what about snowballing

[chrstgtr]

5 hours ago, alre said:

You can't use the same rushing tactics in 1v1 and team games. With or without loot.

Not in the games current state. You could if loot was higher. That’s why I said it would be nice if there was a civ that had high loot, so they wouldn’t be slower as much by rushing. To offset this benefit they would need to be weak in p3 (ie weak siege, fewer upgrades to have weaker units, or something like that). That would create a true raiding civ that is meant to fight early and often

[alre]

you want a loot level so high that if you manage to do a succesfull rush (and it doesn't even have to be extraordinarily succesful, you said) then you have an economy that can compare with that of other players that have been booming instead?

now imagine this in a 1v1 setting, could be a school example of what snowballing is: you have been raided and you managed to reduce damage a little, so you may think you are fine, or at least that you can catch on, but no, the other player is so much forward (as if he was booming) that he can keep rushing without fear of being reset, because just a small margin snowballs more and more again.

[Gurken Khan]

14 hours ago, wowgetoffyourcellphone said:

I really don't know who pissed in your Wheaties. Maybe you were just born this way.

Ja ja, sunshine.

[LetswaveaBook]

2 hours ago, chrstgtr said:

Not in the games current state. You could if loot was higher. That’s why I said it would be nice if there was a civ that had high loot, so they wouldn’t be slower as much by rushing. To offset this benefit they would need to be weak in p3 (ie weak siege, fewer upgrades to have weaker units, or something like that). That would create a true raiding civ that is meant to fight early and often

I think loot does not really make that much of a difference. I just rushed with 3 cavalry and killed 5 women early, for an amazing 50 food early.

The main thing it did is probably cause significant idle time and set my opponent back. The loot is not a consideration, even not if it was tripled.

Also extra loot is nice if you get good engagements, but it does not really help to get those good engagements in the first place. A faction good for rushing will probably be one that helps to get those good engagements and the loot can be disregarded.

[chrstgtr]

11 hours ago, LetswaveaBook said:

I think loot does not really make that much of a difference. I just rushed with 3 cavalry and killed 5 women early, for an amazing 50 food early.

The main thing it did is probably cause significant idle time and set my opponent back. The loot is not a consideration, even not if it was tripled.

 

Three cav raiding doesn't sound like a rush. That sounds like a harass. Killing five women with 3 cav is also not a successful rush--at best you are even after that exchange if the defender doesn't panic and create a bunch of idle time. Loot would help--saying having extra res isn't helpful is clearly misguided. 

11 hours ago, LetswaveaBook said:

Also extra loot is nice if you get good engagements, but it does not really help to get those good engagements in the first place. A faction good for rushing will probably be one that helps to get those good engagements and the loot can be disregarded.

 

13 hours ago, alre said:

you want a loot level so high that if you manage to do a succesfull rush (and it doesn't even have to be extraordinarily succesful, you said) then you have an economy that can compare with that of other players that have been booming instead?

No, clearly a good play should be rewarded, and a bad play should be punished. 

 

13 hours ago, alre said:

now imagine this in a 1v1 setting, could be a school example of what snowballing is: you have been raided and you managed to reduce damage a little, so you may think you are fine, or at least that you can catch on, but no, the other player is so much forward (as if he was booming) that he can keep rushing without fear of being reset, because just a small margin snowballs more and more again.

The point of what I proposed is that you would have to keep rushing (and keep doing it well) or else the rushed player will eventually recover and their natural late-game strength will overwhelm the rushing civ, which is naturally weaker late game. It could also be the case that that rushing civ provides a lot of loot if they do a bad rush, so that one bad engagement could flip the game back or even or upside down. 

[Ceres]

If I may chime in: It seems that loot is a "hot" topic.  Maybe it's possible (and worth?) keeping look but allowing users to toggle it on/off on the map etc. selection screen before the game starts? I don't know whether this helps to find out if and how much loot contributes to a snowballing effect, but in the end, you could let players decide what they like more. Again, I don't know if this is technically easy to accomplish and whether it would add any value, but maybe this is something to consider instead of thinking binary, only.

[LetswaveaBook]

2 hours ago, Ceres said:

If I may chime in: It seems that loot is a "hot" topic.  Maybe it's possible (and worth?) keeping look but allowing users to toggle it on/off on the map etc. selection screen before the game starts? I don't know whether this helps to find out if and how much loot contributes to a snowballing effect, but in the end, you could let players decide what they like more. Again, I don't know if this is technically easy to accomplish and whether it would add any value, but maybe this is something to consider instead of thinking binary, only.

