Information necessary to choose between armours

[ker]

Let's say I'm trying to decide which armour my character should be wearing out of the ones she found in the dungeon so far. What information (values and/or how those values interact with the greater system of crawl) do I need to make an "informed" decision?

Bonus questions: Are all of the values/relationships necessary to give the choice its depth? (i.e. If there are too many numbers to display, do we need all of them?) And, more radically, is this an interesting choice to ask the player at all? (i.e. Does the decision actually have depth?) (I suspect the answer is "yes" but I think it's worth keeping in mind.)

Spoiler: show

P.S. I realise this is a bit brainstorm-y but I wanted to focus the discussion on the problem rather than my solution (which I'll be posting just below). Apologies if this shouldn't be in GDD.

[ker]

What I would like the game to do is every time I put on body armour is to print something like:
"...Finished putting on X.
Your stats have changed!
AC:12(+3); EV:12(-3); ... Stealth:200(-24); etc...
Press ? to find out more."
Where "?" would take you to a screen which would explain what "AC" stands for and why it's not the same thing as EV. Maybe it would need to be broken down in to "?"->"a" for "What is AC?", "?"->"b" for "What is EV?", etc. No idea how hard that would be to implement but one can dream

As for the second question: I know, GDR is not literally the same as more AC, but is it practically any different from just more AC? Also, I'm not a fan of EV penalty of encumbrance because it seems like I'm sacrificing some defence (EV) for... some defence (AC). I realise that AC and EV are more/less effective against different enemies, but does it matter in a practical sense? Could I just have the encumbrance penalty apply directly to AC? That way it's easier to tell if you are too weak to get a defensive boost from a particular armour.

Alright, tell me why I'm stupid

[dowan]

Well, armor selection seems fairly straightforward, at least up to the point of looking for the biggest AC+EV number + acceptable spell failure %. The AC vs EV debate can complicate things a bit there, I guess.

Is there an interesting choice between 10ac 10ev vs 12ac 8ev? I don't know, there's probably a better choice between the two. Of course the particular AC and EV values of any given armor will depend on your stats and skills...

So maybe the choice of armor itself isn't terribly interesting, but the strategic choices that lead one armor to be better than another are. At least, that's how I see it... I think I'm answering questions you never asked at this point though...

NINJAEDIT: Seeing the change in your stats summarized like that would be nice, even better if you were told before you spent the turns to try it on. I have yet to find a way to incorporate GDR meaningfully into the decision making process, especially given the fact that it's only useful for melee attacks. (At least, I think it's only melee attacks, I can't remember if it also works on ranged physical attacks like bows...)

[VeryAngryFelid]

As for the second question: I know, GDR is not literally the same as more AC, but is it practically any different from just more AC?

Yes, it reduces damage spikes. You literally cannot get hit by max monster damage.

Also, I'm not a fan of EV penalty of encumbrance because it seems like I'm sacrificing some defence (EV) for... some defence (AC). I realise that AC and EV are more/less effective against different enemies, but does it matter in a practical sense?

Yes. Being hit with every corrosive attack or poison attack matters a lot. As far as I know those apply even with 0 damage.
I remember how I almost starved vs curse skull as a tree because hunger ghost was hitting me every single time (also it stopped my berserk in about 2 attacks).

Could I just have the encumbrance penalty apply directly to AC? That way it's easier to tell if you are too weak to get a defensive boost from a particular armour

What do you mean as too weak? Str does not affect AC bonus, you just lose ability to cast spells and/or avoid being hit. It cannot be replaced with changes to AC.

[ker]

I think I'm answering questions you never asked at this point though...

No, no, that's the sort of answers I was hoping to get! For more context I'm think of this because of the defense of hidden numbers thread. Specifically:

Crawl's math should be transparent enough for players to (a) make intelligent skilling decisions, and (b) know whether a given item or spell is strictly better than another along some given dimension (e.g. raw damage output), without diving into the code or consulting a wiki (or much less using a calculator that someone else has laboriously coded after doing one of these things). Right now, it's not.

[GDR] Yes, it reduces damage spikes. You literally cannot get hit by max monster damage.

Well, yes, I understand what GDR *literally* does. But in *practise* does the reduction in spikes actually matter? Is it even noticeable against the background of spikes in damage?

