Concerning randomness in games, and the question of luck and skill, I recommend you the following book : Characteristics of Games by George Skaff Elias, Richard Garfield, and K. Robert Gutschera.
Basically, Garfield states that randomness and skills are not opposite, they are orthogonal (Tic Tac Toe has no luck at all, but it doesn't make the skillset required to win the game very impressive, while Poker has a huge element of luck, but it is a game that works much better if you are skilled at it).
Still paraphrasing Garfield, randomness allows for :
-More different situations (hence the Fischer version of chess where the initial positions are randomized), and less focus on memorizing the perfect sequence of actions.
-It makes the game harder for the better player, and easier for the lower skill one, which results in less one sided boring games (the better player needs to play a conservative strategy to avoid luck playing too big a role, while the weaker one can still win from time to time, especially when adopting a high risk/high reward strategy). On the other hand, playing a stronger or weaker chess or RTS player than yourself is pretty boring and very often a foregone conclusion.
This is very important in playing against an AI in a game that would be too complicated to be brute forced.
They also make comebacks more likely than in a purely deterministic game.
- It introcues an element of risk management in the strategy
-It gives an excuse to lousy players to blame their defeats on.
End of my paraphasing of the book
The problem is not randomness in itself but the way it is used :
If it is used at the end of a game (like one scenario of Space Hulk, where you evacuate as many space marines as possible, and roll a dice afterwards. You win if you rolled less or equal to the number of Space marines evacuated), it is pretty lousy because you cannot take any action based on that.
Same usually goes for Computer RPG : Too much randomness can be bad, when you cannot cope with a single bad result, or a likely sequence of results (for instance, in many games, one character down might mean you either lost the game, or have a very reduced chance to complete it, forcing you to reload).
That is one of the problem in the RNG approach for the Battle For Wesnoth campaign actually (losing a high level unit can be enough to doom the whole campaign), but it harldy a problem in multiplayer(it is highly unlikely to field, let alone lose a high level unit in MP, so replacing casualties usually just require drafting a newbie warrior).
But if used sparingly it allows for games that stay interesting longer (the endgame in Chess is hardly what anyone could call interesting for instance), and with a wider range of opponents. It forces the player to rely more on strategy and less on memorization (and other skills very specific to a single game).
So how does it apply to Drums of War?
I think Drums of War strikes a good balance between randomness and skill :
You cannot guarantee to go through a battle without losing a single guy, but you can make sure your key characters don't die, and you can insure that you will win the battle. You don't have to replay the same turn several times to get a favorable RNG sequence(you cannot btw, as you can only save between battles).
Basically, Garfield states that randomness and skills are not opposite, they are orthogonal (Tic Tac Toe has no luck at all, but it doesn't make the skillset required to win the game very impressive, while Poker has a huge element of luck, but it is a game that works much better if you are skilled at it).
Still paraphrasing Garfield, randomness allows for :
-More different situations (hence the Fischer version of chess where the initial positions are randomized), and less focus on memorizing the perfect sequence of actions.
-It makes the game harder for the better player, and easier for the lower skill one, which results in less one sided boring games (the better player needs to play a conservative strategy to avoid luck playing too big a role, while the weaker one can still win from time to time, especially when adopting a high risk/high reward strategy). On the other hand, playing a stronger or weaker chess or RTS player than yourself is pretty boring and very often a foregone conclusion.
This is very important in playing against an AI in a game that would be too complicated to be brute forced.
They also make comebacks more likely than in a purely deterministic game.
- It introcues an element of risk management in the strategy
-It gives an excuse to lousy players to blame their defeats on.
End of my paraphasing of the book
The problem is not randomness in itself but the way it is used :
If it is used at the end of a game (like one scenario of Space Hulk, where you evacuate as many space marines as possible, and roll a dice afterwards. You win if you rolled less or equal to the number of Space marines evacuated), it is pretty lousy because you cannot take any action based on that.
Same usually goes for Computer RPG : Too much randomness can be bad, when you cannot cope with a single bad result, or a likely sequence of results (for instance, in many games, one character down might mean you either lost the game, or have a very reduced chance to complete it, forcing you to reload).
That is one of the problem in the RNG approach for the Battle For Wesnoth campaign actually (losing a high level unit can be enough to doom the whole campaign), but it harldy a problem in multiplayer(it is highly unlikely to field, let alone lose a high level unit in MP, so replacing casualties usually just require drafting a newbie warrior).
But if used sparingly it allows for games that stay interesting longer (the endgame in Chess is hardly what anyone could call interesting for instance), and with a wider range of opponents. It forces the player to rely more on strategy and less on memorization (and other skills very specific to a single game).
So how does it apply to Drums of War?
I think Drums of War strikes a good balance between randomness and skill :
You cannot guarantee to go through a battle without losing a single guy, but you can make sure your key characters don't die, and you can insure that you will win the battle. You don't have to replay the same turn several times to get a favorable RNG sequence(you cannot btw, as you can only save between battles).
Last edited:
