12 January 2018

WildStar: Assault & Support Power On Weapons

Hey, do you still remember the calculation of health and assault power or support power on items in WildStar? If you do, that's great! If not you can check it out now and now.

Assault Power and Support Power on Weapons

As you may have noticed. If you compare the assault power and support power on gear with weapons you may notice some huge differences. Some weapons have both support power and assault power on them. Secondly, the amount of assault and support power on weapons is so high it doesn't even compare to other items. So obviously these use a different calculation than the rest of the items.

However, before we're going to go into details of how it's calculated, let's see if there's a relation between them on said items.

Relation Between Assault and Support Power

We have a high number and a low number. If our weapon has support power as primary stat this one is the higher one else its assault power. The interesting part is the same item level rewards the same amount of assault and support power. So the only difference on the same item level is whether or not it's a healer or tank weapon, or if it's a damage dealer's weapon.

If one would assume that the support power and assault power are related to each other in a way that each represent one part of some kind of power rating stat that scales up with the item level this would mean that both support power rating and assault power rating will most likely represent a typical percentage. When I'm talking about typical percentages I'm talking about numbers like 
  • every 10% (10%, 20%, 30% ... 90%, 100%) 
  • every third (33%, 66%) 
  • every quarter (25%, 50%, 75%, 100%)
and similar. The idea behind this is simple. If you see a clean percentage, no floating-point and doesn't seem odd, chances are high it has been set as a fixed value. So let's see what we have here.

In this example, I'm using Foreman Krause's Last Resorts Spellslinger pistols, though any weapon should do. I've got several of these, so let's start with an item level 100 weapon that has 8353 assault rating (rounded) and 2784 support rating (rounded). The good thing about an item level 100 weapon is that dividing it through the item level to get the increase of each item level is literally just dividing by 100. This means the number gets smaller but the digits don't change. This can be really helpful as sometimes odd numbers may result from these calculations.

If we take the total power rating of this weapon we sum up both support and assault rating which should give us 8353 + 2784 = 11137. Now we can calculate how much weight each rating has on this total value by dividing the respective rating through the total rating like so: 8353 / 11137 = 0.75. Now, of course, the actual number is 0.7500224476968663. Though we used the rounded value that stands on the weapon - technically speaking the value on the weapon is not actually rounded but cut off after the floating point instead. So we have to assume a certain error in our calculation. Taking this error into account we expect the assault power rating which is the main stat of this weapon to be 75% of the power rating. In return, this means the support power rating value must be 25%. Let's check it to make sure: 2784 / 11137 = 0.25. Again rounded up from 0.2499775523031337.

Even though we can't say for sure this relation exists this is a valid proof that it might.

Getting the Exact Assault and Support Power

If you don't trust those floating-point numbers let's calculate the exact assault and support power rating. If I haven't mentioned how to do this I'm gonna repeat the procedure now. It's quite simple you just have to follow these steps:
  • Unequip your current weapon
  • Switch to a build that doesn't push or increase assault power rating (the AMPs)
  • Use the chat command for assault power rating:
  /eval Print(GameLib.GetPlayerUnit():GetAssaultRating())  
  • Alternatively for support power rating use:
  /eval Print(GameLib.GetPlayerUnit():GetSupportRating())  
  • Equip the respective weapon and use the same command again
  • Subtract the greater number from the smaller number
So let's do this.
Using the command with no weapon equipped I've got a 5312 assault power rating. No AMP bonus, using the healer action set right now. Equipping the weapon and redoing the command I now have 13665.125 assault power rating. This means the weapon should have 13665.125 - 5312 = 8353.125 assault power rating. Next to the support power rating, but the same procedure. Make sure to switch to a build that does not have the support power rating AMPs. I've got 1100 support power rating currently with no weapon and 3884.375 with the weapon. So that's 3884.375 - 1100 = 2784.375 support power rating on the weapon.

Now let's go for the total power which is 8353.125 + 2784.375 = 11137.5. That's a rather round number, which is good. It means we might be on the right path. Now let's calculate the percentages and see if they're round.

Assault power rating percentage: 8353.125 / 11137.5 = 0.75
Support power rating percentage: 2784.375 / 11137.5 = 0.25

Now that's more how I like it. If this is solid proof for a relationship then I don't know.

But How Does It Scale?

This is the last question. However, for this, we need more weapons. I've sold a lot of guns. I've collected all except four. Additionally, my current weapon is the same item with a different item level. As you might be accustomed already - I'm gonna create a table with all the samples we have. Four samples and a bonus one.

