I was just posting in another thread and mentioned the concept of a "Fair Dice Outcome" generator, and I got to thinking that it could be a design option.
The way it would work is that instead of the engine throwing dice, it would generate expected outcomes in as smooth/even a fashion as possible. Given that standard WG dice aren't fair in the first place, this would result in a WLWLWL.. pattern with the occasional extra W thrown in based on how many dice are attacking/defending and what the border dice mods are.
I'm not sure that I would use this as a designer myself, but I thought I'd throw it out there.. For instance, some designers might want to offer it as a Scenario option, and I suppose there might be some designs out there that might be better using as fair dice as possible.
It could be a game creator option in unranked games so that players could play test games to identify strongholds and weak areas in boards, or if replay (unranked, of course) is ever an option, it could be used to test for the expected outcome of a game.
Are you suggesting that luck would be removed from the game?
So now it's just 50/50 attack vs defend, so when you make any attack from any country, you'll know, before you try the attack, whether you'll take the country?
I like this idea a lot, and in fact it is something I have been thinking about recently. In my case though it is something I just wanted to apply to certain areas of the board, where I would rather luck have less of an effect. So ideally I'd like this to be on a per border basis.
One thing I was wondering, and maybe someone with better math skills can tell me - can you fake this by changing the dice sides? If I create a border where dice are 11v12. The overall odds are very close to a normal 6v6 border. I was wondering though, does 11v12 have less randomness/spread to the outcomes than 6v6. I have no idea if it does, but if so, you could get closer to a 'fair dice outcome', by just increasing the # of sides of the dice in the right ratio.
BorisTheFrugal wrote:Are you suggesting that luck would be removed from the game?
Pretty much, yes. It would turn a game into a pure skill proposition. Outcomes with good players would mostly be determined by luck of the draw. I.e, who goes first, who starts with which territories, etc.., so its uses might be limited, but there may be some interesting uses.
Ozyman wrote:
I was wondering though, does 11v12 have less randomness/spread to the outcomes than 6v6?
If I understand your question correctly; off the top of my head, no. The odds change, but I would conjecture that the occurrences of skewed outcomes wouldn't be any less common or volatile.
Ozyman wrote:
One thing I was wondering, and maybe someone with better math skills can tell me - can you fake this by changing the dice sides? If I create a border where dice are 11v12. The overall odds are very close to a normal 6v6 border. I was wondering though, does 11v12 have less randomness/spread to the outcomes than 6v6. I have no idea if it does, but if so, you could get closer to a 'fair dice outcome', by just increasing the # of sides of the dice in the right ratio.
Without doing any of the appropriate analysis, I can say that subjectively I feel like 11v12 is more "streaky" than 6v6. That is it is more likely to have long streaks of wins and losses.
Alpha wrote:Without doing any of the appropriate analysis, I can say that subjectively I feel like 11v12 is more "streaky" than 6v6. That is it is more likely to have long streaks of wins and losses.
My Spidey-senses have me disagreeing with Alpha on this one. I'm thinking that for all practical purposes, 100v99 is no more streaky than a coin flip.
Ozy - Even if you could choose a dice odds that would give a 50/50 dice odds even within the realm of tie goes to the defender, that won't be implementing M's suggestion. His suggestion would not just make it an equal likelihood that each team will win any given dice roll (50/50 odds) but would actually force a 50/50 outcome.
Translation: You can flip a coin 50 times, and though both are equal odds, you aren't guranteed 25 heads and 25 tails. M's suggestion is to make the output be THE ACTUAL output. So if you've got 51 armies, and an opponent has 50, you'll ALWAYS win.
M - Sorry, my friend: Do not want.
@Boris,
We are definitely thinking about two very different things. I'm not saying i like my idea. I just thought of it, that's all.
Your idea is interesting. I'm assuming that in a 51 v 50 battle, the victor only ends up with 1 army. My first instinct is to suspect that this would encourage blitzing. Players would deplete their attack stacks down to 1. Reducing as much of their opponents bonuss as possible. There might also be the potential for stalemate with some boards.
This could be true for my idea as well, so..
