How Common Should Low-Attribute Penalties Be?

[Hotrod]

Branching off from a discussion on how we roll up attributes. I threw together a spreadsheet-based attribute roller for human attribute basic rules, including the bonus dice rolls and the extra D6 of the first bonus results in a 6. I did not include attribute bonuses from having 1 or 2+ stats below 7. After rolling up 100 stats, the two top-scoring results had the same combined attribute score, but one would have been wildly more effective than the other in most situations.

But I missed something. You see, one of those characters had an I.Q. of 8. In RUE, penalties kick in for an I.Q. of 8 or lower (your O.C.C. Related skills get cut in half, and you get 1D4 extra secondary skills; a pretty severe penalty, and they get worse the lower you get). For all other attributes, penalties kick in at a score of 7 or lower. If you get 6 or lower, you get to roll 1D4+3 and add a bit to a non-penaltied attribute of your choice.

I hadn't really considered this before, but a roll of 7 or 8 is not at all unusual. In fact, you're much more likely to have pretty severe penalties for a character's low attributes than you are to have a modest bonus for exceptionally high attributes.

Well, I re-ran the spreadsheet and rolled up 100 random sets of attributes, but this time I counted up the number of penalized stats per character under the rules as written (8 or lower for I.Q., and 7 or lower for the rest).

21 had no attribute penalty
33 had 1 penalized attribute
24 had 2 penalized attributes
16 had 3 penalized attributes
5 had 4 penalized attributes
1 character had 5 penalized attributes.

Now there's plenty of uncertainty here. With some matrix manipulation, I could produce an exact probability solution, or I could script out the dice rolls and increase the number of characters sampled high enough that the uncertainty would be negligible, but this little 100-character sample tells me that weak spots are the norm in rules as written. If this sample reflects the actual probability distribution, then roughly half the population has multiple attribute-based disabilities.

That carries some amazing implications for the world of Rifts. For instance, I've been giving my disposable N.P.C. generators a default value of 10 for all attributes. That puts them in a select top 20% of ability. That might be appropriate for N.P.C.'s expected to pose a meaningful challenge, but it doesn't reflect the general population. When we consider other species with a lower attribute than the human, say 2D6, this means that a majority of that species is disabled in that attribute by human standards.

It's so strange though, how rare it's been for me to see a published N.P.C. or player character dealing with these disabilities as written. Unless I got freakishly bad results in my sample, and I don't think I did, then under the rules as written, characters with disabilities are the norm, and characters without disabilities are the exception.


Is this normal for an RPG? Is it intentional? Is it how it should be?

[MyDumpStatIsMA]

Is this normal for an RPG? Is it intentional? Is it how it should be?

Thanks for all that data. I made manual rolling groups, but ultimately came to similar conclusions.

I do believe it's intentional. Given how KS writes about low-attribute characters in RUE, and how to deal with them. He evidently wants people to roleplay more than powergame.

It's an admirable goal, but I wager that most players want at least a couple of bonus-conferring attributes. Why? Well, we probably want to feel both special and specialized. Not much point in making a combat-oriented character with no PP bonuses. Likewise, why would you make a charming rogue with a mediocre MA score?

I think having every attribute high enough to give a bonus is as hollow as having all with penalties, or having all without bonus or penalty. Ideally, there should be a mix. I find the notion of playing characters who are neither good nor bad enough to get a bonus/penalty for anything, rather dull. If all our attributes are between 9 and 15, say, those numbers are utterly meaningless most of the time (with the exception of PE determining HP, and how long we can swim, etc). They act as placeholders, functionally. It's also hard to roleplay such 'averageness.'

I like the idea of having 1-2 strong attributes (with bonuses), most others middling, and at least one low attribute. Not necessarily low enough to give a penalty. Why? Well, as you noted, even the least severe penalties are relatively much more character-altering than the lowest levels of bonuses.

Having an ME of 19 isn't going to break your character by making it too strong. But having an ME of 7 will make you very weak. Forgetting all other penalties to skills, horror factor and such, a score of 7 will mean that as a non-psychic, you'll need to roll a 16 to avoid being crippled by psionic attacks. You're going to be losing a lot of rolls.

Having an ME of 19 will lower that roll difficulty to 13. You're still going to be losing more than half your rolls.

I feel like most bonuses don't start breaking the game until the attribute's 24. That's when most of the +5s start to kick in. And of course, 30+ is way overboard for all but a few OCCs and RCCs.

By the way, as an aside: Speed is way too easy to increase to the high 20s and low 30s. Running plus General Athletics plus an average initial 3d6 Speed roll, will often give you 25+ Speed. Other day I rolled up a Crazy and had several awful Speed-related rolls (that I didn't re-roll), so now his Speed is actually lower than several other OCCs I have that don't get any OCC bonuses to Speed. I've got a Robot Pilot with a Speed of 31. Not cherry-picked, just straight rolling.

[desrocfc]

<snip>

Well, I re-ran the spreadsheet and rolled up 100 random sets of attributes, but this time I counted up the number of penalized stats per character under the rules as written (8 or lower for I.Q., and 7 or lower for the rest).

21 had no attribute penalty
33 had 1 penalized attribute
24 had 2 penalized attributes
16 had 3 penalized attributes
5 had 4 penalized attributes
1 character had 5 penalized attributes.

[/b]

Is this normal for an RPG? Is it intentional? Is it how it should be?

Bearing in mind game design only takes you *so far* into modelling reality; the simulationist approach will always find flaws in the system.

From a statistical standpoint, I'd love to see what happened if you ran this test around 2000 times, which is closer to what would be used to model a population of hundreds of millions, like the USA. In addition to looking at penalized attributes, ones that generated a bonus and the median attribute scores per set would be valuable data points. Even with just 100 data sets, you have something like 75% sitting at 2 or fewer penalized attributes, which could be a 3 or an 8, which is a big difference in terms of penalties. Not sure how the script is written, but the identification of those 3s to 8s and to which attribute would be a factor as well. Even then, over half have 1 or no penalties, which anecdotally seems more realistic.

I'm definitively NOT a fan of the penalties for "Low" attributes and the overly punitive aspects, particularly given the higher ceiling for bonuses. I'm re-writing them to be a more holistic approach, where penalties only apply for 7 and below, and the penalty from 8 to seven is not the cliff it currently is. There is also a conflation of playing heroic characters with "average persons" of reality that does not translate well to RPG Attributes. Hey, a new blog article for me - Thanks!

NPC issues aside, I think this was likely too small a sample with a couple of key data points lacking to make meaningful implications; as presented though, your closing statement "under the rules as written, characters with disabilities are the norm, and characters without disabilities are the exception" certainly could hold water, until you dig a little deeper. Welcome to the dangers of statistical analyses. I mean, you're far and away better than the idiot-stick Lancet article that launched a Playboy Bunny to infamy for propagating some highly erroneous conclusions......

