Skip to Content
 

Help with some mathematical comparisons

5 replies [Last post]
Afrofrycook
Offline
Joined: 05/27/2013

So I'm in the middle of designing a board game that will have deck building and dice rolling elements in it. After some brain storming, I came up with a dice rolling system that had room for different methods of influencing it.

As far as Dice Rolling, the system works by having a target number. Hitting this number exactly provides you one success. Rolling higher gives additional successes equal to the difference between the number rolled and the target number. For example, if you had to roll a five and you rolled an 8, you'd have three successes. In this game, we're identifying these additional success as Threshold Success.

A related mechanic is that you start rolling a D6 in this game, but you can upgrade the dice you roll a "leveling up" system. You start with a D6, but you can end up with a D12 through upgrading.

Now we have three major types of effects that modify how successful you were.

1. A bonus to your dice roll. This is the most power, and as such, has an inherently lower number. We call this a Straight Bonus
2. A bonus that is capped by your Dice's highest number. So it's really good to make sure you do well, but won't provide you Threshold Successes. We call this a Natural Bonus.
3. A bonus to Threshold Success if the roll was successful. So in the above example, if you played a card that was +2 Threshold Successes, you'd have six Threshold Success instead of just 3. We call this a Threshold Bonus.

Now players will choose characters when they start, each belonging to different groups. Originally, each group was going to have access to the their own version of the above cards. But then I wondered if each of the three groups could each have one of the top three be an optimal Primary, with a less useful Secondary, and a near useless Inferior.

So for example.
Group A: Excels at Straight Bonus, Adequate at Natural Bonus, Inferior at Threshold Bonus.
Group B: Excels at Natural Bonus, Adequate at Threshold Bonus, Inferior at Straight Bonus.
Group C: Excels at Threshold Bonus, Adequate at Straight Bonus, Inferior at Natural Bonus.

I really like the idea of having each group be better at certain types. It makes the choices players make very apparent. But the question becomes, how do we make that balanced? Should I instead go with the easier "Everyone gets an Primary, Secondary, and Inferior version of each" method instead? Anyone have any thoughts on the matter? If you have math or sites where I can plug in the numbers myself, I'd be very appreciative.

Thanks!

laperen
Offline
Joined: 04/30/2013
im not very clear about your

im not very clear about your bonuses, so just trying to clarify, correct me if my guesses are wrong

Straight bonus - directly increasing your dice roll to reach the target number

Natural bonus - lowering the target number?

Threshhold bonus - additional effects should you roll higher than the target number

kos
Offline
Joined: 01/17/2011
Mathematical challenge

I appreciate a good mathematical challenge.

The thing that complicates this challenge is the variable target number and variable dice size. The relative value of each of the 3 types of bonuses varies depending on the target number compared to the dice size.

Definition: "Value" in my analysis below is the expected number of Threshold Successes that a +1 card would give you (on average) above what you would have expected to receive without it. So a Value of "1" means you can expect to get +1 success from a +1 modifier, while a value of "0.5" means you can expect to get an extra 0.5 success from a +1 modifier.

I put these definitions into a spreadsheet and came out with the following results for 1d6.

Target Straight Natural Threshold
0 1 0.83 1
1 0.83 0.67 1
2 0.83 0.67 0.83
3 0.67 0.5 0.67
4 0.5 0.33 0.5
5 0.33 0.17 0.33
6 0.17 0.17 0.17
7 0.17 0 0

(Apologies for the formatting. Paste it into your favorite spreadsheet to see it better.)

Straight and Threshold come out almost exactly the same, with the only differences being at 1 and dice_size.
Natural is almost always 1/dice_size worse than Straight and Natural, again with exceptions at 1 and dice_size.

Note that the same analysis where "value" was defined as "the chance of getting at least 1 success" would yield different results. For example, although Straight and Threshold yield the same average number of successes over time, we know that on any individual die roll Straight has more chance of getting at least 1 success whereas Threshold has more chance of getting 2 successes.

I also expect (but have not verified mathematically) that there will be an increasing gap between Straight and Threshold with higher modifiers. E.g. I expect that +3 Straight would be substantially better than +3 Threshold becoming even more pronounced as the target number gets higher.

I also expect that the difference between the effect types would be less pronounced as the dice size increases. E.g. The differences between +3 on 1d6 will be the same as +6 on 1d12.

Overall, I would rank the 3 bonuses as Straight > Threshold > Natural, where for small bonuses Straight = Threshold.

Regards,
kos

Afrofrycook
Offline
Joined: 05/27/2013
laperen wrote:im not very

laperen wrote:
im not very clear about your bonuses, so just trying to clarify, correct me if my guesses are wrong

Straight bonus - directly increasing your dice roll to reach the target number

Natural bonus - lowering the target number?

Threshhold bonus - additional effects should you roll higher than the target number

Very close. The Natural Bonus is instead the same as the Straight Bonus, only it is limited by the Dice. The Threshold Bonus basically just makes your successes extra successful, but only if you hit the target number.

Sorry if the above was confusing.

Knicksen
Knicksen's picture
Offline
Joined: 05/18/2011
Dice website

Have you seen this...

http://anydice.com

I use it to test probabilities and the site owner is helpful.

Afrofrycook
Offline
Joined: 05/27/2013
Knicksen wrote:Have you seen

Knicksen wrote:
Have you seen this...

http://anydice.com

I use it to test probabilities and the site owner is helpful.

I've been poking at this since you posted. Still trying to figure out if it can show what I need. But overall, it looks neat. :D

Syndicate content


forum | by Dr. Radut