Page 1 of 2

Proposal: Make Str better

PostPosted: Thursday, 20th June 2013, 23:47
by Siegurt
The proposal in a nutshell is this: Make str impact weapon speed.

What I have in mind roughly is that either str, or your str weighted stat impact how quickly your trained weapon skill gets you to minimum delay, so that at say, 20 str, it's exactly as things are now, at 10 str it takes twice as many weapon skill points to get you to minimum delay and at 40 str it takes half as many weapon skill points to get you to min delay.

This would mean that large, heavy weapons would *not* always be the optimal choice for low-str characters, and increasing your Str would have a measurable impact on the game, without changing the ultimate disposition of the damage curve significantly.

I've always felt like a Str 10 elf swinging around a huge two handed triple sword just as quickly and easily as a monstrous Str 40 ogre was kind of silly anyway.

Using "Weighted Str" would mean that Dex would be a limiting factor as well, so that races with a strong disposition towards Str over dex (e.x. ogres) wouldn't have an over the top runaway bonus towards getting say, GSC online and at min delay super early.

Note that I'm not suggesting that additonal Str have any *other* impact on weapon damage (having 40 str should not effectively double your weapon skill for damage-doing purposes)

Alternately the penalty side could be implemented without any of the bonus side (So the *best* you could do is the current training regime) which would effectively mean there was a effective minimum str to use any given weapon optimally (This could perhaps be implemented as "You can only reduce a weapon's delay by a half your Strength") Which would have some of the same effect without giving any training discounts to people who push str further than the optimum level.

The second possibility is probably easier to balance, but I don't like it as well, it doesn't reward pushing Str to higher levels, as Dex and Int are rewarded currently in the realms that they give you bonuses.

It would also mean (and I think this is a good thing) that not all characters can get any weapon in the game to minimum delay with proper training, forcing people to branch out into lighter-but-less-damaging weapons as the best solution.


***Edit pulling some of the latest formula and code suggestions up from the thread below, just to make it easier for anyone who hasn't been following along and doesn't want to read the whole thing:

Current formula for weapon delay is base delay - (Skill+10/20) (Minimum half or .7) proposal is that Str influence the denominator of that formula, in the general form of:
  Code:
Skill*10 /
(20*(percentage unimpacted by Str)) + ((20*(percentage impacted by str)*(normal str attribute)) / (current weighted str attribute))

Where currently "percentage" is 50%, "normal" is 20, and "current weighted str attribute" is the value returned by calc_stat_to_dam_base (Strength weighted towards dex, using weapon weighting)

Using our current suggested values, that reduces down to:
  Code:
Skill*10/
(10+(200/calc_stat_to_dam_base()))

As the reduction in weapon deay (with the normal weapon minimum delay)

(Effectively increasing/decreasing the effect training has on weapon speed, but not increasing the current min/max delay, other than potentially making min delay be out of reach for certain weapons without increasing strength)

To effect this change, one line of code would need to be made in player_weapon_speed() in melee_attack.cc:
From
  Code:
attack_delay -= div_rand_round(constant(you.skill(wpn_skill, 10)), 20);

to
  Code:
attack_delay -= div_rand_round(constant(you.skill(wpn_skill, 10)), 10+(200/calc_stat_to_dam_base()));

Re: Proposal: Make Str better

PostPosted: Thursday, 20th June 2013, 23:54
by Siegurt
Yes, although with Dex as the primary and Str as the secondary characteristic.

I was proposing something less complex than the ranged code (because that stuff's hairy as all get-out) but with similar effects.

Re: Proposal: Make Str better

PostPosted: Friday, 21st June 2013, 07:25
by DracheReborn
I like this idea. Right now it seems a bit of a no-brainer to simply go for the biggest, baddest weapon on a melee fighter.

It would be nice if size is factored in too. It would make sense for example, that it would take less strength for an ogre to wield a great mace than it is for, say, a deep elf. Properly formulated, it might be possible to get rid of the artificial restriction on giant clubs/GSCs - an ogre/troll of average strength can wield a GSC, but then you'd think an exceptionally strong minotaur can do so as well. On the flipside, a particularly weak ogre should have trouble wielding a big club.

