Weapon Strength Weighting Removal

[svendre]

I saw this in the change log:

--------------------------------------------------------------------------------
d466a24 | Chris Campbell | 2016-01-12 19:11:51 +0000

Remove weapon strength weighting
Makes to-hit be purely determined by dex, and to-dam by strength. This should
make the effects of dex/str slightly more noticeable, and more importantly
removes the mostly insignificant differences between weapons with different
strength weights.

I really dislike this change. I don't understand the reason for the removal of the weighting system, was it just to be easier in the code? Even though the weighting system may have been "mostly insignificant", I thought it made a lot of sense. Why was it necessary to remove it? I thought it was a very clever way to distinguish weapons from each other and tied in with planning the development of a character. Now, if you want to achieve a balance you what... train dex to use an executioner's axe because it's not very accurate... and train strength to use a short blade so it will hit harder where it's already more accurate... ??

If the weighting system was mostly insignificant, and the effects of dex/str needed to be more noticeable, is there a reason the weighting system couldn't remain but be turned up to a greater more significant effect? I don't particularly want the mechanics to be dumbed down. As for the weights being shrouded in mystery, that could have been easily fixed by simply displaying them on the weapon description page. Apologies if I'm not understanding something about this change.. but if it's what it sounds like, please reconsider.

[Tiktacy]

From what I understand, the only argument you have against it is more-less flavor related. To be honest, I'm still confused as to what your objection is, I'll try my best to guess though.

Training dexterity for accuracy and strength for damage is the way games traditionally function(at least in my experience). I think that is a big improvement over the status quo, I have been playing this game for a while and I STILL don't understand the actual effects of dex and strength on damage output, I just put it all into dex because I heard it was the best choice from someone a long time ago.

Either way you are going to get a boost in overall damage no matter which one you choose. Dexterity may not increase damage directly like strength does, but it increases the amount of times you get to deal damage to begin with(accuracy) as well as allows you to stay in battle longer(and therefore get more hits in) due to its effect on dodging.

Its easier to code, more intuitive, gets rid of an essentially hidden mechanic in the game, and at the cost of nothing(not even flavor, its only changed slightly). It seems like a net gain no matter how you look at it, at least to me it does.

[Hurkyl]

I don't understand the reason for the removal of the weighting system, was it just to be easier in the code?

It's already coded; it takes additional work to remove it. (albeit, probably minimal work)

Why was it necessary to remove it?

From my perspective:

• The effect was hidden
• The effect is misleading; e.g. people who hear of it often believe that damage-per-hit for a quick blade is improved more by Dex than Str, when in fact they have the same effect (prior to the change).
• I suspect the magnitude of the effect is insignificant enough that it's not worth the added complexity
• I think the general opinion is that it is misleading in that people who hear of it think stats have more of an effect than they actually do

Removing the mechanic, of course, is the cleanest way to eliminate these problems. And, IMO, it has minimal downsides -- the worst is that maybe the low strength races might need an extra point of Str to rebalance weapon damage or somesuch.

Note that the mechanic can be reintroduced in the future should someone come up with a way to make it a good one; in the meanwhile, removing it is better than the status quo.

I thought it was a very clever way to distinguish weapons from each other and tied in with planning the development of a character.
...
I don't particularly want the mechanics to be dumbed down.

I don't think there are any comparable weapons that have significantly different Str weighting; I'm unconvinced that it actually ties into the development of a character if one realistically assesses the effect.

Now, if you want to achieve a balance you what... train dex to use an executioner's axe because it's not very accurate... and train strength to use a short blade so it will hit harder where it's already more accurate... ??

This looks like an example of what I said above; even prior to the change, Dex improves the accuracy of an Executioner's axe moreso than Str, and Str improves the hit-damage of a short sword moreso than Dex. (although only slightly in both cases)

I suspect even after the change, training Str would increase overall damage for an Executioner's Axe than training Dex, although I would have to wizmode it to be sure.


If the weighting system was mostly insignificant, and the effects of dex/str needed to be more noticeable, is there a reason the weighting system couldn't remain but be turned up to a greater more significant effect? As for the weights being shrouded in mystery, that could have been easily fixed by simply displaying them on the weapon description page. Apologies if I'm not understanding something about this change.. but if it's what it sounds like, please reconsider.[/quote]

[johlstei]

In my eyes, it was just a complicated mechanic that didn't pay rent in terms of interesting gameplay effects to actually warrant it.

