Random Stat Implementation

[Viktoriusiii]

Okay, im home now and here is what is under the hood.

 

Pistol code from items.xml

 

So if I read that right (and i might be wrong as I never changed much in the xml) I was right... wasn't I?

8-20% variation on both ends.

Right? Otherwise please explain. BC that is what I am reading out of this.

[Laz Man]

So if I read that right (and i might be wrong as I never changed much in the xml) I was right... wasn't I?

8-20% variation on both ends.

Right? Otherwise please explain. BC that is what I am reading out of this.

 

Not sure what you previously stated but here is an example.

 

Let's say a pistol has a base damage of 100 (hypothetical). Based on these calculations a t1 pistol could have a damage value from 85 to 115.

 

Because there is a specified bonus for higher tiered versions an extra 10 to 50 more damage is possible for t2 to t6 pistols. Therefore a t2 to t6 pistols could have a damage spread of 95 to 165. There is nothing that indicates higher tiers have a higher chance at higher damage rolls when compared to lower tiers (except t1).

 

If this is true, it is theoretically possible to get a max damage roll on a t2/orange tier item.

 

So my takeaway is dont be so fast to scrap/sell t2 to t5 items unless all you care about is having the most mod slots possible.

[Psychodabble]

So if I read that right (and i might be wrong as I never changed much in the xml) I was right... wasn't I?

8-20% variation on both ends.

Right? Otherwise please explain. BC that is what I am reading out of this.

 

Not sure what you previously stated but here is an example.

 

Let's say a pistol has a base damage of 100 (hypothetical). Based on these calculations a t1 pistol could have a damage value from 85 to 115.

 

Because there is a specified bonus for higher tiered versions an extra 10 to 50 more damage is possible for t2 to t6 pistols. Therefore a t2 to t6 pistols could have a damage spread of 95 to 165. There is nothing that indicates higher tiers have a higher chance at higher damage rolls when compared to lower tiers (except t1).

 

If this is true, it is theoretically possible to get a max damage roll on a t2/orange tier item.

 

So my takeaway is dont be so fast to scrap/sell t2 to t5 items unless all you care about is having the most mod slots possible.

 

So I was right. Tiers correlate to mod slots, not stats. Higher tier items with better stats are pure RNG. That makes a lot of sense with what I've observed in game.

 

Now the question is, is that good design?

 

@Gazz: you went quiet after my last response to you. Have you got an answer to the question?

[Viktoriusiii]

Not sure what you previously stated but here is an example.

 

Let's say a pistol has a base damage of 100 (hypothetical). Based on these calculations a t1 pistol could have a damage value from 85 to 115.

 

Because there is a specified bonus for higher tiered versions an extra 10 to 50 more damage is possible for t2 to t6 pistols. Therefore a t2 to t6 pistols could have a damage spread of 95 to 165. There is nothing that indicates higher tiers have a higher chance at higher damage rolls when compared to lower tiers (except t1).

 

And my takeaway is, that T1 has a base of 100, T2 110, T3 120, ... , T6 150

So that the range is

T1:

85-115 (100)

T2:

93-126 (110)

T3:

102-138 (120)

T4:

110-150 (130)

T5:

119-161 (140)

T6:

127-173 (150)

 

This is the most intuitive way to interpret this data, and the one that makes the most sense, coinsides with my first guess and is what I actually observed in game!

So a T2 CAN'T have better stats than a T6 (in this pistol example) but a T3 can have better stats if the T6 has rolled REALLY bad (127-138 with the medium of 150)

 

This is very well done if this is correct and props to the fun pimps!

 

IF Psychodabble is correct and values of T2-T6 are just random, that is BAD. It would defeat the purpose of levels... but I don't think that is the case. As nearly all my higher tiers are better than the lower ones, with only few exceptions.

 

 

 

Any official statement?

[Gazz]

I think you're getting close to reading what's written there in plain sight. =P

[Viktoriusiii]

I think you're getting close to reading what's written there in plain sight. =P

 

nice to hear.

But not everyone is well versed in XML editing

So I just go off what I can read from a basic programming logic

 

Can you elaborate further on how it is meant exactly? BC if I'm close I still seem to have overlooked some stuff ^^

[Gazz]

What? And ruin all my fun? =P

[Viktoriusiii]

What? And ruin all my fun? =P

 

Nice to see you haven't lost your defining characteristic over the last year

[meganoth]

I think we can all agree that the comparison user interface has to improve independant of any changes to the color system. Even if you only needed to compare blues with blues you don't want to subtract effects from mods and different ammo to get at a result.

 

Apart from that:

 

I like that you can find weapons with great damage and abyssal magazine size. That is a wonderful choice to make. And if damage is the only stat you care about then fine, just ignore the other values and don't complain that they are there for comparison.

