Unit values

posted on March 1st, 2011, 2:35 am

After being curious about why the hitpoints and attack numbers are all long decimals, I finally found this awesome unit value calculator Excel file somebody made available (thanks a lot)

I notice the DPS-to-offensive values are still decimals though (0.4113 for phasers).  Is there a reason for this?  My brain likes whole numbers so I'm struggling to understand why it needs to be like this 

Also, it looks like this calculator isn't being used, unless there are other factors involved.  My own Excel spreadsheets show a wide variance in this ratio, for example Martok B'rels deal 50% more DPS/offensive value than Charghs (I've only put in Klingons so far)

According to the spreadsheet, Martok B'rels have an offensive value of 16.5 instead of their listed value of 15.

posted on March 1st, 2011, 2:56 am

I'm sure Blade would thank you - and I definitely will 

I have not adjusted the unit values calculator in awhile, so I would need to ask Optec if some of the variables have changed

However, when I've just plugged in the values now, I'm getting a value of 14.67 - which is rounded up to 15

What do you mean DPS to offensive values are decimals? Are you talking about the unit-values calculator or...?

posted on March 1st, 2011, 5:07 am

I mean damage/shot delay, to get the DPS.  Then I figured the offensive value would be

Offense = DPS*0.5

instead of

Offense = DPS*0.5365

What's the purpose?  It's maddening!!!

I checked my numbers for the B'rel and I made a mistake.  You're right about the 14.67...but I don't see why they can't just make it 15.  It's not hard to reverse-engineer from an offensive value and a set ROF.

Going back to my spreadsheet, can you double-check my extreme number on the other end of the spectrum?  I see the Chargh as having an offensive value of 17.1305 instead of 23.  But perhaps the devs wanted to slant the numbers since it has the highest damage and lowest rate of fire of any ship I am aware of.

posted on March 1st, 2011, 5:16 am

The Unit Values Calculator is unfortunately a very simplified version of the tools that Optec uses to calculate values in Fleet Ops. What this means is that vessels which have some special functionality will not be taken into account. This includes multi-targeting weapons and variable damage weapons (like the Chargh's ion torpedo). This is why the Chargh's offensive value is incorrect according to the calculator .

If you would like, I'd recommend reading this page: Editing Tooltips - The Hitchhiker's Guide to Fleet Operations

Offensive Value in FO is dependent on the range, weaponry, speed of the vessel, and accordingly I just used an average value of all torpedo, all pulse, and all phaser weapons to come to a relatively decent value for each (this is why torpedoes and pulses do roughly 20% increased damage compared to their offensive value). This is why the values from the calculator are not even, but instead have to be rounded

posted on March 1st, 2011, 5:38 am

Okay, I'm understanding more and more.  I suspected there were more underlying calculations, but then I figured it was still needlessly complicated.  Like why on earth is the hitpoint calculation:

Shield Hitpoints = [(Shield Defensive Value * (Percent of HP made of Shields))^1.1] * 18

Raising it to the power of 1.1 serves no purpose at all!  It's needless complexity!

A few more questions while I'm learning?

1. In the physics files, all the speeds are multiplied by 1.2.  A ship with listed speed 130 uses 156 as its gamespeed.  Why not just give us the same numbers the game engine uses?

2. When a weapon has damagevariance = X, does that mean the range is:

D to D+X, or

D-X to D+X?

Or maybe D-X/2 to D+X/2?

[Optec]

posted on March 1st, 2011, 7:19 am

Last edited by Optec on March 1st, 2011, 7:21 am, edited 1 time in total.

raising something by the power of something serves the purpose of making it a non-linear relation. in this example, shield hitpoints increase with defensive value, but they increase more and more the larger the defensive value gets

Try using matlab or any graph ploter to make the difference more visible. Most of the concepts behind the Fleet Operations balancing were developed in matlab before they were built into a scripting system

[Tryptic]

posted on March 1st, 2011, 8:20 am

Last edited by Tryptic on March 1st, 2011, 8:23 am, edited 1 time in total.

