Jump to content

  • Log In with Google      Sign In   
  • Create Account

Interested in a FREE copy of HTML5 game maker Construct 2?

We'll be giving away three Personal Edition licences in next Tuesday's GDNet Direct email newsletter!

Sign up from the right-hand sidebar on our homepage and read Tuesday's newsletter for details!


We're also offering banner ads on our site from just $5! 1. Details HERE. 2. GDNet+ Subscriptions HERE. 3. Ad upload HERE.


[Monthly Discussion] on RPG flaws - Month 5: "Battle Encounter Design"


Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.

  • You cannot reply to this topic
17 replies to this topic

#1 n00b0dy   Members   -  Reputation: 103

Like
-2Likes
Like

Posted 11 September 2012 - 06:20 AM

All rpgs have taken the assumption that given a good gear and skill level they should be able to win a fight 100% of the time. They adopt the holy trinity concept (healer, tank, dps) and end up with 5-10 min boss fights where they follow a predetermined macro they read from a guide. These fights allow 0 player creativity. Same combo spammed every battle, same outcome every battle, they follow a battle chorography like bots. A bot would be more efficient than a human in this case, 0 lag, perfect coordination, 100% ability to dodge aoe effects.

What if this wasn't true ? What if battles were luck based. And chance to win was based purely on luck. Better gear or playstyle will increase the chance to win but there would always the risk to outright lose regardless of your actions because luck failed.

An example from dnd that doesnt follow determined battles:

3 kobolt mages (each deal 40 dmg with ice ray every 2 sec). Aggro immunity.

vs you (3 members)
Tank: 110 hp. 3shotted.
Dps/healer: 50 hp. 2 shotted. Healer: heals 20-40 hp. Dmg dealers kill kobolts in 2-4sec.

Party compositions.
a) Random : If you get lucky, you manage to spread the damage and win the battle. Else you lose.

b) 3 Rogues : stealth, stunlock them and kill them before they react. if you get spotted before you stun, you die.

c) Holy trinity : with luck, if you spread the dmg you defeat them.

What would this allow is more chaotic situations, a party member dies, and then your party has to improvise in order to win the fight. You lose a party member, they lose a party member too.

Sponsor:

#2 Ashaman73   Crossbones+   -  Reputation: 7801

Like
2Likes
Like

Posted 11 September 2012 - 06:30 AM

All rpgs... adopt the holy trinity concept (healer, tank, dps)

You talk solely about MMORPGs ?

and end up with 5-10 min boss fights where they follow a predetermined macro they read from a guide. These fights allow 0 player creativity. Same combo spammed every battle, same outcome every battle, they follow a battle chorography like bots. A bot would be more efficient than a human in this case, 0 lag, perfect coordination, 100% ability to dodge aoe effects.

The same could be said about football or any other team sport.

What if battles were luck based. And chance to win was based purely on luck.

Then I would never see a reason to play this game...

#3 NaturalNines   Members   -  Reputation: 334

Like
0Likes
Like

Posted 11 September 2012 - 08:06 AM

5-10 min boss fights where they follow a predetermined macro they read from a guide

Dominant strategies are always going to emerge, and players will always be able to look them up. The only option you have here, I believe, is to make the game interesting enough and not too challenging so that the player will enjoy their unique builds or strategies without feeling the need to power play or that they're being handicapped.

n00b0dy, on 11 September 2012 - 07:20 AM, said:
What if battles were luck based. And chance to win was based purely on luck.

Then I would never see a reason to play this game...

I agree with Ashaman73. The only thing I hate more than winning based solely on luck is losing based solely on luck. If anything RPGs need to shift away from random chance, not towards it.
Good news, everyone! I have a signature now!

#4 PyroDragn   Members   -  Reputation: 404

Like
1Likes
Like

Posted 11 September 2012 - 08:54 AM

would always the risk to outright lose regardless of your actions because luck failed.


Losing based solely on luck is (almost) always going to be a bad thing. Imagine a game of Tic-tac-toe. If you play perfectly against another player who also plays perfectly, no-one will ever win. Now, if you played and at the end you rolled a die, if you rolled a 1-5 the result stayed as it was (a draw), if you rolled a 6 you lost. Most of the time, because of your ability you're going to draw, but sometimes you're just going to lose.

Apply the same to any skill game. "I was about to kill the boss, but I rolled a six." "I was going to cross the finish line, but I rolled a six." "I played a perfect game, but I rolled a six."

The problem is "losing regardless of your actions." Randomness could affect the encounter, but not determine the outcome outright. In fact some of the most memorable encounters I have had in current MMOs is from things randomly going wrong and having to adapt to them.