 

Since other stats besides damage are (for most players surely) much less important I would even make their ranges bigger AND shift their chances after the damage roll. So that higher damage more likely coincides with small magazine or low durability and vice versa.

 

This system is unusal by the way (another advantage in my book). And it sure has the disadvantage that you can't judge a weapon by its "cover". Nothing a better GUI can't repair.

 

Like Steve Balmer I want to jump on a stage now and shout "GUI GUI GUI". We need a better GUI here.

[Adam the Waster]

I kinda like it but i wish it could more random stats like

 

more/less durability , Faster/slower attack speed , more/less entity or block damage etc

[Nucleus24]

I've gone into my servers items.xml and removed all of the "random stats" on all of the weapons, tools, armor, and clothing.

 

I like this much better to be honest, because I don't want to spend a bunch of time looking at numbers to figure out if something is actually better than what I have. Unequipping mods in order to do a comparison is a huge pain in the butt.

[Kalarro]

I've gone into my servers items.xml and removed all of the "random stats" on all of the weapons, tools, armor, and clothing.

 

I like this much better to be honest, because I don't want to spend a bunch of time looking at numbers to figure out if something is actually better than what I have. Unequipping mods in order to do a comparison is a huge pain in the butt.

 

But then once you get an item, thats it. With random stats the fun of searching and finding lasts more.

[Adam the Waster]

But then once you get an item, thats it. With random stats the fun of searching and finding lasts more.

 

yeah i just wish it was a bit more random stats then just damage!

[Nucleus24]

There is plenty to do in this game besides go searching for a pistol with +1 block damage. There is gear to get yourself, things to loot, bases to build, quests to do...

 

Not to be offensive, but I don't have a surplus of free time to spend searching for minor numerical bonuses when I could be out building a base, teaching new people how to play, making gear to give out, etc.

 

Plus, unequipping mods to compare if an items is better or not is a huge pain.

[Nucleus24]

I think you're getting close to reading what's written there in plain sight. =P

 

I'm literally a systems engineer, and the XMLs are a little...tough to figure out. I had to mess around before I realized that the clothing/armor weren't modified by a -.15,.15 mod or anything, but the base values were defined as a range like (4.5,7.5).

 

Speaking of which, why do a lot of the values have strange rounding? Some clothing gives .2 temperature resistances, armor effect resistances average to .103 a lot, etc.

[Viktoriusiii]

yeah i just wish it was a bit more random stats then just damage!

 

Damage

Rounds per minute/attacksperminute(for melee)

Magazine Size

Range

Durability

 

Armor, effect resistance, sound detection, staminausage, mobility and durability for armor

 

+ Modslots

 

 

Sooo... yeah

[Laz Man]

And my takeaway is, that T1 has a base of 100, T2 110, T3 120, ... , T6 150

So that the range is

T1:

85-115 (100)

T2:

93-126 (110)

T3:

102-138 (120)

T4:

110-150 (130)

T5:

119-161 (140)

T6:

127-173 (150)

 

This is the most intuitive way to interpret this data, and the one that makes the most sense, coinsides with my first guess and is what I actually observed in game!

So a T2 CAN'T have better stats than a T6 (in this pistol example) but a T3 can have better stats if the T6 has rolled REALLY bad (127-138 with the medium of 150)

 

This is very well done if this is correct and props to the fun pimps!

 

IF Psychodabble is correct and values of T2-T6 are just random, that is BAD. It would defeat the purpose of levels... but I don't think that is the case. As nearly all my higher tiers are better than the lower ones, with only few exceptions.

 

 

 

Any official statement?

 

If it works the way you say it does I'd be fine with it. An easy way to verify this is to open the cm and take a sampling. For example, If you get a t2 pistol with damage similar to a t6 on the high end that would confirm it is random across t2 to t6. What it wont tell us is if the frequency is higher/lower based on a specific tier unless you take a huge sample .....but whose got time for that lol...

[FA_Q2]

I'm definitely here. I'm listening.

But the degree of stat variance is the important part.

 

Well, since you are here

 

All I really want is for the interface to display base values to compare from. It is never a good thing when I need to pull out a pen and paper to keep track of the base values for my items

[Psychodabble]

@Gazz: I'd still like an answer to my question. What is the advantage to this distribution over one that separates tiers AND provides more variance at the high end?

[Viktoriusiii]

@Gazz: I'd still like an answer to my question. What is the advantage to this distribution over one that separates tiers AND provides more variance at the high end?

 

although I am not gazz, I am pretty sure that it is to give a weapon a character.

 

Just because you have gamestage 500 and 5 points in lucky looter should not mean you fin that super op weapon.

You find a lot of better weapons, but not each and every one.

