Desert Traders [1.12 SP scenario]

[skeptical_troll]

This one is quite experimental, if the mechanics work out well I want to reuse them in the future.

The player controls the khalifate, and he needs to keep open trade routes in the desert and let caravans reach the city. The way it works, briefly:
- routes are open if the players controls all villages along them
-while a route is open, caravans will be (randomly) spawned at its farthest end and start crossing the desert, if they reach the base they'll give you some gold (30 at 'easy' and 'medium', 25 on 'hard')
-you can't earn gold in any other way (except the base +2 income)
-Enemies are mostly randomly spawned on different spots, also chosen randomly.
- To allow better control of the map, owned villages along trade routes work as 'roadhouses'(not sure if this is the correct word) where mounted units can switch horse and hence move further (MP restored). Each road house can be used 3 times to change horse each turn (4 times on easy).

Some issues I am not sure about, on which I'd like to hear opinions:
1) Does the spawning method make the gameplay too luck-reliant?
2) would a different objective work better? Currently you have to let 30 caravans arrive (possibly too high) 4 caravans from each route arrive t0 the city.
3) Balancing: different starting gold, gold and income of the bandits, price of mounted units, number of horses per station are tuned depending on difficulty. Let me know if you think the difficulty you played corresponds to the expectations.
4) Did you find that going Rami-heavy is a too-trivial solution?
5) Stations are currently 7-8 hx from each other, so that a unit with 8+ MP can freely travel around the map, if enough horses are available. Do you think this should be prevented?


This aspects will likely change in the future, together with the dialogues which are somehow silly ATM.

[Inky]

Desert Traders v1.0

Hey, this was a really fun scenario; very original idea!

Replay summary:
Recruited a horde of Ramis (and a couple Khaiyals for melee damage), and opened the northwest and southeast trade routes while letting the bandits fight the local wildlife. Once the gold from the caravans started coming in, formed a massive Rami gang and stormed the center, killing the bandit leader on turn 28 and opening the two central routes soon afterwards. Kept some units stationed in the villages around all the trade routes to prevent spawned enemies from stealing villages.
The rest of the turns were just waiting around for the remaining caravans to arrive; finished turn 45.

DsTr-Desert_Traders_replay.gz

Scenario version 1.0 / Wesnoth 1.12.5
EDIT: there have been major changes to the scenario since version 1.0; scroll down for a more recent replay

(70.27 KiB) Downloaded 640 times

-Objectives: 30 caravans does feel like too many; it's likely that by the time you get 15-20 caravans you already dominate the map because of all the units you can buy with the caravans' gold. Maybe an alternate objective could be to open up all 4 trade routes.

-Spawning frequency: Maybe I just got lucky, but in my playthrough I almost never had to defend the southeastern route (some scorpions spawned in the far southeast just once; otherwise the enemies mostly spawned near the center and went after the bandits.)

-Balance: Due to the move refreshing for mounted units and the log village spacing (7-8 hexes apart), there is no reason to ever recruit anything besides Rami (which have 7 MP, or 8 if quick). If you advance the nonquick Ramis to Saree (which get an extra MP) and quick Ramis into Qanas, then you'll have an entire army of 8 MP mounted units which can all move across the entire map in a single turn, which is pretty overpowered.
Maybe there could be some kind of limit to how many times a unit can move again (after all, riding does get tiring after a while). Another possibility is to make some of the log villages 9+ hexes apart (maybe change some to tent villages).

-A minor annoyance was that caravans from the eastern route would often move onto the log village at (42,45) and prevent your mounted units from using it - maybe the caravans could avoid village terrain.

-Bug: Routes remain open even when enemies take one of the villages (on turn 30 of the replay, a caravan spawned in the northwest route even though a naga had captured a coastal village.)

[skeptical_troll]

Hi Inky, thanks for your comments!

-Objectives: 30 caravans does feel like too many; it's likely that by the time you get 15-20 caravans you already dominate the map because of all the units you can buy with the caravans' gold. Maybe an alternate objective could be to open up all 4 trade routes.

I am thinking of making it 'let 4 caravans arrive from each route', otherwise just opening them could be too easy. I can add more challenge with turns or when the routes open. ATM, nagas spawn if the coast route is open, and trolls if at least one of the central ones is. Higher level enemies after some turn could be a solution, rather than cluttering the maps with more units.

Spawning frequency: Maybe I just got lucky, but in my playthrough I almost never had to defend the southeastern route (some scorpions spawned in the far southeast just once; otherwise the enemies mostly spawned near the center and went after the bandits.)