If you get unlucky, and you still pull through and win. Then it feels rewarding.

If you get unlucky, and you try to cope, but you don't manage to quite pull it together - then you appreciate the challenge, and you see what you should have done.

If you get unlucky and there's nothing you could have done, then you just feel cheated.

#5 Orymus3   Crossbones+   -  Reputation: 9959

Like
3Likes
Like

Posted 11 September 2012 - 09:14 AM

I feel that my thread (even down to the title and syntax) has been hijacked :S

Dominant strategies are always going to emerge, and players will always be able to look them up. The only option you have here, I believe, is to make the game interesting enough and not too challenging so that the player will enjoy their unique builds or strategies without feeling the need to power play or that they're being handicapped.

Chess?

Edited by Orymus3, 11 September 2012 - 09:15 AM.


#6 NaturalNines   Members   -  Reputation: 334

Like
0Likes
Like

Posted 11 September 2012 - 10:24 AM

I feel that my thread (even down to the title and syntax) has been hijacked :S

I knew something about this one looked different...

Chess?

Ok you got me there. I don't know of any dominant strategy in Chess, so that's a good exception.
Good news, everyone! I have a signature now!

#7 PyroDragn   Members   -  Reputation: 404

Like
0Likes
Like

Posted 11 September 2012 - 10:44 AM


I feel that my thread (even down to the title and syntax) has been hijacked :S

I knew something about this one looked different...

Chess?

Ok you got me there. I don't know of any dominant strategy in Chess, so that's a good exception.


I don't think it's really an exception. It's just not as obvious. There isn't a (known) perfect way to play since there are so many variables (including the fact that is a PvP situation), but it is definitely possible to make a wrong move. A grandmaster is going to beat a run of the mill player because he knows the better strategies for different situations. There are commonly used openings because they've come to be known as close to current best practice.

It is possible to study chess and learn better moves for different situations. We're just not at the meta-gaming stage where anyone knows "this is what you -must- do" because there's so much to consider.

#8 NaturalNines   Members   -  Reputation: 334

Like
0Likes
Like

Posted 11 September 2012 - 01:22 PM

It is possible to study chess and learn better moves for different situations.

Agreed, though I don't know chess tactics nearly well enough to debate their merits. Still, chess serves as a great example of a game resistant to dominant strategies, since you won't find any tactics/moves nearly as powerful or widely applicable as Halo 1's pistol or FF6's morph-nuking bosses (excluding the 4 [5? i forget] move check-mate, but that doesn't work on pros, they'll just laugh at you).

Still, the harm caused to gameplay by the existence of a dominant strategy is that it can be applied with advantage to every situation, making variance worthless. I'm pretty sure chess avoids this problem.
Good news, everyone! I have a signature now!

#9 Orymus3   Crossbones+   -  Reputation: 9959

Like
0Likes
Like

Posted 11 September 2012 - 01:52 PM

The resistance of chess can be found in a lot of well-balanced games such as Starcraft 2. If the player has multiple viable options, and an ability to change strategy to adapt, then you get something going on. Small balancing mistakes quickly ruin this though.

#10 bwight   Members   -  Reputation: 165

Like
1Likes
Like

Posted 11 September 2012 - 02:02 PM

Chess only lacks dominant strategies when both opponents are equally skilled at playing. If a professional is playing a novice the professional has a few dominant strategies up his sleeve that give him the upper hand. In MMOs you can look at it the same way. The NPCs are not as skilled as the players and have no knowledge of how to counter the dominant strategies and level the playing field again. For this reason its generally more difficult to find dominant strategies in PvP situations. Of course there will always be dominant players and gear but its not the strategy as much as it is the player. If the AI for the game could be improved to learn some basic counters to common strategy then the game would seem less predictable.

Lets take a few examples
If you're in a PvP arena and you're facing a healer + tank + dps what are some strategies?

1. Try to reduce the damage output of the dps player and stay alive until the healer runs out of mana
2. Try to burst kill the healer and then take out the other players
3. Try to spread damage out across everyone and see if the healer can keep everyone alive
4. Try to burst the dps player and use interrupts on the healer at strategic times to make healing difficult

There are many more strategies that can be inserted here but my point is that the dominant strategy will often vary depending on your group makeup and how the opposing players are effectively countering your strategy. Players will generally try one strategy and make a decision if it will work or not. The NPC players have one strategy for all fights they attack the person who is doing the most damage or healing until a tank taunts them. I know that in a PvP environment unless the tank was being disruptive to my offensive or healing abilities I could care less about him, so why should a NPC treat a tank any different. You have to remember that the computer could be easily be programmed to be way to dominant.


