Skip to Content

An alternative to bonus that cancel each other

4 replies [Last post]
larienna's picture
Joined: 07/28/2008

In a game I played and in a design of mine, I have a situation where something gives bonus to a die roll for each player.

Each player rolls a die and add their bonus. The highest roll + bonus is the winner.

For example, Player one has a total of +2 and player two a total of +5.

Now the +2 bonus of the first players is somewhat canceling the bonus of the 2nd player. So a +2/+5 bonus is the equivalent of a 0/+3 bonus or even a +252/+255 bonus. In the end, what matters is that there is a difference of 3 points.

Now I find this pretty dull, since for example, giving your self a +1 bonus, or removing 1 point on your opponent has the same results.

Now I was looking for alternative ways to calculate linear bonus which gives more interesting results and possibilities. I have found a few solutions which are not that interesting and would like to hear more suggestions.

A- Bonus is triangular, or another pattern. For example, 1 bonus point does not equal directly +1 to your die roll. You need to accumulate more and more points before you get a bonus. For triangular numbers it could give something like:

1 = +1, 3 = +2, 6 = +3, 10 = +4,

This is somewhat interesting because you cannot get very high bonus if you do abusive combination. You can also be in a situation where a player has exactly 6 points, and making him lose point make his bonus drop by one rank.

B- Maximum bonus: There is a roof to the maximum bonus you accumulate. When you reduce your opponent's level, you reduce the maximum instead of the bonus. So if the limit is +5 and a player has +7, then he gets +5, but if you give him a -1, the maximum drop to +4. So there is an advantage to drop your enemy value even if he exceed the maximum.

C-Not a direct contested roll: If players roll against a target number, giving yourself a bonus increase your chances to reach your target number, while a minus to your opponent reduce his chances to succeed. So the bonus does not directly cancel each other.

Any other ideas?

Relexx's picture
Joined: 05/31/2010
I personally Like A

I personally like A as a solution, however it does add more accounting. The other option is to adjust the dice being rolled. Realistically the bonus is creating a phase shift in the results. Thus changing the probability to of dice rolls (eg 1d12 vs 2d6) the phase shift is perhaps a little less linear.

Or you could use multiple dice and count the results, opposed to straight roll+bonus. Thus having +1 give 1 additional dice. The intent is to gain more dice of greater value than your opponent, or whoever has the greatest sum. That way +'s to you can increase your dice count, and -'s could decrease your opponents dice pool.

rcjames14's picture
Joined: 09/17/2010
Dampening or Amplifying Variance

It seems to me that when you look for non-linear scaling, you will either dampen or amplify differences created by bonuses. In your geometric scale example, if it takes 1 card to roll 1 die, 3 cards to roll 2, etc... you have introduced a logarithmic curve to the equation that functions like diminishing marginal utility to encourage generalization and return to the mean. Whereas if you roll 1 die for 1 card, 3 dice for 2 cards, etc... you have introduced an exponential curve that functions like positive feedback loops to encourage specialization and deviation from the mean.

Depending upon what you hope to achieve in the gameplay, some non-linear scalars will work for you while others work at counter-purpose to your goal. So, do you wish to dampen or amplify variance? Or is this just a theoretical exercise?

What I could not quite envision, but I think would be fascinating is a non-linear scalar that could be applied to your linear bonus which has a vacillating effect on variance. So, rather than being asymptotic, it fluctuates wildly, sometimes providing a great bonus, other times producing little benefit at all. I imagine that a supply/demand dynamic creates this fluctuation between low and high returns over time because of information assymetry and complexity. So, an actual agent/player based model would need to create a similar condition if it wishes to have the consequent effect.

Joined: 05/25/2010
I guess it really depends on

I guess it really depends on what you do with the rolls.

For instance if the rolls are not directly compared (as in risk combat), a +1 to a roll can mean different things.

To recall from an old thread. If you are rolling for damage and each attack has a different hit % then +1 to damage makes a much bigger splash on something with a 50% chance to hit than on something with a 10% chance to it.

I think you are wanting something simpler, but simply by adding any other dimension to the rolling (% of success, critical chance of any kind, thresholds for minimum success, etc), a simple +/- bonus can go a long way, especially if you give players options to use either/or.

But yes, this idea needs another mechanic paired with it to have any relevance.

irdesigns510's picture
Joined: 06/24/2009
drag racing.

larienna wrote:

Now the +2 bonus of the first players is somewhat canceling the bonus of the 2nd player. So a +2/+5 bonus is the equivalent of a 0/+3 bonus or even a +252/+255 bonus. In the end, what matters is that there is a difference of 3 points.

this, plus some other parts in your post, sounds like you are trying to equalize ranks.
they do this in drag racing by giving the slower car a shorter track, or an earlier green light.
what you have in "C" gives kinda the start of that, where the target number is the finish line.

one thing to work with might be a variable target number, where bonuses go to that instead of the dice, and eliminate that +0/+3 edge.

Syndicate content

forum | by Dr. Radut