Shield reform proposal discussion

[crate]

I think people get way too hung up on "won't block more than one attack" (the same way they get way too hung up on some things like GDR and spell hunger). If you're getting hit more than once per turn, move so you're not. This is seriously not a big deal unless you're with chei, or if you're a naga. I think mechanically shields work fine as-is. The fact they only are good at blocking one attack per turn is fine.

The situations where you are getting attacked by more than one enemy per turn and you cannot move to a place where this won't happen and you are actually in danger of dying are rare and tend to be really bad situations anyway. You might run into a centaur pack in open space (and in this case a shield is still much better than no shield until you close the gap). That's about it. Stuff like standing next to multiple yaktaurs is fine: yaktaur melee is not very scary unless they found a good melee weapon. Standing next to an orc knight and five orcs is fine: the orc knight is the only dangerous enemy there, and if a single orc knight is scary enough to threaten to kill you then you can move to somewhere where he's the only enemy attacking. So this leaves multiple-attack enemies: mainly various trolls, occasionally hydrae (most characters don't have a shield in lair anyway), some uniques. Personally I think it's ok if shields are less effective against the later attacks, the strongest attack always comes first anyway in a single monster's attack order.

Even in open spaces like much of lair as long as you have corners and obstacles you can usually end up fighting one enemy at a time, though you might be moving a dozen steps between attacks.

The main drawbacks to wearing a shield are these: 1) you cannot use a twohander 2) they hurt spellcasting. These are obviously balanced by the extra defense. The extra defense is noticeable, but so is the loss of damage.

The fact that there is not a lot of incentive to use non-buckler shields is probably a problem but this proposal doesn't seem to me to do much to address this ... big shields will still have the same problem of being awful for characters who kill things unless you get a lot of skill.

My suspicion is that changing shields so that they are equally effective against all attacks in a turn but not telling anyone wouldn't even get noticed for quite a while.

[jejorda2]

Could it make sense to reduce the penalty to blocking multiple attacks per turn when wearing a medium odd large shield and no weapon? You could use both hands to control the shield while using spells for offense. Maybe you have to wield the shield or wear it twice so you can't make unarmed melee attacks.

[Bloax]

Could it make sense to reduce the penalty to blocking multiple attacks per turn when wearing a medium odd large shield and no weapon? You could use both hands to control the shield while using spells for offense. Maybe you have to wield the shield or wear it twice so you can't make unarmed melee attacks.

This

make higher shield skill reduce the rate at which effective SH value gets reduced per blocked attack in a turn, and maybe even implement a GDR-like lower cap % based on shield size

plus this

Instead of EVP = base_evp - skill / 5, how about EVP = base_evp * 27 / (4 * skill + 27)

and optionally, one or two new shield brands, would collectively be an excellent way to improve shields.

I much prefer this solution to about all in this thread. (Though the Check EV -> AC -> SH would probably also help a lot systematically.)

[BlackSheep]

Note that the part that shield get worse against certain attacks is not true, shields with my proposal would still block weak attacks if those are the only attacks you may ever get that turn, the damage treshold would only be applied if there are multiple attacks. Is this unintuitive? Maybe a description of the skill saying "This skill improves your ability to block incoming attacks and your capacity of judging which attacks are best to block in case you get attacked more than once" would dissipate any doubt, maybe not.

I don't know crawl internals enough to see if this is hard to calculate (but I've seen very complex computing done in some code parts), or if it is needed to change the order of something, for instance, predicting potential damage could mean that crawl has to figure out attacks in parallel instead than one after another regardless of how these are presented to the player. This is more easily said by current developers who can assert that the result would not be worth the effort, but this I could understand.

I can read the code and have written (very minor) additions to it. I can tell you that there is no mechanism in place that would allow the code, while evaluating one attack, to anticipate or evaluate other attacks. As it stands now, each actor (player or monster) takes their entire turn independently. Parallel computation of turns by itself would be a massive undertaking. Even if you handwaved that part, the number of corner cases you'd have to handle in order to predict which attack would be the best one to block would be huge. And all of this work would be in the name of making shields work a little better?

If you're seriously interested in a developer's input on the feasibility of your idea, go to the crawl-dev IRC channel and ask them about it.

[palin]

As it stands now, each actor (player or monster) takes their entire turn independently. Parallel computation of turns by itself would be a massive undertaking. Even if you handwaved that part, the number of corner cases you'd have to handle in order to predict which attack would be the best one to block would be huge.

