What you call a horrible idea, I call a fundamental aspect of game design: The Short vs. Long term trade-off, that the player has a choice between something good now, or to hold off and get something better later. The golden standard of design is to make choices rewarding for the player, not restrict player choices. The current Multi Pool system fails on both metrics. It does not make it rewarding to the player to sacrifice their only 2 stat points, and massively reduces the variety of play.
To try and go into the underlying issues, I’ll break down what you’ve said:
Instant, non-grindy, start:
I actually like the early game. It’s the only time I feel particularly challenged, since it’s the time I have the fewest resources and the hardest time acquiring more as I scour every house for skill books and try and balance eating and drinking enough to survive with staying out of trouble long enough to learn. In the Cata design document I believe this is referred to as the “survival” phase of the game, when the player must struggle to earn their place in the post-cataclysmic environment.
While I don’t disagree with allowing a player to bypass some of this if they choose to, I do not consider enforcing bypassing it through game mechanics to be good design. While I agree there’s a problem in balance, I consider it to be one of cost/reward, not a core system issue.
Brainstorming: To alleviate early game grindiness.
[spoiler]The ideal would be to have any progress the player makes to be organic, without painful, artificial walls to progress.
One possibility would be to open up some of the earlier recipes to all characters, and to remove artificial barriers to player progression that make the system feel “gamey”. Mechanics is a good example:
A starting player can coke up and remove advanced cybernetics from their own rib cage at 0 First Aid, even if they will probably fail.
A player who has read an introductory Mechanics text book up to level 1 cannot change a tyre, or even attempt to.
A player with tailoring less than around 4 is probably going to have a bad time trying to do so much as darn a doily.
By allowing the player to attempt at lower levels you increase risk (aka fun), while increasing training opportunities. I can barely make toast, but I can try to craft lasagna, maybe even succeed without producing a burnt mess, and probably get a lot more experience cooking by overextending myself than making a dozen pieces of toast.
Likewise, in reality when we wish to achieve a skill, we practise. In game, the only way to do this is to expend resources (fine) and time (fine for the character, not so for the player) trying to achieve something specific, yet simple, over and over again.
Why not abstract this? Simply have a command or recipe “Practise: Sewing/Electronics/Fabrication.”, that takes up resources and time to simply improve the skill in question.
The player creates a garish abomination with nine sleeves, a gaping hole in the front where the stitching came undone, and the zip is lopsided, and they cut it back up into rags when they’re done, losing materials in the process.
The player has more motivation to scavenge (danger = fun) for materials to continue practising.
Books are still useful, because they let you learn faster, and without the resources (and progress to higher skill levels, and recipes you can try even at a lower level with a high risk of failure…).
Most of the grindiest parts are abstracted into simply sourcing the appropriate materials and trying your luck failing to create things with a small chance of success (which gets double mileage of risk = fun and unlikely success against all odds = fun).
Skills at start gimping the character long term is a bad choice.
I could go on ad nauseam about short vs. long decision making design choices as I mentioned at the start of my post, but I’ll just point out that pretty much every single pick in character creation is short vs. long.
Martial Arts traits? Short term gain, the books are everywhere.
Bionics starting character? Short term gain, since you can find the parts elsewhere, or one of the nastiest starting disadvantages with Broken Cyborg.
Skills? Any skill is temporary unless it’s level 7+ dodge or something (which is more or less a permanent, unreplicable bonus, just like stats, considering how hard it is to reach higher levels of dodge).
Night Vision? Quick? Fleet-footed? There’s mutagen for that.
You are theoretically gimping your character by choosing any of these things, and there are precious few things other than stats, MD and Parkour Expert that aren’t just there to make your early game less random, and that’s 100% fine, there’s no problem with traits because traits are better balanced and can make a serious difference to early game consistency, and that advantage (and inherent advantage of being more reliable long term) is worth the trade-off in long term power.
Brainstorming: Narrowing the gap between short and long term skills vs stats balance.
1: To reduce balance problems between stats vs skills.
[spoiler]There are two ways to approach this. Either stats are worth more than skills, or skills are worth less than stats. In a perfect system you would be prepared to sacrifice long term benefits for short term benefits, because that’s already largely the case for Traits.
My first suggestion was a 1:3 exchange rate for stats into skills (this matches pretty well to what professions give on top of gear, bionics et al). You suggest that this would introduce a flaw to allow more points for skills, but I believe you are mistaken here. There is no way to reduce your skill points below 0, so the player could only buy extra skills with stats, not the other way around.
Also note that with a 3:1 point variety, the “strictly better” issue diminishes. With a 1:1 system taking 3 points of Dodge equates to maybe 25% of your total stat (assuming 12). With a 3:1 system, it’s 8.3%. Still, short term vs. long term strategy, but better balanced.
