183 réponses à ce sujet

[soteria]

Sylvius wrote...

Isabela is a dagger user. Present tense.

So that's what she is when we get her. Can we change that going forward?

It doesn't sound like it from my understanding. I copied a partial transcript of the podcast here. This is the relevant part:

Weaponry though, I need to be quite clear, weaponry can be changed out. It’s important to note for followers that the followers will stick to the weapons that they prefer. Isabela is a duelist at heart, she was in Origins, she taught you how to be one, she would teach you how to fight. She’s not exactly what you’d call a “bow girl.” She does have her preference for her daggers, but beyond that, it’s really up to you. So if you find a particular set you like, a set you’ve enchanted or you think are particularly good, you’re able to swap those in and out so you’re able to change the model, change the appearance, that kind of thing from a weapon standpoint.

I don't know how to read "you can spec her as a bow user" from that. I don't see any point in saying "she's not a bow girl," "followers will stick with weapons they prefer" when in fact you can have her use whatever weapon you like. When asked about companion talent trees, Mr. Laidlaw said (my emphasis), "they get the majority of the talents from their base class" and a unique tree.

I don't want to be doom and gloom, but the most optimistic reading I get from that is we can have her equip a bow but not actually develop archery talents.

[In Exile]

tmp7704 wrote...
To use your analogy, if the friend that tells you not to wear blue is fashion designer and you don't know first thing about fashion, then it's possible you'll just accept their advice without delving into why's. And as long as there actually are valid reasons not to wear blue then hey, you got useful advice.

Sylvius would say that unless you can understand it, you can't take it as sensible advice. At least I think.

So the friend could be a world renown authority; Sylvius would never believe you have a good reason to listen unless you understand the reason for the advice.

While yes, knowing how the game calculates the ratings can make it overally more useful (especially for advanced player) it doesn't exclude the mechanics being useful without that, like Sylvius has stated.

I let my opinion bleed in for a second there. Sylvius honestly believes it's all or nothing, as I understand. Just wanted to point out his attitude, given how idiosyncratic it is. This way we can ideally avoid several pages? Anyway, I'm really going to leave it to the two of you to resolve; I should but out and not speak for a third person.

[Sylvius the Mad]

tmp7704 wrote...

It's "suggest" because the game doesn't know if player isn't using their own, custom criteria to determine whether item is an upgrade for them -- you brought up example of that yourself earlier, with theoretical situation of player who is looking for armour with specific attribute.

It doesn't make the feature not useful, merely not always useful.

but each time I make the decision, what's useful is whether the star-rating is relevant right now.  Whether it's usually relevant uis only valuable if I'm willing accept an error rate above zero.  Then I can just make educated guesses and be content.

But if my maximum accepted error raet is zero (and it usually is) then I need to know - each time I ask - what the star rating means in this particular circumstance.

If the game doesn't answer that, then the star-rating is just an error-generator.

[Sylvius the Mad]

soteria wrote...

I don't know how to read "you can spec her as a bow user" from that. I don't see any point in saying "she's not a bow girl," "followers will stick with weapons they prefer" when in fact you can have her use whatever weapon you like.

Neither do I.  But maybe Mike does.

I don't want to be doom and gloom, but the most optimistic reading I get from that is we can have her equip a bow but not actually develop archery talents.

The only sensible reading I see is that we just don't know the answer yet.

[Sir JK]

Sylvius the Mad wrote...
but each time I make the decision, what's useful is whether the star-rating is relevant right now.  Whether it's usually relevant uis only valuable if I'm willing accept an error rate above zero.  Then I can just make educated guesses and be content.

But if my maximum accepted error raet is zero (and it usually is) then I need to know - each time I ask - what the star rating means in this particular circumstance.

If the game doesn't answer that, then the star-rating is just an error-generator.

But the star system is not designed to tell you if something is good. It's just designed to be a quick reference to what items are low ranked for your level. So you can asses your inventory at a glance and then go for an in depth study.
If you're carrying a level 5 ring at lvl 15 because you think it fits perfect in the build, then the star system will still mark it as low levelled. In that case the stars aren't very useful because you think the ring is worth it. But if you're also accidentally carrying a low leveled belt that you've simply forgotten to replace, then it'd be very useful to see that suddenly it haven't got any stars any more.

This is essentially an ease-of-access feature. No more. No less.

[upsettingshorts]

I used the Football Manager example for a reason. A 2.5/5 star player can seem like a weak link and not be, but the information to discover if he actually is or not remains. The star ratings just provide an at-a-glance way to figure out where you should start probing for weaknesses or building around strengths. It doesn't replace the core process of that evaluation if you don't want it to.

[Sylvius the Mad]

Sir JK wrote...

But the star system is not designed to tell you if something is good. It's just designed to be a quick reference to what items are low ranked for your level.

Which, as you just pointed out, has no gameplay relevance.

[Sir JK]

Sylvius the Mad wrote...
Which, as you just pointed out, has no gameplay relevance.

Yes it does, it has a ease-of-access relevance. It helps players at a glance assess their inventory. Kind of like labelling of files with paperwork. It helps you at a glance evaluate the item and decide if yo ought to look a bit more closely at it.
Think of it this way: I open up my inventory to check something and notice that my belt now only have 2 stars. Indicating that it's time to start looking for upgrades since it's not as good it could be.

Now, if you're the kind of player that regularly examines your inventory in detail, then no it's not very useful. If you're not however, and there are plenty of players like that, then it's a very useful feature to see what items are below the designed level range and which ones aren't.

Modifié par Sir JK, 08 novembre 2010 - 05:27 .

[tmp7704]

Sylvius the Mad wrote...

but each time I make the decision, what's useful is whether the star-rating is relevant right now.  Whether it's usually relevant uis only valuable if I'm willing accept an error rate above zero.  Then I can just make educated guesses and be content.

But if my maximum accepted error raet is zero (and it usually is) then I need to know - each time I ask - what the star rating means in this particular circumstance.

If the game doesn't answer that, then the star-rating is just an error-generator.

Out of curiosity and because it seems relevant -- what is your view on the concept like WoW Item Level?

It's basically very similar thing -- a number of stats and possible attributes attached to the gear are assigned values, and a single number is calculated from that as a result.  Then certain brackets of item levels are associated with player levels, to determine what sort of gear can be considered "optimal" for player at given level.

The main thing i'm curious about it, does knowing the details behind the formula used to calculate the item level allow you to accept any easier on the face value that item with level of 213 has generally better stats compared to item with level of 200? Even though due to complexity of the formula you aren't going to derive any useful information about individual attributes from that single item level number, and you'll have to compare tooltips of items in question to determine exact differences between them?

[Sylvius the Mad]

tmp7704 wrote...

Out of curiosity and because it seems relevant -- what is your view on the concept like WoW Item Level?

Until I followed that link, I was entirely unfamiliar with it.

It's basically very similar thing -- a number of stats and possible attributes attached to the gear are assigned values, and a single number is calculated from that as a result.  Then certain brackets of item levels are associated with player levels, to determine what sort of gear can be considered "optimal" for player at given level.

The main thing i'm curious about it, does knowing the details behind the formula used to calculate the item level allow you to accept any easier on the face value that item with level of 213 has generally better stats compared to item with level of 200? Even though due to complexity of the formula you aren't going to derive any useful information about individual attributes from that single item level number, and you'll have to compare tooltips of items in question to determine exact differences between them?

Not having the formula makes the resultant single-value exactly worthless.  Having the forumla allows you to draw a variety of demonstrably true conclusions about the item based on the single-value number.  The single-value still wouldn't be as informative as the complete list of stats, but at least knowing how it was calculated would preclude some possible combinations of stat values.

[Sylvius the Mad]

Sir JK wrote...

Sylvius the Mad wrote...
Which, as you just pointed out, has no gameplay relevance.

Yes it does, it has a ease-of-access relevance. It helps players at a glance assess their inventory.

Only if they're willing to accept errors.

What do you mean by "assess"?  If we can't derive any specific information about the item from the star-rating, then the star-rating alone allows zero reasonable conclusions.  How then could anyone possibly use that rating to "assess" the contents of his inventory?

Think of it this way: I open up my inventory to check something and notice that my belt now only have 2 stars. Indicating that it's time to start looking for upgrades since it's not as good it could be.

It's not as good as it could be according to the formula that creates the star-rating (which might have no relevance for you - you can't know if it does unless you have that formula), assuming that there's better equipment available (again, "better" according to the formula).

Do you really not see how many implicit assumptions you're making?

[upsettingshorts]

Sylvius the Mad wrote...

What do you mean by "assess"?  If we can't derive any specific information about the item from the star-rating, then the star-rating alone allows zero reasonable conclusions.  How then could anyone possibly use that rating to "assess" the contents of his inventory?

It's directive, not informative.  The stars are meant to draw your attention, not inform upon your eventual conclusion regarding their worth.  

Modifié par Upsettingshorts, 08 novembre 2010 - 06:33 .