Is it "normal" is a little loaded, but I get what you're implying. Intentional or not, it's the baseline for PB Chargen, and I know enough folks either ignore the penalties or allow re-rolls so it's a moot point. Come to think of it, I can't remember ever playing a PC that rolled an Attribute less than 8.... <shrug>

[EliBenedict]

Exact distributions aren't too hard to work out. Turns out, approximately 41.94% of rolled characters have multiple penalties.

Dark Fortress has 3d6 probabilty distributions already worked out to four digits.
(https://www.thedarkfortress.co.uk/tech_ ... _rolls.php)

The chance of rolling an 8 or higher is .8379, and the chance of rolling a 9 or higher is .7407. So the chance getting a character with no attribute penalties is .8379^7*.7407=.2148.

The chance of a character with one penalty (only) is .8379^6*(1-.8379)*7*.7407+(1-.7407)*.8379^7=.3658.

Making the chance of a character with two or more penalties 1-.2148-.3658=.4194.

So about 41.94% of characters have multiple penalities, though 78.16% of rolled characters have at least one.

But, here's the other side of it.

As you noted, per RUE, p.282, a character with one score of 6 or lower (true for 37.48% of all characters) lower adds 1d4+3 (average of 5.5) to another score. If they've got two abilities 6 or lower (true for 16.48% of characters) they add 1d4+5 (average of 7.5) to one score, and 3 to another.

So almost 54% of rolled characters get a compensation bonus. Now, I'm running out of time on my break, so I won't calculate these probabilities, but, given that the average roll for a 3d6 is 10.5, and the average magnitude of a compensation bonus is 5.5, or 7.5 and 3 if you have two weak scores, and given that the character can assign the bonus where they choose, and assuming reasonable character with no exceptional scores would use it to push a higher score over the line so they can get at least some attribute derived bonus, we can conservatively guess that a majority of these characters will end up with at least one exceptional score.

Which means (likely) a plurality of all characters (at least 25%) will be in the situation of having at least one exceptional score (with a derived bonus) and at least one weak score (with a derived penalty.) Which has the net effect of making them all more "charged": both gifted and challenged.

[Hotrod]

<snip>

Well, I re-ran the spreadsheet and rolled up 100 random sets of attributes, but this time I counted up the number of penalized stats per character under the rules as written (8 or lower for I.Q., and 7 or lower for the rest).

21 had no attribute penalty
33 had 1 penalized attribute
24 had 2 penalized attributes
16 had 3 penalized attributes
5 had 4 penalized attributes
1 character had 5 penalized attributes.

[/b]

Is this normal for an RPG? Is it intentional? Is it how it should be?

NPC issues aside, I think this was likely too small a sample with a couple of key data points lacking to make meaningful implications; as presented though, your closing statement "under the rules as written, characters with disabilities are the norm, and characters without disabilities are the exception" certainly could hold water, until you dig a little deeper. Welcome to the dangers of statistical analyses. I mean, you're far and away better than the idiot-stick Lancet article that launched a Playboy Bunny to infamy for propagating some highly erroneous conclusions......

Is it "normal" is a little loaded, but I get what you're implying. Intentional or not, it's the baseline for PB Chargen, and I know enough folks either ignore the penalties or allow re-rolls so it's a moot point. Come to think of it, I can't remember ever playing a PC that rolled an Attribute less than 8.... <shrug>

Your comment about my little study's lack of statistical strength is 100% valid; there's a reason why I'm not using these methods for doing epidemiology studies on vaccines. I could write a script to iterate this a few hundred thousand times in MATLAB, but I don't have MATLAB handy, and I don't know Python well enough to write it out. The different threshold for low-I.Q. penalties make this a little complicated, so I just manually iterated the dice roll and counted it out myself. 100 iterations was my limit for a little forum post.

With the approach I took, I could easily be off on any of those groupings by as much as 10. To do this properly, I'd have to characterize each attribute penalty tier for each attribute in terms of its effects on a character, tie that into the difficulty associated with raising that attribute (very easy for speed, pretty easy for Strength, hard for P.P., and very hard for mental attributes) and create some histograms of outcomes with nice, tight error bars. If you want to do a deep dive on this, I'd be curious to see how you'd look at this question.

For the purposes of this forum post, a rough hack at the question with fairly crude methods was enough to surprise me. With character generation being a somewhat broken process, this aspect of Rifts is something I just haven't seen applied, not to N.P.C.'s and not to player characters. It just weird to realize that what the system as written produces does not match up at all to what the game's writers, GMs, and players tend to create.

There's a lot to explore about the attribute system as written, and I've seen much of it explored at length on these forums. I haven't yet come across any effort to dive into just how common low-attribute penalties actually are, or what they should be. I don't have any definitive answers on any of that; I mostly just wanted to raise the topic and hear some thoughts, and I very much appreciate yours.

[EliBenedict]

. With the approach I took, I could easily be off on any of those groupings by as much as 10. To do this properly, I'd have to characterize each attribute penalty tier for each attribute in terms of its effects on a character, tie that into the difficulty associated with raising that attribute (very easy for speed, pretty easy for Strength, hard for P.P., and very hard for mental attributes) and create some histograms of outcomes with nice, tight error bars. If you want to do a deep dive on this, I'd be curious to see how you'd look at this question.

I think tackling this with standard data analysis/statisical modelong tools is probably backwards. It's not like a real world inductive population study (or even the RPG equivalent where you'd have a sample of pre-created characters and try to discern statistically the underlying mechanics.)

The generative mechanics are already defined mathematically, so the proper analysis is arithmetic and probabilistic. Assuming d6s are equally likely to deliver any given 1-6 result, the distributions can be derived directly and exactly. No need for error bars.

[MyDumpStatIsMA]

Which means (likely) a plurality of all characters (at least 25%) will be in the situation of having at least one exceptional score (with a derived bonus) and at least one weak score (with a derived penalty.) Which has the net effect of making them all more "charged": both gifted and challenged.

Agreed, but then you're in the awkward position of hoping that you don't get low scores in the 3 mental attributes that can't be raised except by extreme means (like taking the Ley Line Walker OCC explicitly to counteract a penalized mental stat).

The absolute best possible (semi-plausible) roll would be something like 13 average on mentals, 6-13-13 on physicals, whatever for PB, and 6 for Speed. Speed and PS are both very easy to raise. So then a 1d4+5 and a +3 on any of the 13s would guarantee a bonus.

If we get a sub-7 roll on any mental attribute, we're going to need to use the +3 on it to attempt to raise it out of penalty status, in which case we've only really got a chance of using the 1d4+5 on a single stat to get a bonus.