Re: Proposal: Make Str better

PostPosted: Friday, 21st June 2013, 12:42
by Tiber
Related: https://crawl.develz.org/tavern/viewtopic.php?f=8&t=4502

In my mind, I'd like it to be a formula similar to spell power and int, where fighting and str/dex come into play. There's also already a str/dex rating, so rather than just using str, you could also use: (str * strWeight + dex * (1 - strWeight)). The race's size factor might come into play, but that seems somewhat redundant since base strength is already correlated to size.

Re: Proposal: Make Str better (Please don't)

PostPosted: Friday, 21st June 2013, 14:30
by rebthor
Please don't make Str better. The encumbrance change is enough.

Re: Proposal: Make Str better (Please don't)

PostPosted: Friday, 21st June 2013, 19:19
by ElectricAlbatross
rebthor wrote:Please don't make Str better.

The OP proposes to make low strength worse as well as high strength better. Seems like the proposal is actually "Make Str matter". Do you disagree, or have an actual argument for why we shouldn't do that?

Re: Proposal: Make Str better (Please don't)

PostPosted: Friday, 21st June 2013, 19:26
by battaile
Based on having read the post you quoted from, I will take a stab at these!

ElectricAlbatross wrote:The OP proposes to make low strength worse as well as high strength better. Seems like the proposal is actually "Make Str matter".

It would appear that he understands this.

Do you disagree,

I believe he does.

or have an actual argument for why we shouldn't do that?

I would reference the line of his post which you omitted from your quote, "The encumbrance change is enough."

Re: Proposal: Make Str better

PostPosted: Friday, 21st June 2013, 19:57
by Siegurt
Yes, really my topic probably should be "Make Str more important, particularly for melee builds" not "Make Str better"

And I disagree that the encumberance change is enough, but can certainly see that there's room for an opposing opinion :)

Re: Proposal: Make Str better (Please don't)

PostPosted: Friday, 21st June 2013, 20:38
by ElectricAlbatross
battaile wrote:I would reference the line of his post which you omitted from your quote, "The encumbrance change is enough."

You are right. I must have missed that the first time since it is neither substantiated in the least nor at all relevant to the OP's concerns, but I must now agree that it is at least technically an argument. Thank you for the correction.

Re: Proposal: Make Str better (Please don't)

PostPosted: Friday, 21st June 2013, 21:11
by battaile
ElectricAlbatross wrote:You are right. I must have missed that the first time since it is neither substantiated in the least nor at all relevant to the OP's concerns, but I must now agree that it is at least technically an argument. Thank you for the correction.

I have no idea what you want him to substantiate or why you think an argument that a thing doesn't need further buffing is not relevant to a request for a buff to said thing, but you're welcome.

Re: Proposal: Make Str better

PostPosted: Friday, 21st June 2013, 22:20
by ElectricAlbatross
I don't know, if Siegurt agreed that the encumbrance change was enough, then he wouldn't be posting this proposal, right? Since Siegurt obviously disagrees that whatever recent changes happened to Str are enough to make it a worthwhile stat to raise, this thread could benefit from somebody backing up that statement beyond simply stating it.

Also, I have noticed no change in the way encumbrance works and can't find anything relating to that in the commit logs. What are these changes, exactly, and why do they improve Str to the point that further buffs are unnecessary?

Re: Proposal: Make Str better

PostPosted: Friday, 21st June 2013, 22:27
by Siegurt
Specifically Encumberance in .13 replaces the old EV penalty for armour. Str acts as a divisor for the penalty, so the more Str you have the lower the penalty is, it's a smooth curve penalty which replaces the old stairstep-penalty-if-you-don't have a minimum number.

This means effectively that more Str always reduces your armour penalty by some margin, and that the amount it reduces your penalty is decreasing-to-infinity.

the numbers themselves are balanced out so that if you have about the same str you used to need to eliminate the penalty, you'll still have some, but it will be pretty small.

Re: Proposal: Make Str better

PostPosted: Friday, 21st June 2013, 22:30
by Siegurt
Well, I should say the numbers are balanced out that way at the "Scale mail and less" armours, the heavier armours do need increasing amounts more Str to get the penalties to the same levels (IIRC GDA now needs roughly 35 str to get the same penalty that you get from 9 Str and scale mail, rather than 27)

I could go look to get you exact numbers, but I don't feel like it right now.

