what about snowballing

[alre]

I want to share a video that I found some time ago, thanks to another post by @Lion.Kanzen:

 

Before watching this, I hadn't realised how much attention should be put on snowballing when designing a RTS game, and I think it's important for us to discuss this because 0AD is a game that snowballs an awful lot: after losing a fight, you may come up with a plan for coming back and turn the tables, but in 1v1, it's better just to resign, because you know it's gonna be pointless anyway. In fact, 0AD would be a lot more fun if it it wasn't so easy to escalate any advantage so quickly.

I think that 0AD could be a lot better in terms of anti/pro snowballing mechanics both in economy and in warfare, but one thing that really stands out, and thus I'd like to discuss immediately, is loot: the author of the clip above says that pro-snowballing mechanics are not necessarely bad, because they can be very fun to benefit from, like veterancy in many games, included 0AD, while anti-snowballing mechanics should be more hidden to avoid feeling punishing. Well, loot in 0AD is a mechanic that is hidden (not fun at all, almost impossible to notice in fact) but favours snowballing: it has it all wrong.

I think looting could be a fun thing, if it was more evident, or even explicitly commanded by players (like plundering enemy buildings, or maybe even collecting resources from corpses if you gain control of the battlefield) but they way it is now, it's just a free gift for players who are already winning, and a strong factor towards making it impossible for losing players to come back from a bad position.

[Feldfeld]

In 0 A.D., even in 1v1 it's possible to comeback, I myself managed a small comeback in a 1v1 yesterday, but I agree it can be hard and the game is quite snowbally.

In order to comeback from a disadvantage in 1v1 I see 2 main strategies possible:
- In early game, if you are around equal in military might, you can try training more soldiers and overrunning the opponent in an all out attack
- If it is going to late game no matter what, or if you have less military than the opponent, then play very conservative and try to defend early p3. Hopefully after some time you will have stabilized the economy situation.

I think the promotion system is nice and the snowball is not its only use: for example, making use of healers in the army benefits a lot from promotion system, and by that I mean not just the healers, since the units getting healed don't die they get promoted, also healing strong units is better than healing weak ones.

For looting however, although I believe right now it's only a small factor and has very small effect even in snowballs, I agree that the looting of fighting army is quite hidden and difficult to feel in the game.
Killing traders also loots the resource they're transporting, which his however quite nice, but if the defender microes well he can delete the trader before which is unnice.

[BreakfastBurrito_007]

And in 4v4, it is possible to come back, but it requires a good strategy and helpful teammates. In a24, I observed that comebacks were sort of enforced by the gameplay mechanics and balance, such as building garrisoned arrows being very powerful. In a 4v4, a player could be nearly finished, but then rebuild somewhere else quite easily because of how hard it was to move around the map. In one game, @Dizaka built his city faster than @chrstgtr could destroy it, he built something like 10-15 forts in a counter-clockwise direction around the edge of the map.

I agree that we don't want to have as many player-independent snowballs, like loot. I think a great example of something good that can be called a snowball is healers like @Feldfeld said. Keep in mind that in a24 attacking was like building a snowball in 50 degree heat.

[borg-]

Just an idea. wouldn't it be interesting to avoid snowball, give a population increase by cc? For example, if you are playing 1v1 with max 300 pop, then houses and other buildings can give you max population 200, the rest can only be achieved by building new ccs (like 50 max pop per cc). I think this would encourage more expansion/fighting for territory.

[alre]

2 hours ago, Feldfeld said:

In 0 A.D., even in 1v1 it's possible to comeback, I myself managed a small comeback in a 1v1 yesterday, but I agree it can be hard and the game is quite snowbally.

In order to comeback from a disadvantage in 1v1 I see 2 main strategies possible:
- In early game, if you are around equal in military might, you can try training more soldiers and overrunning the opponent in an all out attack
- If it is going to late game no matter what, or if you have less military than the opponent, then play very conservative and try to defend early p3. Hopefully after some time you will have stabilized the economy situation.

I think the promotion system is nice and the snowball is not its only use: for example, making use of healers in the army benefits a lot from promotion system, and by that I mean not just the healers, since the units getting healed don't die they get promoted, also healing strong units is better than healing weak ones.

For looting however, although I believe right now it's only a small factor and has very small effect even in snowballs, I agree that the looting of fighting army is quite hidden and difficult to feel in the game.
Killing traders also loots the resource they're transporting, which his however quite nice, but if the defender microes well he can delete the trader before which is unnice.

could be that looting is that unimportant, I'm not sure. can't be unless we try the game without I think.

it's also possible that I'm exaggerating the difficulty of comebacks, but to me it seems that if your army starts to break trough in its way to the enemy city center, there's no way out for that player, even if he/she was ahead in eco and maybe is able to raid your own or is able to play smart some other way.

borg- is right about expansion potentially playing against snowballing, although that would be a huge shift for 0AD, as CCs are currently very costy and also not that necessary for economic growth.