It's all very messy and rather convoluted for my tastes. Last night I came up with the following simplification.

Eliminate all bonus dice on 16+ rolls. Eliminate compensation for low attributes. Choose 3 attributes to give +2, before you roll them up. Choose 3 attributes to give -2. And finally, choose 2 attributes to give +6. It's imperative to choose the attribute bonus/penalty before the roll, otherwise it's too easy to cherry-pick what goes where.

At face value, this might seem too generous to some people. It raised my average (manual rolls, 10 complete characters rolled up per test group, 3 prior groups) from 11.21 to 12.65.

However, even getting a 3d6+6 does not guarantee a bonus. I chose ME and PP for my +6, and though my PP rolls were unusually high (I never rolled under 12, and got two 17s, before the +6 was applied), I only got a bonus to ME on 4/10 rolls.

I tried eliminating the +2 element, and it resulted in gaining 6 more penalized attributes. It only created 2 bonus attributes, but it also generated multiple 15s on my PE that could easily reach bonus levels with skills applied.

With the +2 and -2 and +6 spread, I had 17 attribute rolls that reached bonus levels, and 11 that had penalties. Some as low as 2-3. Getting 11 penalties out of 80 rolls is typical for my 4 complete test groups. Eliminating the +2 would bring that total up to 17 penalties, which is rather high, but at least it's effectively balanced out by also having 15 bonus attributes.

[MyDumpStatIsMA]

It just weird to realize that what the system as written produces does not match up at all to what the game's writers, GMs, and players tend to create.

It honestly doesn't strike me as that weird, in the context of (speculation on my part) KS seemingly wanting to set the bar pretty low for character ability. He established that's the baseline, and we're free to push that baseline way up. Which, evidently, most of us do by various means.

My main goal is to find the most direct and probabilistically reasonable way to do so.

Re-rolling feels hollow to me. Might as well just write in whatever attributes you want, rather than waste time re-rolling to get the desired number.

[eliakon]

This also sets aside the fact that while you can roll up a character with five stats under 7... you are probably not going to actually play them which messes up the distribution curve a lot.

[drewkitty ~..~]

Looking at this as a lazy GM I would have the player re-roll ether the whole char or the 7- attributes.

[ShadowLogan]

Since someone mentioned that 100set was too low, I setup a Python script to generate 10,000 sets of 8-attributes using 3d6 +1d6 on a roll of 16+ with an exploding die of +1d6 if the bonus roll was a 6. I will define Penalty Attributes universally as 7 and below (7-), interestingly enough in RUE IQ of 8 should also qualify for this (the only attribute where 8 gives a penalty), but keeping it a universal 7 is easier. The scores where tracked two different ways: w/assigned attribute (in order as they appear on the character sheet IQ to SPD) and just generic score. This means for example a score of 8 would be recorded generically (in one file) and then again after being given its attribute partner (ex IQ) in a second file. The files then made their way to a Database program (LibreOffice Base in this case, would have preferred an old version of MS Access or Kexi but both seem to be giving me crashing trouble) where I ran SQL queries.

The attribute order would track 3 separate data points beyond the 8 attributes. Each of those extra Data points counted the number of dice scores that falls within a given range (penalty, bonus, normal), which means I would know how many instances of each type occured within a given set. So for example a set of 10, 13, 24, 8, 30, 10, 11, 7 would result in 2 bonus, 1 penalty, and 5 normal.

Of those 10,000sets:
-the maximum number of instances of Bonus Attributes (16+) was only 6 (8 should be possible, surprised it did not appear given the other two categories DID reach
-the maximum number of instances of Penalty Attributes (7-) was 8
-the maximum number of instances of Normal Attributes (8-15) was 8
-all 3 dice roll score zones did have instances of 0 (meaning scores are all penalties, normal, or bonus w/norm or pen)
-the computer's Random Number Generator (in this instance) resulted in 42 different combination of scores (by bonus, normal, penalty NOT individual attribute combinations)
--1x was 6 attributes with bonus and 2x normal
--1x was 6 bonus and 1 penalty
--1x was 6 bonus and 2 penalty
--1x was 1 bonus attribute with 7 penalty attributes
--1003x (the most common) of which had 1x bonus attribute with 2x penalty attributes
--surprisingly individual attribute combinations did not repeat (though maybe not really as there are like over 282 BILLION combinations IINM for 8 attributes with values between 3-30)

of those 80,000 attribute rolls (10,000sets * 8 attributes):
-14,974 where Bonus Attributes (18.71%)
--of which only 2,144 could be achieved by the Exploding Die (25+, technically 16+6=22 would allow for it, but that falls within the range of 3d6+1d6 so there could technically be more exploding die I just did not track exploding die instances intentionally)
--only 153 instances of a score of 30 was achieved (only a score of 23 had fewer at 131)
--4,258 instances of a score of 18 (the most frequent number of any bonus attribute score)
-40,049 where Normal Attributes (50.06%), technically 41 of these are an IQ of 8 (and should be counted as Penalty, BUT ARE NOT)
-24,977 where Penalty Attributes (31.22%)
--5,017 of which are 3s, surprisingly 7s (@5,063) where the most common penalty attribute score

I would be remiss if I did not point out that these results apply to this specific instance of using the random number generator to produce dice results. It is entirely possible that if I ran it again I might end up with a different break down. A larger data set would be easy enough to do with the python script, I'm just not sure how large I can go before the database programs I am familiar with might break (Base was getting slow in returning values) or even run out of disk space (I typically use small partition sizes).

If anyone wants to see the code for the script I will post it AS IS and with a WARNING TO USE AT YOUR OWN RISK disclaimer.

[MyDumpStatIsMA]

of those 80,000 attribute rolls (10,000sets * 8 attributes):
-14,974 where Bonus Attributes (18.71%)
--of which only 2,144 could be achieved by the Exploding Die (25+, technically 16+6=22 would allow for it, but that falls within the range of 3d6+1d6 so there could technically be more exploding die I just did not track exploding die instances intentionally)
--only 153 instances of a score of 30 was achieved (only a score of 23 had fewer at 131)
--4,258 instances of a score of 18 (the most frequent number of any bonus attribute score)
-40,049 where Normal Attributes (50.06%), technically 41 of these are an IQ of 8 (and should be counted as Penalty, BUT ARE NOT)
-24,977 where Penalty Attributes (31.22%)
--5,017 of which are 3s, surprisingly 7s (@5,063) where the most common penalty attribute score

Both the chances of rolling a penalty and a bonus are much, much higher than my manual rolling probabilities. Granted, my sample is much smaller (10 characters rolled per test, 4 tests, 320 manual rolls in total), but the results are consistent between tests. It's about a 5% chance to get a bonus, and 12.5% to get a penalty.