Re: Proposal: Make Str better

PostPosted: Friday, 21st June 2013, 22:45
by Siegurt
Incidentally, to make this change as I originally envisioned it requires changing this line:
  Code:
attack_delay -= div_rand_round(constant(you.skill(wpn_skill, 10)), 20);

in player_weapon_speed() in melee.cc

to this:
  Code:
attack_delay -= div_rand_round(constant(you.skill(wpn_skill, 10)), 400/you.strength());

to change it to "weapon weighted strength" you'd change that line to this:
  Code:
attack_delay -= div_rand_round(constant(you.skill(wpn_skill, 10)), 400/calc_stat_to_dam_base());


Min speed is uneffected, this only effects how many points you need to put in to get a weapon there.

Re: Proposal: Make Str better

PostPosted: Friday, 21st June 2013, 23:17
by Siegurt
Compiling and playing around with those values in wiz mode with the fsim results in me deciding that's too large of an impact, something closer to:

  Code:
attack_delay -= div_rand_round(constant(you.skill(wpn_skill, 10)), 10+(200/calc_stat_to_dam_base()));


Seems a little closer to the mark.

(That means at 30 str it takes 21.6 skill to get min delay on a weapon with 20 base delay (Instead of 26) and with only 10 str, if you max your skill at 27, you'll get your delay on a 20 base delay weapon to about 11)

With 10 str you'll be able to get a weapon with a base delay of about 180% to min delay given max skill, and with only 5 str you'd only be able to get weapons with a base delay of 12 (or 11 for short blades)

However using calc_stat_to_dam_base, you're dex figures in there too, so with most light weapons, your dex will bring that number up considerably (A short blade with str 5 and dex 14 uses 9 as it's "calc_stat_to_dam_base", which lets you get min delay on a sabre at skill 19)

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 00:14
by rchandra
ElectricAlbatross wrote:Also, I have noticed no change in the way encumbrance works and can't find anything relating to that in the commit logs. What are these changes, exactly, and why do they improve Str to the point that further buffs are unnecessary?


I often find that even in fire/pearl dragon armour gauntlets of str are preferable to gauntlets of dex, for the effect on both EV and spell success. This is a bit of a buff to high Str characters, I'm not at all sure when various characters should choose str vs dex now.

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 01:41
by Siegurt
It doesn't seem like it requires lines of spoilers to understand.

"Strength and to a lesser extent dexterity makes you attack more quickly" sums it up nicely.

To find how how quickly you attack, hit something and look at the value in the parentheses next to the turn counter.


To roll it into an easily wiki-able formula for the obsessively detail oriented who want spoilers:

"Your attack stat is your strength, modified towards the average between your strength and dexterity by a weapon dependent amount between 10% and 90% of the way towards the average"
"Your attack delay will be reduced by half of your weapon skill plus or minus 1% for every 2% your attack stat is above or below 20"

I'd probably also put the "this weapon is better for the dexterous" and "this weapon is better for the strong" description lines back in. Since str weight of the weapon would have an actual noticeable in game influence and you'd want to know even un-spoiled whether you'd be better off with an eveningstar or a demon whip.

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 01:41
by crate
str seems close to ok to me now

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 03:19
by Siegurt
Well, I sat down and did some charts on how Str effects casting and EV for given suit of armor in increments of 5 from 5 to 40.

Looking at the breakdown, it seems like each encumberance penalty has a fall off point where further Str raising will confer minimal additional benefit.

from my testing it looks approximately like:
Leather: 7
Ring: 10
Scale: 15
Chain: 20
Plate: 25
CPlate: 30
GDA: 32

Is where the fall off point lies roughly.

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 09:12
by Bim
'Well, I sat down and did some charts on how Str effects casting and EV for given suit of armor in increments of 5 from 5 to 40.' - Ladies, I have arrived.

In all seriousness though, this would be a welcome improvement in making strength matter, and it makes sense that strength should confer better weapon speed.

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 14:13
by Klown
The wiki even tells you Strength is bad.

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 15:09
by Sporkman
Klown wrote:The wiki even tells you Strength is bad.