[Sylvius the Mad]

Upsettingshorts wrote...

It's directive.  Not informative.  The stars are meant to draw your attention

But if we don't know what they mean, they can't draw our attention.  There's no reason for us to think that 2 stars or 3 stars means anything.

[upsettingshorts]

Sylvius the Mad wrote...
But if we don't know what they mean, they can't draw our attention.  There's no reason for us to think that 2 stars or 3 stars means anything.

It means "you should look at this" relative to your level.  A two-star item draws your attention to its potential inadequacy, a four-star item draws your attention to its potential superiority.  It's still up to the player to determine, in each specific case, the particular usefulness of the item in question. 

The precise relation between star value and relative levels I imagine will be contained within the game's documentation. 

Modifié par Upsettingshorts, 08 novembre 2010 - 06:37 .

[Sylvius the Mad]

Upsettingshorts wrote...

The precise relation between star value and relative levels I imagine will be contained within the game's documentation.

We can hope.

Modifié par Sylvius the Mad, 08 novembre 2010 - 06:37 .

[Sir JK]

Sylvius the Mad wrote...
Only if they're willing to accept errors.

What do you mean by "assess"?  If we can't derive any specific information about the item from the star-rating, then the star-rating alone allows zero reasonable conclusions.  How then could anyone possibly use that rating to "assess" the contents of his inventory?

Assess as in at a glance evaluation form which I can decide wether to investigate something in depth or not.

The only thing this system does is tell me: More than 3 stars - designed for this level range. Less than 3 stars - Not designed for this level range. As such it has no error margin at all.
The decision wether the item is good, needs to be changed soon or ought to be replaced immediatly is one I should do based on the stats, not the star value.

It's not as good as it could be according to the formula that creates the star-rating (which might have no relevance for you - you can't know if it does unless you have that formula), assuming that there's better equipment available (again, "better" according to the formula).
Do you really not see how many implicit assumptions you're making?

It is no different from the colour coding certain games does, or adding fancy names to indicate that they're "magical" (thus "better"). Nothing in those systems can tell you if an item is good for your build or not. It can however indicate that a new item is available that -might- be better.
All the stars do is tell you if the item was designed for your level or not. Quickly and easily. The actual evaluation of them is something you should do yourself. But the stars can help you pinpoint which ones might need evaluating soon or just help you keep track of your items.
It doesn't need to give you objective detailed information since that information will be available on a mouseover. It's not something you should base your equipment decisions on. Just something quick and easy to assist you a bit if you need it.

[Meltemph]

Sylvius the Mad wrote...

Upsettingshorts wrote...

The precise relation between star value and relative levels I imagine will be contained within the game's documentation.

We can hope.

It would be weird not to assume that... Specially if the intent is to get people not used to RPG's to start paying attention to that sort of thing.  Kinda a "Hey I got a 4 star weapon and my current one is a 2", it would not be hard to assume someone unfamiliar with certain aspects of RPG's would think that the 4 star is better.  Then if they use it and it ends up not being quite as good in certain situations, then it would not be hard to imagine that the person will try and figure out why.

All they(developers) can do is use supposition, and how they think they can best deal with situations and they have to do it, with all types of people in mind.  Also, I think it would be safe to say, most people don't defy things that don't mesh perfectly with ones personality or way of thinking quite like you, so you have to look at it from a "newbies" perspective.

[upsettingshorts]

Sylvius the Mad wrote...
I highly doubt that.  BioWare seems to go out of their way to avoid including equations anywhere in documentation.

I think it's as simple as a question of accuracy versus precision. 

The star ratings strike me as meant to be accurate, "This item is potentially not good enough to use anymore relative to your current level, have a look." So if at level (X), the game is balanced with the expectation that characters are using equipment with a value of (X +/- 5), it will display items that qualify with an appropriate amount of stars. 

Precision, something I think you're hinting at by raising the issue of the equation behind the mechanic, is not the goal of the ratings.  Precise data on the usefulness of the item is revealed in the item's traditional stats.

Still, in acknowledgement of your edit, I would not be against them revealing details of the equation.

Modifié par Upsettingshorts, 08 novembre 2010 - 06:50 .

[AlanC9]

soteria wrote...
I don't know how to read "you can spec her as a bow user" from that. I don't see any point in saying "she's not a bow girl," "followers will stick with weapons they prefer" when in fact you can have her use whatever weapon you like. When asked about companion talent trees, Mr. Laidlaw said (my emphasis), "they get the majority of the talents from their base class" and a unique tree.
I don't want to be doom and gloom, but the most optimistic reading I get from that is we can have her equip a bow but not actually develop archery talents.

