Latest Posts

Topic: Economic Part versus Heroes

hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 1607
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2020-04-01, 10:44

Native wrote:

I think one of the most annoying things about the hero stuff is the "automatic" training of the heroes, as the game tends to train a lot of "a little bit trained" soldiers instead of one really good one (which is what you want). Players that know how to click on stuff and micro the soldiers around can achieve this, but for my part I don't have the patience for that and widelands should more be about caring about your eco than microing one specific soldier into your arena and making him stay there.

Also another point is that the randomness of fights is too big in my oppinion. More often than not a full rookie defeats a half trained soldier, which shouldn't happen when looking at how difficult it is to train them.

this depends on the definition of half trained, but should be (and is to my experience) very rare. This might happen if you only train in evade (especially barbarians and empire) as there we have some luck. in case you have at least 2 other promotions it gets very rare e.g. a health promotion should give you an advantage

So my proposal/wish would be to reduce the randomness and make the game automatically train full heroes instead of the current "train all soldiers a little bit".

with just a few clicks you can increase the chance to make this happen. basically reducing the number of soldiers to train to 1 in every training facilty might do the trick over time, if you continously can supply the trainingsite. Microing is just necessary to overcome shortage of supply (stopping the facility until fully supplied) or making the first hero very fast. In any other case heroes should be trained automatically after a while. (it is easier to have this automated for the frisians as they have linear traaining in their training sites.)


Top Quote
the-x
Avatar
Joined: 2019-01-19, 13:23
Posts: 674
Ranking
One Elder of Players
Posted at: 2020-04-01, 13:15

Native wrote:

I think one of the most annoying things about the hero stuff is the "automatic" training of the heroes, as the game tends to train a lot of "a little bit trained" soldiers instead of one really good one (which is what you want). Players that know how to click on stuff and micro the soldiers around can achieve this, but for my part I don't have the patience for that and widelands should more be about caring about your eco than microing one specific soldier into your arena and making him stay there.

Also another point is that the randomness of fights is too big in my oppinion. More often than not a full rookie defeats a half trained soldier, which shouldn't happen when looking at how difficult it is to train them.

So my proposal/wish would be to reduce the randomness and make the game automatically train full heroes instead of the current "train all soldiers a little bit".

+1

Yes, you have exactly the point.

Training is very lots of micromanagement, which is not bad in the first place but since you repeat it every game in exactly the same way it gets boring. At least we need a system that the stronger upgrades are trained first and in optimal case an economy that aims at training highest upgrades by a system.

Edited: 2020-04-01, 13:15

Top Quote
the-x
Avatar
Joined: 2019-01-19, 13:23
Posts: 674
Ranking
One Elder of Players
Posted at: 2020-04-01, 13:20

hessenfarmer wrote:

the-x wrote:

Yes, indeed there are 2 things which make lots of unbalance

  • Multiplaying of upgrades: 1,3x1,3x1,3x1,3 = 2.856 which should be 1,3+1,3+1,3+1,3 - 4 = 1,2 -> the difference between 2,8 and 1,2 is what a hero is to strong. These upgrades should be added, so indeed also strategys like going for attack value are possible - the actual concept leads to high unbalance, since 2,8 is more than 100% more than 1,2

This formula is simply not correct, as upgrades have a different effect. To understand this you need to understand the current fighting system. this has been explained in a lot of threads already but I'll try it again.
1. a fight is round based
2. attacker strikes first (although I am not 100% sure who will be the attacker, as it might be that an intercepting soldier might be the attacker as well)
3. first of all it is evaluated by random function if the attacker hits taking into account the evade probabilty of the defender. if he misses nothing happens
4. if he hits then the strength of his hit is calculated by random between his upper attack value and lower attack value
5. this calculated attack value is reduced by the defenders defence percentage.
6. the remaining attack value is subtracted from the defenders remaining health value.
7. if the defender has at least 1 health point left it his his turn to attack.

so there are multiple values where you can influence this algorithm but neither of them is simply multiplying or addition.

  1. increasing attack is adding a fixed value to the upper and lower attack value between which the real attack value will be chosen for each hit.
  2. health directly adds healthpoints.
  3. evade increaes the chance to evade a hit completely there by a multiplying factor is increased.
  4. defense increaes a multiplayer as well.