The wiki is not an accurate source of information. Especially for trunk.

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 15:46
by mumra
It was my understanding that Str affects weapon damage?

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 16:20
by mikee

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 16:20
by Siegurt
Yes, str and dex do effect weapon damage.. by a near un-noticeable amount.

Specifically the same "attack stat" I mention above modifies a weapon's max base damage by "(attack stat-10)/78" percent (roughly 12% for every 10 points you are above 10)

On a triple sword (19 damage) with 20 str,that's an average of about 1.2 damage/hit,

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 16:28
by mumra
Siegurt wrote:near un-noticeable amount.


"Near un-noticable" actually translates as "noticable", right? If you don't think a damage increase is good then I'm not sure what more would impress you...

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 16:32
by neil
Siegurt wrote:Yes, str and dex do effect weapon damage.. by a near un-noticeable amount.

Specifically the same "attack stat" I mention above modifies a weapon's max base damage by "(attack stat-10)/78" percent (roughly 12% for every 10 points you are above 10)

On a triple sword (19 damage) with 20 str,that's an average of about 1.2 damage/hit,


That's /39 in trunk, so it has twice as much effect.

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 16:39
by Siegurt
Ah, yes, I missed that commit.

Looking at it now that's more significant than it used to be, although I contend that 2-4 average damage for all 9 of your stat bumps isn't really a huge difference over the 1-2 it used to be.

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 16:44
by mumra
Stacked with other potential damage modifiers, weapon speed, haste, and so on, that can be more effective damage output per turn than it sounds. And it all adds up over a long time and a lot of fights...

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 16:47
by Siegurt
So it it enough that you will (in the absence of other modifiers, like the desire to wear heavier armour) take str over dex/int on a melee fighter for that damage bonus alone? Because it's not for me.

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 20:07
by dpeg
Regardless of the current balancing situation (is Str too weak, too strong, or just right?) I find Siegurt's approach to make strength matter for weapons to be ingenious and very elegant.

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 20:24
by Lokkij
Sporkman wrote:The wiki is not an accurate source of information. Especially for trunk.

(Sorry for being kind of off-topic)
Well, at least we have version tags on almost every page now, so you can at least see if something is up-to-date now :P

Re: Proposal: Make Str better

PostPosted: Saturday, 22nd June 2013, 20:43
by mumra
I'm not sure if it makes sense for Str to impact speed: I would say strength for power (i.e. damage), Dex for speed. It might not make complete sense physically speaking but it's a clear separation and easy to understand.

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 06:02
by Siegurt
Dexterity has very little to do with how quickly you can swing a sword. It contributes to your ability to swing it precisely to cause damage to the optimum location, and get your blows to land when and where they should, and how you move your body to get into position to hit someone, but the actual speed of the weapon moving through the air is strength based. The more force you can apply to overcome inertia (Which applies to both swinging a weapon, and to stopping it so you can swing it again) The faster you can swing it. And yes, weapon moving more quickly does equate to more power, since power is a function of deceleration times mass.

Now you can argue that this is just a game, and a simple "Strength is damage, Dexterity is speed" breakdown is one that's easy to understand and would make the tradeoffs for one vs the other obvious, and that's fine if that's your opinion and if that's the way things shake out down the road I won't cry in my beer. Of course, that's not the way it works currently. Currently it's "Weapon type and weapon skill is speed" and "Strength and Dex together contribute to the base damage and accuracy in a not-obvious way"

However, my objective in making this suggestion isn't to make things simpler or easier to understand (The relationship between strength, dexterity and weapon damage are already pretty cloudy for the average non-source-digging user, and I wasn't attempting to make things clearer for them)

What I was shooting for, was two fold:

1. Make Strength a more important factor for a melee user, particularly one who wanted to use the big "ultimate" weapons.
2. Make the choice of which weapon to use not be "always use the biggest one", having lighter, faster weapons be better choices for some character builds.