Since rogues are bow-eligible this implies an actual lockout on all but one talent tree. As opposed to DAO where it was possible but typically not worth doing.

[Sylvius the Mad]

Upsettingshorts wrote...

Sylvius the Mad wrote...
I highly doubt that.  BioWare seems to go out of their way to avoid including equations anywhere in documentation.

I think it's as simple as a question of accuracy versus precision. 

The star ratings strike me as meant to be accurate, "This item is potentially not good enough to use anymore relative to your current level, have a look."

I find it interesting that this is always the example offered.  Because you're right, if the star-value is low it will create incentives to look at the stats more closely.

But if the star-value is high, it does the opposite.  A 4-5 star item will likely be assumed by the player to be at least adequate, and that might be untrue.  Players might also seek out items based on their star-value alone.

[upsettingshorts]

Sylvius the Mad wrote...
I find it interesting that this is always the example offered.  Because you're right, if the star-value is low it will create incentives to look at the stats more closely.

But if the star-value is high, it does the opposite.  A 4-5 star item will likely be assumed by the player to be at least adequate, and that might be untrue.  Players might also seek out items based on their star-value alone.

The two-star examples, if dealt with improperly, will also lead to players who are not curious to discard still useful items in the same way that four-star examples might lead to players presupposing the item's superiority.  I believe this is intentional, as it fits in with the stated development goal of making the game easy to learn but (just as) hard to master. 

A new player could, given how simple the idea of stars is to understand at a glance, potentially play the game basing their entire inventory management style on it, especially if they are based on the idea that level (X) toons are assumed to be having gear evaluated around (X +/- 5) for the purpose of balance.  That makes it easy to learn.

What makes it (just as) hard to master, for those of us used to RPG mechanics, is learning to integrate the new way of displaying potential value with our usual thought process of evaluation.  In that way the star mechanics may help guide veteran players as they evaluate old gear that might be obsolete, or new gear that might be an upgrade.  Mastering the mechanics of the system is the same as ever - and we'll know that even if the stars contradict our evaluation of an item, our experience with the system gives us the confidence to make the choice that suits us.

New players to a genre don't make choices with confidence, so they're going to be approaching the stars differently than you or I.

Modifié par Upsettingshorts, 08 novembre 2010 - 07:03 .

[Sylvius the Mad]

Upsettingshorts wrote...

A new player could, given how simple the idea of stars is to understand at a glance, potentially play the game basing their entire inventory management style on it, especially if they are based on the idea that level (X) toons are assumed to be having gear evaluated around (X +/- 5) for the purpose of balance.  That makes it easy to learn.

Exactly.  But if the star-values don't actually contain any information, then the players aren't actually learning anything.

New players to a genre don't make choices with confidence

This is not my experience.  In my experience, most people do make choices with confidence, even when that confidence is not justified.  They'll simply base those decisions on the most obvious information available, and should that information prove to be inaccurate they'll get frustrated and angry.

[upsettingshorts]

We can trade predictions of how most players will react, I'm just not sure how we'll ever be able to prove that we're right. Certainly not until after the game is released, and probably not even then.

I think I see what they're trying to do, and that's enough for me.

Modifié par Upsettingshorts, 08 novembre 2010 - 07:15 .

[Sylvius the Mad]

Upsettingshorts wrote...

I think I see what they're trying to do, and that's enough for me.

I'll agree that their objective appears to be sound.

I just can't think of an example where such a thing ultimately worked out well.

[upsettingshorts]

Sylvius the Mad wrote...
I'll agree that their objective appears to be sound.

I just can't think of an example where such a thing ultimately worked out well.

Well and this is slightly off topic, Football Manager's system seems to work - but the way in which it is different sets it apart in that the star rating evaluations are subjective in the sense they are based on what is more or less the evaluation of an ingame NPC with his own stats.

Your assistant manager is part of your staff and can be replaced.  You can see how good he is at judging talent and using his evaluations can be done in-character (as the team manager).  Combined with scouting reports there are a great deal of players who approach FM in precisely this in-character way - despite OOC ways of evaluating talent, so the fallibility of the assistant is of tremendous importance to their approach to gameplay.

The goal of objectivity means that DA:2's system will ultimately be judged differently, especially since if I don't like how star ratings are being judged I can't simply replace the entity that's giving them to me with something I think will be better or more precise - like I can in FM.

Modifié par Upsettingshorts, 08 novembre 2010 - 07:42 .