Exactly, but indeed the values are multiplied - with a little of randomness, in fe evade if the hit hits or not. Also since it is round based the higher upgrades get a lot stronger than just levelling from zero upwards.

I mean, yes we can let it that way but then multiplayer might stay rarely played cause so much difference

Edited: 2020-04-01, 13:20

Top Quote
Native
Avatar
Joined: 2020-03-04, 22:11
Posts: 11
Ranking
Pry about Widelands
Location: Bietigheim-Bissingen
Posted at: 2020-04-01, 13:26

the-x wrote:

Exactly, but indeed the values are multiplied - with a little of randomness, in fe evade if the hit hits or not. Also since it is round based the higher upgrades get a lot stronger than just levelling from zero upwards.

That's exactly my feeling... Soldiers that are not fully trained are a lot weaker than fully trained soldiers (which is what I meant with "a little bit trained"). It feels training a skill to lvl 2 is not nearly half as good as training it to lvl4. So it's more an either train them to full level or don't train them at all and produce masses of soldiers instead. Imho the levelling should be linear not exponential to also reward training to lower levels. And as training to the highest level is (especially for players that don't know the game too well) really difficult and annoying the solution I tend to do is to train more soldiers instead of levelling them, which really isn't a good solution in my oppinion (and doesn't really have to do anything with a good or bad eco management).

I mean, yes we can let it that way but then multiplayer might stay rarely played cause so much difference

Agreed


Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 1607
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2020-04-01, 14:07

Native wrote:

the-x wrote:

Exactly, but indeed the values are multiplied - with a little of randomness, in fe evade if the hit hits or not. Also since it is round based the higher upgrades get a lot stronger than just levelling from zero upwards.

That's exactly my feeling... Soldiers that are not fully trained are a lot weaker than fully trained soldiers (which is what I meant with "a little bit trained"). It feels training a skill to lvl 2 is not nearly half as good as training it to lvl4.

You can't tell this generally, cause it depends on various factors. e.g. one level attack upgrade for frisians is 1006 points starting at a meridian of 1450 (lower value is 1300 upper value is 1600) so level 2 increase is nearly doubling the attack value while the next 2 are only adding 2/3. the effect is that you need less succesful hits with the higher values so by this you have less rounds which reduces the effect of evade a bit perhaps. However it is dependent on the opponents tribe as well if you fight amazons with frisians and they are fully trained in evade they will evade each blow with 74% chance while your frisian soldier will only evade with 35% chance so you have a suitable chance to kill even a lvl4 attack soldier of frisians with a lvl 3 evade soldier of amazons if you mange to evade often enough. The different values for the tribes have been chosen for giving them some character. balancing has ensured so far that lvl 0 vs lvl 0 of each tribe vs each other is more the less a fair chance. and max lvl vs max lvl is a fair chance (statistically with a basis of 10000 fights) max lvl vs one lvl down should result in a 75% to 80 % chance by this, but it strongly depends on which value has been not fully upgraded.

So it's more an either train them to full level or don't train them at all and produce masses of soldiers instead. Imho the levelling should be linear not exponential to also reward training to lower levels. And as training to the highest level is (especially for players that don't know the game too well) really difficult and annoying the solution I tend to do is to train more soldiers instead of levelling them, which really isn't a good solution in my oppinion (and doesn't really have to do anything with a good or bad eco management).

training them (from the 2nd hero onwards) to the full level needs a very strong economy with a very good road system to supply the trainingsites, or a lot of micromanagement. So in longer games the economy should decide the game. Problem arises if you always play atlanteans vs atlanteans on crater or a similar small map.

Only thing I could agree on is to set the trainingsites to only 1 soldier by default to help the player not waste ressources. However this setting will probably increase the chance of soldiers getting kicked out.

Agreed


Top Quote
niektory
Avatar
Joined: 2019-06-03, 20:06
Posts: 185
Ranking
Widelands-Forum-Junkie
Location: Poland
Posted at: 2020-04-01, 14:27

hessenfarmer wrote:

Native wrote:

I think one of the most annoying things about the hero stuff is the "automatic" training of the heroes, as the game tends to train a lot of "a little bit trained" soldiers instead of one really good one (which is what you want). Players that know how to click on stuff and micro the soldiers around can achieve this, but for my part I don't have the patience for that and widelands should more be about caring about your eco than microing one specific soldier into your arena and making him stay there.