That is meant to be the 'easy' route to protect. The spawning frequency of caravans however is a bit lower (1 every 3 turns) than the coastal one. I am inclined to play with caravan spawning to make some route more 'valuable' than others. Another option was to have different gold income from each route, but that may overcomplexify the mechaninc.
Scorpions should appear both South and North of the SE route from time to time, but in the north they are often attracted by the bandits if you don't control the center.

there is no reason to ever recruit anything besides Rami

This was indeed my concern I think it's the biggest flaw of the scenario now. They are even quite cheap, for mounted units. More pierce-resistant enemies could encourage the player in recruiting other units, but Ramis also have impact melee. I could brake the roadhouse net into 'branches' so that they can't really go everywhere in one turn, as you suggest. Or better I may penalize the use of the roadhouses in some way, for example each time you use it you gain MP but you also lose ~3 HP cause running in the desert is exhausting. What I would ideally aim at is the player managing 'stationary teams' along each of the routes with infantry and specialized units like naffat (who are amazing against scorpions), plus a flexible supporting team of mounted units (mostly rami+advancements)

-A minor annoyance was that caravans from the eastern route would often move onto the log village at (42,45) and prevent your mounted units from using it - maybe the caravans could avoid village terrain.

-Bug: Routes remain open even when enemies take one of the villages (on turn 30 of the replay, a caravan spawned in the northwest route even though a naga had captured a coastal village.)

Thanks, fixed both of those (forgot that the capture event does not trigger if the capturing faction has no leader)

EDIT: Just uploaded a refined version with significant changes
1)cost of mounted units is higher (depending on difficulty)
2)objectives changed to 'get at least 4 caravans from each route'
3)The stations can be now used only 3 times each turn (4 on easy). Drawback of this is that I didn't find a way of implementing it in a way that you don't lose the bonus on undo. I understand that this will be possible in wesnoth 1.13. 'Realistic' explanation behind this is that there is a limited number of horses available in each roadhouse.
4)I toned up a bit the enemies, especially at 'hard' level. I am thinking about modifying the trolls in order to give them better movements on the sands, they are painfully slow now.

[Inky]

Desert Traders version 1.2

-I like the new roadhouse mechanic; together with the extremely high cost of Rami it's definitely worth it to use the foot units now. You also need some naffats to deal with the increased number of scorpions and some healers for the foot units, so the unit variety is much more balanced now.

-Spawns: In general I liked the increased spawns because it kept things interesting, though I felt the scorpion spawn rate is maybe too high now (in my attempt, the eastern trade route got swarmed with several groups of scorpions near the beginning). It also makes the trade routes much harder to defend; it seems that the best/easiest strategy now is to focus on one route at a time and abandon a route once you get 4 caravans from it.

-Income: It's very easy to go negative on income, and actually you're more likely to have income problems on easy/normal since you can recruit more units initially, so I think on easy/normal there could be more starting villages or higher income.

-The AI caravans sometimes behave strangely when enemy units are nearby - they will go out of their way to "attack" (move next to) a nearby enemy now. I don't think this happened in version 1.0 so it seems a recent change is causing it.

-I couldn't tell the northern and northeastern route caravans apart since they all mixed together, so I wasn't sure which routes I still needed to keep open; maybe the caravans could be named based on which route they're from (e.g. "Northern Route Caravan").

Replay summary:
Recruited about 1/2 Ramis and 1/2 Jundis (because they're fast and cheap); in retrospect I badly needed naffats to use on the scorpions. Opened the northwest and southeast routes, but quickly lost them to hordes of nagas/scorpions.
Fortunately some caravans had made it through, so I got a bunch of reinforcements (including lots of naffats), eventually took the border routes back, and then abandoned them once I got 4 caravans since I was deeply negative on gold anyway.
Slowly advanced to the center of the map and opened the two central routes around turn 50. After that it was relatively easy to defend those two routes until the 4 caravans arrived; finished turn 75.

[skeptical_troll]

Thanks for trying it again, I'm glad to see that things are going in the desired direction! My impression is that it is a bit too long now. It is also true that if one is unlucky with enemy spawning at the beginning (like I think you were), he might need some time to get back in good shape (while enemies are hopefully killing each other). One possibility I am considering is setting a rather easy objective (e.g. 10 caravans) and, when it is accomplished, let the player the option to keep playing for some extra time or extra caravans, if he's enjoying the scenario and wants to try to open all the routes.

I felt the scorpion spawn rate is maybe too high now

I toned it down a little bit on hard, more significantly on 'easy' and 'normal'

Gold: I increase the income on normal and easy by 2 and 4, respectively. However, at these levels there are less enemies, so there shouldn't be the need of using all the gold. Also, I put a lot of villages in the center as an incentive to fight the bandits as soon as possible, but I understand that could be impossible if the units are too busy in protecting the two peripheral routes to guarantee some initial revenue.

The AI caravans sometimes behave strangely when enemy units are nearby

I hope it is fixed now. I was previously using a microAI, but to make the caravans avoid the stations I had to move to the default AI and did something wrong in setting it.

I followed your suggestion of 'naming' the caravans with 'origin: E/NE/Coast/Mountains', and also add labels to the routes in case the correspondences weren't obvious. Also, the status of a route is now specified in the objectives.

All these (rather minor) changes are now out in the 1.3 version.

[Xara]

Played on Normal.

I like the idea, but the game took too long. It was really a pain to command 30 units for 50+ turns. By the time I got 4 caravans from Mountain Route and NE Route, I already got about 25 caravans from the sea route and south route. So it was slower than 30 caravans in total for me. After turn 20, the game cease to be challenging, it just felt like doing repetitive chores.

1) Does the spawning method make the gameplay too luck-reliant?
2) would a different objective work better? Currently you have to let 30 caravans arrive (possibly too high) 4 caravans from each route arrive t0 the city.
3) Balancing: different starting gold, gold and income of the bandits, price of mounted units, number of horses per station are tuned depending on difficulty. Let me know if you think the difficulty you played corresponds to the expectations.
4) Did you find that going Rami-heavy is a too-trivial solution?
5) Stations are currently 7-8 hx from each other, so that a unit with 8+ MP can freely travel around the map, if enough horses are available. Do you think this should be prevented?

1) No particular feeling.
2) I feel the game should end in no more than 40 turns. Better to apply a time limit.
3)Normal was not hard. Sorry, I just meant the difficulty was okay to me. I played casually and still won without difficulties. Didn't feel anything wrong with the balance.
4) It is okay.
5) I feel a unit shouldn't utilize the stations more than once per turn.

[skeptical_troll]

Hi Xara, thanks for your comments. I must convene that the game takes too long now. I made the following modifications in the last uploaded version:

-objectives now are easy: 15 caravans in total OR 2 from each route. I think this is doable in ~<30-40 turns.
-I adopted a more liberal approach: once the objectives are reached, the player is asked if he wants to play for 10 more turns. The scenario can be extended as many times as the player likes.
-the first two times that the game is extended, the scenario gets more difficult, either by increasing the probability of spawning enemies (first time) or by spawning higher level units (second time).
-Nagas are now spawned even if the route coast is closed, but with a lower probability than when the route is open.

BTW, what do you mean with

3) Normal was not hard.

? I'm unsure if you want to say that it is well balanced (normal should be normal, not hard), or that it lacks challenge.

[taptap]

Made two attempts, both on hard. 1st one saw massive naga spawns (9 until I abandoned), a spider rampant among the outlaws and some scorpions right away slowing takeover of the SE route. I lost due to overrecruiting Rami. Second attempt saw no naga, no scorpions, but a concentrated attack by outlaws (that could however be defeated by the forces at hand), only a few bats spawned in the north this time. I abandoned to look how a third attempt may play from very winnable position. Difficulty changes a lot between runs, but gameplay too.

If you want simpler objectives why not make it saving a certain amount of gold (better not, due to upkeep)? The mechanic would work splendidly in a "Cities of the Frontier" like setting as well, with caravans substituting village gold.

Something you may want to contemplate is a "standing down" mechanism (again as in Cities of the Frontier) as otherwise you end with increasingly negative upkeep throughout the scenario.

Edit: Added a replay. Kept losing my cavalry. Used SE route + pushed the center with infantry. Most caravans that arrived came from SE, a few from mountain route, only at the very end coastal road was under control with sufficient troops. Finished before fighting a large troll wave. Fighting did not become easy (as I feared when first defeating blue). Until the very end I had to fall back, regroup, play tactically sound to get it working.

[skeptical_troll]

Hi Taptap, thanks for your feedback! By looking at your replay, the impression I got is that there are in general too many scorpions/nagas, and too much randomness in general. I don't know what's up with trolls, they should have 1/4 of probability to appear once one of the central routes are opened, but from your replay it seems that they are spawned every single turn. If it's a bug, I can't find it, and didn't remember a similar problem in other replays. It also still takes too many turns, probably more than what an average player would consider fun to play. In your case, I think you were penalized by the fact that only few caravans arrived from the east route at the beginning, which is crucial to get some initial revenue and set up an effective task force.

I'm thinking of an alternative randomization in which rather than giving a probability for spawning in each turn, I randomly extract the number of turns between two different spawning, within some reasonable interval. This way I should avoid the strongest statistical fluctuations in the random spawning. It will need some testing though.

