attack delay

[galehar]

More than the fact that the attack delay stops improving at a certain skill level, I am annoyed by the fact that the speed doesn't increase linearly. Going from delay 6 to 5 is a much bigger speed increase than going from 13 to 12. It also means that raising your M&F skill from 11 to 12 will have a much bigger impact on the damage output of your demon whip than your great mace.

So I suggest that instead of having the delay decreasing linearly, we have the speed increase linearly. This can be achieve by using this formula for weapon delay:

  Code:

delay = base_delay * k / (skill + k)

With k being the skill level at which the speed stops raising. Of course, ideally, k should be set at 27, but it would be a huge nerf to all fast weapons which would probably need to be balanced. I see 2 ways to balance such a change:

• Increase the effect of weapon enchantment. Make +x actually mean +x, not +random2(x)
• Increase the effect of stats. Ideas for formulas are much welcome.

Here are a few graphs of the effect of the change on speed for various values of k and a weapon with base base delay of 12 as an example.

This is a big change, but now is a good time to do such a change. We have time to test and balance before the next release.

[galehar]

After a quick chat on irc, I think the change wouldn't be worth it if we don't set k at 27. So consider this as the proposal:

  Code:

delay = base_delay * 27 / (skill + 27)

[bobross419]

Not arguing with the idea, just pointing this out. This will increase the min delay for some of the slower weapons.

Exe Axe for example - Wiki says current min delay is 7, under the new formula it would be 10. I'm not able to put any effort into the maths right now, but couldn't this push Demon Whip even more into the realm of only weapon that is worth it?

[galehar]

Right. I guess I'll go back to my drawing board to figure out the proper formula so that speed increase linearly until minimum delay is reached at skill 27.
Since this would wreck balance hard, plenty of other stuff would need to be adjusted, including weapon stats. It would be an opportunity to bring more balance here. At the very least, demon whip would need a much higher skill investment to reach their former awesomeness.

[Galefury]

Big change, huge balance implications.
Main benefits: finally some clarity in weapon speed and weapon skill effects, high level weapon skills will be more useful.
Disadvantages: huge balance change, both early/late game power and fast/slow weapon balance get changed a lot.
IMO it's worth trying, but if this is done it should not be done half-assed. These balance changes will need to be looked at and possibly compensated for. This should not come at the expense of clarity, which IMO is the most important benefit of this proposal. Coming up with changes that actually address the problems caused by this might not be as easy as simply changing enchantment effects (which in itself is not a bad idea IMO).

If the devs want some weapons to grow more with skill than others (as is currently the case) the min delay exceptions could be incorporated into the new formula by reducing k for certain weapons (without capping speed growth at skill level k). This stays close to the status quo, but also keeps all the special cased baggage and reduces clarity.

Instead the base delay could be set to 2x min delay for all weapons, which would keep endgame power the same but improve slow weapons early on (by reducing their base delay). I think buffing 2h weapons is much better than nerfing them, as they get very little use. But since the new speed formula would mostly come at the expense of fast weapons I'm not sure the implicit nerf to slow weapons even needs to be compensated. If enchantment gets improved (which strengthens fast weapons more) something might need to be done.

Edit: one version of reduced k for slow weapons that keeps most of the clarity would be setting it to one specific value (like 18) for all twohanded weapons. k could depend on the wearing of a shield for 1.5handed weapons. Possibly some other benefit of using 1.5handed weapons without a shield would have to be removed.

[Frandu]

delay = base_delay - (base_delay - min_delay)*skill/27 ???

[galehar]

I think k must be equal to

  Code:

(27 * min_delay) / (base_delay - min_delay)

[evilmike]

Regarding this proposed change to attack delay, I'm not against the idea adjusting how melee works. However, I agree with the sentiment that this should be treated very carefully. If this goes the wrong way, it could work out to be a massive nerf to melee - this would be a really bad thing, and would be reminiscent of the changes to armour in 0.6 (the massive nerf that led to almost everyone playing EV-based characters for two versions). You wouldn't want something like that happening again.

[dpeg]

evilmike: Overnerfing simply happens. However, since I was mostly responsible for the AC mess, I'll refrain from giving numerical advice here.

The list of wishes is long. We want
1) that high skill investment is more useful, and more obviously useful;
2) to use the opportunity to make enchantments and stats relevant.

Perhaps this is too ambitious, though.

Regarding balance, I'd be surprised if the early game was affected so much: all the numbers are rather small, and changes should amount to little.

[evilmike]

evilmike: Overnerfing simply happens.

Yeah, it's understandable. And not really problematic as long as it stays in trunk versions... the good thing is that 0.10 is still pretty new. My worry is mostly for 0.10 stable, whenever that may be. It would be unfortunate if pure-melee characters got relegated to "challenge game" status.

[dpeg]

minmay: was too imprecise, sorry. As far as I understand, the to-hit enchantment is not meaningful enough. I'd be glad to hear if it is, but there is some indication to the contrary.

[evilmike]

As far as I understand, the to-hit enchantment is not meaningful enough.

To-hit is a bit lackluster - or at least, it seems lackluster, which is probably just as important. It's not that it does nothing, it's just that damage is what you want out of a weapon. In my opinion, the easiest way to solve this would be to just roll both enchantments into one. Thus, no more double enchantments. A +3 weapon would give a bonus both to-hit and to damage. This would be easy to balance, I think - just roll the EW scrolls into one and adjust how common they are. I think something like this has been proposed before, and I know to some people this idea is probably blasphemy.