Boris - I know that it's not exactly the same, but I was wondering if you could approach it by modifying the dice sides. For my particular desire it would be if it wasn't 100% deterministic, but if I could get rid of half the randomness it would be enough for me. I think I might write a simulator later to see empirically how changing the dice sides affects the 'randomness' levels.
I could also see an exactly deterministic version being interesting.
For 1v1 unit battles, how would the winner be determined fairly?
tom wrote: For 1v1 unit battles, how would the winner be determined fairly?
..not only that, but how would border mods be treated? The more I think of it, the more I do not like this method of battle resolution.
Perhaps we should back up the train and ask ourselves what the uses of such a feature are. I get the sense that Boris simply wants a different battle resolution engine. I, on the other hand, would be much more interested in finding a way to force the expected outcome in as "non-streaky" a way as possible. This is much more likely to respect a given board's game-play characteristics and may present opportunities for players to explore strategies, etc.
I just thought of a way my proposed method could be modified in a way to make it much less deterministic and only occasionally predictable. There are probably a number of ways to do this, but the simplest is..
The designer (or possibly the game creator) designates a Streak Maximum. Dice are rolled as per usual and the game plays out normally until the number of wins or losses in a row reaches the maximum. No matter what dice are rolled next, the outcome is altered such that the very next battle is resolved in contrary fashion.
The cool thing about this idea is that it is dynamic. Low streak settings setting create tight controls on streaking at the cost of more predictable outcomes, and higher streak settings puts moderate reigns on streaking while minimizing the deterministic qualities of the feature.
This could be a global setting or even a territory or border attribute.
Specifics regarding wether or not the feature resets with every new territory attacked or every turn could be assignable.
A setting of 1 essentially turns the game into an alternating Win/Loss battle.
I suspect that depending on how tight the Streak Maximum is set, the actual distribution of outcomes will not be representative of the normal expectations for a given border. For example, I would guess that the expected 1.07:1 win ratio of a 6v6 dice border might be reduced to something like 1.05:1 with the streak minimum set to 4, but that may be the price to pay for putting the brakes on streaks.
Determinism could be further reduced if the Streak Maximum is a range. A setting of 3 - 5 (randomly produced by the engine) would have the feature kick-in in a more random fashion (although if a player notices 5 wins or losses in a row, they know for certain what the next outcome would be). Doubtless, there are ways to completely eliminate deterministic properties if this avenue was further explored. Basically, it's just a matter of placing the setting on an independent (and possibly adjustable) probability curve.
..Thoughts?
Ozyman wrote:One thing I was wondering, and maybe someone with better math skills can tell me - can you fake this by changing the dice sides? If I create a border where dice are 11v12. The overall odds are very close to a normal 6v6 border. I was wondering though, does 11v12 have less randomness/spread to the outcomes than 6v6. I have no idea if it does, but if so, you could get closer to a 'fair dice outcome', by just increasing the # of sides of the dice in the right ratio.
One way to 'even out' dice results is to add the number of dice required to get the maximum result desired. For example,
Dice Min Max Avg Chance to roll 6, 7, or 8
(1) d12 1 12 6.5 25%
(2) d6 2 12 7.0 44%
(3) d4 3 12 7.5 53%
In general adding dice will decrease the range, increase the avg, and make it more likely to roll near the average. This doesn't even out the distribution, it makes it more likely to roll the average which is what I think you were going for - it decreases the chance to roll low or high. Now, this won't help if you are needing to compare 1 die to another die, but it could help if you can compare (3) d4 versus (3) d4.
As an aside, is there a way to post a chart from MS Excel? I created a distribution chart of the dice rolls but could not figure out a way to post it.
SquintGnome wrote:As an aside, is there a way to post a chart from MS Excel? I created a distribution chart of the dice rolls but could not figure out a way to post it.
Can't you just copy and paste? (you may have to change the color of the font so that it shows against a black background).
SquintGnome wrote:One way to 'even out' dice results is to add the number of dice required to get the maximum result desired. For example,
This is an interesting idea, but I wonder that it might be too complicated for players to understand. Regardless, I have my doubts that it would prevent streaking. Wouldn't defender have the same "advantage"?
