Planned Feature Replace "Luck Skill" with "Sneak"

[Pieter Boelen]

Of course "being lucky" is not "a skill that you can improve". Either you have luck, or you don't. It is purely random chance.
So having a skill named "Luck" makes basically no sense.

In the code it is actually called "SKILL_SNEAK" which make a whole lot more sense.
After all, "being sneaky" IS something that you could be bad or good at.

In some parts of the code, this skill is actually used as "Sneak", but in others it quite literally functions as "Luck".
I would like to see this changes so that ONLY the "Sneak" uses remain.
The "Luck" uses should probably be eliminated altogether and/or be replaced by pure random chance.
And of course the displayed name in the interfaces should be changed to match.

I wonder if someone might be interested in searching the PROGRAM folder for "SKILL_SNEAK" to find the spots that may need tweaking?
Perhaps this is something @pedrwyth might be interested in looking into?

 

[Grey Roger]

There are at least three types of skill currently covered by "Luck". Sneaking - the chance of being caught by a guard, who if you're lucky doesn't happen to turn his head in your direction just as you're making your move. Gambling - tying that to sneaking would basically mean you're cheating, as opposed to knowing the odds, reading your opponent's face and looking for a tell, remembering which cards have already been played. Opening chests - you don't need to sneak up on a chest but you do want to be able to open its lock, then be lucky enough to find something valuable inside. All of those could perhaps be replaced by skills if there's room to replace one general "Luck" skill with separate "Sneak", "Gambling" and "Lockpicking" skills, along with whatever else "Luck" is used for in various storylines.

 

[Pieter Boelen]

There are at least three types of skill currently covered by "Luck". Sneaking - the chance of being caught by a guard, who if you're lucky doesn't happen to turn his head in your direction just as you're making your move. Gambling - tying that to sneaking would basically mean you're cheating, as opposed to knowing the odds, reading your opponent's face and looking for a tell, remembering which cards have already been played. Opening chests - you don't need to sneak up on a chest but you do want to be able to open its lock, then be lucky enough to find something valuable inside. All of those could perhaps be replaced by skills if there's room to replace one general "Luck" skill with separate "Sneak", "Gambling" and "Lockpicking" skills, along with whatever else "Luck" is used for in various storylines.

Hmm... valid points. Doesn't simplify things though...
Creating three skills instead of one is definitely beyond the scope of what I had in mind; that would require far more work.

My main point is that the skill named luck should NOT be "Luck", because that doesn't make sense as a skill.
I have seen it used in some parts of the code where it REALLY serves that purpose, such as in the "chance that your block is succesful" and such.
By my reckoning, SKILL_SNEAK should have nothing to do with that. That is either "Defence" or "Fencing" (or both?).

I do think Lockpicking goes together with Sneak quite fine; they're both some sort of "thief's skills".
Not sure how to handle "Gambling" though. My original thought was to replace that with 100% chance, so that to consistently win at gambling, you'd have to be lucky for REAL.
But for some games, such as Poker, the factors that you mention do, of course, very much apply.
And those are things you indeed COULD get better at. But for "rolling dice", there should be no skill involved. Right?

Also, who says that beating the odds on gambling doesn't involve cheating?
Especially when you start consistently winning at "Dice", there must be some foul play involved on the player's part, no?
Actual skill can only bring you so far....

 

[Grey Roger]

I agree that "Luck" / "Sneak" / whatever else it's to be called has nothing to do with blocking, that should indeed be down to your "Melee" skill. As for "Defence", isn't that the one you or your surgeon use to resist ship and crew damage from cannonfire?

Perhaps you've noticed that the dice aren't exactly even and if you start them with 3 upwards, they're more likely to end up 6. It's no more cheating than the person who supplied the dodgy dice. But the point about dice is that it's a quick way to play the gambling in PoTC, which is why I tend to prefer it to the card games - get it over with, collect the money, earn the skill, and most importantly get the "Girl Won in a Card Game" quest. Once I've done the quest I don't bother any more.

In reality, if there are dice involved, the game of choice is probably going to be Liar's Dice, which certainly does involve reading your opponent more than reading the dice. (I've played it for real - not for money, just for fun! There's one person who, some of the time, doesn't even look at his dice before bidding, and still does well.) Beating the odds on gambling doesn't have to involve cheating, otherwise the famous faces you routinely see on TV poker would be banned from every casino in the US and Europe.

 

[Black Bart]

Some people are just lucky, so there is that Is it an actual 'stat' or just some magic beans/fairy dust we get sprinkled with as babes? No idea, but certainly i've 'experienced' those that are more lucky than others, and a unlucky person, especially in the context of history and this period of Age of Sail, was seriously shunned by others. You didn't want an unlucky sailor (or Captain!) on your ship!

