[en] Edain Mod > [Edain] General Suggestions

Snowballs and Sieging - Why Sieging sucks in edain

(1/2) > >>

Goodfella:
Sieging and Snowballs

Hi everyone, I’d like to make a post about sieging and how it currently functions in edain

I will not suggest solutions to this problem in this post (I will make a separate post about that), instead I will simply analyse the problem

If we can all agree that this problem is real, we can look at ways in which to solve it.

My definition of sieging:

When I talk about ‘sieging’ I’m talking about:

‘The endgame destruction of the enemy’s fort. The checkmate of edain.’

I am not talking about the early/mid game ‘raids’ on the base, which are much less affected by snowballing because these can (and do) happen when the balance between players is equal or even when the raiding player is losing.

The problem:

Sieging is boring in 4.4.1.

But why?

My answer:

The ‘snowballing’ effect makes games pre-determined before the siege begins. A pre-determined game is boring. Therefore, sieging is boring.

What’s more, there’s currently no incentive for the attacker to begin the siege. It is in his favour to wait, max-out, then siege. As for the defender, there’s no real incentive to do anything (except perhaps to harass) and so in most cases the defender resigns before the siege begins. This is, again, because of the snowballing.

Snowballing is not a bad thing. In fact, it is often necessary to end a game. However, I feel there is a unique problem with snowballing when we implement sieging because the siege inevitably starts when one player is already losing.

Adding more unique and interesting sieging units is great, but it does not solve the core problem. These units will only shine if a siege actually happens!

What is snowballing?

Snowballing is a positive feedback loop.

It is present in any game that involves resources that can be used to gain more resources. It goes a little something like this:

‘I win a battle, so I have more troops. I can now win more of the map and get more money. More money gives me more troops, more troops give me more money, more money gives me more troops etc. etc.’

Unchecked snowballing leads to pre-determined games.

‘My economy is so big that even if I play like trash I can still easily win, I can buy back all I lose, you can’t.’

The opportunities for a comeback are diminished by the second when one player snowballs.

You can read more about snowballing (and some proposed fixes) here:

https://gamedevelopment.tutsplus.com/articles/the-snowball-effect-and-how-to-avoid-it-in-game-design--cms-21892

Why does it matter in Edain?

Almost every game has snowballing, and for good reason: someone needs to win! Stalemates are even more boring than sieges! All counters to snowballing in game design therefore reduce snowballing, instead of removing it completely.

So why even care about it in edain?

Snowballing becomes a uniquely problematic issue when we introduce end-game sieging into the game because we start from a severely imbalanced position.
A siege will only begin if the attacker has already won the map. The attacker will therefore have a huge economic and/or military advantage over the defender. The snowball is already in full swing! What’s more, it keeps getting bigger and bigger.

The defender hides in his base whilst the attacker builds an unstoppable force. There is no reward for the attacker to attack! The game is decided, and the win is only becoming more secured the longer he snowballs, until he maxes-out, at which point he demolishes the fort with ease.

The defender for his part is also racing towards maxing out, but he’s so much slower than his enemy, so time is working against him. The best tactic for the defender is to leave his fort as early as he can and try to wrestle back control of the map, striking whilst the snowball is small. If he is ever forced to stay in his fort the game is over. This is the exact opposite of what we want for sieging to exist in online play!

Does the problem exist and what does it look like? Evidence from games

So, what happens? The battle for the map (and therefore the game) is won by one player. Both players then sit there waiting for the snowball to be big enough to end the game. Perhaps the defender tries to harass, but the attacker has a superior army that can stop this harassment, he can recapture lost territory more easily. The units he loses fighting for said territory are less valuable to him thanks to his superior economy:

‘Whatever you can do I can do better – and even if can’t, f*ck it I’m rich!’.

The snowball strikes again.

I’ve experienced this so many times in my own games and when observing others. I’m sure others can relate to this problem too. Just look at some of my videos, or your own replays or any game you play in the future.

Also, look to see how many times the ‘gg’ is given before (and in some cases WAY before) the siege starts. Why does this happen? Because we simply ask ourselves ‘What’s the point?’ and what is the point of sieging when the game is already over?

Conclusion

Sieging will not work in online edain unless we combat snowballing – some of the ways we can do this may seem a little extreme, but that is for my next post :D

Caveat

This problem of snowballing only exists if we want sieging to exist in online play. Offline there is no problem. The A.I does not care if the game is pre-determined. It’s actually quite nice to completely steam-role the enemy :D . The problem also does not exist if we don’t care about sieging being played online. We could just accept that it is an optional extra instead of an essential, and competitive, finish to a game.

NoldorSithLordsShipwright:
I think you'll be glad to hear (if you haven't already) that the next update (4.5) focuses specifically on this issue!

Walküre:
He must have surely noticed that, being him a Balance Tester. I'm quite certain about it ;)

I think the topic of the thread aims to gather additional feedback even from the public boards of the forum. Balance is usually deemed the toughest aspect to tackle in the management of a Mod; in this case, the issue really deals with so many different themes and it's commonly believed that balance itself seldom provides definitive answers. Hence, further debating on the matter is always more than welcome, whether siege is overhauled or not.