Oh.  I had no idea.  This makes a lot more sense now.

I mean no offense, but that's not the best way to do it.  A non-linear relationship will make things simpler in the short term, but it means you can't communicate universally.

A non-linear relationship means that anybody who does not have your matlab graphs in front of them will get incorrect information by reading the defensive value.  Giving the number to the community without explaining it is deceiving the customer, even if that customer isn't paying and you don't owe them an accurate report.

You should consult your graph and give larger ships higher defensive values, rather than making a point of defense have different value in different situatios.  In the same way, a ship with 20 offense and long range should have the exact same DPS as a ship with 20 offense and short range.  The people who play Fleet Ops and read the values on this site know how range and speed affect a ship's offensive ability, you should trust them to figure it out on their own instead of abstracting the data.

I know I'm not actually in a position to say everything I just said, but I feel it's the truth and you can benefitfrom it.  I still want to help develop Fleet Ops, if you decide another pair of hands and a new point of view is worth the trouble of having me around.

posted on March 1st, 2011, 9:40 am

The unit attributes were introduced to make it easier for people to compare the usefulness of vessels without tinkering with the math behind them. I would be happy to use simple linear systems for them (I actually did that in beta1 and beta2), but sadly, life isn't a linear equotation. Conversion from numerous stats of a vessel (and a ton of stuff which is not easily predictable with math formulas) to an abstract concept like "usefulness" will always lead to some math.
I'm quite happy with the current attribute system and I think they express the actual usefullness of a craft quite well.

posted on March 1st, 2011, 1:27 pm

I think I can only add to Optec's explanation by saying that if a linear system were utilized, balancing would be a nightmare - especially as the Fleet Ops destroyer is not so far away in stats and abilities from a battleship.

Damage variance is damagebase +/- the variance

posted on March 1st, 2011, 4:26 pm

Do you mind sharing the actual equations with us then?

I think the decision to inversely relate a ship's range and its speed has been a big step to minimize the problems you're talking about.

For example, the Chargh:  I believe the idea is since it fires one big shot every 4.75 seconds, its weapon is extra powerful and should deal less damage in return.  My experience from Eve has been that this is not the case.  Weapons with high burst damage in small numbers don't have any special advantage over rapid-fire guns.  When you increase numbers, you reach a point where you can 1-shot enemies, but if you get even more numbers you all fire on one enemy and the rest of the damage is wasted.  And we're talking 25 second reload times on some of these guns.

Another game with long-reload units is the Warzone 2100 project.  In that case they deal the exact same level of DPS as other units and have a 15 second reload time(!!)  It balances out fine because sometimes you oneshot enemies, and sometimes your damage gets wasted.

Now in Fleet Ops a high damage number is good for killing subsystems, but it can also overkill subsystems and damage is wasted.  As a result the Chargh's burst damage is a much smaller advantage than you've given it credit for, and the ship actually punches at 17, way below its listed value of 23 offense.  Anyone who tries to use the Chargh as a combat ship without its special will notice the discrepancy but won't know why it is there.

@Dom, the relation between Defensive value and hitpoints has nothing to do with balancing.  Whether it's 10^(1.2) vs 20^(1.2) or 16 vs 36, the balance remains the same.  Using a non-distorted display value won't make it a nightmare to balance.

posted on March 1st, 2011, 5:24 pm

Most of the concepts behind the Fleet Operations balancing were developed in matlab before they were built into a scripting system

You make my day mr. scientist

[Dominus_Noctis]

posted on March 1st, 2011, 6:23 pm

Last edited by Dominus_Noctis on March 1st, 2011, 7:51 pm, edited 1 time in total.