[Gurken Khan]

12 minutes ago, alre said:

could be that looting is that unimportant, I'm not sure. can't be unless we try the game without I think.

I don't think loot has much of an impact. I just looked at my latest summary and I used ~70k res and had ~1k loot.

 

22 minutes ago, borg- said:

Just an idea. wouldn't it be interesting to avoid snowball, give a population increase by cc? For example, if you are playing 1v1 with max 300 pop, then houses and other buildings can give you max population 200, the rest can only be achieved by building new ccs (like 50 max pop per cc). I think this would encourage more expansion/fighting for territory.

I fail to see how that would counter snowballing. If you got an increasing advantage, how would that not help in expanding too? And how would Military Colonies count toward that? (Which, btw, I think are already very strong, while the ability seems to be arbitrary. Carthage and Rome couldn't build military colonies? Maybe I just don't know enough history.)

[Dizaka]

I liked how catapult splash dmg prevented snowballing.

[Ceres]

This is an interesting topic. Maybe the balancing advisors could share their thoughts on it. However, I don't know by the top of my head who that is. Maybe @Stan`
can help?

[borg-]

24 minutes ago, Gurken Khan said:

 

 

I fail to see how that would counter snowballing. If you got an increasing advantage, how would that not help in expanding too? And how would Military Colonies count toward that? (Which, btw, I think are already very strong, while the ability seems to be arbitrary. Carthage and Rome couldn't build military colonies? Maybe I just don't know enough history.)

Imagine that you have gained an advantage of 30 soldiers against your enemy who is playing Sparta, and your economy is slower because you are producing skiritas. In the time you need to build a new center, it has a chance to equalize in number, and even make a sneak attack on your new center under construction or even on other strategic areas. Colonies would also have a population increase, less of course. It would also make the fight for territory something real and necessary.

Edited August 26, 2021 by borg-

[Gurken Khan]

Hm, I guess. Although I probably wouldn't consider that already snowballing in effect.

[borg-]

The biggest problem with 0a.d is that you can easily reach your population limit without having to do anything. No need to expand territory, no special buildings, nothing, all the resources you need are in your reach. I think making the territory something more valuable would be interesting, also future additions of camps to ally, treasures among others, can make this snowball diminish.

[Gurken Khan]

45 minutes ago, borg- said:

The biggest problem with 0a.d is that you can easily reach your population limit without having to do anything. No need to expand territory, no special buildings, nothing, all the resources you need are in your reach.

I believe you are a better player than me, yet I can't agree to that in its absoluteness. If you don't advance the enemy/enemies will bite you in your behind or elsewhere; to advance you need special buildings, advancing will automatically expand your territory. A lot of games I watched, for example on the popular Mainland, included the intermediate goals of reaching resources like woodlines or minerals, and/or gaining a strategic position.

[hyperion]

42 minutes ago, borg- said:

The biggest problem with 0a.d is that you can easily reach your population limit without having to do anything. No need to expand territory, no special buildings, nothing, all the resources you need are in your reach. I think making the territory something more valuable would be interesting, also future additions of camps to ally, treasures among others, can make this snowball diminish.

It seems all it takes is to play a different map than the one and only mainland with low size. Valid reason for blocking expansion is the distance between civ centers and possibly high cost. Also people do expand by means of planting structures at the border, some times house chains to cross half the map.

Also some snowballing is needed or the game won't ever end. Me think snowballing isn't excessive in 0ad.

[Jofursloft]

15 hours ago, alre said:

I think looting could be a fun thing, if it was more evident, or even explicitly commanded by players (like plundering enemy buildings, or maybe even collecting resources from corpses if you gain control of the battlefield) but they way it is now, it's just a free gift for players who are already winning, and a strong factor towards making it impossible for losing players to come back from a bad position.

I don't agree with this for what concerns 1v1 matches. I watched the replay of some 1v1 I did against very good players recently and I found that the total number of kills each players gets before the end of the match is around 300. Using skirms and spearmen as an example: 300 kills means something like 5 food and 5 wood per soldier killed = 30 units you can train thanks to the ones who you kill. In my opinion it's a totally fair number.

The main problem in my opinion is that when playing some players don't remember that when a unit dies the enemy player loots the resource they're transporting. I've see this mistake so many times. Always remember to pass from storehouses before going into a fight.  

12 hours ago, alre said:

it's also possible that I'm exaggerating the difficulty of comebacks, but to me it seems that if your army starts to break trough in its way to the enemy city center, there's no way out for that player, even if he/she was ahead in eco and maybe is able to raid your own or is able to play smart some other way.