I bet some euristics can be done to avoid all that complexity, but if the stuff it's not worth the effort, let's discard the proposal not because it's bad per se but because it requires a lot of development.

[Psieye]

The main critic about shields is that's not going to block more than one attack per turn (and you can read this opinion in almost any thread in tavern mentioning shields), and my proposal tries to fix than by concentrating on blocking dangerous attacks.

Disagreed. The main critic about shields is that it's useless against numerous highly threatening attacks. "I can't block everything if I'm surrounded or taking flak from multiple missiles" to me sounds like it translates to "can I have something to let me survive being sloppy with my positioning?" Even with a big 2h, even with Cleave, the basics of (melee) tactics in crawl don't change: fight in corridors or at least with minions to avoid being surrounded and fight as few opponents as possible at a time.

With a serious investment in Shields, it's not hard blocking 2~4 attacks in one turn. It's just that by the time you make that investment, the threats you care about don't get blocked by shields: stuff like beams (e.g. bolt of fire), mutagenics, paralysis, hellfire and torment. AC/EV do a good enough job against physical attacks and you care less about them unless you have neither.

[palin]

With a serious investment in Shields, it's not hard blocking 2~4 attacks in one turn. It's just that by the time you make that investment, the threats you care about don't get blocked by shields: stuff like beams (e.g. bolt of fire), mutagenics, paralysis, hellfire and torment. AC/EV do a good enough job against physical attacks and you care less about them unless you have neither.

So I think shields should receive some kind of buff, but not improving SH against a single opponent, otherwise they would be too convenient.

And I feel they should have some sort of sinergy with heavy armour.

But you're entitled to disagree

[Psieye]

With a serious investment in Shields, it's not hard blocking 2~4 attacks in one turn. It's just that by the time you make that investment, the threats you care about don't get blocked by shields: stuff like beams (e.g. bolt of fire), mutagenics, paralysis, hellfire and torment. AC/EV do a good enough job against physical attacks and you care less about them unless you have neither.

So I think shields should receive some kind of buff, but not improving SH against a single opponent, otherwise they would be too convenient.

And I feel they should have some sort of sinergy with heavy armour.

But you're entitled to disagree

You haven't directly responded to my point. What is it that you want? For shields to give you some slack for bad positioning? For shields to take less EXP investment so you don't feel bad about ditching them when you hit lategame? To run around feeling awesomely invincible?

I once grinded Hell for EXP and ended up with a MD that could do 'everything' - 1h melee, shields, crossbow, earth magic, buffs, Chei invocations. It was awesome being able to peak at 50+ AC, 50+ EV and 50+ SH simultaneously with the amount of Pan/Zig scumming I did. My (looted) eudemon blade wasn't for killing fiends, it was for killing popcorn. Point-blank LRD, LCS, slouch and shatter (yes, with CPA and large shield) were my fiend-killing sources of damage. I think that was also the character where I bored myself scumming so much that I even had Necromutation online. My point is: you can run around feeling amazingly invincible if you scum for EXP and loot long enough. You can still get killed easily if you get careless but that's crawl what do you expect? Shields don't need a tweak to bask in that feeling of invincibility. Just as relevant: you don't need to be 'invincible' to win.

[palin]

You haven't directly responded to my point. What is it that you want?

That's stated in the proposal. To let melee character get a meaningful choice between one-handers and shields. Currently there is no choice, you go two handed. Also to make SH skill have sinergy with Armor not by improving the aptitude but by taking into account potential damage.

I think this thread is to be closed, since BlackSheep asserted with some confidence it's to much code to do.

[Psieye]

You haven't directly responded to my point. What is it that you want?

That's stated in the proposal. To let melee character get a meaningful choice between one-handers and shields. Currently there is no choice, you go two handed.

Again, disagreed. Just because 1h+shield means you can't (optimally) rely on melee to kill the serious threats doesn't mean you can't be doing melee the majority of the time. If your character literally is incompetent at doing ANYTHING OTHER THAN MELEE (and Trog summons, because you need no skill investment for that) then sure, you go 2h. Going 1h+shield+other skill means you're making some parts of the game easier and other parts harder. 2h Trog is a proven and popular style but nothing says 1h+shield+other is inherently bad: it just takes more work and pretty darn good loot. But hey, some races are easier than others yet we still consider this a meaningful choice. Just because a choice isn't 50:50 doesn't mean it's meaningless.