 

As I've said before:

My trusty boomstick with max dmg that I found when I was lvl 5 has become a sort of... legendary on its own.

When I found a T6 that was better... I was kinda sad... but the completionist in me knew it was time to let my boomstick go for an allround better weapon.

 

And to answer your question with less personal opinion and more gamedesign in mind:

It is to simulate a world. You aren't in a computer game where everything is bound by levels.

Also it means that you aren't garantueed a great weapon, just because you skilled it to the max.

 

 

And if crafting wasnt a thing the "hidden level" would also be a great way to do it, but because crafting works with levels, levels are necessary... to keep it from beeing so stale as it was in A17 this overlapping range gives a bit of flavour to this otherwise static mechanic.

 

Remember A17? Once you got a better tier, you threw the old one away. And you want that back? :-?

It was one of the bigger complaints of that alpha (with the removal of gunparts and the dumbing down), and all your proposal would do is go back to A17 but instead of scrapping all weapons of same and lower levels instantly, you only scrap lower levels instantly and you still have to compare stats.

 

 

 

 

As I see it, there is no negative effect of this system that would be outweighed by the amount of "engagement" it produces for the player.

Its more on the personal opinion side... but objectively (from a gamedev standpoint) it is not a bad design in any way.

[Morloc]

I am actually PRO even wilder variation (lower chance of it happening but higher possible diviation from the norm stat)

 

 

Steep, but also wide bell curve distribution.

 

 

-Morloc

[Psychodabble]

although I am not gazz, I am pretty sure that it is to give a weapon a character.

 

Just because you have gamestage 500 and 5 points in lucky looter should not mean you fin that super op weapon.

You find a lot of better weapons, but not each and every one.

 

As I've said before:

My trusty boomstick with max dmg that I found when I was lvl 5 has become a sort of... legendary on its own.

When I found a T6 that was better... I was kinda sad... but the completionist in me knew it was time to let my boomstick go for an allround better weapon.

 

And to answer your question with less personal opinion and more gamedesign in mind:

It is to simulate a world. You aren't in a computer game where everything is bound by levels.

Also it means that you aren't garantueed a great weapon, just because you skilled it to the max.

 

 

And if crafting wasnt a thing the "hidden level" would also be a great way to do it, but because crafting works with levels, levels are necessary... to keep it from beeing so stale as it was in A17 this overlapping range gives a bit of flavour to this otherwise static mechanic.

 

Remember A17? Once you got a better tier, you threw the old one away. And you want that back? :-?

It was one of the bigger complaints of that alpha (with the removal of gunparts and the dumbing down), and all your proposal would do is go back to A17 but instead of scrapping all weapons of same and lower levels instantly, you only scrap lower levels instantly and you still have to compare stats.

 

 

 

 

As I see it, there is no negative effect of this system that would be outweighed by the amount of "engagement" it produces for the player.

Its more on the personal opinion side... but objectively (from a gamedev standpoint) it is not a bad design in any way.

 

I don't know how I've managed to miscommunicate my wishes so thoroughly, but apparently I have because you seem to think I want things I have NO desire for.

 

1) I never said anything about wanting super OP weapons, nor did I say anything about wanting every weapon to be better than the last. I don't care what the actual numbers are, I just want to know that if I find a T6 it's going to be better than T2-4. If that means lowering T6 drops, fine, I don't care how it's balanced, I just want the tier system to be more meaningful and intuitive.

 

2) There's no reason why my suggestion would take away "character" from weapons. With more variance at the high end, every T5/6 item would be MORE unique than at present if those broader ranges were applied to all the statistics, not just damage/armor.

 

3) The current implementation does nothing for realism or world simulation. This IS a computer game and the levels are an abstraction meant to represent quality. If we really wanted to "simulate a world", we'd hide the actual numerical statistics and remove the color system entirely. Both the ability to examine stats and the color system are gamey abstractions. The only really issue here is how well do these systems enable and enhance gameplay.

 

4) I don't see anything in my suggestion that guarantees anyone anything. In fact, my suggestion would mean that even if you find a T6, there's no guarantee that it will be particularly good by T6 standards...only that it will be a clear upgrade from a lower tier.

 

5) I had a lot of complaints about A17, but scrapping low tier items was never one of them. That issue is also rendered moot by the new item part system. The point is that items you don't choose to equip should still retain some value and the part system combined with my suggestion absolutely do that, especially once the player can craft T5. Unless the player already has a T6 of that item type, every item found still contributes parts to the strockpile, allowing another chance at a T5 craft that could be statistically superior or unique in an interesting way - a weapon/item with "character". It would also keep stat comparing to an absolute minimum as only equal tier items would merit close examination.