In my personal opinion, pre-determined games have become a recurrent and quite persistent woe of the Edain Mod. The cause of most of the concern in the community. I really hope that people will understand and get acquainted with the rationale of the forthcoming update.

Mogat:
I posted already a reply in the internal forums, but since we decided to discuss it publicly, here is the copy of what I have written:



Very nice analysis, it basically sums up the sieging-problem in competitive environment.

Before I give my opinion on sieges, some things about the snowballing you adress.

Snowballing as you say exists in every rts game, however there are some points, that make it exponentially stronger in Edain:
1) Hero levels:
If you manage to level up your heroes and your opponent doesn't, it is almost irrelevant what he does and how well he microes, as long as you keep them protected and don't completely brainfreeze.

2) The battalion mechanic:
As long as at least one soldier in one battalion survives you lose nothing.

3) Leveling mechanic of units:
In Edain, you almost always either completely crush your enemy in a fight or get completely beaten up. Anything inbetween almost never happens.
The consequense is, that while you keep every unit of which atleast one soldier survives + you keep the gained levels of you crushing that fight, your opponent does not only lose his units, which is bad enough, but even made you stronger through it.

Combined with the fact, that spellbook spells have been massively nerfed (and either require a big army in support, or just damage units/leave single soldiers alive) the snowball keeps rolling like an avelance, getting bigger and bigger.

Sieges

I have mentioned it countless times already, but imo the burden of winning lies on the defender, not on the attacker. The defender has to win through skill and not through artificial punishments on the attacker.
I agree however, that this is much more difficult to accomplish than in any other rts (starcraft for example), because of the humongous snowballing in Edain.

I always repeat myself, but bfme has given us the best comeback mechanics aleady: the spellbook.
But instead of allowing it to stand on its own, like in bfme1 and 2, the strength of spellbook spells in edain is proportional to the army supporting it, which is not necessarily bad, but makes comeback via spellbook very hard to achieve.
This means, that spells mostly benefit the snowballing player. (of course that is not the case with every spell, but that is the general direction they are going).

If we  try to give the defender some kind of artificial advantage by all available means (be it stronger defenses, more economy etc.) there are only two outcomes possible imo:
It will either not award the winner of the fight for mapcontrol, or just make sieges infinitely long and boring.


So how would I personally solve that situation?

I think first of all we have to put the very specific problem of sieging aside and adress the general possibility of comebacks, exacly as Goodfella states. If we manage to reduce snowballing to an equal level as in most other rts games (sc for example), it would be a big step in the right direction already.
If however we only focus on "sieging" and ignore the basic mechanics that lead to the problem, it will risk infinite sieges or the unimportance of the battle for mapcontrol.

After talking with Goodfella yesterday about this whole problem, I would agree, that introducing an active mechanic, that makes the attacker want to attack the base now and not sitting around and snowball out of control would be nice and at the same time still requires the defender to use it properly for it to help him (comeback through skill and not through sitting around and getting buffed up).

Mainly however, before anything else gets discussed, those conservative changes must be taken. That is the most important point in my eyes!

Julio229:
I'm not the best player myself (and more of an offline one than an online one), but I'd like to state my opinion on the snowballing  and comebacks issue, which I think I've seen a bit. I agree with what has been said, and in my opinion comebacks are very difficult to make because once a player controls the whole map, he has the most resources and henceforth, he will have the superior army and more capabilities to stop the efforts of the defender to come back, so all that is left is waiting until the attacker has enough troops to destroy the defender because there's no advantage for the defender to try to attack the enemy (as the enemy will have a bigger army to stop the harassing), and staying in the fort isn't helpful either because the enemy will continue to get a bigger army. Spellbook powers, as has been said, should help in this, but they are usually not strong enough to help make a comeback, and work more when the state of the game is more balanced and not in the later stages.

Another problem (that will probably be fixed in 4.5, I suppose) is the strength of towers in 4.4.1 (I haven't got enough experience with the megafix, but they seemed to be quite weaker and nicer to bear with there). While in theory the OPness of towers is amazing for the defender, when the stage of one player controlling the map and the other being almost defeated is reached, towers in their current state can unnecessarily make the length of the game longer, when there's no comeback possible because of snowballing and one player controlling all of the resources.

I hope that with the changes already stated to be in 4.5 to sieging, spellbook or economy buildings, it will be easier to make a comeback and such, make the later stages of the game, sieging and snowballing, less of an issue, and specially making sieging more fun and comebacks possible.

The tower thing is less related to it, but I think it is sort of, in the sense that, at that stage of the game, both a possible comeback and not taking thirty minutes to finish off a mostly defeated enemy should be possible.

So, in conclusion, I think there are two big problems: the snowballing and impossibility for the defender to come back once the other player controls the map, and the slog that sieging can suppose sometimes, that leads to snowballing as the attacking player would want to be sure to be able to get through the siege quickly. Sure, towers could defeat the armies, but the map control would give the attacker so many resources that, if he just escapes with at least one soldier of his battalions alive and keep the heroes alive aswell, they lost nothing but time. Finding a solution to this would not only help improving the siege in multiplayer, but also the game experience in general.

I hope I made sense, as I'm not exactly an expert on multiplayer, but I thought I could state my opinion on this and add a bit to it :P

Navigation

[0] Themen-Index

[#] Nächste Seite