Skip to main content

DFK Blogs

Combat Meditation Part 2: HP by Class


Introduction


Hello fellow adventurers its MeanMachine with part 2 of my exploration into the intricacies of how to maximize HP and MP on your DFK Heroes.

There are two major influences on HP, your heroes stat growth and their class. Each class has a unique growth distribution that feeds into the HP growth formula.


Each class has a specific percentage chance of rolling a small, regular or large HP gain, and those chances play just as large a roll in HP increases as stat growth. Each heroes chances by main class can be found in the Hero Base Stats Sheet and are shown below:


This post will dive into the details of how these growth percentages combine with heroes native stat growth to determine what is a high HP for a given hero class and which classes can produce the highest HP heroes. Before we start simulating data, we can see right away there are 8 classes whose highest % chance is the large HP growth option, making them the most likely to have large HP gains: The Knight, Monk, Legionnaire, Paladin, Shapeshifter, Dragoon and Dreadknight. Some of these are classic tank heroes and these %’s make sense but more interesting are the monk and shapeshifters, who I would not have predicted to be in the upper echelons.


Methodology


Process

The process used to generate the following results is the same as part 1 of this series. I wrote some python code to mimic all aspects of heroes’ level-up and simulated 1000 heroes leveling up to the specified level with a specific set of parameters. This gives us a probability distribution to use to determine how good a hero’s rolls have been, and a way to calculate the expected value of a hero’s stats, including HP and MP, for that set of parameters. Since HP and MP have a heavy right tail in terms of outliers in the growth distribution, I’m using the median of the 1000 heroes HP as their expected HP value for the result of their level-ups. The difference in expected values from a set of level-up criteria represents how effective that level-up strategy is at boosting HP.


DFK Hero Assumptions

There are a lot of variables that influence stat growth, and the goal of this post is to highlight variability from the small/med/large growth chances, so I’ve made some simplifying assumptions that are consistent where not otherwise specified throughout the blog post. When comparing across classes I’m looking at a common level 15 hero with matching main/sub class who is leveled up according to their highest growth % stats, with ties resolved towards VIT. No enhancements are added to the summon and no crystals used on the leveling. While level-ups by highest growth % will not result in equal stats boosted for each hero, it should match what we are most likely to see in different hero classes, given devs have stated these the highest growth stats are the most important for that hero. There have been a number of hints that 15 is the first level a hero will be able to begin combat, so I’m choosing it as my benchmark to make equal comparisons.


Variables Explored

All the classes are compared together to see the range of HP growths available to them, and are then grouped in trends to take a deeper dive in HP growth patterns across classes. The impact of subclass on hero HP growth is briefly examined. Finally, a summary of each classes individual distributions for different levels of VIT prioritization is presented. The appendix includes the full look at each classes individual range of scenarios for HP growth to level 15.


Magic damage dealers (and the bard) have a universally low health growth, all with median values around 500-550 HP. One thing that stands out to me about these heroes is how HP is low across all levels of classes, summoner and sage do not have a demonstrable health difference from the basic classes. For any of these caster classes, HP values around 650 or higher are likely to be around the 75% in outcomes, and thus represent good HP growth and a competitive hero HP-wise at level 15. Given that these heroes overall growth is lower, due to both stat growth and higher % of small and medium Health rolls, efforts to boost their health result in less HP per stat point or crystal than other classes.


Physical Damage Dealers

Physical damage dealers health growth distributions are shown below. Right away we see there is a tier break between the more sneaky/ranged classes and the more close combat oriented heroes.


The sneaky/ranged heroes have HP growths similar to the caster classes, there is about a 150 HP difference between them and the close combat heroes. There is a slight HP bonus afforded to the advanced and elite classes of 20 HP, but a hero with 700 HP would be in the top quartile for any of these squishier physical damage dealers.

In contrast, a great HP growth on the ranged heroes is the equivalent of a median/ slightly below average hero. These heroes have fairly consistent health outcomes, averaging around 700-750 with any hero around 850 or higher. What these 4 classes, monk, pirate berserker and dark knight share is higher native vit growth percentages and respectable chances of rolling high hp growth. These lend themselves to a focus on HP growth.


Tanks

Finally, here are the HP distributions for the tank classes:


What stands out right away is that all tanks and not created equal, under this leveling scheme Knights, Legionnaires and Paladins stand out as the heroes with the most HP, while warriors, DreadKnights, shapeshifters and dragoons fit it better in the close-up physical crew with berserkers and dark-knights. For those heroes, an HP of 900 represents good growth, whereas for knights, legionnaires and Paladins, over 1000 HP at level 15 stands out among the distribution of potential outcomes for those heroes.

For these heroes, how they are leveled starts to make a big difference in these outcomes. Knights and paladins got VIT boosted guaranteed every level, whereas heroes like the warrior and DreadKnight only had a 50% of the player selected stat bonus. This result is more reflective of the fact that Knights, Legionnaires and Paladins are designed to have the most HP by stat growth %’s, rather than that these other heroes can’t be competitive high HP monsters. One other result worth discussing here is the Dragoon, which has not only the lowest average HP of the tanks but also of the all the close-up physical damage dealers, and a smaller range of HP outcomes. It was of the only heroes whose primary stat % growth don’t indicate leveling it for VIT, which says to me its skills will not be supportive of it as a tank.