I agree with this if we are talking about 1v1 matches (and still not medium/low level ones). However, being ahed in eco doesn't justify you from not having trained a proper army. If the enemy player attacks you with a big army and you have no way to fight that army back I think it's your fault and you should lose. 

I also remember a 1v1 @chrstgtr (kushites) vs @ValihrAnt (maurya) in which Valihrant failed his archer ele P2 rush but even in terrible pop and eco disadvantage he could win the game thanks to micro and building positioning. The replay would be really interesting (even if I think that game was played on svn) 

 

Edited August 27, 2021 by Jofursloft

[alre]

all loots are 10% of the value of the unit/building. it means that a battle that gives you an advantage of 50 men, for instance, also gives you a resource advantage for 5 more men. why would that be a fair number? why not 0? you now have a men advantage, isn't that a sufficient reward? this kind of situations can happen when a player manages to group all together a very big army and overwhelm defending armies if not properly prepared. a single mistake by the defender can throw the whole game away for him/her.

[alre]

man, sometimes I literally loose whole groups of men to the lag! because commands don't pass trough it, and so my people makes suicide charges instead of retreating.

[borg-]

17 hours ago, Gurken Khan said:

I believe you are a better player than me, yet I can't agree to that in its absoluteness. If you don't advance the enemy/enemies will bite you in your behind or elsewhere; to advance you need special buildings, advancing will automatically expand your territory. A lot of games I watched, for example on the popular Mainland, included the intermediate goals of reaching resources like woodlines or minerals, and/or gaining a strategic position.

I'm referring to special buildings, buildings scattered around the map that you can capture to win the game, for example.
What I mean is that 0 a.d doesn't challenge you to leave your small town because you have everything you need in a small piece of territory. What if we had structures that would give you victory if captured? Or nomadic tribes capable of producing mercenaries very quickly and cheaply to surprise their enemy, or even valuable treasures that give some sort of military/economic advantage if captured. This would encourage another style of play besides the snowball.

Edited August 27, 2021 by borg-

[Micfild]

2 hours ago, alre said:

all loots are 10% of the value of the unit/building. it means that a battle that gives you an advantage of 50 men, for instance, also gives you a resource advantage for 5 more men. why would that be a fair number? why not 0? you now have a men advantage, isn't that a sufficient reward? this kind of situations can happen when a player manages to group all together a very big army and overwhelm defending armies if not properly prepared. a single mistake by the defender can throw the whole game away for him/her.

I always thought that loot was a fair mechanic (if not a bit underwhelming). The situation I'm considering is this. If you are the agressor, that means that you have to move your army from your base to the enemy base and that takes time. Time that you're not collecting resources with said army and time your enemy is collecting resources with his. So to attack, you're basically accepting the risk that you'll fall behind in economy.

Secondly, you'll probably meet the enemy in their own territory, where they could have static defense (like turrets) and base layouts that will benefit the defender, giving him a greater advantage while fighting your army. Also, he is in reinforcement range, which means he can continuously boost his forces while the fight is going on, while your reinforcements (if any) will have to come all the way from your base.

There is an argument that using horses or mercenaries to attack, instead of citizen-infantry, could lessen the economic loss of having your workforce go to war. The problem is that horses are more expensive than infantry and can only gather food (so they are less useful to have in comparison to infantry in an economic sense), while mercenaries can't gather resources and cost large amounts of metal (which might slow down your technological advancement).

In case the agressor loses the fight, then the loot he gathered can actually help him rebuild, offsetting a small amount of the costs of going to war (which can be considered a feature to limit the potential snowball the defender will have).

If the agressor manages to stalemate or even win the fight then his advantage will be great (well, depending on how convincingly he won) and that is to his merit. After all, attack is already a huge risk with no guarantees.

============================

As for loot being a sort of hidden feature, i do agree that it could use a bit more visibility. To that end i have a suggestion. Since loot is small per unit we could have some sort of artificial (hidden) bank that starts collecting loot when the fight starts. After X seconds of no combat, the bank will dump all the respective looted resources on the players, probably with a satisfying sound like a "ting". If you want to make it more visual, you could always have the number of resources looted appear as a notification to the player.

[Gurken Khan]

13 minutes ago, borg- said:

I'm referring to special buildings, buildings scattered around the map that you can capture to win the game, for example.

I don't know if I would like that, also seems pretty unrealistic. But there is the 'capture the wonder' winning condition.

15 minutes ago, borg- said:

valuable treasures that give some sort of military/economic advantage if captured.

There are already treasures. If, for example, I manage to claim 6 treasures on Sahara Oasis, that's 1800 resources that clearly give me an economic advantage/lets me build my military faster.

[alre]

18 minutes ago, Micfild said:

In case the agressor loses the fight, then the loot he gathered can actually help him rebuild, offsetting a small amount of the costs of going to war 

how so, if the defender also gains loot he can immediately spend to counterattack? 