You're getting bonuses and penalties at almost 3x the rate of my manual rolling. So either I have weird luck, or there's something off in how those numbers are being randomly generated.

[ShadowLogan]

There might be a tiny bit of how I coded the initial 3d6 roll (python allows you to set min/max values so I set 3 and 18 instead of rolling 3 die at 1 and 6), but I do think the bulk of it is the nature of which is more truly more random a computerized random number generator vs manual dice and the differences in sample sizes (a larger sample size is going to be more accurate).

[Killer Cyborg]

Is this normal for an RPG? Is it intentional? Is it how it should be?

No.
No.
Different.

[EliBenedict]

[quote="ShadowLogan"]There might be a tiny bit of how I coded the initial 3d6 roll (python allows you to set min/max values so I set 3 and 18 instead of rolling 3 die at 1 and 6), but I do think the bulk of it is the nature of which is more truly more random a computerized random number generator vs manual dice and the differences in sample sizes (a larger sample size is going to be more accurate).[/quote]

Yeah, that's not going to work. 3d6 gives you bell curve distribution with the middle values being much more common than the extremes. There's at 12.5% chance of rolling either a 10 or 11 (meaning one quarter of all rolls generate one of these two values.) Whereas there's only a 0.46% chance of rolling a 3 or an 18. Your system gives equal probability to all values, making both penalties and bonuses much more common.

That said, I both deeply respect and am utterly flumoxed by the effort of using the exactly defined probabilities (as described in the rules) to code a population creating model, and then analyzing that results of that model to determine the distribution of results for the initial probabilities.

You can just calculate the distributions directly.

[MyDumpStatIsMA]

There might be a tiny bit of how I coded the initial 3d6 roll (python allows you to set min/max values so I set 3 and 18 instead of rolling 3 die at 1 and 6), but I do think the bulk of it is the nature of which is more truly more random a computerized random number generator vs manual dice and the differences in sample sizes (a larger sample size is going to be more accurate).

As EliBenedict said before me, a simple 3-18 random generator isn't how dice probabilities work. It is absolutely imperative to separate them into 3 groups of 1-6, if you want to simulate attribute rolls.

Forgetting my own testing with manual rolls, just think about it this way:

There's only one combination of dice values that will create an attribute of 3: you must roll three 1s.

Whereas to get a 12, you can roll the following: 6+4+2. 6+5+1. 5+5+2. 5+4+3. 4+4+4.

The numbers with more dice combinations will probabilistically occur more often. That's why random number generation of 3-18 isn't equal.

If you have the time and inclination, you can plug in the different values and see how wildly divergent the 3d6 method is from the 3-18 method.

[ShadowLogan]

After adjusting my initial code from rolling 3d6 as one random number to 3 separate numbers (and totaling) I found another error in the code, the bonus die was only being added IF it also came with an exploding die. I have since fixed the code for both and re-ran it, the results are with a new dataset and not the same as the initial.

From 80,000 attribute rolls (8 attributes * 10,000sets) listing the total score (from die) and how many times they came out (again this is for this dataset, another run could result in different distribution):
3 came out x384
4 came out x1,123
5 came out x2,206
6 came out x3,739
7 came out x5,600
8 came out x7,856
9 came out x9,162
10 came out x10,044
11 came out x10,147
12 came out x9,068
13 came out x7,791
14 came out x5,520
15 came out x3,669
16 came out x0 (which makes sense as it is the lowest trigger score for a bonus die, though skill modifiers can raise it to this value)
17 came out x392
18 came out x561
19 came out x615
20 came out x581
21 came out x633
22 came out x226
23 came out x124
24 came out x84
25 came out x90
26 came out x128
27 came out x91
28 came out x113
29 came out x44
30 came out x9

For the 10,000 attribute sets there where a total of 23 different combinations of bonus/normal/attribute (and sets that are possible but missing, like straight one type). There where still instances of ZERO for the three different zones. Of note:
-2554x resulted in 0B, 7N, 1P (most common)
-1842x resulted in 0B, 6N, 2P (2nd)
-1514x resulted in 0B, 8N, 0P (3rd)
-2x resulted in 0B, 2N, 6P (the worst combination achieved though 8P is a possible combination not reached)
-1x was 4B, 4N, 0P (the best combination achieved though 8B is possible combination not reached)
-The largest Bonus Set was 4, the largest Normal was 8, and the largest Penalty Set was 6 (though hitting 8 for all should be possible)
-there was a total of 3,691 Bonus Attributes, 63,257 Normal Atributes, 13,052 Penalty Attributes

These results are not probabilities, they are the quantity out of 80,000 times X happened (sometimes presented as a percentage of the total) when I did the rolls. This is what the OP described in their initial post rolling 100sets of attributes. The probability of a given number appearing is going to vary by the actual number as some results can only be achieved with 1 combination of dice and others have more (this I know), then factor in combinations (of multiple rolls) and order (do you roll the attributes straight or do you pick and choose). The Probabilities of X in this case are fixed (barring something like loaded die), but you generate X-number rolls of 8-attribute sets it is possible to get resultant results that would suggest the probabilities are wrong if the set is too low (which was brought up).

[MyDumpStatIsMA]

-there was a total of 3,691 Bonus Attributes, 63,257 Normal Atributes, 13,052 Penalty Attributes

So, that means a 4.6% total of all attributes rolled came up as a bonus, and a 16.3% total of all rolls came up as penalties.

That's about in line with what I'd expect.

The tricky part would be to figure out how many were within range of being brought up to a bonus, or elevated above penalty range.

Realistically, only PS, PE, and SPD can be improved by 4+ points with skills. We should not factor in OCC bonuses, as that's too restrictive. Only normal (widely available) skill bonuses should apply in trying to figure out the final bonus/penalty amounts.

For example, if we take the rolls from 12-15 (i.e, those within bonus range after skills are applied), it comes out to around 33% of all rolls. However, this needs to be reduced, since only 3 of 8 attributes can reach 16+ with skills. Not that SPD gets a bonus there, but it does become more useful in general.

The same would apply for figuring out how many penalty stats can be easily brought up to neutral.

Using the crude and unscientific method of simply multiplying the total number of penalties by 3/8, we can roughly estimate that, after skill bonuses are applied, the number of penalties could be reduced to 10% of all rolls.

Likewise, if we take the total number of rolls that range from 12-15, take 3/8 of that, we find that 9768 of those could be within range of reaching 16 or 17 with skills. Which then, theoretically, brings us to the possibility that almost 17% of all rolls could end up as bonus attributes after skills are applied.

[ShadowLogan]

The same would apply for figuring out how many penalty stats can be easily brought up to neutral.