Ozyman wrote:One thing I was wondering, and maybe someone with better math skills can tell me - can you fake this by changing the dice sides? If I create a border where dice are 11v12. The overall odds are very close to a normal 6v6 border. I was wondering though, does 11v12 have less randomness/spread to the outcomes than 6v6. I have no idea if it does, but if so, you could get closer to a 'fair dice outcome', by just increasing the # of sides of the dice in the right ratio.
I hesitate to say that it's impossible, but several experiments indicate that 3v2 Risk dice don't like a high split percentage. For 6-sided, it's 37%-34%-29% for attacker lose 2-split-defender lose 2.
For 11v12-sided it's 36%-34%-30%. For 41v47-sided it's 37%-34%-29% also.
For close to even, like 9v10-sided, you get 34%-34%-32%. You get smaller variation from flipping coins twice: 25%-50%-25%.
M57 - A couple of thoughts:
1) I think there was confusion from my last post: I don't want 50/50 resolution, or to remove dice odds. I WANT dice odds, I WANT randomness, and I WANT luck to be involved. Without it, the game would feel pre-determined to me. My comment "M's suggest" was meant a mistype of "My Suggestion," it was me trying to paraphrase where I thought your suggestion was going. Your first post said something to the effect of the output turning into WLWLWLWL with a few extra W's put in because of modifiers. My description was (I thought) a redefinition of what you were proposing, which I didn't like. You apparently see them as different however, and so I need you to explain.
2) Reiterative: I do not like making the output of a 51 v 50 always go to the team with 51, because it takes the luck/chance/gamble out of the game. Strategy goes out the window if you can look at a run you want to make at an opponent and KNOW you can knock out his 3 continents. Strategy shows up when you DON'T know if you'll be able to break him, but you know you need to because he's stronger than you.
3) But as for your idea of a streakiness quotient....I need to ponder that one for a while before I jump in.
M57 - A couple of thoughts:
1) I think there was confusion from my last post: I don't want 50/50 resolution, or to remove dice odds. I WANT dice odds, I WANT randomness, and I WANT luck to be involved. Without it, the game would feel pre-determined to me. My comment "M's suggest" was meant a mistype of "My Suggestion," it was me trying to paraphrase where I thought your suggestion was going. Your first post said something to the effect of the output turning into WLWLWLWL with a few extra W's put in because of modifiers. My description was (I thought) a redefinition of what you were proposing, which I didn't like. You apparently see them as different however, and so I need you to explain.
2) Reiterative: I do not like making the output of a 51 v 50 always go to the team with 51, because it takes the luck/chance/gamble out of the game. Strategy goes out the window if you can look at a run you want to make at an opponent and KNOW you can knock out his 3 continents. Strategy shows up when you DON'T know if you'll be able to break him, but you know you need to because he's stronger than you.
3) But as for your idea of a streakiness quotient....I need to ponder that one for a while before I jump in.
BorisTheFrugal wrote:M57 - A couple of thoughts:
1) I think there was confusion from my last post: I don't want 50/50 resolution, or to remove dice odds. I WANT dice odds, I WANT randomness, and I WANT luck to be involved. Without it, the game would feel pre-determined to me. My comment "M's suggest" was meant a mistype of "My Suggestion," it was me trying to paraphrase where I thought your suggestion was going. Your first post said something to the effect of the output turning into WLWLWLWL with a few extra W's put in because of modifiers. My description was (I thought) a redefinition of what you were proposing, which I didn't like. You apparently see them as different however, and so I need you to explain.
@Boris,
Sorry, I didn't catch the subtlety in your post. And yes, I misunderstood your 51 over 50 statement to be a different idea where the larger number wins and no dice are thrown - I.e., the larger number wins, period. (Which addmitedly, is not much different than the WLWL solution, so I should have understood it). Regardless, I agree that the (my) WLWLW solution with the requisite W's thrown in is too predictable to be good as more than a strategy or positional testing tool..
3) But as for your idea of a streakiness quotient....I need to ponder that one for a while before I jump in.
First a disclaimer. Personally, I don't have a problem with "streakiness". I consider it part of the game. So I'm not endorsing my solution. I'm just offering it.
If players/designers want odds, randomness, and luck, but have streak aversion and think it ruins the game, an adjustable cap on streaks looks like a pretty simple yet elegant solution. I think it addresses the larger issues that a lot of players seem to have:
Ozy, Boris, are these the types of issues you're trying to address?