In the context of the game, do we know exactly how much influence 'Luck' has on events (like blocking that Grey Roger mentioned), maybe we can tone it done massively for those kind of 'skills' and less so for things like gambling etc?

 

[Pieter Boelen]

Searching the Program folder will give us the answers to your question.
It just requires someone who has the will and time to that.
But it isn't technically hard to find out.

 

[pedrwyth]

I'm not personally against the concept (in a game) of innate "luck" giving a character better than normal chance odds of a favourable event (nor indeed its converse of a character who is consistently unlucky). I agree that such a factor is not a skill to improve BUT could be placed in appropriate situations. Following the ideas of differentiation expressed above I would suggest 3 sort of elements say subterfuge, acuity and innate luck. only the first two of which can be enhanced by practice and experience (unless a "lucky" talisman remains to be found).

Some activities may benefit from more than one element. Thus sneaking past someone would involve a combination of subterfuge and acuity. Pretending to be who/what you are not when ashore (as currently controlled by your flag) is mainly subterfuge. Leaving innate luck as a concept would allow it to remain for say - blocking - but those that wished could reduce its impact to zero if that is their real world bubble. Gambling could be helped by acuity and innate luck. Lockpicking is a direct skill which might be helped by better acuity to "sense" the levers but what you find inside is pure chance (thus tempered by innate luck).

Subterfuge would be partially linked to the master of disguise perk, acuity ruined by substance abuse (be that alcohol or whatever) etc etc.

Different starts could have different innate luck levels (eg luck of the Irish but a castaway is down on his luck I suspect), but if the effects are all settings in internal settings all could be individually tweaked to taste.

As @Pieter Boelen suggests the first step is to search for sneak as defined and try and separate out the elements that should /could be in play at each point. I will take a look through the program folder to make a list to allow discussion on which elements (or none) should be applied to which.

 

[Pieter Boelen]

Thanks, @pedrwyth!

While I'm not opposed to splitting it, I do fear that could make what was meant to be a relatively small change a REALLY large one.
Would other skills then need splitting as well?

Your suggestion of linking it to abilities does sound very promising!
If indeed the main SKILL by default serves only one of the purposes, maybe the separate abilities (Disguise/Gambling) could serve the others.
Or those abilities could unlock the "Sneak" skill for the other uses as well?

 

[Black Bart]

I'd think it 'easier' to just find where and exactly by how much the current system of Luck influences things like blocking (where real world skill is the preferred outcome) and tone those instances down to the most minimal of effect, while things like gambling could be left as is (again IF the influence of Luck can be found on them).

Rather than add complexity (extra code) you can hone the current system to better reflect the desired outcome/influence of Luck in the game code perhaps? I would say that in general i do not notice the effect of my Luck skill in my game, so it is not obvious (which is a good thing) how much it is influencing the outcomes of various situations. Having said that i don't think i've ever had a char with more than 5 or 6 in Luck (Officers have had more, i think i even had an Officer with 10 luck at some point).

 

[Pieter Boelen]

I'd think it 'easier' to just find where and exactly by how much the current system of Luck influences things like blocking (where real world skill is the preferred outcome) and tone those instances down to the most minimal of effect, while things like gambling could be left as is (again IF the influence of Luck can be found on them).

Finding them is easy. Windows Search Index & Notepad++ to search for SKILL_SNEAK will tell us everything we need to know.
See here how how that works: Tutorial - Modding Tips & Tricks | PiratesAhoy!

Rather than add complexity (extra code) you can hone the current system to better reflect the desired outcome/influence of Luck in the game code perhaps? I would say that in general i do not notice the effect of my Luck skill in my game, so it is not obvious (which is a good thing) how much it is influencing the outcomes of various situations. Having said that i don't think i've ever had a char with more than 5 or 6 in Luck (Officers have had more, i think i even had an Officer with 10 luck at some point).

I wasn't imagining to have extra complexity added; just some simple "removing SKILL_SNEAK" where it doesn't belong (in swordfights, for example).
And possibly replacing some uses with abilities that already exist. Maybe....

 

[pedrwyth]

Easiest is leaving it all as is because it is pretty much invisible in its usage so in that sense isn't "broken", however we are looking at a "planned feature" (according to the thread flasher).

So I'm part way through search and examination of both SKILL_SNEAK and "sneak" called as the skill (and listing type of use, where skill is enhanced etc).

I would intend to replace those that are a "luck" element with a new define, set for starters at the same as starting SKILL_SNEAK so any change has initially no effect - except you can't increase your luck through experience. It can then be tweaked to individual taste (I can envisage a lack of agreement in design by committee as to which elements should have a chance modifier and which shouldn't) - perhaps we could end up with a small suite of luck elements but now this/these new define(s) will be static (unless we have "lucky" items). Some cases might be candidates for changing to an ability (existing) but still the first step is sorting through the occurrences because they are opaque in making their various impacts unless you read the code.