Item Name Item Level Total Power Rating Assault Power Rating Support Power Rating
Bardlet's Hair Triggers 65 7239.375 5429.53125 1809.84375
Foreman Krause's Last Resorts 100 11137.5 8353.125 2784.375
Bardlet's Hair Triggers 105 11694.375 8770.78125 2923.59375
Bardlet's Hair Triggers 110 12251.25 9188.4375 3062.8125
Foreman Krause's Last Resorts 135 15035.625 11276.71875 3758.90625

So you should also be familiar with the next step. We can check whether or not this is linear by subtracting the next from the previous. If we get the same value for the same jump of item levels it should be linear at that part. There are other ways to check this as well but that's the most straight forward in my opinion so let's go with it.

12251.25 - 11694.375 should be the same as 11694.375 - 11137.5

<=> 12251.25 - 11694.375 = 11694.375 - 11137.5
<=> 556.875 = 556.875 (true)

Great, that went fast. So the next step would be to check how much power rating we get for each item level. So let's divide the increase by the item level. We had an increase of 556.875 for every 5 item levels. So we get 556.875 / 5 = 111.375 for each item level.

Since we know that the main stat is 75% and the second stat is 25% we know that either assault or support power rating (in our case assault power rating) changes by 0.75 * 111.375 = 83.53125 and the other one (in our case support power rating) changes by 0.25 * 111.375 = 27.84375.

The Formula for Power Rating

Since we know it's linear we can use our lovely linear function:

f(x): y = m * x + t

replaced with our nice labels:

f(itemLevel): powerRating = m * itemLevel + t

m is the increase per step which we've already calculated. It's the 111.375 and the step is each item level. So for the calculation of t, we can replace all the other variables. Let's take the stats of the item level 135 weapon and put it into our formula.

f(135): 15035.625 = 111.375 * 135 + t

As you see we are missing the t. No problem, if we subtract (m * x) we have t left out.

<=> 15035.625 = 111.375 * 135 + t | -(111.375 * 135)
<=> 15035.625 - (111.375 * 135) = t
<=> 15035.625 - 15035.625 = t
<=> t = 0

Oh, so we have a linear function that has no start value, as in, it starts at zero or you could say it has no offset. So to finish our formula off:

f(x): y = 111.375 * x

or

f(itemLevel): powerRating = 111.375 * itemLevel

Other Formulas

So the other formulas are quite simple. We know the main stat is 75% of the powerRating. So we could say:

mainRating = 0.75 * powerRating

Multiplying both sides of our equation by 0.75 we get:

0.75 * powerRating = 0.75 * 111.375 * itemLevel

We can replace the left side with mainRating and calculate a little on the right one. Doing this we get:

mainRating = 83.53125 * itemLevel

Doing the same with the secondary rating which is 25% of powerRating we get:

secondaryRating = 0.25 * 111.375 * itemLevel

Calculating the right side we get:

secondaryRating = 27.84375 * itemLevel

So What's The Current Maximum Power?

As we know currently the maximum item level is 170. So if we put 170 into our formulas we get:

f(170): y = 111.375 * 170 = 18933.75
g(170): y = 83.53125 * 170 = 14200.3125
h(170): y = 27.84375 * 170 = 4733.4375

And now we know a damage dealer weapon of item level 170 will have 14200 assault power rating and 4733 support power rating. A tank or healer weapon of item level 170, on the other hand, will have 4733 assault power rating and 4733 support power rating. 

11 January 2018

I'm Max! Experience? I Don't Care!

I already might have kind of addressed this in my Over Consumption of Luck post, but now I'm going to focus on this and go into more detail.

Experience in Games

If you played any RPG games or games that implement leveling you may have come across this weird thing called experience. So apparently you seem to get it from doing certain things and once you've reached a certain amount your level increases and you need more experience to continue.

As experience shows this will go on for a few levels until you reach one certain level which people refer to as maximum level, max level or level cap.

Your goal until then was to collect experience all the way. Now you don't need experience anymore. It has no further use for you. So let's stop farming it.

Content Loss

This may not be the case for every person but I imagine quite a bunch of people that would actually stop playing content that has been interesting earlier. Luckily there's a simple way to prevent this from happening.

Simple Problem - Simple Solution

If experience is still worth it for these players. They will still have an interest in it, of course. So the goal is to figure out what to do with the experience after the maximum level. You could probably do like anything from a game designer's perspective. Though something I like to do is select and write about examples that have already been used. So let's go over my two favorite games that implemented this.

Guild Wars 2
In Guild Wars 2 experience is still used after the maximum level. Here it plays a role in crafting unique items and item skins or other useful stuff. It's also used to buy items to upgrade materials to a higher tier. So as you can see even as a max level character this is still more than just useful.

WildStar
After reaching your level cap in WildStar your experience is converted to a currency called Elder Gem. This currency is used for several useful unlocks and to buy neat stuff and that's not all. It goes even further. The elder gems are capped weekly. You can get 250 each week. Once you have capped your elder gems all experience is instead converted to gold, silver and/or copper. So even if you max your level and have everything you can buy with elder gems experience is still worth it.