Players employ many different strategies for protecting healers in battlegrounds. Healers are almost always played behind the front line of attack; if they get too close they become an easy target. If they're being attacked by a ranged player they generally try to position themselves in a location that puts the other player vulnerable to their team and hope to stay alive long enough for their team to kill the opposing player. Other strategies include hiding healers in a group of people or on a cliff where maybe you weren’t looking. Line of sight is also a common tactic in combination with quick heals. I could go on but my question is, "Why can't the same tactics apply to the raids?".

Its not about luck, its about adapting and countering attacks or defenses.

Edited by bwight, 11 September 2012 - 02:05 PM.


#11 PyroDragn   Members   -  Reputation: 404

Like
0Likes
Like

Posted 11 September 2012 - 06:53 PM

I'm pretty sure chess avoids this problem.


The NPCs are not as skilled as the players and have no knowledge of how to counter the dominant strategies and level the playing field again. For this reason its generally more difficult to find dominant strategies in PvP situations. Of course there will always be dominant players and gear but its not the strategy as much as it is the player. If the AI for the game could be improved to learn some basic counters to common strategy then the game would seem less predictable.


As bwight noted above (+1 btw) the biggest reason chess is so lacking in dominant strategy is because it's a PvP game, with balanced teams. If you played chess against a computer who made predictable moves as a response then it wouldn't be very long before someone came up with a dominant strategy at all. The random element of the human player is the biggest counter to the dominant strategy.

Getting back to the OP, this doesn't mean that you should just start randomising the chance to win, but adding variation to an encounter can make the encounter more interesting, and even better would be if you could get the boss to react to what the players are doing (in an intuitive way).

#12 NaturalNines   Members   -  Reputation: 334

Like
0Likes
Like

Posted 12 September 2012 - 09:14 AM

it wouldn't be very long before someone came up with a dominant strategy

I think you and bwight might be mistaking what is being referred to by dominant strategy. This link might clear that up.
http://levine.sscnet.ucla.edu/general/cogsci.htm
Good news, everyone! I have a signature now!

#13 Orymus3   Crossbones+   -  Reputation: 9959

Like
0Likes
Like

Posted 12 September 2012 - 10:30 AM

Chess only lacks dominant strategies when both opponents are equally skilled at playing. If a professional is playing a novice the professional has a few dominant strategies up his sleeve that give him the upper hand.

I'm not so sure. A lot of high-ranked chess players have to rake a few games before seeing any actual tendencies. A best of 7 could really end 4,3, which isn't that representative of difference in skill and/or use of a dominant strategy. The strategy would therefore be called a viable option, not a dominant one.

#14 bwight   Members   -  Reputation: 165

Like
0Likes
Like

Posted 12 September 2012 - 10:44 AM


it wouldn't be very long before someone came up with a dominant strategy

I think you and bwight might be mistaking what is being referred to by dominant strategy. This link might clear that up.
http://levine.sscnet...eral/cogsci.htm


You're right, my definition of dominant strategy does not match with what is described in the article you posted. So i'll agree with you that Chess does not have a dominant strategy. A dominant strategy by definition in the article you posted is a strategy that works regardless of what the opposing player chooses to do. I think though that most MMOs do not have a dominant strategy, at least by that definition, maybe some RTS games do.


Chess only lacks dominant strategies when both opponents are equally skilled at playing. If a professional is playing a novice the professional has a few dominant strategies up his sleeve that give him the upper hand.

I'm not so sure. A lot of high-ranked chess players have to rake a few games before seeing any actual tendencies. A best of 7 could really end 4,3, which isn't that representative of difference in skill and/or use of a dominant strategy. The strategy would therefore be called a viable option, not a dominant one.


I was talking about a professional playing against a novice player. If a pro was playing a novice i'd be surprised if the novice even won a single game. However, still there is no dominant strategy because the pro will always change his strategy based on what the other player is doing, this goes against what we just learned is dominant strategy.

#15 PyroDragn   Members   -  Reputation: 404

Like
0Likes
Like

Posted 12 September 2012 - 12:37 PM

I think you and bwight might be mistaking what is being referred to by dominant strategy. This link might clear that up.
http://levine.sscnet...eral/cogsci.htm


Dominant strategy equilibrium – strategy profile in which each player plays best-response that does not depend on the strategies of other players

I'm going to unfortunately disagree with Bwight and say that my definition of dominant strategy did match the above, and I still think that chess (probably) has dominant strategy.