[AC vs EV] Yes. Being hit with every corrosive attack or poison attack matters a lot. As far as I know those apply even with 0 damage.

So would you wear a different armour to avoid being corroded? Or would you approach a corrosive monster differently if you where wearing heavy/light armour?

[AC enc penalty] How? I remember how I almost starved vs curse skull as a tree because hunger ghost was hitting me every single time.
What do you mean as too weak? Str does not affect AC bonus, you just lose ability to cast spells and/or avoid being hit. It cannot be replaced with changes to AC.

I realise Str doesn't affect AC (right now). I'm suggesting that if AC and EV are both defensive stats you could make Str affect on your defences more obvious if it affected AC (and not EV). Obviously that doesn't work if the drop in EV is significant in its own right, as you are suggesting.

Also, so far no one mentioned encumbrance's effect on accuracy. Should that be a factor in armour choice?

[VeryAngryFelid]

Well, yes, I understand what GDR *literally* does. But in *practise* does the reduction in spikes actually matter? Is it even noticeable against the background of spikes in damage?

It is easy to see with Dr. Personally I was surprised that I got so high damage but then recalled that I have no GDR.

So would you wear a different armour to avoid being corroded? Or would you approach a corrosive monster differently if you where wearing heavy/light armour?

I often switch from heavy armour to middle/light armour for Slime.

I realise Str doesn't affect AC (right now). I'm suggesting that if AC and EV are both defensive stats you could make Str effect on your defences more obvious if it affected AC (and not EV). Obviously that doesn't work if the drop in EV is significant in its own right, as you are suggesting.

Also, so far no one mentioned encumbrance's effect on accuracy. Should that be a factor in armour choice?

EV penalty provides extra difference for species. For example, I switch to the heaviest armour for Tr, Og, Gh while Sp, Ko, Ha, Te and even Mi have better defense in middle or light armour.
Yes, accuracy penalty is pretty significant early on, I pick up +0 plate armour on D1 but don't wear it yet.

[ker]

It is easy to see with Dr. Personally I was surprised that I got so high damage but then recalled that I have no GDR.

Is that because you were comparing X AC with 0% GDR with X AC with Y% GDR? Would the amount of damage you are taking be comparable to X/(1+Y) AC?

I often switch from heavy armour to middle/light armour for Slime.

Fair enough. From a point of view of simplifying the display, I'm not sold that this is a decision point that makes the cost in clarity worth it but I'm not one of the developers so I'll let it go

Yes, accuracy penalty is pretty significant early on, I pick up +0 plate armour on D1 but don't wear it yet.

If you never wear plate armour that you find on D1, how is that different from spawning plate armour later on in the dungeon?

I'm asking all these questions not to be difficult but to think really critically about what information a new, unspoiled player is expected to digest to make a sensible decision about what armour to wear. My hope is that we can find some way to make that decision clearer (without losing depth).

[VeryAngryFelid]

Is that because you were comparing X AC with 0% GDR with X AC with Y% GDR? Would the amount of damage you are taking be comparable to X/(1+Y) AC?

I didn't compare anything, I just got unexpectedly high damage, it was close to max monster damage. No AC can help with that.

If you never wear plate armour that you find on D1, how is that different from spawning plate armour later on in the dungeon?

I was trying to answer your question about accuracy penalty being meaningful by "yes, it is meaningful" and providing a real example of situation when it is meaningful.

My hope is that we can find some way to make that decision clearer (without losing depth).

The decision depends on body size, str, int, dex and spells, I don't think you can replace it all with AC adjustments.

[Siegurt]

So there are two answers you could get, theres the "this is close enogh to the correct answer that even when it is wrong, it is wrong by a small enough margin that it doesn't matter in practice" answer which is as given above "the highest total ev+ ac that gives acceptable failure rates" i would say i would go with more ac in the case of a tie.