Diablo III
Reaching the max level in Diablo III allows you to collect paragon experience. So your leveling continues like before, but instead of unlocking new abilities and passives you increase your stats now. Most of your stats are capped at paragon 800. This is not the case for vitality (increases health) and your main stat (increases damage) though.

What Can We Take From This?

Something I already posted in the Over Consumption of Luck post. As game designers, we should make sure that the content of our game is not lost due to missing interest from rewards. Aside from enjoying content with friends, no one likes to run something that brings absolutely nothing. It's the same problem with power creep. Bringing in more and more content having the old content grow mold is not how we should handle these things. Solving this issue like solving the experience problem in a way that is described here or similar will allow even the oldest veterans to still be able to enjoy gaining experience. If done right, maybe they'll even focus on it?

10 January 2018

Guild Wars 2: PvP Pips vs WvW Pips

So you may remember me checking out how many matches the PvP season in Guild Wars 2 takes and you may remember me doing some calculation on how much the WvW pips have taken. Today I'm gonna combine both calculations and research to compare them with each other!

Calculating How Long PvP Takes

We only know the amount of matches PvP takes but not the amount of time. Especially since this value changes according to every single match you may have, we can only assume an average which we can work with, but that should be no problem.

If you remember back in the PvP post we have found out that an average player or a player under platinum needs 61 to 224 matches to complete PvP. Next, we need to know how long a match takes on average. Unfortunately, you can only see your last ten PvP matches so I have to work with this.
These matches took me:

match length 10m 22s 14m 12m 15s 6m 8s 10m 8s 11m 33s 8m 36s 9m 30s 5m 54s 10m 10s
conversion
calculation
10 * 60 + 22 14 * 60 12 * 60 + 15 6 * 60 + 8 10 * 60 + 8 11 * 60 + 33 8 * 60 + 36 9 * 60 + 30 5 * 60 + 54 10 * 60 + 10
time in seconds 622 840 735 368 608 693 516 570 354 610

Now we can add up all the seconds and divide it through the amount of matches we've listed - ten matches. Doing so we'll get:

(622 + 840 + 735 + 368 + 608 + 693 + 516 + 570 + 354 + 610) / 10 = 591.6

Converting this back to minutes we have 591.6 / 60 = ~9.86 minutes per match. Next, we apply the queue time. Queues are on average around 2 minutes. The time, until everyone has readied up, is probably around 5 to 30 seconds on average more like 12 seconds. The map selection takes most likely around 8 seconds and the time until a match actually starts is on average about 55 seconds including the loading screens. Summing all this up we get  12 + 8 + 55 = 75 seconds, which is 75 / 60 = 1.25 minutes plus the 2 minutes from before. So I'd say it's safe to assume 3.25 minutes to wait for each match. Adding this to our previous number we now get 9.86 + 3.25 = 13.11 minutes for each match.
 Now multiplying this with our amount of matches we need, we get between 13.11 * 61 = 799.71 minutes and 13.11 * 224 = 2936.64 minutes to complete the season track. Those numbers seem big, let's convert them to hours by dividing through 60 again. Now it's 799.71 / 60 = ~13.33 to 2936.64 / 60 = ~48.94 hours to complete it. That sounds fairly reasonable for how long a season goes.

Though we haven't gotten the real minimum here. A player who gets the most pips, that's what we have to calculate with to get a total minimum. At least we have our total maximum dependent on average time. We could calculate the maximum time with 15 minutes since that's how long a match can take at most, usually. However, it's extremely unlikely for every match to take that long.

Our next step is to assume the minimum amount of matches which is 45 matches. 45 matches taking ~13.11 minutes per match or 45 * 13.11 = 589.95 minutes, that's 589.95 / 60 = 9.833 hours.

Combining the total minimum and the maximum we get a range from 9.83 hours to ~48.94 hours.

Taking the WvW Time Consumption Into Consideration

We don't need to calculate much here since the prior post has taken care of this already. In the post, we've calculated that it takes between 6.36 hours and 40.28 hours.