Dominant strategy as I understand the above means "what move you make so that no matter what your opponent does you are in a better position to win" or perfectly, en-route to a guaranteed win (or at least a guaranteed non-loss, for example in the case of Tic-tac-toe).

This is not the same moving to put you in a better position than you are currently. Nor is it the same as moving to put you in a better position than your opponent. This'd depend on how you define 'better position' though.

Imagine the following scenario:

Two players want to kill a mob/boss. The boss is a pansy and does no damage, but he has a large amount of health so he takes a while to kill. Once he reaches 80% health he will randomly kill either 'Everyone within 10 feet' or 'Everyone more than 10 feet away.' There is a time limit to DPS him down to 0.

Now, the "Dominant Strategy" is for one person to be within 10 feet, and the other to be more than 10 feet away. Once he reaches 80%, one player is going to die, and the other player is free to DPS him down to 0.

However, if you consider that if you had both stood within 10 feet, and he killed everyone more than 10 feet away, you would be in a better position than the above, since neither of you die.

The dominant strategy isn't oriented towards "the best possible position" but towards "the best possible position considering every situation" or even "in the worst case scenario"

Talking about chess specifically, in the end-game dominant strategies become clearly apparent. In the following example:

(Image and quotes from: http://rjlipton.wordpress.com/2010/05/12/can-we-solve-chess-one-day/ )

Posted Image

Here White can give checkmate—in 99 moves. I don’t think a human would even find the first move, let alone all 99.


Computers have currently 'solved' perfect play for 5 pieces on the board. 3 pieces plus the 2 kings.

...if you got down to 5 pieces, you might as well be “playing chess against God”


Actually up to six pieces, and currently working on seven; the problem is getting to the point where we know every possible move right from the beginning of the game.

Now, none of this proves that there is a dominant strategy for the game of chess. But, if you got a chess board and placed the two kings and any three other pieces, there is definitely a dominant strategy for one of the players. The fact is that, from the start of the game, the teams are perfectly balanced, but one player moves first, and the other responds. This lends itself to a slight bias toward one player or the other, and makes it more probable for a dominant strategy to exist for one or both players.

We'll know in a few years once the computers have figured it out.

#16 NaturalNines   Members   -  Reputation: 334

Like
0Likes
Like

Posted 12 September 2012 - 02:16 PM

But, if you got a chess board and placed the two kings and any three other pieces, there is definitely a dominant strategy for one of the players.

But then you're not playing "Chess", you're playing "Some Game PyroDragn Made Up Using Chess Pieces and a Picnic Blanket". Just because a "strategy" happens to be "dominant" in a particular scenario does not mean that its a dominant strategy for the game in general. I think you're having a semantics issue here.

My favorite, and what I believe to be the simplest, example is the pistol from Halo 1. It became a dominant strategy among experienced players to use the pistol because it could kill at any range in 3 shots, which resulted in a higher DPS than most other weapons. Any weapons that could beat the pistol in DPS suffered from fallbacks that maintained the pistol's dominance: the shotgun could only beat the pistol in DPS at very close range and was useless medium-long range, the sniper rifle and rocket launcher had a much more limited clip/ammo capacity, only 1 or 2 of any existed per map, and all had to be found in-game (whereas players started with a pistol). This makes the pistol a dominant strategy because no matter the scenario, a pistol was not only a good bet, but an advantageous one (if you could land the shots, of course).

Hope that cleared it up.
Good news, everyone! I have a signature now!

#17 PyroDragn   Members   -  Reputation: 404

Like
0Likes
Like

Posted 12 September 2012 - 02:35 PM

But then you're not playing "Chess", you're playing "Some Game PyroDragn Made Up Using Chess Pieces and a Picnic Blanket". Just because a "strategy" happens to be "dominant" in a particular scenario does not mean that its a dominant strategy for the game in general. I think you're having a semantics issue here.


The point I was making was that, if you play a game of chess and you get down to a situation with 6 or less pieces on the board, there is an absolute, demonstrable, perfect way to play, ie, a dominant strategy. Currently we're working on being able to understand with a 7 piece scenario. Then will come 8, then 9, etc etc. When we get to the point of having calculated every possible move from the start of the game, then we'll know how to play "the perfect game" it'll just take a while until we get there.

#18 n00b0dy   Members   -  Reputation: 103

Like
0Likes
Like

Posted 12 September 2012 - 03:58 PM

yeah in about ~100-500 years of computer calculation, but maybe the computer that performs this caclulation may crash or run out of electricity, then its all from start. Of course if they use many computers it could be reduced.




Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.



PARTNERS