Then there's the "close to optimal" answer, "actually optimal" involves a lot of calculations around the maginal effeciveness of a bit of ac or ev, and a whole lot of contextual stuff (where the actual answer will vary depending on exactly what you are fighting) if you don't care to get into that much math (or go dig up large tables) you can probably get away with remembering that ev is most efficient around the hd of the critters you are fighting, and ac is most efficient around the max damage of said critters, that isn't exactly true, but it's close enough for me. Again the "best total protection" is most often the same combination as "most total points" but since optimal ac/ev is worth roughly twice as much as least optimal, it might be worth it to go from say 3>0 ev in exchange for going from 23>25 ac, however if it is, it isnt bt much.

In short, ignore that second paragraph unless you actively enjoy spending time crunching numbers instead of playing the game.

Regarding gdr:

All things being equal, more gdr is better than less gdr, but i would rarely, if ever, sacrifice points of ac for more gdr. Again the effeciveness of gdr is really dependant on what you are fighting, and what your ac already is. It is pretty much never worthwhile to use as part of your decision matrix, because it is only very very rare that it will impact the optimal case, and when it does, it is by a small margin.

[ker]

I didn't compare anything, I just got unexpectedly high damage, it was close to max monster damage. No AC can help with that.

Ok, so if you made monster damage less spiky would GDR still be necessary? Is anything gained by light armour characters getting hit by spiky high damage?

I was trying to answer your question about accuracy penalty being meaningful by "yes, it is meaningful" and providing a real example of situation when it is meaningful.

Cool. I was trying to challenge your definition of meaningful by asking a (hopefully) insightful question.

The decision depends on body size, str, int, dex and spells, I don't think you can replace it all with AC adjustments.

Not once did I suggest that you could replace it all with AC adjustments.

Also, are you suggesting that if we tell the player: your body size is A, Str B, Int C, Dex D and spell failure E....Z, it would be clear to him or her which armour is the best in a given situations? Without looking up the wiki? Colour me sceptical.

[VeryAngryFelid]

Ok, so if you made monster damage less spiky would GDR still be necessary? Is anything gained by light armour characters getting hit by spiky high damage?

A better question would be "is anything gained by heavy armour characters losing access to spells" and the answer is "yes, they get less spiky damage".

Cool. I was trying to challenge your definition of meaningful by asking a (hopefully) insightful question.

Do you think I cannot create an example when you have similar situation on D:6 or where you decide to allow generating +0 plate armour?

Not once did I suggest that you could replace it all with AC adjustments.

That's true, probably you don't have enough details for starting a thread in GDD. It's more like "what's the idea of heavy armour?" and "let's braintstorm how to change it".

Also, are you suggesting that if we tell the player: your body size is A, Str B, Int C, Dex D and spell failure E....Z, it would be clear to him or her which armour is the best in a given situations? Without looking up the wiki? Colour me sceptical.

Wiki won't help here, you should try the armour and see your spells, AC, EV (create wizard mode character if the armour is in shop). It is exciting to die if you are impatient and are trying the armour near unexplored territory, also it is exciting to die while walking to stairs and then you see in morgue something like +13 CPA {MR+, rF+, slay +5}

[goodcoolguy]

What information (values and/or how those values interact with the greater system of crawl) do I need to make an "informed" decision?

Experienced players are often in disagreement and/or wrong about what information is required. In my opinion, this information is not available through the game interface or fsim. Since people are using the word "entropy" on tavern now, I guess would say the right measure is probably related to the average damage and entropy of the damage distributions you get for a dangerous, but plausible encounter. You want low, "predictable" damage.

[ion_frigate]

I wonder if it would be possible to keep a running/rolling total of (damage_taken/damage_directed_at_you), over something like the last 100 or 200 attacks. While it's not the complete picture, it's probably a better number to base a decision off of than anything else.

Of course, there are the related issues of it only giving you the information after you wear the armor for a while, and the fact that this number will vary based on the monsters you're fighting. But both these issues shouldn't be a huge problem, since you'll amass a meaningful sample of monster attacks pretty quickly, and so get an impression of how good your armor is for this particular area pretty fast.

[ker]

That's true, probably you don't have enough details for starting a thread in GDD. It's more like "what's the idea of heavy armour?" and "let's braintstorm how to change it".

The point of the thread seems to be getting away from us so let me restate it.