[...]

So my proposal/wish would be to reduce the randomness and make the game automatically train full heroes instead of the current "train all soldiers a little bit".

with just a few clicks you can increase the chance to make this happen. basically reducing the number of soldiers to train to 1 in every training facilty might do the trick over time, if you continously can supply the trainingsite. Microing is just necessary to overcome shortage of supply (stopping the facility until fully supplied) or making the first hero very fast. In any other case heroes should be trained automatically after a while. (it is easier to have this automated for the frisians as they have linear traaining in their training sites.)

In my experience there are several reasons that added together make training heroes very tricky in a normal game:

  1. Lower level weapons often take less materials to make than higher level ones. Because of this, if the weapon/armor smithy is not fully supplied it often keeps producing the cheap arms and not producing the expensive ones. Having low economy targets for arms and stopping the training site when it's not fully supplied does help with this, but requires constant attention.

  2. Even when you can get a soldier all upgrades from one training site, you then have to get that half-trained soldier into the other site to receive the other half of the training instead of a rookie. If you don't do anything about it the common result is a bunch of soldiers trained in one building and a bunch trained in the other, but few fully trained ones. (This point doesn't apply to Frisians, but the others still do.)

  3. It gets even more difficult if any of your military sites are set to prefer heroes, because they will take the half-trained soldiers too, leaving only rookies to be (half-)trained. And you do generally want to send the heroes to the front line...


Top Quote
the-x
Avatar
Joined: 2019-01-19, 13:23
Posts: 674
Ranking
One Elder of Players
Posted at: 2020-04-01, 14:39

hessenfarmer wrote:

Native wrote:

the-x wrote:

Exactly, but indeed the values are multiplied - with a little of randomness, in fe evade if the hit hits or not. Also since it is round based the higher upgrades get a lot stronger than just levelling from zero upwards.

That's exactly my feeling... Soldiers that are not fully trained are a lot weaker than fully trained soldiers (which is what I meant with "a little bit trained"). It feels training a skill to lvl 2 is not nearly half as good as training it to lvl4.

You can't tell this generally, cause it depends on various factors. e.g. one level attack upgrade for frisians is 1006 points starting at a meridian of 1450 (lower value is 1300 upper value is 1600) so level 2 increase is nearly doubling the attack value while the next 2 are only adding 2/3. the effect is that you need less succesful hits with the higher values so by this you have less rounds which reduces the effect of evade a bit perhaps. However it is dependent on the opponents tribe as well if you fight amazons with frisians and they are fully trained in evade they will evade each blow with 74% chance while your frisian soldier will only evade with 35% chance so you have a suitable chance to kill even a lvl4 attack soldier of frisians with a lvl 3 evade soldier of amazons if you mange to evade often enough. The different values for the tribes have been chosen for giving them some character. balancing has ensured so far that lvl 0 vs lvl 0 of each tribe vs each other is more the less a fair chance. and max lvl vs max lvl is a fair chance (statistically with a basis of 10000 fights) max lvl vs one lvl down should result in a 75% to 80 % chance by this, but it strongly depends on which value has been not fully upgraded.

This is very good, since tzhe character of the different tribes makes a lot more intresting situations. Indeed its not true that the stroger economy wins. For example i played a lot of games against WorldSaviour, i am always ahead in Economy, Wares, Military and every statistic except land (by ahead i mean of course also the ware curve, that it gets the same way low but then getting a high Steigung) Its mostly about 15% to 20% where i have more wares, more increasing numbers, 20-35% more military upgrades. Funnily it turns out every game the same, as soon he has his 1 oder second hero you can through the whole economic part off, cause he attacks heals, changes his hero to attack which is so op and ater 2 min that game is over. Your feeling? Every decision in economic part did not change anything at all.

Edited: 2020-04-01, 18:40

Top Quote
the-x
Avatar
Joined: 2019-01-19, 13:23
Posts: 674
Ranking
One Elder of Players
Posted at: 2020-04-01, 14:44

niektory wrote:

hessenfarmer wrote:

Native wrote:

I think one of the most annoying things about the hero stuff is the "automatic" training of the heroes, as the game tends to train a lot of "a little bit trained" soldiers instead of one really good one (which is what you want). Players that know how to click on stuff and micro the soldiers around can achieve this, but for my part I don't have the patience for that and widelands should more be about caring about your eco than microing one specific soldier into your arena and making him stay there.