The other problem is that stats are simply too cheap to specialise. >3 points in a skill raises the cost to 2 points per pick, while you can raise stat points 6 points before the price increases. You would need 24 points poured into stats before you started losing efficiency. Combined with your starting abilities being more or less unchangeable that makes it significantly better and cheaper to focus on stats, which isn’t ideal.
To alleviate this, we can apply the same restriction: 8-11, 1 point. 12-13, 2 points, 14-15, 3 points, 16-17, 4 points, 18-19 5 points, 20: 6 points.
Together, this means that if a player chooses to take skill points instead of increasing one stat to 12, that’s a 6:1 advantage. That early game advantage becomes as cheap as chips, while the end character is pretty close in power.
For a lower power level early game, instead of making skills cheaper we can simply add 1 point to those modified stat costs and keep skills the same, resulting in a lower power start all around. Selling points below 8: 1 point. 8-11, 2 points. 12-13, 3 points, 14-15, 4 points, 16-17, 5 points, 18-19 6 points, 20: 7 points.
Once again, you have long term investment, but with this system it’s a significant long term investment - getting one of your stats up to 11 costs 6 points that could have got you some decent skills, multiple traits, and a much better chance, overall, of surviving to your first birthday.[/spoiler]
2: To reduce balance problems between professions and anything else.
[spoiler]Many professions are arguably already a pretty good deal, giving rare bionics, or irreplaceable traits (skaters and MDs, for example). Making the non-skill benefits work out is a question of tweaking prices so that it represents both the price of the bonuses and the lost opportunity cost (you could have picked another, different, awesome profession or one that granted bonus points). A profession that grants Karate should be overall cheaper than a trait that grants karate, because you can have twelve traits but only one profession.
The primary issue with skill granting professions is that unless you’re specialising anyway low level skills are not valuable, regardless of pool system - +2 dodge from ballet dancer is either +2 skill points gratis, or +10 and reaching unattainable levels of epicness, depending how many points you mix-maxed into dodge. Feast or famine, your profession is either irrelevant the moment after you start the game, or has granted you something that nobody else can easily get.
One possible way to ensure that professions are always desirable is to have their starting skills go hand in hand with permanent learning speed bonuses in their chosen fields - For example, a Juvenile Delinquent would take as much effort to go from Melee 5->6 as a Survivor would from 6->7.
Meanwhile, to help maintain balance, defining what actions gain skills would be unchanged. A ballet dancer and their +2 starting dodge would not be able to improve against a zombie child, since their dodge would be 2 already, but would never rust (since it rusts as though they had 0 skill) and improve rapidly against a zombie up to level 3 as though they had 0 dodge.
This would mean that your profession not only defined your early game through its gear and skills, it also defined your adventuring career by showing where your character’s personal strengths are as an individual.[/spoiler]
3: Resolving the unresolvable divide between stats and skills:
[spoiler]There are two fundamental differences between stats and everything else:
They cannot be amended, your base stats are all but permanent.
They have affects that cannot readily be replicated.
The most simple would be to allow mundane and extraordinary stat growth. Extraordinary is easy,
a single charge artifact type deally granting a chance to permanently boost a stat by 1, autosaving as it does so, with chances worsening as stats get higher. Effectively Granades, with more of the gradually shaving off pieces of your psyche to dark higher powers in exchange for power. If stat increases become mathematically feasible, if difficult and random the higher you go, then high starting stats become just another short term advantage.
Mundane on the other hand, could come in various ways, the easiest being the inclusion of skills. Stats give flat rate bonuses to the best of my knowledge, so while they could still benefit from these self-same skills (not penalising them for their playstyle), any flat bonus from a non-stat source diminishes the value of the stat itself.
Weight lifting: Adds to the amount the player can carry, trained by being near or over your maximum weight limit.
Toughness: Increases the amount of health the player has, and how readily they overcome diseases. Trained by: Being sick, getting injured, and healing back up again.
Crafting: There’s melee and marksmanship, why not a crafting overskill that gives you overall bonuses of success or speed during manufacturing jobs?
Literacy: Your speed at reading books increases the more you get used to reading.
Trapping: Already increases your chance of finding traps, and could/should give experience by doing so if it doesn’t already, making it less tedious to level if you hang around minefields practising.
Precision: Increases your chance of landing critical hits. Improved by: Landing critical hits and hitting things normally (though slowly).
To retain overall power level, the current bonuses from stats could even be diminished slightly to account for the bonuses from the new skills, leaving overall balance unchanged.[/spoiler]
These or similar approaches would hopefully serve to tackle the root issues that multiple pools aims to address, while retaining the maximum degree of freedom to the player on where to focus their limited resources.