[Hurkyl]

Actually, let me repeat this separately since it was probably the most important part of my post and is somewhat buried.

The most important thing here is that (IMO) removing strength weighting is preferable to the (prior) status quo.

Maybe a way to make strength weighting a good mechanic can be devised in the future; at such time the mechanic could be reintroduced. In the meanwhile, things are better with this change than without.

[chequers]

I don't particularly want the mechanics to be dumbed down.

The goal of DCSS isn't complex mechanics. Using a pen & paper-alike combat system with both to-hit and to-damage rolls is already pretty complex, when you think about it.

[svendre]

OK well you may say that it is preferable, but I do want to know the why. Going back to your statements:

• The effect was hidden
It could have been documented in the weapon description.

• The effect is misleading; e.g. people who hear of it often believe that damage-per-hit for a quick blade is improved more by Dex than Str, when in fact they have the same effect (prior to the change).
The information is available, in code even for anyone who wants to read it if they really wanted to know. The game design based on, baby sitting people to not make incorrect assumptions? I don't know, seems strange. There is a counter argument that people already had a feel for different weapons with stat combinations and now they will be mislead until they adjust to the new settings.

• I suspect the magnitude of the effect is insignificant enough that it's not worth the added complexity
But the devil is in the details. You know what makes me sad, is to see a game with a budget like Diablo3 implement ultra simplistic models. Crawl has been a grand escape from that, I just cringe at the thought of it conforming too much.. losing too many of the slightly more complex details which set it apart from the fully conformed. No single detail matters so much in itself, but the sum of all the details matters so when does it matter when you lose one?

• I think the general opinion is that it is misleading in that people who hear of it think stats have more of an effect than they actually do
I've only ever heard just how totally not worth considering at all it is, several times. I'm not sure where you can find a single person saying what a huge difference they make. Perhaps I missed something. Anyhow, a few points of stats, sure they don't matter. If you're talking about deciding if you want to have a character with 6 str and 50 dex using a triple crossbow or 50 str and 6 dex using a dagger.. well, I personally think that I would notice the difference.

On the point of damage being the same no matter if you increase dex or str, I think that is a bit oversimplified. Perhaps on an average over a number of fights, damage might work out to be the same - but it doesn't mean that there isn't a difference. One reason is because it's quite a bit different thing if you one-shot something versus you needed more hits to kill something, the amount of damage you receive will not likely be the same. Then there are branding effects.. it may make a difference if you say, hit a guy with poison or anti-magic several times in a row, but you weren't even trying to do so much damage, which wind up being a decisive factor in if you win the fight or not, etc. etc.

I guess I'm just don't understand, was it breaking the code somehow? It's been in there for years, was there a major revision and this functionality would have had to have been re-implemented? If anything, it seems it would have taken effort to remove it rather than leave it?

[dpeg]

svendre: You already got very good replies, so I'll just say this: we don't dumb the game down. Instead, we're trying to make the game better and a little tighter by removing superflous mechanics -- for example, no-brainers (providing only fake choice by giving the same thing in different colours, or pretending a choice when there's always a clear best option) and marginal variety.

The following removals were all part of this: species (fake variety), divinations (no-brainer), racial armour/weapon egos (minimal effect). Removal of weapon strength weighting is of the latter type.

[svendre]

svendre: You already got very good replies, so I'll just say this: we don't dumb the game down. Instead, we're trying to make the game better and a little tighter by removing superflous mechanics -- for example, no-brainers (providing only fake choice by giving the same thing in different colours, or pretending a choice when there's always a clear best option) and marginal variety.

The following removals were all part of this: species (fake variety), divinations (no-brainer), racial armour/weapon egos (minimal effect). Removal of weapon strength weighting is of the latter type.

Superfluous:
1a : exceeding what is sufficient or necessary : extrab : not needed : unnecessary. 2 obsolete : marked by wastefulness : extravagant.