The non linear relationships have quite a bit to do with balancing - cost scales with hitpoints, speed, range of weapons, passives etc . Of course, if you changed all the values to be linear, then you could do what you wanted, but it wouldn't provide a better system than the current one, because many of those stats have to scale to compensate for tech tree costs etc - the only thing that would be changing to the 'user' is seeing that the offensive, system, and defensive values are a bit larger than they are now for higher tech ships.

When you are talking about the Chargh you are doing so without regard to micromanagement or gameplay relevance (such as when you are talking about how DPS should be the same between a shortie and a longie - which completely ignores fundamental issues of kiting, gap closing, arc etc)  . Just as importantly, if we take a look at the math, you are using my torpedo values to calculate a value which does not take into consideration range, arc, speed of the vessel, etc, which means that you are using an average across all tiers of torpedo vessels to make an assumption about balance. Most values are 'non-linear' for balancing in FO.

Balance has not changed simply because you're learning how balance is roughly calculated in Fleet Ops . In the end, these are comparison values only - a Negh'var may have more offense than a Sovie, but it's still going to get eaten by that ship.

EDIT: some examples below

Unit A has a regen rate of 10 points per second and 11 hitpoints
Unit B deals 10 damage every second.
Unit C deals 20 damage every 2 seconds.

Same DPS... but when...

Unit A is being attacked by Unit B - Unit A regenerates faster than it can be killed.
Unit A is being attacked by Unit C - Unit A is now killed on the first volley.

And...

Unit B is short ranged with 180 arc while Unit C is long ranged. If Unit B and C duke it out, Unit C will be able to retreat while firing and deal damage for longer while doing so.

Simple examples, but they do explain why there are so many exceptions to unit values all while trying to maintain a comparative atmosphere.

And of course, one other instance - values themselves, even if designed to be totally linear, will betray themselves as having non-linear relationships on the battlefield. Regeneration is more powerful on a stronger ship, than on a weaker ship for instance.

[Tryptic]

posted on March 1st, 2011, 7:45 pm

Last edited by Tryptic on March 1st, 2011, 7:57 pm, edited 1 time in total.

Hold on, nobody said anything about ship cost being tied into attributes.  Perhaps I should have expected it given where this conversation is going, but if you just assume I know how the system works then of course my arguments will look foolish.

What I am suggesting is instead of having ^(1.2) in your defensive value, you make defensive value linear and use ^(1.2) in your equation that determines the strength bigger ships should have.  The result is the same, and it has nothing to do with the relationship between the values and cost.

---

I'm saying the Chargh's listed value is incorrect AFTER taking into account the many gameplay factors involved.  I am comparing your system to the systems of other games I have played and disagreeing with you based on that knowledge.

Of course DPS should mean the same thing for a short ranged ship and a long range ship.  DPS is a discrete orthogonal term, like Kilograms or degrees Celsius: you make the short range ship have a higher DPS than the long range ships to compensate for closing, kiting etc.  But you don't make the term "DPS" mean something different for different ships.  It's like saying a 100-meter long ship is actually longer than twice the length of a 50-meter long ship.

Having all the numbers relate to each other creates a self-balancing system and can make your work easier, but I believe it would be better to keep the tools orthogonal and make adjustments manually.

In an orthogonal system, if you want to increase JUST the attack of a certain ship, you increase it.

To do the same thing in a non-orthogonal system, you need to first increase the attack, then decrease the ratio of attack-to-cost to make it cost the same as before, plus whatever other relationships were tied into attack need to be modified.

The interconnected system is better at making overall changes, while the orthogonal system makes it easier to balance the game manually and respond to specific problems.  I would use the exact same equations you do, but keep the equations separate from the data instead of embedding them into the language.

posted on March 1st, 2011, 7:53 pm

Crud dude, I just added in some more text 

posted on March 1st, 2011, 7:58 pm

Lol yeah, I made several mistakes in that last one, and immediately turned around and edited it a few times.  Let's read each others' posts for a minute and get back on the same page