Start the Comparison

Now that we have our numbers we can start to compare. The easiest way to compare is to calculate the reward we get in the same time unit. To get the same time unit we need to divide the reward through the time it takes to get it. Luckily I've already listed the rewards in the prior posts so here we go:

PvP Rewards

  • 560 Shards of Glory
  • 34 Unidentified Dyes
  • 83 gold 85 silver
  • 44 Transmutation Charges
  • 100 PvP League Tickets
  • 400 Ascended Shards of Glory
  • 3 Boxes of Grandmaster Marks
  • 1 Llama Mini Choice Box
Let's start with per hour:
  • 560 / 48.94 = 11.44 to 560 / 9.833 = 56.95 Shards of Glory every hour
  • 34 / 48.94 = 0.69 to 34 / 9.833 = 3.46  Unidentified Dyes per hour
  • 838500 / 48.94 = 17133 to 838500 / 9.833 = 85274 copper each hour
  • 44 / 48.94 = 0.9 to 44 / 9.833 = 4.47 Transmutation Charges every hour
  • 100 / 48.94 = 2.04 to 100 / 9.833 = 10.17 PvP League Tickets per hour
  • 400 / 48.94 = 8.17 to 400 / 9.833 = 40.68 Ascended Shards of Glory an hour
  • 3 / 48.94 = 0.06 to 3 / 9.833 = 0.31 Boxes of Grandmaster Marks each hour
  • 1 / 48.94 = 0.02 to 1 / 9.833 = 0.1 Llama Mini Choice Box every hour

WvW Rewards

  • 365 WvW Skirmish Claim Tickets
  • 35 Skirmish Chests
  • 76 Memories of Battle
  • 142 Reward Track Progresses
  • 11 Mist-Warped Packet
  • 3 Mist-Warped Bundles
  • 4 Mystic Coins
Next, we do the same with the WvW rewards but using the WvW times:

  • 365 / 40.28 = 9.06 to 365 / 6.36 = 57.39 WvW Skirmish Claim Tickets every hour
  • 35 / 40.28 = 0.87 to 35 / 6.36 = 5.5 Skirmish Chests an hour
  • 76 / 40.28 = 1.89 to 76 / 6.36 = 11.95 Memories of Battle per hour
  • 142 / 40.28 = 3.53 to 142 / 6.36 = 22.33 Reward Track Progresses each hour
  • 11 / 40.28 = 0.27 to 11 / 6.36 = 1.73 Mist-Warped Packet every hour
  • 3 / 40.28 = 0.07 to 3 / 6.36 = 0.47 Mist-Warped Bundles an hour
  • 4 / 40.28 = 0.1 to 4 / 6.36 = 0.63 Mystic Coins per hour

Last Step

Now we either compare each piece with each other like WvW Skirmish Claim Tickets and their item prices with PvP League Tickets or we calculate the total worth of all items and compare that with each other. Another option would be to do both of these options.

So let's start comparing the tickets. In PvP, you need 20 of these for the Record of League Participation and another 30 for the Record of League Victories to upgrade your armor to legendary. In WvW you need 1095 to upgrade armor to legendary. This means you need 50 for PvP and 1095 for WvW, which is  1095 / 50 = 21.9 times more. So, in this case, WvW tickets need to be 21.9 times worth more. Looking at our calculation we see that WvW tickets are 9.06 / 2.04 = 4.44 or 57.39 / 10.17 = 5.64 times worth more. Off to a bad start for WvW.

Compared to PvP, WvW gives instant reward track progress boosts. Each one gives 25 points according to the Guild Wars 2 wiki. We get between 3.53 and 22.33 reward track progress each hour. That's between 3.53 * 25 = 88.25 and 22.33 * 25 = 558.25 per hour. This is something special the PvP reward track doesn't have, and it's hard to compare it. At least we can say you get up to a whole reward more in the reward tracks each hour.

Therefore PvP rewards up to 3 Grandmaster Marks and a Llama Mini Choice box, as well as Transmutation Charges. This is also hard to compare.

I haven't researched the Mist-Warped Packet and Mist-Warped Bundles as they drop differently from the Magic-Warped Packet and the Magic-Warped Bundles. At the same time, it heavily depends on the market. Same with mystic coins and unidentified dyes. So in this regard, WvW might be better than PvP if the market stands good. Unfortunately PvP rewards between 2 gold 27 silver 79 copper and 11 gold 33 silver and 87 copper every hour on the hand with no relation to the market.

Last but not least don't forget that WvW has drops, karma and event rewards, which PvP doesn't have.

Conclusion

PvP and WvW consume nearly the same time for getting the total amount of pips. A huge difference here is however that PvP's seasons tend to be 8 weeks long and WvW is only one week. So you need to do 8 weeks in one week in WvW, at the same time unique currencies like the tickets are rewarded only once the track is finished, so WvW gives you a lot more tickets if you think about it, but you need to work for those tickets, quite a lot. Another huge thing is that WvW rewards are dependent on market values. If certain market values like tier 6 materials or mystic coins are high in value the WvW weekly reward track is also high in value. Same counts towards blueprints of siege weapons. The last point that should be mentioned here is that WvW boosts your normal reward tracks, so if you go for those rewards WvW is gonna be your best bet. Or doing both if you have the time to do so.

About Me

My photo
I'm a M.Sc. Games Engineer and I created this blog to share my ideas, theorycrafting, thoughts and whatever I'm working on or doing.