This has the nice effect of making weapon enchantments a bit more relevant, and also greatly simplifies things for the player (especially new players). Plus, it just gives you less to think about. To-hit would become meaningful in the sense that it's rolled into the overall weapon enchantment.

Slaying should probably keep the double system, because it works differently and it allows for interesting things like (-2, +3) rings of slaying.

One reason I think this is the best approach is because it's very hard to quantify a thing like accuracy in a way that can be easily grasped by the player. Damage is easy, because the player gets a nice number that says "this is how much harder I will hit", and the feedback in combat is pretty clear with how much damage deal per hit. Accuracy is a lot harder to gauge, in part because there are so many more factors at play when it comes to hitting or missing a target. By rolling the accuracy enchantment into the damage one, I think this problem is solved quite elegantly.

[bobross419]

In the most basic sense, to-hit is just as easy to conceptualize as damage. I'd have to say that rolling both enchantments into 1 would make the benefits of to-hit even more vague.

As a brand new player to both Crawl and roguelikes, I assumed that To-Hit is compared to EV and Damage is compared to AC. I also assumed that To-Hit would be compared against SH. A perusal of the code confirms these assumptions... Granted there is a lot more that goes into each calculation than a basic To-Hit vs. EV/SH and Damage - AC, but I don't think that a player needs to really know the specifics to benefit:

Stat, Weapon Skill, and To-Hit enchantment help against EV and SH.
Stat, Weapon Skill, Fighting and Damage enchantment help against AC.

Pretty basic stuff, and the first time I was able to put massive to-hit enchantments on Demon Whips and Exe Axes I was able to see a difference in my hit ratio.

In most of my games I've had enough EW scrolls to put my chosen weapon up to at least +7/+7 by the time I've cleared Lair, Orc, Hive, and a few other branch levels. And I've gotta say there is still some excitement (for me at least) when I find a rare EWIII scroll

@evilmike - It took me exactly 3 scrolls to figure out the difference between EWI, EWII, and EWIII... There are also descriptors to help give an indication of how much you missed just like there are indications of how hard you hit.

[evilmike]

Part of the problem is that EV is just way more complicated than AC. AC is simple: damage is reduced by a number from 1 to AC. GDR makes this a bit more complicated but monsters don't have that, as far as I know. Basically, you can easily tell how much extra damage you'll be dealing to a monster without doing any number crunching. If your weapon has +8 damage enchantment (which works as +1d8), you have a fairly exact idea how much extra damage you'll be dealing.

The accuracy enchantment is a lot more opaque. How much more likely to hit does +8 make you? There's no easy way to express this, unlike damage. Where AC and damage are nice and simple, trying to figure out how EV works results in crap like this: http://i.imgur.com/0RSGG.png (note: this graph is for players, not for monsters... something like this for monsters would look different, but it would be about as confusing). It's just not very easy to grasp how it works. I think a big reason why to-hit is undervalued right now is because almost no one even understands it in the first place.

The reason why I like rolling both enchantments into one is because it allows accuracy to keep having an effect, but the player can value the enchantment for the damage increase (and just think of the accuracy boost as a nice, albeit confusing extra). Instead of wondering how much of an effect that first number has, or trying to figure out if (+5,+1) is better than (+1,+5), players can just know that bigger = better and leave it at that.

[bobross419]

Without spoilers you can't really know that +8 damage enchantment gives +1d8. A player could assume that it gives a flat +8.

I was under the impression that the dev team didn't want players to have to worry about specific formulas. Higher To-Hit results in a higher chance of getting to damage the target.

I just don't see how having just one enchantment value would benefit the game, the basic concepts aren't hard at all. If someone wants to better understand the intricacies they can always do some research on Wiki/Forums. If someone really wants to optimize they can always dig through the code and crunch numbers. In the RPG world this is called min-maxxing.

[Danakh]

For the attack delay proposal : Maybe the nerf to fast weapon may be aleviate with the merge between stabbing and short blade ? Other fast weapon may also get a small stabbing bonus to help balancing.

For the accuracy lack of appeal. Who does critical hit works ? (is there any ? )
As i see the main problem is that high accuracy value is not important as margin is irrelevant. You hit or not, that's all
High damage is on the opposite very important, with armor damage reduction. the more you get, the better (overdamage is not a problem due to high variance)

Making some damage bonus (critical) for high accuracy would raise it's interest.

[Danakh]

To help keeping a difference between the two, critical hit should be a multiplicative damage after armor reduction.
So it won't help much against high AC monster, but a lot more against low AC (both for hitting, as they usually have high EV, and for damage with critical)

High accuracy weapon for fast and low armored target
High damage weapon for heavy armored target

[RFHolloway]

how about

weapon speed = base speed (1 -wsf(Min (skill,dex)/27)
Base weapon damage = base damage *(1+1/3((str-10)/27)*(1+wdf)^skill

wsf = weapon speed factor
wdf = weapon damage factor

wsf might be 2/3 for a rapier like weapon, but 1/4 for a club - then for the rapier the attack delay would be one third of the inital at skill 27, but capped by the dex of the user (you can't move that fast if you are liable to trip over your own feet)

wdf is a straight damage increase for each point of skill - might be 5% for a club, but 1% for a rapier - strength also gives a bonus(formula is probably total inappropriate)