I considered setting a gold goal as an objective, but as you may have realized it is not easy to accumulate money, and it may become tedious in the long run. Plus, I thought of this scenario as more based on territory control than economic management. I did not know 'cities of the frontier', it is quite an interesting add-on, although merging that system with the trading mechanism here may be non trivial. Importing the stand-down option instead should be pretty easy, I can certainly do that.

[taptap]

I guess this is randomness. I was happy with the time it took, a single scenario campaign can last longer than you expect for campaign scenarios otherwise and gameplay was still exciting towards the end. The same system gave me long periods of calm on other routes and if it is too much to handle there is always the option to regroup and let enemies soften each other up and after my experience of massive Naga spawns right away in another match I did not even try to take control of the coastal road at start.

This is a fun scenario as is, the Cities of the Frontier reference was more for future inspiration than as a proposal. An option to stand down units instead of suiciding them in case you end up with an unbalanced army + high upkeep at some point would be appreciated though.

The dialogue upon caravan arrival talks of 30 gold regardless of difficulty, although you get 25 on hard.

[skeptical_troll]

Glad to hear that, then I'll avoid messing up with the spawning mechanism again. I followed your suggestion and now it is possible to discharge (is it the appropriate word?) units by sending them to the 'barracks', just behind the main keep.

The dialogue upon caravan arrival talks of 30 gold regardless of difficulty, although you get 25 on hard

fixed that, thanks!

[Konrad2]

You get a base income of 2 gold, so that's a way to earn gold despite not receiving caravans.

EDIT:
When Morlo kills his first unit, he says something and the enemy responds with 'you moron! [...]'.
you -> You

EDIT 2:
You are 'still' using timeshifting lightshouses. (For you that's only a cosmetic thing though.)
#3808

[enclave]

I don't know what's up with trolls, they should have 1/4 of probability to appear once one of the central routes are opened, but from your replay it seems that they are spawned every single turn. If it's a bug, I can't find it, and didn't remember a similar problem in other replays.

hi, sorry i haven't tried your caravans, (but i have some interest in mechanics, so i might try soon), just wanted to mention that I had some add-on in the past which used randomness (similar like 1/4 or so) to recruit male or female elvish peasants.. and looks like wesnoth random is not so random, 1 of the day i could have nearly all units male (in as many games as i was starting, just whole day), other day it could be nearly all females (whole day, many games in a row), and most days it was really 1/4 or about so.. so i would say that some of the days wesnoth uses some kind of randomness that falls more or less repeatedly at same number... I used {VARIABLE_OP NS_hit rand (1,2,3)} and I was never bothered to change it.. but if you wanted you could use independent of wesnoth randomizer, which I think would be lua math.random function... this should do proper random.. and I think this code should work:

Code: Select all

[lua]
code = <<
		local NS_HIT_X = math.random(1,4)
		wesnoth.set_variable("wml_rand",wesnoth.synchronize_choice(function() return { value = NS_HIT_X } end).value)
>>
[/lua]

This would randomize 4 numbers, 1,2,3,4 and then create a wml_rand variable that you could use in WML as $wml_rand (it would be equal to one of 4 numbers) and the code above you can just insert anywhere in your WML, ask for details if interested. synchronize_choice is a MUST, otherwise there will be oos in multiplayer. Im not sure if lua wesnoth helper.rand("1..4") would be a wesnoth randomizer or an independent one.. I would recommend to use math.random to make sure u get proper random results..
but to be honest.. i dont think it's a big deal... just thought to let you know my similar experiences... (at least that's how it was back in 1.12, no idea if 1.14 random is same or something changed... i never ported my add-on to 1.14, it's too broken due to incompatibilities, so i can't tell if the randomness is different in 1.14) best of luck.

PS. OMG, just ntoiced that i spent so much time replaying to 2016 post... well.. well.. well..

[skeptical_troll]

EDIT 2:
You are 'still' using timeshifting lightshouses. (For you that's only a cosmetic thing though.)
#3808

Ohh, that looks much nicer, thanks! Changed that.

...
PS. OMG, just ntoiced that i spent so much time replaying to 2016 post... well.. well.. well..

don't worry, my doubts haven't vanished in these two years I wonder if anyone have done some test about this kind of correlations, and whether the random generator is the same in 1.14 . If I have some time in the future I'll experiment a bit with the method you are suggesting. Thanks!

[Konrad2]

Just a small thing, but the scenario awards (most times) the wrong laurel/wreath upon completion. (Apparently caused by using the old difficulties syntax.)
Details as to why and how to fix it can be found in issue #3983.