It would be fairly easy (but time consuming) to make a mod that sets all loot of building/units to 0. Another solution might be looking at some hero's with loot bonusses and use that to multiply all loot by 0

[Stan`]

You can also apply a global aura for all players that sets the loot to 0.

Code changes should be similar to those.

https://code.wildfiregames.com/D4250

[alre]

today I was just thinking about how a 1v1 could last more than 20 minutes in this alpha, when this replay came out in the you tube thread:

I don't know about you, but I was amazed by what Lorenz has managed to pull out in this game. I think that, with the exception of the very first minutes, he played better than @LetswaveaBook for all the match. Of course Letswaveabook was also playing very well, he could afford to slip a couple of times because he was really never in danger.

What do you think about it? Do you agree with me here on this particular match? Do you think, more in general, that the next iteration of the game should have late game more decisive?

[wowgetoffyourcellphone]

9 minutes ago, alre said:

What do you think about it? Do you agree with me here on this particular match? Do you think, more in general, that the next iteration of the game should have late game more decisive?

I think that match will go down in 0 A.D. history, honestly. It's that good.

But I wish we could see more champions and ranged siege. More chariots from Britons, more swordsmen from Romans. Units who make these civs unique, and they should be viable.

[LetswaveaBook]

12 hours ago, alre said:

Of course Letswaveabook was also playing very well

You were asking for opinions and I agree with this one.

 

12 hours ago, wowgetoffyourcellphone said:

But I wish we could see more champions and ranged siege.

I have seen some games of borg- and he seems to use champions more than I do. borg- certainly knows what he is doing and it is a legitimate style. My approach is getting good numbers with siege and p3 upgrades early and then advance and see what the enemy has. Such an approach doesn't suit waiting 60 seconds to unlock champions and then training them.

[wowgetoffyourcellphone]

8 minutes ago, LetswaveaBook said:

Such an approach doesn't suit waiting 60 seconds to unlock champions and then training them.

I really hate the champion unlock techs. I feel like they're completely unnecessary, or cost too much/take too long to research as you indicate.

[LetswaveaBook]

Something I noticed about snowballing is the imbalance of ranged/melee units, where ranged units are more valuable/important.

If you defeat an opponent in a battle, he loses some of his ranged units. While the winner only loses some melee units and keeping the more valuable/important ranged units alive. Currently, you only IMHO need enough melee units to tank enemy fire. So having lost some melee units is not bad as long as you have enough to tank opposing fire, In some situations having about 10% melee units can be enough in these situations. If you have most of your ranged (&valuable/important) units alive, then you remain at nearly full firepower.

So here is another way for me to advocate reducing the power of ranged units.

[alre]

I don't think ranged units are always more valuable than melee. If you are losing melee units, you usually want to replace it, then an extra melee is more valuable than an extra ranged in that case. Also melee units have steeper ranking benefits, so increasing the value of melee units may actually result in heavier snowballing probably (kind of what we have seen with melee cav in this alpha: when it starts rolling it just gets stronger).

[LetswaveaBook]

4 hours ago, alre said:

Also melee units have steeper ranking benefits, so increasing the value of melee units may actually result in heavier snowballing probably (kind of what we have seen with melee cav in this alpha: when it starts rolling it just gets stronger).

They indeed have steeper benefits, but it is a lot harder for them to promote, even if ranged units need more xp. Xp is gained per damage dealt and ranged units deal more damage and aren't prone to dying as much a frontliners. Also, promotion is a subject we could tackle in other ways.

I was playing with Leonidas and had hoplite tradition, so you expect to do more damage and get easy promotions for the hoplites. That was not the case in the battles.

[real_tabasco_sauce]

I'd like to revisit this topic to see if there are some ways to limit snowballing. Obviously, if you gain the advantage, it's rightfully yours, but the opponent should have more opportunities to win it back. And I think there would be more  fun/creative play by increasing these opportunities.