[...]

So my proposal/wish would be to reduce the randomness and make the game automatically train full heroes instead of the current "train all soldiers a little bit".

with just a few clicks you can increase the chance to make this happen. basically reducing the number of soldiers to train to 1 in every training facilty might do the trick over time, if you continously can supply the trainingsite. Microing is just necessary to overcome shortage of supply (stopping the facility until fully supplied) or making the first hero very fast. In any other case heroes should be trained automatically after a while. (it is easier to have this automated for the frisians as they have linear traaining in their training sites.)

In my experience there are several reasons that added together make training heroes very tricky in a normal game:

  1. Lower level weapons often take less materials to make than higher level ones. Because of this, if the weapon/armor smithy is not fully supplied it often keeps producing the cheap arms and not producing the expensive ones. Having low economy targets for arms and stopping the training site when it's not fully supplied does help with this, but requires constant attention.

  2. Even when you can get a soldier all upgrades from one training site, you then have to get that half-trained soldier into the other site to receive the other half of the training instead of a rookie. If you don't do anything about it the common result is a bunch of soldiers trained in one building and a bunch trained in the other, but few fully trained ones. (This point doesn't apply to Frisians, but the others still do.)

  3. It gets even more difficult if any of your military sites are set to prefer heroes, because they will take the half-trained soldiers too, leaving only rookies to be (half-)trained. And you do generally want to send the heroes to the front line...

Yes, and this system we could easily make more automatic. Esypecially training the strongest soldier first

Also i like if there are dieffent characters in the game like a +3 soldier, one upgraded in evade maybe someone who is better defending - if we are true to ourselves we have at the moment only 0 and 1 - 0 is the unupgraded soldier, 1 is the hero.

Its true we dont need diversity so often, but in this part yes: it gets boring if in optimum there are only 2 characters on the field, 1 and 0 theres not much strategy except handling heros ... which is microbased and needs the same clicks every game


Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 1607
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2020-04-01, 17:24

the-x wrote:

This is very good, since tzhe character of the different tribes makes a lot more intresting situations. Indeed its not true that the stroger economy wins. For example i played a lot of games against WorldSaviour, i am always ahead in Economy, Wares, Military and every statistic except land. Its mostly about 15% to 20% where i have more wares, more increasing numbers, 20-35% more military upgrades. Funnily it turns out every game the same, as soon he has his 1 oder second hero you can through the whole economic part off, cause he attacks heals, changes his hero to attack which is so op and ater 2 min that game is over. Your feeling? Every decision in economic part did not change anything at all.

you have more wares, because he is using his wares and not producing on stock. If he has more land he almost sure had more buildings producing. I have seen a lot of replays from him to learn about how he is playing and he manages to outproduce every other player nearly every time so it is not all about heroes in his game play. He is just using all his ressources in a very efficient way plus he is doing a lot of micro in the meantime.


Top Quote
hessenfarmer
Avatar
Joined: 2014-12-11, 23:16
Posts: 1607
Ranking
One Elder of Players
Location: Bavaria
Posted at: 2020-04-01, 17:32

niektory wrote:

In my experience there are several reasons that added together make training heroes very tricky in a normal game:

  1. Lower level weapons often take less materials to make than higher level ones. Because of this, if the weapon/armor smithy is not fully supplied it often keeps producing the cheap arms and not producing the expensive ones. Having low economy targets for arms and stopping the training site when it's not fully supplied does help with this, but requires constant attention.

  2. Even when you can get a soldier all upgrades from one training site, you then have to get that half-trained soldier into the other site to receive the other half of the training instead of a rookie. If you don't do anything about it the common result is a bunch of soldiers trained in one building and a bunch trained in the other, but few fully trained ones. (This point doesn't apply to Frisians, but the others still do.)

  3. It gets even more difficult if any of your military sites are set to prefer heroes, because they will take the half-trained soldiers too, leaving only rookies to be (half-)trained. And you do generally want to send the heroes to the front line...

Agreed, this is what makes it difficult. But it should be difficult to a certain extent. I believe we could do probably something about Problem Nr. 2. But Problem 1. and 3. could be solved within current game mechanics, this is just a matter of how good you play. Especially to solve nr. 1 is what the game should be about: Building an econoy that supports your Weapon industry.


Top Quote