 

[Pieter Boelen]

Easiest is leaving it all as is because it is pretty much invisible in its usage so in that sense isn't "broken", however we are looking at a "planned feature" (according to the thread flasher).

Indeed this is no high priority issue and there is no pressure on seeing it done.
Just something that would be nice, since "Luck" as a skill doesn't make so much sense to @Armada and myself.

I would intend to replace those that are a "luck" element with a new define, set for starters at the same as starting SKILL_SNEAK so any change has initially no effect - except you can't increase your luck through experience.

I'm afraid I don't quite understand you there. You mean remove the skill influence and replacing it with a fixed value?
That might work. For starters, I would expect any TRUE luck to be set to "a chance of 50%" or so.
Then you'd have to be genuinely lucky to get the benefit there, right?

So I'm part way through search and examination of both SKILL_SNEAK and "sneak" called as the skill (and listing type of use, where skill is enhanced etc).
[...]
Some cases might be candidates for changing to an ability (existing) but still the first step is sorting through the occurrences because they are opaque in making their various impacts unless you read the code.

Thank you very much for having a look into this one!

 

[pedrwyth]

Here is the result of my initial trawl through PROGRAM folder. It may not be very clear but each line is a file that uses sneak in some way (there are probably others) and a very cryptic summary of how it may be being used. I have made a rough stab at beginning to split out the types of use. There are, of course, also many files where particular characters get a level set for sneak for whatever general or quest purpose.

Anyway food for thought/ further discussion

 

[Pieter Boelen]

Here is the result of my initial trawl through PROGRAM folder. It may not be very clear but each line is a file that uses sneak in some way (there are probably others) and a very cryptic summary of how it may be being used. I have made a rough stab at beginning to split out the types of use. There are, of course, also many files where particular characters get a level set for sneak for whatever general or quest purpose.

Anyway food for thought/ further discussion

That's an awesome reference, that is!

 

[Tingyun]

Since it was brought up in today's discussion, I read over the discussion here, and looked into all the uses of luck in the current game, and I have a few thoughts:

1) Transitioning luck away from magic to a real skill type effect makes sense, but
2) If luck truly became "sneak", it might have too limited a role. Anyone can pick a lock with enough attempts, and how often do players need to pick a pocket later on?

What about a broader planning of luck as a skill: Agility, instead of Sneak? It would of course encompass all the roguish sneaking abilities, but it could have broader effects in concept, including encompassing current luck combat influences.

For example, in the current game, damage deal by the player is heavily influenced by the player's luck skill. Just taking guns as an example:

dmg *= (0.75 + fRand(CalcCharacterSkill(attack, SKILL_SNEAK))/(40.0/3.0));

With a ten luck skill, on average the player will have a 1.125 modifier here, with a 1 in luck .7875, meaning a character with a 10 in luck on average does 43% more damage with a gun than someone with a 1 in luck. Similar formula for blade damage

43% more damage is too much if we think of it as a magical skill, but if it is everything that goes into agility, hand eye coordination, precision of movement, it makes sense.

Last week, I proposed transitioning the armor coverage modifier from luck to fencing: Unrealistic combat influences in the internal settings | Page 2 | PiratesAhoy!

If we want to think of luck as agility including in-combat type effects, then maybe we should reverse that change. On the other hand, if we don't like the idea of luck as broader agility, then the current influence on combat damage is too high, and should be transitioned to another skill.

If it were just a matter of picking one and everything would work, I'd personally say luck as agility makes sense, and fits well with its pickpocket picklocks roles. Most rpgs have rogue type characters have a primary attribute relevant to combat and based on agility.

But that has implications of its own--luck would then become the sort of skill that warriors have defined as their skill importances in officer types, instead of now, where it is more merchants/peaceful types. And they would have to be able to grow in the skill through combat. And maybe we don't want that, given luck's other roles.

Maybe it is cleaner to also remove the combat damage effect, now that the defensive effect is gone. It really turns on whether we think luck could retain enough relevance if focussed on stealth actions or its other roles and removed from personal combat.

But it makes sense to pick one path or the other on that, because we are in a halfway state right now (luck critically important to damage but recently removed from defensive role, and also missing on warrior characters--master at arms, bosun, guards, etc, all have 0 luck importance, and don't have skill growth in it, and their combat damage dealing ability suffers a bit for it, missing out on a very significant 40% bonus to damage).

Neither path is too hard to implement once we have picked one, and I can handle the work. But I would be happy with either path--both have good reasons behind them. Which do people think, luck as broader agility including combat, or instead finish the job of removing it from combat by transitioning its approximately 40% damage boost role to the current combat skills?

 

[Pieter Boelen]

If Luck becomes Agility, what is the difference with Fencing?
And would there be any skill left to affect, say, Tavern games?

I personally like the idea of Luck being something to help you when doing sneaking in enemy towns and undercover missions and the like.

