said, I explored what would happen if we raised the revive rate to various levels.

Minimum % Revivable New Base Glory Value / Troop Worst-case Loser Glory Potential Loss* Notes
10.9% 1.0 0% current!
50.0% .436 35% (7% if evenly matched)
70.0% .159 53% (21% if evenly matched) kind of rough if griefed; big change in glory earned versus current battles (6.2x less)
80.0% .068 62% (28% if evenly matched)
* Ignore revive costs. Add roughly 20% to account for revive costs.

Minimum revive rates of 50%

seem like a good compromise. What do you think? Glory will take about twice as many battles to earn (for the same size battles) as before, but you’ll lose up almost 5x fewer ships than now (they’ll be revivable). Losers will get hurt a bit worse, but it won’t be too catastrophic (just enough to be pissed at your aggressor and want to hit back!).


#12

I think the change is too new to be sure how it will work out.

And before anyone makes any assumptions, NO, I am not a fan of taking away the rescued sailors/troops/hats/whatever. I am down by 120k troops compared to 5 weeks ago, I have been steadily loosing faster than I can repair… but they system is changing every few weeks lately (it feels like) AND we just got new blood in Atlas/beta. Give it a little longer to collect some more data before changing it again so it can be a better informed change, not a “knee-jerk” reaction.

Doc


#13

Glory taking longer to earn in exchange for more repaiable ships seems a good soulution if unwilling to go back to the last iteration. Will be interesting to see it play out. The last adjustment on repairables combined with the glitchy beta base has me only want to defend.


#14

The most jarring change I found was the removal of rescued sailors/peasants. In a previous pvp I actually ended up high on the leaderboard due to not sleeping and staring at the screen all day, killing anything that moved. I was only able to do this because I got peasants back and I could rebuild my ~10k troops over and over. I actually found this to be better than having troops to revive. Clocks and gold I have and can farm or win more of. Only way to get more troops on demand besides using rubies is bullhorns and those depend on being lucky in bazaar


#15

Peasants/hats are more valuable than gold :scream:.
Right now PG assumes we want glory more than we want troops.

It is hurting.


#16

Every revivable ship is like a gift of 0.6 peasants (plus gold and time). We could certainly turn off reviving for attackers and bring back sailor saving. (Saving peasants is still possible, it’s just param’ed to 0 right now and attacker reviving is param’ed up from 0)


#17

All I know is after attacking I’m able to get back MUCH LESS Troops back than I was before the 3.70 update. Also glory payout sucks. Change it back


#18

Revivable ships are awesome. I love having them. Zero question about that.
Extra peasants are awesome.
Both together is an airship/troop building orgy.

I think people miss having some of both.

Of course, our big complaint was people building too many ships too fast before. (Mostly building them out of credit cards). Now the issue is building them too slowly, unless you build them out of credit cards (which has slowed, troop:dollar, but the question is if it slowed relative to how much the free building slowed… I do not have the data to even hazard a guess).

Probably the higher revive, lower glory scenario that PGDave mentioned somewhere (a 50% scenario) will be a good solution, probably. However, I think it is too soon to know how this number set is panning out for sure.

We have to seriously consider: do we value ships/troops more or glory more?
We SAY we want our ships, but then individuals all go scurrying off to get glory points for beta event prizes and waste their ships. Heck, all my glory points recently (190k) came from defending my regions from friendly fire!!! Just because people wanted event points! What a waste of ships! Someone clearly values glory over ships. :sweat:


#19

Thanks for the detailed explanation @PGDave

As discussed elsewhere,

One option with the above is to have instant troops as rewards available for glory. Hence if you do really well and earn a lot of glory during a battle, you can pickup some troops from rewards. If you do poorly in battle you get lots of revives.

Some thoughts

  1. current problem is attacker loses momentum. Yes I might have lots of glory, but without any troops I’m stuck doing nothing for a few weeks.

Example

I have say 20 000 troops and get involved in a war,

All attacks are highly successful by me so 10:1 ratios etc

1st attack: (me vs player 1, team A) I kill 10k troops, lose 1k. Lots of glory, revive 100. Net: i have 19100 troops.
2nd attack: (player 2, team B vs me) I get attacked on taunter , outcome is I lose 8000 attacker loses 20000. Lots of glory, 15% revive =1200. Net is I have 12300 troops.
3rd attack: repeats scenario one. Net is I have 11400.
4th attack: repeats Scenario 2. Net is i have 4600
5th And 6th: net is i now have 0 troops.

So ive gotten maximum glory potential (or near maximum) by excellent execution. I’ve got tonnes of glory, and in theory im great. I’m much better off than the 6 different players from 3 different teams that I fought with in GLORY realised.

However, practically I’m screwed. I now have nothing I can do for two weeks till I’ve rebuilt, and granted those specific 6 players combined are worse off then me, but the thousands of others out there can now wipe the floor with me. Even if they don’t, i cant do any more fighting, even though technically my fights went brilliantly.

Not fun. Feels like punishment. It isn’t really - net glory was i got maximum rewards, but practically I cant play.

So… Solutions options

Glory earned can translate to new troops faster. Eg make troop regen much faster after highly successful fight for the successor.

Or

Troop rewards for glory (exchange). This has to be forwards looking only. Not for tens of thousands of glory stored on easily maxxed primes of first iterations.

Or

Instant summon troop horns for aggressor in battle. With revive for defender. The above weights correctly adjusted to factor all costs. This will solve the agressor momentum thing a little bit.

Lastly, on a side note. The gold"value" of things are too high. Eg 50million gold isnt worth 43k rubies and never will be. Hence why revive numbers are so low. Factoring in too much value for gold


#20

I get why you use Glory as a marker, but I do think an unintended consequence is that it feels like success is being punished. The math might disagree with me, but it felt better pre-3.70 to know I’d get enough sailors back from attacking successfully to build about 70% of my ships back. Now I get the minimum pretty much every time and it feels more alarming to see my troop stock keep dwindling even though technically I come out 20% ahead.

If I recall correctly, you actually got rewarded pre-3.70 for success in attacking by getting more recoverable troops. This now seems flipped on its head whereby Glory is king and ship inventory suffers when Glory is higher.

I’d either like to see troop recovery added back to the game, or some instant build ships awarded for attackers as suggested or something to avoid what is a death spiral for active players–you get zeroed out faster by succeeding. This is a bizarre result.