Ignoring Class AND if you could take (practically) every Physical Skill found in RUE (12x non-HTH in total, though 1x isn't needed):
PS Bonus Total: +13 (a few other skill exist to push this higher like Iceskating and Snow Skiing from WB20 or Juicer Football from WB10 which might be class restricted off-hand)
PP Bonus Total: +2 (WB8 HTH Martial Arts can push it a bit higher, as can some of the WB20 skills)
PE Bonus Total: +12 (MiO, WB8 HTH MAs, and WB20 have options that can push it higher)
SPD Bonus Total: +6-26 depending on the die roll (more from WB10, WB20, and WB8 skills)

There are 2x HTH Skills in WB8 at the grant an ME Bonus, though these skills could be region locked. And since you can't normally take 2x HTH skills, you'd be looking at a small bonus.

This means that for Physical skills, if you are willing to invest the skill slots you could counter a PS or PE or SPD (even on a bad set of rolls) score of 3 for each. SPD being the easiest to raise (w/3x skills), with PS and PE being linked and progressing much slower than SPD (each time gaining 1 or 2 points each). PP is the hardest to fix (only 2 skills).

Class bonuses for ME and MA are certainly possible (along with IQ, but IINM nowhere near as common as the other mental attributes), and PB can be adjusted with Cybernetics (maybe I would think). Megaversally there is the Leadership skill from the 2E RT line (Mac SB) that grants an MA bonus, though I don't know off hand if it was ported to another line/book.

It should be noted that some classes the penalty attributes at initial roll up are meaningless (PS, PP, PE, Spd) as they can be countered by enhancement methods that result in the initial roll being tossed out and replaced (Bionics) or Augmented with a bonus that has a minimum auto score (Juicer/Crazy implants) that if you don't reach you get automatically bumped to the minimum.

[MyDumpStatIsMA]

This means that for Physical skills, if you are willing to invest the skill slots you could counter a PS or PE or SPD (even on a bad set of rolls) score of 3 for each.

Well, how realistic is that really? Countering an initial attribute roll of 3, I mean. Just because X number of physical skills exist, doesn't mean that the vast majority of OCCs have a sufficient amount of OCC-related skills to choose them. Physical secondary skills aren't going to provide the same level of attribute boosting; only ones that boost PE or PS are Body Building, Running, and General Athletics.

Most non-combat OCCs don't allow you to take Boxing, Acrobatics, and Gymnastics. Typically Gymnastics is more commonly available than Acrobatics and Boxing. Wrestling and Kickboxing aren't always available to every OCC, either. Many OCCs won't allow you to take more than a couple of the 'premium' physical skills (which I consider Boxing, Gymnastics, Acrobatics, Kickboxing, and Wrestling to be).

The only physical attribute-boosting skills that every OCC can take as a related, are things like Outdoorsmanship, Physical Labor, Forced March, etc.

Beyond all that, there's also the 'opportunity cost' of being forced to take physical skills to compensate for low attributes. If you need to spend OCC-related skill selections on physical skills, and you're a Rogue Scientist or Operator, then you're sacrificing a lot of your OCC's natural advantages in non-physical skill bonuses.

Class bonuses for ME and MA are certainly possible

Most of those are tied to a minimum attribute threshold. For example, the Rogue Scholar gives you a +2 to MA, but has a minimum MA of 10.

In other words, you're not going to get an OCC boost towards a penalty attribute, because the penalty attribute will be too low to qualify to play that OCC. The only big exception to this (that I can think of, anyway) is the Ley Line Walker, which allows you to put a 1d4 towards any mental attribute, while only requiring a minimum for IQ.

[ShadowLogan]

Well, how realistic is that really? Countering an initial attribute roll of 3, I mean. Just because X number of physical skills exist, doesn't mean that the vast majority of OCCs have a sufficient amount of OCC-related skills to choose them. Physical secondary skills aren't going to provide the same level of attribute boosting; only ones that boost PE or PS are Body Building, Running, and General Athletics.

Oh I agree dumping 11 skills to counter score of 3 for PS and PE or SPD is unlikely. With the right selections though they could turn that 3(s) into an 8 or 9 pretty easily* using between 3 and 5 skills, there really isn't a need to go hog wild since of those 11 Physical Skills in RUE (that offer attribute bonuses):
-8 offer PS bonus and 4 of them are a +2PS instead of +1
-8 offer PE bonus (2 are +2, the rest are +1)
-3 offer SPD Bonus (heck just take Running and you are likely not looking at penalties anymore)
-overlap, all 11 come with a "buy 1 get x free" in terms of additional applicable combat stuff (more attribute bonuses, specified combat bonuses, or SDC)

So, if you goal is to only get out of Penalty territory (for PS/PE/SPD) and not into bonus territory (not going to happen with a score of 3, but for other penalty zone scores...) it becomes more manageable and not as "skill slot expensive". Skill availability might be an issue in Related/Other and it certainly is using Secondary Skills (single master list), but there is no need to buy all the skills possible to compensate.

Beyond all that, there's also the 'opportunity cost' of being forced to take physical skills to compensate for low attributes. If you need to spend OCC-related skill selections on physical skills, and you're a Rogue Scientist or Operator, then you're sacrificing a lot of your OCC's natural advantages in non-physical skill bonuses.

But no one is forcing anyone to compensate for those low attributes by spending skills to raise them up. And as we know not all Attributes have the option of "skill" improvement. So, in some cases you might be stuck with them. There are other ways to address "low" attributes beyond just take skills like:
-select an augmented OCC like 'Borg, Crazy, or Juicer (if you are trying to get rid of low attributes in the PS/PP/PE/SPD area) as I mentioned
-use tech (firearms instead of melee, MD melee might not be subject to the low PS penalty rules, get an exoskeleton suit, become a test subject for Robot TI or Robot VR, etc)
-become a mage and use spells to counter your weak attributes (assuming you can meet the requirements for)
-roll for psionics (or select a class or get implants to trigger)
-Genetic Engineering (you'll have to likely seek out someone to do the work like Atlantis, Lone-Star, Gene-Splicers, Gene-Tech, one of the Silver Republics in SA, maybe even Tritonia and if we're talking the Megaverse at large...)
-Symbiotes/Parasites, not just Bio-Wizard (IINM there are non-Bio-Wizard partners one can bond with)
-use a supply of Drugs or Magic Potions
-talk to the GM about having low power magic items (or even high-power ones like in HU that can turn frail into powerful) one could have acquired to compensate

And I'm sure there are probably a few more I'm not thinking of.

[MyDumpStatIsMA]

So, if you goal is to only get out of Penalty territory (for PS/PE/SPD) and not into bonus territory (not going to happen with a score of 3, but for other penalty zone scores...) it becomes more manageable and not as "skill slot expensive". Skill availability might be an issue in Related/Other and it certainly is using Secondary Skills (single master list), but there is no need to buy all the skills possible to compensate.