Context: Crawl, as is, does a shit job of telling you the information you need to make an informed decision about all sorts of things. This is bad. The naive solution is to dump all of that information to the player. This would case analysis paralysis/information overload. This is bad. This is also too big of a problem to solve in one thread. So -
Problem: Crawl does a shit job of telling an unspoiled, new player the information they need to make an informed decision about what armour to wear.
Proposed Solution (as outlined in the second post): Print relevant stat changes when swapping armours (i.e. provide that information when it's relevant) and provide an opportunity to dive deeper and find out what those stats actually mean. Furthermore, (and this might have been too much for one thread) let's have a serious think about wherever all of those stats actually matter *in a practical sense*.

So to be clear - I am not looking for advice on how I should pick my armour. I am looking to discuss UI changes (possibly) backed by changes to the game rule-set to allow a nonspecific player to make more informed decisions.

Do you think I cannot create an example when you have similar situation on D:6 or where you decide to allow generating +0 plate armour?

To put my earlier question in a different way:
Let's say you found +X plate armour after finishing Lair. Is there a character that would say no to it because of the accuracy penalty as opposed to any other affects that encumbrance has? I don't think so. Maybe I'm wrong, and I'm almost certainly playing sub-optimally, but I've never made a decision on wherever to put on a particular armour based on its affect on my accuracy outside maybe the early game (as you rightly pointed out). But even in the early game the decision isn't terribly interesting as I think it's essentially the same thing as spawning plate armour later on in the dungeon. *If* all that is true, then why not remove encumbrance rating affect on accuracy to streamline the UI and make the decision between armours clearer?

A better question would be "is anything gained by heavy armour characters losing access to spells" and the answer is "yes, they get less spiky damage".

Eh, this seems just wrong to me. The thing gained by heavy armour characters, in my experience, is a greater AC+EV total (i.e. more defences). Maybe I'm way off base here...

Anyway, all of that is besides the point because when I was talking about "gain" I meant gain in terms of depth to the game. I'm not sold that light armour characters getting spiked by damage is an interesting facet of the game. This is why I'm suggesting that something can be done to streamline the UI and make the armour decision clearer.

@goodcoolguy and ion_frigate:
That's an interesting idea! Maybe the solution then isn't providing data but giving the player better tools to analyse the effect a decision has had? The morgue files already tell you how often you dodged/blocked/etc. What if that data was presented as an average somewhere on the armour description page?

[Sprucery]

Remember, there is no GDR, only GMDR (guaranteed melee damage reduction).

[VeryAngryFelid]

So to be clear - I am not looking for advice on how I should pick my armour. I am looking to discuss UI changes (possibly) backed by changes to the game rule-set to allow a nonspecific player to make more informed decisions.

I think it won't be implemented, you are supposed to try different armours and see how they work. Which spells to cast is even more important but my patch for displaying chance to hit has not been accepted https://crawl.develz.org/mantis/view.php?id=9437

To put my earlier question in a different way:
Let's say you found +X plate armour after finishing Lair. Is there a character that would say no to it because of the accuracy penalty as opposed to any other affects that encumbrance has? I don't think so. Maybe I'm wrong, and I'm almost certainly playing sub-optimally, but I've never made a decision on wherever to put on a particular armour based on its affect on my accuracy outside maybe the early game (as you rightly pointed out). But even in the early game the decision isn't terribly interesting as I think it's essentially the same thing as spawning plate armour later on in the dungeon. *If* all that is true, then why not remove encumbrance rating affect on accuracy to streamline the UI and make the decision between armours clearer?

Again, I think it won't get implemented. I had a similar thread about weapon accuracy (viewtopic.php?f=8&t=21314) which is even less relevant than accuracy penalty from armour and has no context and yet nothing changed.

Eh, this seems just wrong to me. The thing gained by heavy armour characters, in my experience, is a greater AC+EV total (i.e. more defences). Maybe I'm way off base here...

Less spiky damage is still a nice bonus. I switched from leather to GDA on a stabber of Dith in one of my games because I was tired of retreats after spiky damage.

Anyway, all of that is besides the point because when I was talking about "gain" I meant gain in terms of depth to the game. I'm not sold that light armour characters getting spiked by damage is an interesting facet of the game. This is why I'm suggesting that something can be done to streamline the UI and make the armour decision clearer.

Spiky damage is interesting, devs love to put player in hard situations (see traps).