19 minutes ago, Micfild said:

After all, attack is already a huge risk with no guarantees.

I don't think attacking is that bad as a risk, especially if you are near to a base of yours

[BreakfastBurrito_007]

30 minutes ago, Micfild said:

In case the agressor loses the fight, then the loot he gathered can actually help him rebuild, offsetting a small amount of the costs of going to war (which can be considered a feature to limit the potential snowball the defender will have).

Usually it is better to leave behind enough units on the various resources needed so that you are able to train replacements to your army. 

I think it is more risky (to your team) to not attack. If you have 100 extra units in your base gathering res instead of fighting, you will accumulate lots of extra resources which you are unable to use since you are already at the pop cap (assuming you have all the upgrades you want). For this reason the units have more value if they are used to fight. Also, if you attack first, you can usually decide where to fight, since it is now your enemies' goal to limit the damage you do to them.

Also, attacking does not necessarily mean moving directly to someones base, it can also mean flanking their army, gaining a position close to their base that they don't like, building offensive buildings, or causing an economically unfavorable reaction from your enemy.

[Micfild]

32 minutes ago, alre said:

how so, if the defender also gains loot he can immediately spend to counterattack?

Well, units will take time to produce and then move to the opponents base so the counterattack will have to start with only the surviving units. Also, since those units will also have to move there first, it gives time to the opponent to rebuild some units.

39 minutes ago, alre said:

I don't think attacking is that bad as a risk, especially if you are near to a base of yours

it is as you say. It's hard to consider all situations and of course there will be some that makes attacking be less risky. Having banked resources, being near the opponent, only trying to harass instead of a full on clash, etc will have an impact on the results. That is also why i don't think looting (as it is currently implemented) is such an snowballing mechanic. It's effects are so meagre that it's hardly felt (as was also pointed out).

 

25 minutes ago, BreakfastBurrito_007 said:

Usually it is better to leave behind enough units on the various resources needed so that you are able to train replacements to your army. 

Agreed, but that also means that your opponent will have an easier time repelling it. (Just to be clear, I'm not advocating in favor of all-ins, i just want to point out the issues the attacking force will have to deal with).

 

28 minutes ago, BreakfastBurrito_007 said:

If you have 100 extra units in your base gathering res instead of fighting, you will accumulate lots of extra resources which you are unable to use since you are already at the pop cap (assuming you have all the upgrades you want). For this reason the units have more value if they are used to fight.

Agreed, but at that time, the loot bonus kind of becomes inconsequential (since you'll probably already have enough resources or infrastructure to comfortably rebuild such). I was trying to focus on specific moments where the loot bonus can make a difference (even if its a small one).

31 minutes ago, BreakfastBurrito_007 said:

Also, attacking does not necessarily mean moving directly to someones base, it can also mean flanking their army, gaining a position close to their base that they don't like, building offensive buildings, or causing an economically unfavorable reaction from your enemy.

0AD is indeed a nuanced game and as you point out, your mere presence near the enemy base might inflict some economical damage without the need to fight.  But then again, in those cases, loot doesn't factor in so it kind of detours from the point of this discussion.

[ChronA]

0 AD's Pro-Snowball Features:

• Lack of strong AOE damage sources that would require tight control to counter
• Strong ranged units that remain effective with minimal control
• (Effectively) Frictionless-less pathing, allowing arbitrarily dense unit packing (negating the effect of range differences)
• Citizen soldiers that remove the need to choose between eco and military growth
• Veterancy promotions
• Capturable buildings
• Loot

0 AD's Anti-Snowball Features:

• Abundant resources that permit a player to survive falling behind on expansions
• Very strong static defenses
• Very durable buildings that can garrison and protect large numbers of units of arbitrary type
• Limited anti-building counters
• Soft-counter based balancing
• Lack of major unit and faction differences

You will notice a lot of those anti-snowball features line up with many of the top complaints about 0 AD's design...

One idea that might fortify the anti-snowball side a bit more, without sacrificing established features or historical authenticity, is to give cavalry an attack-and/or-defense-debuffing aura against ranged infantry. This would provide something adjacent to catapult AOE as a hard counter for dense ranged infantry, and shift the balance a bit more in favor of melee.

[BreakfastBurrito_007]

1 hour ago, borg- said:

What if we had structures that would give you victory if captured? Or nomadic tribes capable of producing mercenaries very quickly and cheaply to surprise their enemy, or even valuable treasures that give some sort of military/economic advantage if captured. This would encourage another style of play besides the snowball.

A "capture the wonder" or "king of the hill" gamemode could be quite fun. I feel like these are features that would be great to have as optional add-ons for games of 0ad rather than core features. 

[BreakfastBurrito_007]

4 minutes ago, ChronA said:

and shift the balance a bit more in favor of melee

I can't say for the other players but it is my feeling that melee/ranged balance is quite good at the moment.