I prefer to give Gymnastics to as many characters as I can, as it has so many separate skills it encompasses. Albeit at rather low competency percentages, but it still gives you a chance to do quite a few things without investing into each skill (climbing, prowl, etc) separately.

The best physical skills to me, are Boxing, Gymnastics or Acrobatics, General Athletics, and Running. Everything else I can live without. And if you have an OCC that allows you to take all those, you're looking at not only substantial stat bonuses, but also bonuses to dodge/parry, SDC, and an extra attack.

*Edit: oh wait, I almost forgot. Juggling. As sad as it feels to pick that skill just for the initiative bonus, I sometimes do.

[MyDumpStatIsMA]

So, if you goal is to only get out of Penalty territory (for PS/PE/SPD) and not into bonus territory (not going to happen with a score of 3, but for other penalty zone scores...) it becomes more manageable and not as "skill slot expensive". Skill availability might be an issue in Related/Other and it certainly is using Secondary Skills (single master list), but there is no need to buy all the skills possible to compensate.

I prefer to give Gymnastics to as many characters as I can, as it has so many separate skills it encompasses. Albeit at rather low competency percentages, but it still gives you a chance to do quite a few things without investing into each skill (climbing, prowl, etc) separately.

The best physical skills to me, are Boxing, Gymnastics or Acrobatics, General Athletics, and Running. Everything else I can live without. And if you have an OCC that allows you to take all those, you're looking at not only substantial stat bonuses, but also bonuses to roll, dodge/parry, SDC, and an extra attack.

*Edit: oh wait, I almost forgot. Juggling. As sad as it feels to pick that skill just for the initiative bonus, I sometimes do.

[Hotrod]

I prefer to give Gymnastics to as many characters as I can, as it has so many separate skills it encompasses. Albeit at rather low competency percentages, but it still gives you a chance to do quite a few things without investing into each skill (climbing, prowl, etc) separately.

The best physical skills to me, are Boxing, Gymnastics or Acrobatics, General Athletics, and Running. Everything else I can live without. And if you have an OCC that allows you to take all those, you're looking at not only substantial stat bonuses, but also bonuses to roll, dodge/parry, SDC, and an extra attack.

*Edit: oh wait, I almost forgot. Juggling. As sad as it feels to pick that skill just for the initiative bonus, I sometimes do.

One thing to keep in mind is that the set of physical skills available as secondary skills is limited:
Hand to Hand : Basic
Aerobic Athletics
Athletics (General),
Body Building & Weightlifting
Climbing
Running
Swimming

Of these, Hand-to-Hand: Basic is the most important if your character has no hand-to-hand skill otherwise. After that, choices become either situational or subtle in their benefits.

Situationally, Running is great for nullifying the impact of a low Speed attribute roll. Swimming and Climbing matter if your GM puts you in situations where they matter.

Both athletic skills and body building amount to marginal stat bonuses. Of these three, Athletics (general) has the best set with its parry/dodge bonus, Aerobic Athletics is helpful for characters that kick a lot (its +2 to kicking damage might arguably apply to M.D.C. kicks), and Body Building is basically useless unless you're close to the threshold needed to effectively use a heavy weapon.

Looking past secondary skills and the ones you mentioned, hand-to-hand upgrades are again quite useful (especially if Hand-to-Hand: Assassin or Commando are upgrade options). Fencing is a nice skill for characters who use swords (the +1D6 bonus arguably applies to M.D.C. weapons). Kick boxing is good for characters with low supernatural strength giving them respectably damaging attack options.

Prowl is almost a requirement skill for many players; sneaking is a frequent element of many games.

While many dismiss other physical skills available as marginal stat boosts, don't write that off out of hand. Some physical stats can become far more important than you might think. Consider psi-stalkers, for instance. Anyone playing a psi-stalker should at least consider increasing P.E. with skills. You can add up to +9 to your P.E. by taking 7 skills (more if you use physical skills from Rifts: Canada). Psi-stalkers already start with 4D6 P.E.; it's not impossible that such a player could get up to a P.E. of 30 with a really good roll and the right skill selections. This does two things:
1. Improves the psi-stalker's bonus to save vs magic up to somewhere between +9 and +12, giving the character a huge edge over magic users, their preferred enemy and food source. With a few other bonuses to save vs magic thrown, such a character could become effectively immune to all but the strongest magics.
2. Increases the psi-stalker's hit points and M.D.C. in combat with the supernatural and near ley lines, giving them a greater likelihood of surviving even if their armor gets destroyed.

Similarly, human-scale P.S. usually doesn't mean much, but characters who convert their strength into supernatural strength (magic power tattoo, for instance) can scale up their mega-damage punches with physical skills, making them quite valuable to such characters. In melee-combat S.D.C. settings like Palladium Fantasy, P.S. is my preferred physical stat to boost up as much as possible.

tl;dr, I wouldn't write off most physical skills.

[MyDumpStatIsMA]

tl;dr, I wouldn't write off most physical skills.

I've noticed a shift in how I make my characters over the years. When I first started, I gave all the non-combatants Basic HtoH skill, and focused on non-physical skills. My logic was that if you can't be a great fighter, you shouldn't even try at all.

I've since transitioned into trying to make most of my characters at least semi-competent fighters. Not so much to attack, but for defensive reasons. Dodge, roll with impact, and run speed, are all important defensively even if you have low initiative and aren't a primary damage dealer in the party.

That being said, the huge drawback is that choosing HtoH Martial Arts for a non-combat OCC is going to eat up a lot of your skill selection. Consequently, this means your physical skill selection is going to be restricted. Most OCCs I use only have 3-5 choices left after Martial Arts takes up 2-3 skills. That's why I needed to pare down physical skills to 'must have' and 'can live without.'

[Hotrod]

tl;dr, I wouldn't write off most physical skills.

I've noticed a shift in how I make my characters over the years. When I first started, I gave all the non-combatants Basic HtoH skill, and focused on non-physical skills. My logic was that if you can't be a great fighter, you shouldn't even try at all.

I've since transitioned into trying to make most of my characters at least semi-competent fighters. Not so much to attack, but for defensive reasons. Dodge, roll with impact, and run speed, are all important defensively even if you have low initiative and aren't a primary damage dealer in the party.

That being said, the huge drawback is that choosing HtoH Martial Arts for a non-combat OCC is going to eat up a lot of your skill selection. Consequently, this means your physical skill selection is going to be restricted. Most OCCs I use only have 3-5 choices left after Martial Arts takes up 2-3 skills. That's why I needed to pare down physical skills to 'must have' and 'can live without.'