In exploring the concept I learned a couple things:

1. Str and dex are more important now they they used to be.
2. My initial attempt at a formula needed some tweaking. (Having it's effect seemed about right)
3. It works better to use "Weighted strength" than it does to use just straight strength (Otherwise Str becomes too dominant and it's too easy to make a character that's so strength-heavy they can swing top-end weapons with very little training, also using weighted strength means high-dex/low-strength characters have some good viable options for better weapons)


I'm not as certain as I once was that Str needs more tweaking, but honestly I *really* like the idea and how it plays out in your training regime for a character, it gives a whole set of give and take that makes it so that it's longer a "gimmie" that a melee user will just train their weapon skill up to minimum delay, with the intention of using the largest weapon in the weapon class they've chosen.

I like it that some classes of weapons will be harder or easier for builds with higher/lower Strength/dexterity balances, and that playing through the game, you'll be required to sacrifice to get access to the top tier of weapon at optimum speed.

I can also understand the opposing point of view (That this makes the whole strategy/training/character building thing more complicated, and so it's not obvious what to do at the outset which makes it harder for novices)

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 19:41
by Wahaha
2. Make the choice of which weapon to use not be "always use the biggest one", having lighter, faster weapons be better choices for some character builds.
This is already the case.

Furthermore, to use a triple sword, you need 24 in swords. This is the investment. It's a big investment. You don't need to make it harder to use big weapons, they already have an appropriate cost. What you're doing is (severely) limiting what weapons certain race/class combos can use for no reason.

The proposal linked below is much more elegant in my opinion, but it still suffers from the same problems. Read mikee's post in particular.
viewtopic.php?f=8&t=7007&p=93984

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 21:44
by Siegurt
2. Make the choice of which weapon to use not be "always use the biggest one", having lighter, faster weapons be better choices for some character builds.
This is already the case.


There is currently no case for which anything less than the biggest weapon of the type you're using is optimal in the long run, no matter what build you start with eventually you can train to min delay with the largest weapon of a class.

The only build in which it's better to have a lighter/faster weapon is a stabbing build, and that's only because short swords give a huge bonus to stabbing and are all smaller and lighter, however ultimately with everything maxed out it's *still* better to have the biggest, heaviest short sword you can find.

Furthermore, to use a triple sword, you need 24 in swords. This is the investment. It's a big investment.


It's true 24 sword is an investment, however it's a time investment, not one that actually involves any long term sacrifice of one ability over another.

20 Str isn't over the top ridiculous to achieve, I've had spriggan enchanters with 20 strength without investing any points into it. And 20 strength is all it takes to make it so the cost to get to minimum delay is the *same* as it is right now. You can still get there with 16 strength, it just takes more points.

You don't need to make it harder to use big weapons, they already have an appropriate cost.


I disagree, I think we do need to make it harder to used big weapons, and the cost isn't appropriate because the only sacrifice to use them is time.

What you're doing is (severely) limiting what weapons certain race/class combos can use for no reason.


Well, first of all let's talk about "severely limiting":

What this change would mean, is that you'd need to have a minimum of 16 strength to train to *optimally* use a triple sword, with 14 Strength it's still the best sword in the game, with 12 Strength you're probably better off downgrading. Note that this does *not* limit which weapons you can use, nor does it increase the base delay of the weapon, nor does it alter the damage or accuracy of the weapon. It also means if you invest more into strength (either with equipment bonuses or stat bumps) you'll be able to get to minimum delay earlier, letting you train a wider variety of skills (So that say, "Ogre mage" as an archtype isn't the slowest most grindingly painful experience on the planet) Getting to 14 Str from any race/class starting combo is possible, certainly it's harder for some than others, but if you select "Tiny wimpy guy who is awesome at spellcasting" as your starting package, I don't think it's unreasonable that you expect it'd be harder to attain "Able to optimally swing the biggest and best melee weapons in the game" without some difficulty and sacrifice.

Now let's talk about "for no reason":
I've got several reasons for making the suggestion which I've outlined above, however to put them in a nutshell; I think you should need to invest more into your physical attributes at the expense of spellcasting attributes to become optimally proficient with larger weapons. I don't think you should be able to take every stat bump as "Int" and then go grind for a while in Pan and have the ability to use a triple sword optimally *and* have all the spellcasting bumps that come from Int, I think you should have to choose one over the other, or even choose to go half and half and still have that be a reasonable choice.