 

[Grey Roger]

Short of defining separate skills, "Sneak" has to cover lockpicking, actual sneaking, gambling, and any other situation in which luck and/or sneaking is involved. As I previously said, gambling involves skill as well as luck, and not necessarily for cheating.

Personally, I'm quite happy with it as it is. Perhaps reduce but do not entirely eliminate it as a factor for damage and defence.

If you're planning to rearrange the skill importances in officer types, try to keep them reasonably level with each other. At the moment, I believe it is the Surgeon who gets Sneak because he doesn't get much else compared to other types. (Which means he's getting his combat bonuses by knowing where on the opponent to hit for maximum effect, while the boatswain gets his bonus by better knowing how to score the hit in the first place.)

Of course, if "Shared XP" hadn't been weakened, that would provide a way for all your officers, including the combat types, to get "Sneak" increased and thereby gain the combat bonuses which it provides.

 

[Pieter Boelen]

Of course, if "Shared XP" hadn't been weakened, that would provide a way for all your officers, including the combat types, to get "Sneak" increased and thereby gain the combat bonuses which it provides.

Since when has Shared XP been weakened? NOTHING HAS BEEN DONE YET!
Like... at all.

 

[Tingyun]

Pieter, that makes sense. Also, it would be much easier to remove it from combat rather than redifine how AI roles work.

Grey Roger, I don't think it will ever be possible to get things properly balanced with luck as a combat skill (unless warrior officertypes got it, which would water it down as a covert skill, so is only appropriate if we went fully in that direction).

Here is the current weirdness about combat balance:

A) The AI opponents we want the toughest in melee have 0 luck importance, so they get a damage reduction.

B) Meanwhile, the player does advance in luck. Result is the player gets a damage boost, AI doesn't.

Which operates in opposite direction with

C) the player gets a difficulty adjustment reducing their damage, increasing AI damage.

Which in turn doubles down with

D) difficulty dependent HP increases for enemies, and difficulty dependent Hp decreases for player officers

Those are all essentially effects of the HP and damage done, and they are being pulled in different directions with no sense to it.

Balancing is currently a mess like at across the board, and (the worst part) completely opaque to the player.

How many players realize the best companion warrior they could possiblly develop would be hiring a doctor, and then immediately changing him to a bosun for the HP boost? And why would that possiblly make sense? The idea of "he knows anatomy and knows where to hit" is Hollywood/tv stuff, pretty sure any swordsmen knows better than a doctor where and how to kill someone with a sword.

Moreover, the results are very strange to the player. I remember on switching to swashbuckler, wondering why suddenly I was doing less damage. Also, when I recently started a new freeplay with a high luck character, I was shocked at how much damage I was doing.

Meanwhile, in game, we display the remaining HP of the enemy to the player, as well as damage of weapons, but we never tell them we are fudging those numbers in 3 different directions.

What I am trying to do long term in my work for the general mod is to bring the general balancing in line with 3 principles:

Principle 1) never achieve the same effect with 3 different mechanics. If you want enemies to be able to take more damage at higher difficulty levels, don't first increase their HP then decrease the damage they take then have a luck monopoly to increase player damage once again--just raise their HP the appropriate level.

Principle 2) make mechanics clear and understandable to the player. Don't show them sword damage numbers and enemy hp and then render those only loosely related to anything in game because they are being modified by skills the player isn't told has anything to do with combat.

Principle 3) make sure the mechanics are set up in a way consistent with the way the AI is set to play. If luck is a ccritical part of damage dealing, and the AI types that are supposed to be warriors don't get it, but the peaceful merchants and such do, then that is a problem (and the simplest solution is to replace it in direct combat)

This is a wonderful mod, and such incredible fun, but on those three principles the one limited aspect of personal combat is a mess. I think with some conservative and limited substitutions and transitions, we can preserve balance while solving these issues.

However, I will admit that replacing luck with combat skills in damage effects would be a difficulty increase, because AI warrior lacking luck was essentially a cheat in favor of the player previously. There is a broader project here of replace luck-remove difficulty depedent damage-adjust enemy hp difficulty bonuses that would be configurable to maintain balance while solving all the issue. I'm going to be enagaging with that project in my experimental version anyway, I'm just wondering how much of it to do there and how much to do directly for the main mod (because my preference is to always do work for the main mod if possible, and not just go off on my own path).

 

[Grey Roger]

This might be a good point to advise not to concentrate too much on balance and realism. Realistically, a medium size ship such as a frigate or large xebec could never hope to board and capture a large man of war. If you achieve a perfect balance between players and NPC's then this may end up being the case in the game, too. And that would wreck the finale battle in "Tales of a Sea Hawk" as well as the mission to capture La Couronne in "Assassin".

At the moment combat may not look pretty to a programmer but it works and is fun.