Ultimately, each character is different, and it all depends on what you're going for. There are some skills that are profoundly powerful in ways that most players don't seem to realize:
Read Sensory Equipment + Portable Scan Dihilator turns you into a freakishly effective scout/detector of just about everything of interest.
Breed dogs: dogs are some of the most effective detectors of supernatural evil in the game.
Horsemanship: Horses are as good at detecting supernatural evil, and using this with exotic animals gives you a fighting companion
Pilot: Jet Pack provides about the fastest and safest way to travel for the least cost.
Chemistry: Pharmaceutical provides probably the safest and most reliable way to subdue and get people to talk.
Field Armorer: Repairing 20 M.D.C. of damage for the coast of a single skill selection is pretty efficient, especially since this automatically gives you Basic Mechanics. It can also allow you to make some basic body armor.
Leather Working could allow you to make leather body armor. If you're using M.D.C. hide, that armor can also be M.D.C.
W.P. Shotgun and Archery have loads of options not just for doing damage but for inflicting specific useful effects on a target. I rarely see them used.
W.P. Targeting can also apply to grenades, which can also produce a variety of useful effects.
W.P. Paired Weapons is phenomenally useful in close combat. Against a single opponent with a single weapon, use simultaneous attack/parries to get in free attacks. Or, if the big beastie is distracted, you can duo-strike it for more damage.

Those versatile W.P.'s are pretty awesome, actually; a timely smoke grenade or flash bang can often swing a fight your way more than a boom gun by degrading or disabling an opponent. If I'm playing in a party of heavy hitters and I have a choice between doing a lot of damage or disabling the enemy/enabling my allies, I'll choose the latter every time. Knockdown attacks, net guns, and blinding/stunning/paralyzing/disabling effects provide huge tactical advantages that make fights more interesting and fun for me. I rarely need more than one or two skills to be an effective enabler, but optimizing for combat can frequently take up most to all of the skill choices I have.

What I generally like to do is try to pick a few things I want a character to be able to do and build that character to have as many options as I can. If I want a persuasive/communicative character, I might take Seduction, Wardrobe & Grooming, Interrogation, Impersonation, Disguise, Imitate Voices and Sounds, Electronic Countermeasures, Undercover Ops, and assign my highest two attributes to M.A. and P.B. (if the GM allowed me to do so). I'd also go for pharmaceutical chemistry as a last resort. With that set of skills, I'd have all kinds of ways of gaining and planting information. I might take Targeting to add in some choice effects of grenades, or Shotgun/Archery W.P.'s for the same reason. This kind of character could be incredibly useful outside of combat without being useless in a fight.

Having a few tactical skills is always a good idea in Rifts, but it doesn't take much to contribute: just a hand-to-hand skill and a single W.P. is all you need. I'll take a character using M.D.C. weapon with a respectable punch who keeps a low profile during combat and finds other ways to be useful to a group any time.

[MyDumpStatIsMA]

Ultimately, each character is different, and it all depends on what you're going for. There are some skills that are profoundly powerful in ways that most players don't seem to realize:
Read Sensory Equipment + Portable Scan Dihilator turns you into a freakishly effective scout/detector of just about everything of interest.
Breed dogs: dogs are some of the most effective detectors of supernatural evil in the game.
Horsemanship: Horses are as good at detecting supernatural evil, and using this with exotic animals gives you a fighting companion
Pilot: Jet Pack provides about the fastest and safest way to travel for the least cost.
Chemistry: Pharmaceutical provides probably the safest and most reliable way to subdue and get people to talk.
Field Armorer: Repairing 20 M.D.C. of damage for the coast of a single skill selection is pretty efficient, especially since this automatically gives you Basic Mechanics. It can also allow you to make some basic body armor.
Leather Working could allow you to make leather body armor. If you're using M.D.C. hide, that armor can also be M.D.C.
W.P. Shotgun and Archery have loads of options not just for doing damage but for inflicting specific useful effects on a target. I rarely see them used.
W.P. Targeting can also apply to grenades, which can also produce a variety of useful effects.
W.P. Paired Weapons is phenomenally useful in close combat. Against a single opponent with a single weapon, use simultaneous attack/parries to get in free attacks. Or, if the big beastie is distracted, you can duo-strike it for more damage.

I always take Field Armorer if I can, but unfortunately it's usually tied to pure combat OCCs. Paired Weapon proficiency is one reason I try to take Assassin HtoH whenever possible. Archery is something I ignored until recently, but now I have probably a half dozen characters drawn up that use it. I take Targeting mostly when I want to pair it with the Throw Stones spell; it gives any mage an emergency cost-effective MD attack in any situation.

Those versatile W.P.'s are pretty awesome, actually; a timely smoke grenade or flash bang can often swing a fight your way more than a boom gun by degrading or disabling an opponent. If I'm playing in a party of heavy hitters and I have a choice between doing a lot of damage or disabling the enemy/enabling my allies, I'll choose the latter every time. Knockdown attacks, net guns, and blinding/stunning/paralyzing/disabling effects provide huge tactical advantages that make fights more interesting and fun for me. I rarely need more than one or two skills to be an effective enabler, but optimizing for combat can frequently take up most to all of the skill choices I have.

Disabling over destruction is what I prefer magic for. Not that tech can't do it as well, but the nice thing about magic is not having to worry about carrying anything with you in sufficient quantities.

[Killer Cyborg]

Leather Working could allow you to make leather body armor. If you're using M.D.C. hide, that armor can also be M.D.C.

RUE 324
Skilled at tanning, preserving, and working with animal hides, fur, and leather to make leather goods such as clothing, capes, caps bags, purses, boots, shoes, belts, and even SDC leather armor (the character can repair leather armor too).

Using the skill to make MDC leather armor or goods would be a house rule, although it would be a perfectly reasonable one that I fully encourage.

[ShadowLogan]

Leather Working could allow you to make leather body armor. If you're using M.D.C. hide, that armor can also be M.D.C.

RUE 324
Skilled at tanning, preserving, and working with animal hides, fur, and leather to make leather goods such as clothing, capes, caps bags, purses, boots, shoes, belts, and even SDC leather armor (the character can repair leather armor too).

Using the skill to make MDC leather armor or goods would be a house rule, although it would be a perfectly reasonable one that I fully encourage.

Then what skill is used in the creation of the various MDC leather/hide armors/items scattered across various Rifts titles like the various "Dinosaurs" mentioned in WB14 (in their stat blocks), or WB21 (pg34 details a shop that sells the stuff), a other examples can be found (WB21, WB26, SB3 also in stat blocks) and then you have the armor the Pogtailians made from Dragons and the Fury Beetle Armor. Nor would I be surprised if there where more that I am not aware of.

