Improvement rolls rethought

[Deleriad]

I've been thinking about IRs recently and wondering about a different mechanic. IRs are derived from older versions of RQ where you had a chance of learning from experience each time you successfully learned a skill. Rather than tracking skill use, IRs are awarded at a certain point.

The basic premise is that the higher the skill, the slower it advances. IRs, like older experience rolls, do this by reducing the chance of gaining an increase as your skill gets higher. However, if you do get an increase then the amount of the increase stays constant. So I'm wondering about changing this.

My proposal is that you no longer have to make a roll for an IR, instead, the amount you get for an IR decreases as a skill increases. E.g.
|Current skill| Increase|
|1-50%|1d6+1|
|51-100|1d6|
|101-150|1d6-1*|
|151-200|1d6-2*|

*minimum increase of +1.

That is just an example - I've made the increase quite rich because IRs are quite stingy when played RAW. It might also work better with the training options in the GM guide as you would no longer need different mechanics.

What do people think? One disadvantage might be that some players would need to look at a table to remind themselves about what their increase is but it's a maximum of once per session so probably not that much of a problem.

 

[Mugen]

I like this idea.

However, I would use random values that are more in line with current method.

I mean that with the original system, when you're at X% in a skill, you have :

[X + (100-X)/6)] % to get 1 skill %
(100-X)/6 % to get 2 skill %
(100-X)/6 % to get 3 skill %
(100-X)/6 % to get 4 skill %
(100-X)/6 % to get 5 skill %
(100-X)/6 % to get 6 skill %

So, a skill improvement roll with a skill of X gives on average :

[X + 3,5*(100-X)]/100 skill %

= 3,5 - (2,5*X)/100

 

[soltakss]

I like this idea.

However, I would use random values that are more in line with current method.

I mean that with the original system, when you're at X% in a skill, you have :

[X + (100-X)/6)] % to get 1 skill %
(100-X)/6 % to get 2 skill %
(100-X)/6 % to get 3 skill %
(100-X)/6 % to get 4 skill %
(100-X)/6 % to get 5 skill %
(100-X)/6 % to get 6 skill %

That's the kind of formula I like - copy and paste but forget to change the details. I do it all the time in my programs (but don't tell my boss).

 

[Mugen]

That's the kind of formula I like - copy and paste but forget to change the details. I do it all the time in my programs (but don't tell my boss).

Could you tell me exactly what is wrong ?
I'm the kind of people that is unable to find errors in its own code...

 

[Deleriad]

Well the idea behind the numbers chosen was that I personally find the current IR increases a very low given that you get so few of them. Another possibility might be:
roll 1d6 for each portion of 50% of the current skill. Your increase equals the lowest number rolled. E.g. If your skill is 73% you roll 2d6. If you roll a 2 and a 5 then your skill increase = 2.

I've also been wondering about a better way to do characteristic increases. For example, Dark Trolls with a STR of 3d6+6 (av STR 17) find it more difficult to increase their STR than an average human (av STR 11). Haven't come up with anything too compelling. A math free way to do it might be:
Roll the character's normal characteristic dice and add three. If the result is more than or equal to their current score then the characteristic increases by 1. E.g. Rogrot the troll is STR 19. Normal dark troll str is 3D6+6. Roll that and add three. If the total is 19 or more, Rogrot gets +1 STR. (Otherwise increase a STR-related skill instead.)

 

[Mugen]

I've also been wondering about a better way to do characteristic increases. For example, Dark Trolls with a STR of 3d6+6 (av STR 17) find it more difficult to increase their STR than an average human (av STR 11). Haven't come up with anything too compelling. A math free way to do it might be:
Roll the character's normal characteristic dice and add three. If the result is more than or equal to their current score then the characteristic increases by 1. E.g. Rogrot the troll is STR 19. Normal dark troll str is 3D6+6. Roll that and add three. If the total is 19 or more, Rogrot gets +1 STR. (Otherwise increase a STR-related skill instead.)

I see a problem with the "bell-curved" nature of characteristics rolls. Increasing low stats would be easy and increasing high stats very difficult, perhaps not even worth spoiling 3 IR.

Also, it would be unbalanced for species with characteristic rolls different from 3D6+Y.

 

[Deleriad]

I see a problem with the "bell-curved" nature of characteristics rolls. Increasing low stats would be easy and increasing high stats very difficult, perhaps not even worth spoiling 3 IR.

Currently if you have a STR of 17 you have a 15% chance of getting an increase and that's definitely not worth spending 3 IRs on.

Also, it would be unbalanced for species with characteristic rolls different from 3D6+Y.

Well consider the odds of improving your STR to your species maximum from 1 point below species max under MRQ.
Halfing STR 2d6. Max STR = 15. Chance of improving to STR 15 = 30%.
Dwarf STR 4D6, max=27. Improve to 27 = 5%.

Another way of putting it. Imagine a character who rolled max STR. They can improve by 3 more points. Look at the cost in IRs to do so (on average).
Halfing. STR 12 to STR 15. 3/.4 + 3/.35 + 3/.3 = 19IRs.
Dwarf. STR 24 to STR 27 = 3/.05*3. = 180IRs.
i.e. it costs a Dwarf nearly 10 times as many IRs.

This is true of any species with a maximum rolled stat of 19 or more. Any creature with an *average* rolled stat of 19 or more might as well give up on increasing any of those stats.

The other way to do it is to say that your chance to increase your stat = (species maximum - your current stat)*5%. So a Dwarf that is STR 15 has a (27-15)*5% chance of an increase (60%). To be honest, that's the way that I would probably do it *except* it's still far too expensive in IRs for relatively little gain. Goodness knows how all the NPCs that Mongoose publish with their stat increases manage to afford them.

 

[soltakss]

That's the kind of formula I like - copy and paste but forget to change the details. I do it all the time in my programs (but don't tell my boss).

Could you tell me exactly what is wrong ?
I'm the kind of people that is unable to find errors in its own code...

I just didn't understand it (I'm just a Maths Graduate).

You seemed to have the same line repeated several times with just the percentage gain changed.

Perhaps a couple of examples would help to see how it works.

 

[Mugen]

That's the kind of formula I like - copy and paste but forget to change the details. I do it all the time in my programs (but don't tell my boss).

Could you tell me exactly what is wrong ?
I'm the kind of people that is unable to find errors in its own code...

I just didn't understand it (I'm just a Maths Graduate).

You seemed to have the same line repeated several times with just the percentage gain changed.

Perhaps a couple of examples would help to see how it works.

Sorry.

So, the logic behind the formulae above is that under MRQ, with a skill of X%, you have the following odds after an improvement roll :

*X % to fail the IR and get 1% increase
*(100-X) % to get 1d6 % increase

Which means :

*You have X % +(100-X)/6 chance to increase your skill by 1%, and
*(100-X)/6 % to increase your skill by any number between 2 and 6.

Yes, I did a mistake as you get 1d4+1 and not 1d6 in case of a succesful IR...

 

[Cowboy]

What I've considered is: roll a d10 over the tens value of your current skill rating. If you succeed, you increase by a number of points equal to the number you exceeded the threshold by. You always gain at least one point if you roll a 10.

That way, your skill level affects both the chance and the amount of the increase.