 

6) The negative of the current implementation is clear and has been stated by several other posters. It could be ameliorated by a better stat comparison interface OR a new numbering system that makes comparison more intuitive by returning to the color hierarchy while also adding more variance to higher tiers. I have yet to hear any potential negative to such a numbering shift as it would offer the exact same possibility for player "engagement".

[Laz Man]

I don't know how I've managed to miscommunicate my wishes so thoroughly, but apparently I have because you seem to think I want things I have NO desire for.

 

1) I never said anything about wanting super OP weapons, nor did I say anything about wanting every weapon to be better than the last. I don't care what the actual numbers are, I just want to know that if I find a T6 it's going to be better than T2-4. If that means lowering T6 drops, fine, I don't care how it's balanced, I just want the tier system to be more meaningful and intuitive.

 

2) There's no reason why my suggestion would take away "character" from weapons. With more variance at the high end, every T5/6 item would be MORE unique than at present if those broader ranges were applied to all the statistics, not just damage/armor.

 

3) The current implementation does nothing for realism or world simulation. This IS a computer game and the levels are an abstraction meant to represent quality. If we really wanted to "simulate a world", we'd hide the actual numerical statistics and remove the color system entirely. Both the ability to examine stats and the color system are gamey abstractions. The only really issue here is how well do these systems enable and enhance gameplay.

 

4) I don't see anything in my suggestion that guarantees anyone anything. In fact, my suggestion would mean that even if you find a T6, there's no guarantee that it will be particularly good by T6 standards...only that it will be a clear upgrade from a lower tier.

 

5) I had a lot of complaints about A17, but scrapping low tier items was never one of them. That issue is also rendered moot by the new item part system. The point is that items you don't choose to equip should still retain some value and the part system combined with my suggestion absolutely do that, especially once the player can craft T5. Unless the player already has a T6 of that item type, every item found still contributes parts to the strockpile, allowing another chance at a T5 craft that could be statistically superior or unique in an interesting way - a weapon/item with "character". It would also keep stat comparing to an absolute minimum as only equal tier items would merit close examination.

 

6) The negative of the current implementation is clear and has been stated by several other posters. It could be ameliorated by a better stat comparison interface OR a new numbering system that makes comparison more intuitive by returning to the color hierarchy while also adding more variance to higher tiers. I have yet to hear any potential negative to such a numbering shift as it would offer the exact same possibility for player "engagement".

 

Yes, I think your main question has been diluted after many others have posted. Based on viks posts, based on his experience, he feels the higher tiered items are generally better than lowered tiered versions (e.g. overlapping dmg ranges).

 

It doesnt help that Gazz chooses not to answer directly lol. I'm with you tho, the color / tiers should be a high level indicator if something is better or not. Have you actually found a lower tiered item that x2 more powerful then its higher tier? (E.g. t2 > t4) If not, your concern might be just unnecessary worry.

[Psychodabble]

Yes, I think your main question has been diluted after many others have posted. Based on viks posts, based on his experience, he feels the higher tiered items are generally better than lowered tiered versions (e.g. overlapping dmg ranges).

 

It doesnt help that Gazz chooses not to answer directly lol. I'm with you tho, the color / tiers should be a high level indicator if something is better or not. Have you actually found a lower tiered item that x2 more powerful then its higher tier? (E.g. t2 > t4) If not, your concern might be just unnecessary worry.

 

I don't know that I've ever found such an item, but I also don't know that it matters. Outliers will exist, but the much larger issue to me is finding so many T6s that are inferior to T4-5s. It just feels bad when you're hoping and hoping for an upgrade, you manage to find a T6 of the item type you need and start to get excited, then you check the stats and see it's crap. Then it happens again. And again.

 

I don't know, man. I'm just trying to help improve the game, but Gazz only replies with pithy quips and ignores the substance of the point. Then having to post lengthy monlogues to prevent that point from getting twisted...it's just exhausting and makes me not want to bother. Maybe I'll just wait for final release and mod to fix the stuff that doesn't make sense.

[Laz Man]

I don't know that I've ever found such an item, but I also don't know that it matters. Outliers will exist, but the much larger issue to me is finding so many T6s that are inferior to T4-5s. It just feels bad when you're hoping and hoping for an upgrade, you manage to find a T6 of the item type you need and start to get excited, then you check the stats and see it's crap. Then it happens again. And again.

 

I don't know, man. I'm just trying to help improve the game, but Gazz only replies with pithy quips and ignores the substance of the point. Then having to post lengthy monlogues to prevent that point from getting twisted...it's just exhausting and makes me not want to bother. Maybe I'll just wait for final release and mod to fix the stuff that doesn't make sense.

 

Yeah no kidding, stay strong and hopefully it all gets sorted out. 👍