Before looking at trends in hero HP growth by class, let’s briefly look at the impact of hero subclass on their growth.


Subclass Impact


Subclass’s only direct impact on HP growth comes from the secondary growth %’s during level-ups, there is no other influence like the main class. Looking at the Knight, which as one of the highest HP growth classes will be the most sensitive to subclass impact, there is a minor but mostly negligible contribution to HP growth from subclass. There can be as much as a 30 HP swing between a Paladin and wizard subclass, which does not drive HP growth or outweigh the significant variability in each hero’s level-up process.


Main Class Health Growths


Example Class Health growth chart:


The appendix summarizes the growth charts for each hero class under the same modeling conditions. Below I’ve attempted to summarize with the three archetypes I see in health growth. Each class is simulated with 5 different level-up strategies to showcase their potential health outcomes: the Base, Average, Vit-Pref, Vit-Priority, and Vit-Maxxing strategies. The goal here is to show much potential the player has to boost HP for a given hero class. The 5 scenarios are:

  • Base: A common hero with no vit boost on level-ups and no other enhancements
  • Average: An uncommon hero with a gaia boon (50% chance) of vit on level-up and no other enhancement
  • Vit-Pref: A rare hero with a blue VIT gene, lesser vitality stone, VIT gaia’s blessing (100% chance) and lesser VIT crystal through level 5, and then a VIT boon through level 15
  • Vit-Priority: A legendary hero with a green VIT gene, vitality stone, VIT gaia’s blessing (100% chance) through level 15, and VIT crystal through level 5
  • Vit-Maxxing: A mythic hero with a purple VIT gene, greater vitality stone, VIT gaia’s blessing (100% chance) through level 15 boosted with a greater VIT crystal

Squishy Heroes


Most of the heroes listed in the casters section (Priest, Wizard, Summoner, Sage, Seer, Scholar) along with support and ranged heroes (Thief, Archer, Ninja, Bard, Spellbow) have growth charts that look like the priest:


These heroes all have mediocre VIT growth and HP growth roles on the lesser side. Most can gain an average of 150 HP by level 15 with the modest stat bonuses given in the VIT preference level-up strategies, although given the focuses of those heroes its fair to wonder if the benefit of the VIT stone outweighs boosts to other stats, particularly INT or WIS given their similar impact on MANA growth. These heroes have tight distributions, indicating there won’t be much variability in their health stats that occur naturally.


Stout Heroes


The middle tier of health heroes (Monk, Pirate, Shapeshifter, Berserker, DarkKnight, Warrior, Dragoon) have a growth percentage best characterized by a berserker. Before looking at the breakdowns, One of the most interesting parts of this analysis is the structure of this tier. There are some heroes, like the Shapeshifter and Pirate, that I am surprised are HP competitive with the rest of the heroes in this tier, and other heroes, like the Darkknight and Dragoon, who I am surprised do not slot into the Tank tier. Its notable to me that there isn’t more stratification between these heroes, and I think they can be lumped together as melee damage dealers, who can serve as off-tanks with the correct HP leveling and skills.


These stouts heroes can add an average of approximately 175 health to their level 15 HP total with minimal VIT investment, and for these heroes a VIT build may fit in with the greater PVP strategies. It’s notable that the median HP difference between this tier and the tank tier of heroes shrinks the larger the Vit investment is. It’s also notable that there is more variance in the outcomes for these heroes, particularly at the advanced and elite levels.


Tanks

Finally, the tank heroes (Knight, Legionnaire, Paladin, Dreadknight) all have monster health growth. Their HP growth distributions are all very similar, with the DreadKnight being the highest median health hero.


All of the tanks have strong HP growth that responds well to boosting VIT stats, adding a median of close to 200 HP with minimal VIT boosting, and additional 200 HP with a dedicated effort such as the Vit-Maxxing leveling strategy. One thing worth taking note of is the width of the expected variance even at level 15 for these high HP growth heroes. While the median heroes HP increases at predictable rates similar to that of the other stout heroes, 75%+ outcomes for tank heroes can be close to 200 HP greater than the average, meaning Vit-Maxxed tanks can out-HP their non-HP buffed DPS colleagues by 500-600 HP.


Summary


At a high level, while there are certainly minor differences between classes, when looking at HP they can be grouped into 3 groups based on their built-in stats. Heroes who we expect to have strong HP based on their archetype have the potential for it and the squishier hero by and large do not. Moreover, efforts to boost the health of squishier heroes have modest but notable success, but pale in comparison to the impacts of prioritizing HP on high HP and VIT growth heroes. For all classes, there is room to differentiate your heroes both inside and amongst comparable classes by prioritizing VIT early.

To keep this article somewhat manageable I didn’t go through each hero class, but I did do all the work. HP distributions for each class can be found in the appendix. Sort through there to see the results for each individual hero class. Let me know if you see trends I missed, there’s a lot here to sort through!


MeanMachine



Appendix

























[April 20, 2023]