Here are some mechanisms that contribute to snowballing:

• Citizen soldiers: This means that gaining a military advantage is often accompanied by a proportional economic advantage. Also, by training economy units from military buildings (and military units from the CC), the military infrastructure is largely the economic infrastructure, and vice-versa.
• Loot: Resource and experience loot leave the victor with a military and economic advantage. However, promoted units are somewhat slower to gather, which partially counteracts the other advantages.
• There are other features that could arguably contribute to snowballing as well, like preventing construction with ranged units, but these contribute much less.

Ideas to reduce snowballing:

• Unit diversification and counters: by implementing more sophisticated relationships between the units, we can decrease snowballing by allowing smaller armies to beat larger ones if they have a better composition.
  • we will need to vary more factors that are often held constant for whole unit classes in 0ad, like price, train time, and HP.
• Shift economic contributions more towards civilians (formerly women): This would benefit 0ad by partially deblurring the line between military investment and economic investment, and by decreasing the economic opportunity cost of sending citizen soldier (CS) units to battle. Alternatively, a new eco unit could be introduced to complement civilian gathering abilities, such as a more metal/stone specialized gatherer.
• Unit-specific technology tree: By providing players ways to selectively buff particular units, strategy will emerge in the prioritization of "teching into" certain units based on the civ matchup, map, and resource availability. Developments in this area will compound nicely with unit diversification.

feel free to comment some thoughts and suggestions.

[chrstgtr]

Couple thoughts:

Other factors contributing to snowballing:

1. Promotion feature--this makes it very hard to fight back even if you have equal numbers. This is especially true if there are healers, which aren't much a factor in this alpha but have been in the past. 
2. Weak defensive building--I think this is the biggest factor in the current alpha. The current alpha has weak defensive buildings. As a result, players have a more difficult time going back to their safe space to regain strength.
3. Cav dominance--cav are the dominate meta for a lot of reasons. But it's difficult to repop with cav because of cost, slow train times, and the likelihood of continued harassment. 

I would focus efforts on mitigating the factors that contribute to snowballing. 

14 minutes ago, real_tabasco_sauce said:

Citizen soldiers

You're right that it has a snowball effect but I don't find it to be problematic. This seems to be one of things where the advantage is earned. It is also most pronounced in early/mid game, which I am fine with since it is hard to kill a CC before p3. 

16 minutes ago, real_tabasco_sauce said:

Loot

I don't find this a problem at all. It's relatively low numbers. (I actually think loot should be a bigger part of the game)

22 minutes ago, real_tabasco_sauce said:

Shift economic contributions more towards civilians (formerly women):

Agree.

22 minutes ago, real_tabasco_sauce said:

Unit diversification and counters

22 minutes ago, real_tabasco_sauce said:

Unit-specific technology tree

Without commenting on the merits of these two, I think these are entirely separate proposals that may or may not have some impact on snowballing. You should also consider that, while these will be available to defending players, they will also be available to attacking players, so the snowball effect may become greater with them too. 

[Deicide4u]

1 hour ago, real_tabasco_sauce said:

Unit diversification and counters: by implementing more sophisticated relationships between the units, we can decrease snowballing by allowing smaller armies to beat larger ones if they have a better composition.

The current release (A27) already made some strides towards this, by increasing the melee damage. Units need further diversification with pikemen having more HP but lower damage, spearmen having average HP and average damage, while swordsmen have slightly above average HP and good hack damage.

Or we introduce counters to shift the current meta away from mass ranged units with melee meat shields.

1 hour ago, real_tabasco_sauce said:

Shift economic contributions more towards civilians (formerly women)

Needed to be done a long time ago. CS can already build and repair, in addition to fighting. Why do they contribute to the economy more than civilians? Where's the logic here?

1 hour ago, real_tabasco_sauce said:

Unit-specific technology tree

Some upgrades to specific unit lines can't hurt. For example, in StarCraft, there is a huge difference between an un-upgraded Zergling and a Zergling with both Metabolic Boost and Adrenal Glands upgrades.

[Atrik]

Not sure 0ad is having too much of a snowballing problem. The only case where it does feel like being the case this alpha is with fast, strong, cheap, melee units like champ spear cav and fanatics, because they have no counters and counter everything themselves.

[Deicide4u]

6 hours ago, Atrik said:

Not sure 0ad is having too much of a snowballing problem.

We're not talking about the balance here. We're talking about what happens when one player loses a lot of his units, either defending or attacking.

Does he, assuming he has the necessary macro skill, have a chance at coming back into the game? As things stand now, he doesn't. That is snowballing. 