I don't dispute the text of the Leather Working Skill, I just point out that it seems to be the closest skill one would use to make such items that we are told canonically exist (I'm also not sure when the skill was introduced to Rifts setting prior to RUE).

[Killer Cyborg]

Leather Working could allow you to make leather body armor. If you're using M.D.C. hide, that armor can also be M.D.C.

RUE 324
Skilled at tanning, preserving, and working with animal hides, fur, and leather to make leather goods such as clothing, capes, caps bags, purses, boots, shoes, belts, and even SDC leather armor (the character can repair leather armor too).

Using the skill to make MDC leather armor or goods would be a house rule, although it would be a perfectly reasonable one that I fully encourage.

Then what skill is used in the creation of the various MDC leather/hide armors/items scattered across various Rifts titles like the various "Dinosaurs" mentioned in WB14 (in their stat blocks), or WB21 (pg34 details a shop that sells the stuff), a other examples can be found (WB21, WB26, SB3 also in stat blocks) and then you have the armor the Pogtailians made from Dragons and the Fury Beetle Armor. Nor would I be surprised if there where more that I am not aware of.

I don't dispute the text of the Leather Working Skill, I just point out that it seems to be the closest skill one would use to make such items that we are told canonically exist (I'm also not sure when the skill was introduced to Rifts setting prior to RUE).

You're falling into the trap of believing that there are specific skills for creating everything.
Maybe it takes specific machines?
Or spells?
Or psychic powers?

Whatever it is, we're never told, and there is no canon answer.

Which is why I always just go with Leather Working as applying to MDC gear too.

[eliakon]

Leather Working could allow you to make leather body armor. If you're using M.D.C. hide, that armor can also be M.D.C.

RUE 324
Skilled at tanning, preserving, and working with animal hides, fur, and leather to make leather goods such as clothing, capes, caps bags, purses, boots, shoes, belts, and even SDC leather armor (the character can repair leather armor too).

Using the skill to make MDC leather armor or goods would be a house rule, although it would be a perfectly reasonable one that I fully encourage.

Then what skill is used in the creation of the various MDC leather/hide armors/items scattered across various Rifts titles like the various "Dinosaurs" mentioned in WB14 (in their stat blocks), or WB21 (pg34 details a shop that sells the stuff), a other examples can be found (WB21, WB26, SB3 also in stat blocks) and then you have the armor the Pogtailians made from Dragons and the Fury Beetle Armor. Nor would I be surprised if there where more that I am not aware of.

I don't dispute the text of the Leather Working Skill, I just point out that it seems to be the closest skill one would use to make such items that we are told canonically exist (I'm also not sure when the skill was introduced to Rifts setting prior to RUE).

As per Rifts Australia page 206 its... Field Armorer.

[Hotrod]

Leather Working could allow you to make leather body armor. If you're using M.D.C. hide, that armor can also be M.D.C.

RUE 324
Skilled at tanning, preserving, and working with animal hides, fur, and leather to make leather goods such as clothing, capes, caps bags, purses, boots, shoes, belts, and even SDC leather armor (the character can repair leather armor too).

Using the skill to make MDC leather armor or goods would be a house rule, although it would be a perfectly reasonable one that I fully encourage.

Then what skill is used in the creation of the various MDC leather/hide armors/items scattered across various Rifts titles like the various "Dinosaurs" mentioned in WB14 (in their stat blocks), or WB21 (pg34 details a shop that sells the stuff), a other examples can be found (WB21, WB26, SB3 also in stat blocks) and then you have the armor the Pogtailians made from Dragons and the Fury Beetle Armor. Nor would I be surprised if there where more that I am not aware of.

I don't dispute the text of the Leather Working Skill, I just point out that it seems to be the closest skill one would use to make such items that we are told canonically exist (I'm also not sure when the skill was introduced to Rifts setting prior to RUE).

As per Rifts Australia page 206 its... Field Armorer.

Rifts Australia homemade armor is ramshackle junk armor a la Mad Max, as I recall. I don't remember seeing anything like a M.D.C. leather armor. The only instance of homemeade M.D.C. leather armor I'm familiar with is the armor worn by Dragon Juicers, but no crafting system goes with it.

Player crafting systems aren't well established in Rifts, and this is one area where I think some development would be nice, both from a world-building and a player development perspective. I've been working up a Rifter submission with a system for player characters making and repairing their own armors, where the skills taken and their proficiency levels impact the M.D.C. and movement penalties. I've also created a class that focuses heavily on crafting unique body armors to go with it. I'll be curious to see if Kevin and Sean have any interest.

[eliakon]

Leather Working could allow you to make leather body armor. If you're using M.D.C. hide, that armor can also be M.D.C.

RUE 324
Skilled at tanning, preserving, and working with animal hides, fur, and leather to make leather goods such as clothing, capes, caps bags, purses, boots, shoes, belts, and even SDC leather armor (the character can repair leather armor too).

Using the skill to make MDC leather armor or goods would be a house rule, although it would be a perfectly reasonable one that I fully encourage.

Then what skill is used in the creation of the various MDC leather/hide armors/items scattered across various Rifts titles like the various "Dinosaurs" mentioned in WB14 (in their stat blocks), or WB21 (pg34 details a shop that sells the stuff), a other examples can be found (WB21, WB26, SB3 also in stat blocks) and then you have the armor the Pogtailians made from Dragons and the Fury Beetle Armor. Nor would I be surprised if there where more that I am not aware of.

I don't dispute the text of the Leather Working Skill, I just point out that it seems to be the closest skill one would use to make such items that we are told canonically exist (I'm also not sure when the skill was introduced to Rifts setting prior to RUE).

As per Rifts Australia page 206 its... Field Armorer.

Rifts Australia homemade armor is ramshackle junk armor a la Mad Max, as I recall. I don't remember seeing anything like a M.D.C. leather armor. The only instance of homemeade M.D.C. leather armor I'm familiar with is the armor worn by Dragon Juicers, but no crafting system goes with it.

Player crafting systems aren't well established in Rifts, and this is one area where I think some development would be nice, both from a world-building and a player development perspective. I've been working up a Rifter submission with a system for player characters making and repairing their own armors, where the skills taken and their proficiency levels impact the M.D.C. and movement penalties. I've also created a class that focuses heavily on crafting unique body armors to go with it. I'll be curious to see if Kevin and Sean have any interest.

It can be ramshakle junk yes.
but it states that one of the components you can use to make the suit is MDC leather.
You don't need anything other than the leather as well.
Thus we are left with
1) the book says that making a suit out of MDC leather is possible
2) the only rule we have on the subject in any of the 100+ volumes is... Field Armorer.
I would myself rule that you need both Field Armorer and Leatherworking but thats just me.