Currently what kind of armour you can use optimally is limited by Strength (By a larger degree than this proposal does to weapons), and no-one complains about that, they just use a lighter armour or train strength.

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 21:52
by Siegurt
Oh and actually I did read this post:
  Code:
The proposal linked below is much more elegant in my opinion, but it still suffers from the same problems. Read mikee's post in particular.
viewtopic.php?f=8&t=7007&p=93984


It has a similar net effect, although instead of making the player choose whether to use a smaller lighter weapon or the bigger heavier one, it simply makes the bigger weapons behave like smaller ones. (I just personally happen to like the idea of picking through those mid-level weapons and having one of those be the best for you) But really I think one implementation vs. the other comes down to personal preference.

I also proposed a strength based "speed reduction cap" above, which also has similar net effects with a different mechanic.

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 22:08
by dpeg
When I have some time to do the proposal justice, I'll try to advertise it a bit on ##crawl-dev (perhaps it's been dealth with over there already, don't know yet).

Two more random comments by me:
1. Those most interested in strong melee weapons have a tendency to be characters with lots of experience on their hands. For them, putting stat points into Str might be straightforward already (for armour purposes) or at least won't be a strange choice. On the other hand, for a character interesting in skilling a lot (i.e. at least part-time casting), Siegurt's proposal offers a few new decisions -- I fail to see how that can be a bad thing.

2. The proposal is still quite mild because there is no hard penalty (as there is with strength requirements for body armour). This is similar to how intelligence affecting spell hunger is not a hard penalty: you can (in principle) just eat more food to deal with high level spells and low Int. Having high Int allows you to cast spells more easily (food-wise) *now* but you can work around that. Having higher Str would allow you to reach mindealy earlier but you will still arrive their eventually. Of course, Int offers a lot more than that -- so also from this point of comparison the proposal is quite timid. (It is true that faster weapon speed is of great value in itself, so getting their quickly might be worth something.)

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 22:27
by Siegurt
minmay wrote:So you want picking strength to be a no-brainer?

Oh, and I missed this post in the fray :)

Actually I think currently not picking Str is a no-brainer, I'm trying to make it at least as viable a choice as raising int.

In my head at least, currently if you're a primary spellcaster you pick int, if you're a ranged weapon user, you might pick dex, but also possibly int if you are aiming towards spellcasting, if you're a melee user you pick int if you think you might *ever* use spells, unless you don't have sufficient str to use whichever body armour you're going to be working towards (probably the scale-mail evasion candidates), after which you always pick int, or dex if you are trying to not ever cast spells.

I'd like to see at least for people who get most of their damage out of melee, that Str is always a viable choice, and is at least as strong as Int of a candidate.

In my head, in an ideal world, mostly-meleers would tend to prefer str, mostly-ranged-or-stabbers would tend to prefer dex, and mostly-spellcasters would prefer int, and Hybrids would have to choose which aspect they'd want to add facility with.

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 22:28
by galehar
Siegurt wrote:It's true 24 sword is an investment, however it's a time investment, not one that actually involves any long term sacrifice of one ability over another.

It's a skill investment which definitely involves the sacrifice of not training other skills.

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 22:29
by Siegurt
Unless you just train them later.

I just realized that this is a "is it a closed or open ended game" kind of discussion. if you perceive there to be an ultimately limited amount of skill-raising to be done before you finish the game (or retrieve the orb, or whatever your perception of 'end' is) this will look like a bigger deal than it is.

I come from the perception that the amount of exp *ultimately* available in the game is far larger than you will ever need, and if you need to go to 27 instead of 24 to get a sword to min delay, that's just not a big deal, since you would train those points in the long run anyway. However I can totally see that there's another point of view.

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 22:35
by Siegurt
Also (And I'm not sure if this is clear or not given the types of replies i'm getting from some quarters) is it clear that with high strength this actually makes it *easier* to hit min delay with large weapons?

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 23:10
by mumra
Siegurt wrote:In my head at least, currently if you're a primary spellcaster you pick int, if you're a ranged weapon user, you might pick dex, but also possibly int if you are aiming towards spellcasting, if you're a melee user you pick int if you think you might *ever* use spells, unless you don't have sufficient str to use whichever body armour you're going to be working towards (probably the scale-mail evasion candidates), after which you always pick int, or dex if you are trying to not ever cast spells.