But, this is a game... not an assembly routine responsible for analyzing packets. Exceeding what is necessary can translate into better in the context of a game. I'll play it out and see how it feels some more, but I can't shake the feeling now looking at a pile of weapons thinking: it now matters less which weapon I choose, and why is that better? If you mean to say that it falls into the category of minimal effect, such as racial armour... that really hits home because I remember very well how much that change impacted hybrid elves. It was quite a large nerf, as I recall elven ring mail was the sweet spot that was deemed by many as overpowered because you could cast easier at a critical stage. I always thought it was removed because it was overpowered, and as much as I liked it (and so did many people race to find it asap) I too thought it was overpowered.

[ontoclasm]

If you mean to say that it falls into the category of minimal effect, such as racial armour... that really hits home because I remember very well how much that change impacted hybrid elves. It was quite a large nerf, as I recall elven ring mail was the sweet spot that was deemed by many as overpowered because you could cast easier at a critical stage. I always thought it was removed because it was overpowered, and as much as I liked it (and so did many people race to find it asap) I too thought it was overpowered.

The fact that you think this (for clarity: it's pretty much completely incorrect, racial-ness had hilariously small effects and the "overpowered" elven ring mail was basically just a +2 robe for elves) is kind of why such mechanics get axed. People hear about them, read far too much into them, and then get bizarre notions about their being super important secret techs, when in fact the opposite is true. Vis GDR.

[Hurkyl]

To rephrase the problem, str weighting, as it currently is, is complicated but not complex. It's a detail that affects things, but (allegedly) not to an extent that actually affects any of the decisions you make, except in contrived (or maybe extremely unusual) cases. Thus, its presence obscures things without actually adding to gameplay.

On the point of damage being the same no matter if you increase dex or str, I think that is a bit oversimplified.

You're replying to my comment about quick blades? It's not simplified at all; the quantity that feeds into the damage roll calculation for a strength weight 0% weapon is quite literally (proportional to) the sum of strength and dexterity.

anti-magic several times in a row, but you weren't even trying to do so much damage

Aside: last I heard, the antimagic effect depends on the damage you do, not the number of times you hit.

[svendre]

Wait, how was elven ring mail the same as +2 robes? It provided more mitigation potential than robes, yet did not impede spellcasting as much as standard ring mail.. that was what made it good, not simply the stats.
GDR is also in the same category as not making a difference... what??? So, what's next then, removal of GDR? Just because something isn't perhaps, use this and get +50 damage and hit 4x faster, doesn't mean it doesn't matter.

~~

I wasn't replying to your comment about quick blades, but the second reply on the thread who was saying they take all dex all the time because it's the same exact outcome as taking strength due to the fact they would hit more often. I understand what you mean about the quick blade. I'd probably take dex over str simply because it's a decent stabber and that would benefit more from the dex overall... but on a straight damage calculation it would not matter.
I didn't realize anti-magic used the damage done, that's interesting.. thanks, I'll look that one up. Now I'm wondering about draining as well.

My sentiments do not seem to be too popular, but thanks to those who politely explained their point of view on the topic. Maybe it was mostly flavor, I suppose I liked it anyhow. If a plan ever emerges to re-introduce it in a more meaningful way, I'd be down to brainstorm on how it might work. Also, I have not worked on anything for awhile, but I have an extensive IT background including coding. The last time I looked at crawl, it seemed it could benefit from a switch over to an OOP model.

[asdu]

What impact does this change have on low/very low strength characters?
I vaguely remember reading that the damage formula was special cased to account for low str-high dex characters, and the damage penalty for low str is larger than the bonus for high str afaik.
Is it now worth raising str just to avoid damage penalties on a character that starts with 8 and has no interest in wearing heavy armour? What about a 4 str SpEn?

[TeshiAlair]

It seems to me, though I may be wrong, that this change makes flame/frost brands even better on very high str/low dex characters, and things like dist and elect even better on high dex/low str characters.

[duvessa]

if you think this change makes an actual difference then you don't actually know what strength weighting did before
the commit message is misleading, it wasn't "mostly insignificant", it was completely insignificant

stats did effectively nothing for short blades before (shite base damage multiplied by a small stat bonus is still shite) and stats do effectively nothing for short blades now
the weapons that actually have good base damage were already highly weighted towards str, so the effect of stats on those wouldn't have changed remotely significantly even if str weighting was significant in the first place - which, again, it wasn't

all this thread demonstrates is that
1. a lot of people don't understand melee damage at all, but still feel qualified to comment on it
2. people will complain about literally ANYTHING getting removed, holy shit