Edited 11 hours ago by Deicide4u

[Outis]

12 hours ago, real_tabasco_sauce said:

Unit-specific technology tree

This is very welcome, but Tlthe main emphasis should be civ differentiation.

[Atrik]

4 hours ago, Deicide4u said:

That is snowballing.

Snowballing happens when there is no diminishing return on something you can accumulate, like champions.

Champions are also more resource efficient then their CS counterpart, as well as population efficient. There are no diminishing return with them, quite the opposite.

Fast melee champions units have no counters, and can hardly be "outplayed" because they can pick battles they want, therefor securing an advantage you have with them is easy.

Fast units contribute more to map control, therefore it is harder to recover if the enemy can just easily find and crush anything you try to rebuild.

There is almost no limit to how much strength you can coil up with champions, as you replace workers / CS with champions, your strength can grow despite the population limit.

The players that do play on snowballing don't send their CS to battle, they keep them on eco and replace them (even delete them) with fast units, be it champ cav, or cav. That makes me say that CS as workforce don't contribute to snowballing, instead they often provide an opportunity for defenders to keep eco, while the attacker is losing an opportunity cost of moving his army around.
Turtling is often done like this : build forward defenses, force the enemy to move his army and waste time on defenses, while you keep doing eco with most of your population; therefore you can catch up with any economic disadvantage you had.

[Deicide4u]

21 minutes ago, Atrik said:

The players that do play on snowballing don't send their CS to battle, they keep them on eco and replace them (even delete them) with fast units, be it champ cav, or cav. That makes me say that CS as workforce don't contribute to snowballing, instead they often provide an opportunity for defenders to keep eco, while the attacker is losing an opportunity cost of moving his army around.
Turtling is often done like this : build forward defenses, force the enemy to move his army and waste time on defenses, while you keep doing eco with most of your population; therefore you can catch up with any economic disadvantage you had.

The only reason people do this is because you can defend well with CS and use them as the backbone of your economy. CS gather 3 resources better than civilians, so there is little stopping people from just massing them and then gradually switching to champions when they bank on resources, as you've said.

Let's see how you'll turtle when forced to expand and make actual military units...or risk the opponent out-booming you with better eco units.

[guerringuerrin]

18 hours ago, real_tabasco_sauce said:

• Unit diversification and counters: by implementing more sophisticated relationships between the units, we can decrease snowballing by allowing smaller armies to beat larger ones if they have a better composition.
  • we will need to vary more factors that are often held constant for whole unit classes in 0ad, like price, train time, and HP

IMHO one of the things that contribute to snowballing is the lack of hard -or stronger- counters between units.
 

18 hours ago, chrstgtr said:

Weak defensive building--I think this is the biggest factor in the current alpha. The current alpha has weak defensive buildings. As a result, players have a more difficult time going back to their safe space to regain strength.

Agree with this too. I guess the difficult part is find the right balance to avoid encouraging turtling too much

Edited 5 hours ago by guerringuerrin

[Atrik]

49 minutes ago, guerringuerrin said:

IMHO one of the things that contribute to snowballing is the lack of hard -or stronger- counters between units.

Yes, definitively, more so, stronger and faster units should have also greater weaknesses.

49 minutes ago, guerringuerrin said:

18 hours ago, chrstgtr said:

Weak defensive building--I think this is the biggest factor in the current alpha. The current alpha has weak defensive buildings. As a result, players have a more difficult time going back to their safe space to regain strength.

Agree with this too. I guess the difficult part is find the right balance to avoid encouraging turtling too much

I think players say that because they expect turtling to be : "build a fort then you are safe".

Turtling this alpha is a tone of fun and it's balanced. The attacker CAN make progress, and you have to think of your defenses as layers, instead of just relying on just having invested a very low amount of resources that would make you immune forever to attacks.
Defending should be dynamic (and this alpha, it kinda was). There would be 0 fun if any players could just set up a base with a fort with swords garrisoned and a couple towers, and be rewarded with immunity without him having to do more.

I would point out that defenses are really cheap and currently are already very easy to make worth their cost:
A single tower can have 25+ kills over the course a game very easily. A fort with 80+ kills is common.
A tower cost 200 resource or 2 CS, a fort 900 resource or 9 CS.
Obviously not weak at all.

Edited 4 hours ago by Atrik