I should point out that already this does not qualify as a "no-brainer", there are a few things to think about here, and a few decisions to make: "am I ever going to cast spells? which armour do i want?" etc.; but that's not to say things couldn't be made more interesting (but also I don't think you're taking every single factor into account here).

It's also worth mentioning that I've been working on an idea to make Dex a bit more relevant for evocations, so as a fighter it could be a worthwhile choice over Int if you want to back your melee up with evokables instead of spellcasting.

Overall, though: I don't personally find agonising over which stat to upgrade to actually be a very fun decision. It's nice to have the choice here being fairly clear-cut for most classes/playstyles, so you can take your stat point and get on with playing the game.

Having said that, I don't hugely like the thing of "if you are doing any magic at all, choose Int every time". This is common advice, and I don't know how much this has changed since the last improvements to Str. It sounds like the sort of very simple advice that is easy to give people because it will work and avoids having to go into far more complicated explanations. It would seem natural to me to have the thought process "I'm a fighter, and I want to use spells a little bit, but I also want to do damage, so I'm going to put some stats into str and some into int". If doing that is a strictly bad choice right now then maybe there's a problem.

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 23:20
by Siegurt
Not bad, simply suboptimal.

Optimally (Again this is my perception, not fact) as a meleer-who-wants-to-cast-a-few-spells you raise Int, because it will reduce the amount of exp you need to invest in your spellcasting skills to get your spells to an acceptable failure threshold, so you can concentrate on your weapon skills and fighting, which are much more important for "how much damage my weapon does" than Str is. (And again, that's *after* you've gotten enough str to nullify the really large part of your body armour's spellcasting penalty)

Most melee-ers (IMHO) choose to go with less str, and lighter body armour (mid-range, scale/FDA/PDA/IDA) so they can invest less into Str, so they can put more points into Int, so they can cast spells with less spellcasting skills, so they can use their weapons more effectively.

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 23:22
by Siegurt
(This proposal actually just makes a reverse path for that trend so investing into Str means you can invest less into your weapon skill, and in turn have more skill points available for spells to cancel out your lower int)

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 23:29
by rchandra
before the recent Str changes it was not at all unusual for me to seek approximately equal int/dex on characters like necromancers, or melee guys that want haste/charms, etc. I suspect that that sort of character would now be fine with int/str instead, depending on armour.

Re: Proposal: Make Str better

PostPosted: Sunday, 23rd June 2013, 23:37
by Wahaha
Unless you just train them later.

Later you already won. Being able to scum Pan and get infinite skills is not a valid argument for anything, especially considering most winning characters don't even enter Pan.

About weapon variety: There are a few best weapons for every weapon type because those are the best weapons. They're supposed to be good. They're supposed to be better than a morningstar or quarterstaff. I don't see why you're trying to make people use halberds or whatever, no one will do that because halberds are bad. This proposal will create less weapon choices, not more. Most weapons are already useful as upgrades when you don't have the best weapon. If I'm training polearms, I will use a halberd before I get a bardiche. But if using a bardiche isn't an option I won't use polearms at all.

Besides, faster but weaker weapons are desirable if they have the elec or pain brands. Small races can't even use the biggest weapons except for a lajatang. Some characters find a good weapon that isn't the best base weapon type. Some characters use one handed weapons. I might use a demon blade on my conjurer of Vehumet but I'm certainly not going to train 24 swords for a triple sword if I plan on casting fire storm.

About stat choice, this is for mumra too: Dex is about as good as int currently. A melee weapon user most likely wants dex, not int. The advice of "always pick int if you cast any spells" is wrong in most cases where it is applied.

I wouldn't mind str being more useful, I feel like there's still not enough incentive to pick str. This proposal is not the right way to do it in my opinion. Besides what I mentioned above about weapon variety, this proposal causes balance problems that aren't easy to fix. Characters who get more than 16? str naturally are buffed for no reason. I don't think Minotaurs need to be better at using weapons than they are already. Characters who get less than 16 str are nerfed for no reason. If I start a Tengu Necromancer I most definitely want to use a weapon, but my choice is either more limited than before or I have to raise str, thus making this combo weaker.