Mana speed is a factor here, but if he didnt dispel before he hit he wouldn’t be that interesting.
1 Like
True but, being Very Fast he is usually among the first to strike so dispelling is more important on his second hit. Unless you use mana pots, which should have no bearing on hero power.
Everything about GG , from the name on, is just absurd in my opinion. I hope they start from scratch on that one.
1 Like
You get more good boards than I do 
Actually I don’t have Kage, I only get to fight against him. If I don’t kill him right away he just fires off at me over and over and over …
Hopefully, this is the impetus for skins for classic heroes. I just hope you don’t have to make a hero look stupid so his/her stats can keep up. I like the look of many of the classics. Thorne is actually one of my favorites in that regard.
1 Like
Thanks all for the kind responses. As I briefly mentioned I had been compiling this alongside other game related projects that I hope to be able to add soon of which this is a part but after plugging in the stats for the new Challenge Event heroes first release in Beta and saw the magnitude of their stats creep (as elaborated in the discussion) I thought it best to bring forward posting this part of the work for community discussion whilst they’re at a point when they’re more easily adjusted if it is deemed a concern for the wider community.
Well absolutely hero stats is just one piece of the puzzle, together with the special skill it informs the complete picture of the hero. Up until now most judgement on a heroes performance has primarily focused on just the hero’s special skill and when judging stats done on a bit of gut instinct. This information is meant to be taken together with what people already know about the hero’s special skills.
I for one wouldn’t ditch Alberich for Lady of the Lake (why she isn’t Ice for the lake I don’t know) has his heal seems more useful and I love that Mana Regeneration but where it is clear that some heroes on the wrong end of the toilet brush in both special skill and stats then there is a strong case to be made for them getting a buff the next time SG is looking to do some re-balancing. Examples being comparing Thoth to Seshat or Atomos who along with having a widely derided special is bottom of the Atlantis table.
It is also worth considering whether we’ve had less “blow you away” amazing skills of HOTM recently in comparison with the earlier HOTM because when you factor in the increased stats, elemental link, immunity etc etc that the margin for how amazing the skill can be before the hero is too OP is slimmer. Raises an interesting question as to whether the community would prefer HOTM continuing like as is or almost have a reset and drop back down in stats a bit and lose the additional bonuses like immunity to have Hel-grade specials. Though then again some of the upcoming HOTM specials are looking pretty hot.
2 Likes
Continuing on from reasons in above post in that it is to clarify the stats side to be used in conjunction with judgement of special skill to judge heroes I have opted not to include mana speed in the analysis. Including mana speed here I don’t think can be fairly done without some judgement on the whole skill as the quality of the skill itself as whilst mana speed is a factor in how good a special is but it isn’t totally independent of the quality of the special itself. Thoth may be Fast but his special isn’t as desirable as Hel’s Average speed special. If mana speed were to be factored in the quality of the special would also need to be quantified and that’s a minefield I’d rather not wade into, especially as the usefulness of a special to individual players can also depend on your bench so varies from person to person. As such separating the quantitative stats to help inform people’s own assessment on the qualitative quality of the special and its speed seems the most appropriate way forward.
There’s two factors I believe are important to consider when viewing the numbers, firstly that as we are all aware by now from Titan battles that as the difference between an attacker’s attack and the defender’s defence grows the damage done grows quite significantly to the point that with all relevant +attack and -defence modifiers in the equation that tile damage hits is in the thousands. Whilst the stats creep here appears low on the surface its effect on damage done and damage taken will be more pronounced by virtue of how the damage formula works.
The second factor to consider is Troops, attack and defence stat modifier effects and most immediate heals are based on percentages of the heroes stats that the stat boosts from stat effects and troops is amplified by the higher stats as is the amount of health you recover from a percentage heal as health stat increases.
The stats creep is just one more thing that needs considering alongside special skill improvements and new features such as immunities and family bonuses when determining whether a new hero may be OP or not. Kingston vs Athena would be a good example to look at as they both have the increasing stat debuff for each tile hit mechanic. Comparing the two Kingston’s stats have risen by 0.72% on Athena, his skill become faster, damage done to one hero, debuff expanded to all enemies, gained elemental link and an immunity. Considering there has been more than 2 years of game progress between the two heroes Kingston’s improvements I would probably say were justified – especially considering the stats increase between the two is much less than the overall stats creep over the two years. The case however for the new Challenge Event heroes in beta is harder to make based on their current stats – what is essentially more than two years of stat creep between them and the last Atlantis 5* releases just a few months ago is a good deal out of step with the so far established rate of stats creep.
As @Garanwyn has noted already going up by the +Attack, +Defence and +Health nodes each comparatively has the same effect on Stats Power. If you have two heroes of the same Stats Power to start with and regardless of whether you take +Attack, +Defence or +Health routes provided both heroes are at the same class level they would have an equal Stats Power to each other. Where it makes a difference is for the xAttack, xDefence and xHealth as they work of the heroes maxed stats before embleming. As heroes of the same Stats Power can have different spreads of stats taking the xBoost path that accentuates your stronger stats can result in one of your two heroes pulling ahead in Stats Power.
Where there is a great deal of potential with the Class Trees is the potential to rebalance stats for heroes with good specials who suffer from a glaring weakness in their stats. To take an example from 4* heroes Sir Lancelot has a high Attack stat, does reasonable damage to a single target for a 4* and gives nearby allies both a decent Attack boost and Mana Generation boost. The Mana Generation boost now having additional potency as any Average hero with that has the +2% Mana Boost emblem will have 2 tiles shaved off their charging speed giving Sir Lancelot the potential to turn three average heroes temporarily into fast heroes. In a 4* Very Fast Raid Tournaments he can shave a tile off himself and any hero next to him and Buff Booster he’s applying 2 buffs to three heroes. He is however his downside is he is a tad squishy and because of that generally overlooked however the Fighter Class Tree probably rewards high Attack, low Defence heroes thanks to the x4% Attack node at the end and the ability to go potentially go through 7 Defence nodes and a xDefence node. As such it is possible to give Sir Lancelot a near 25% defence boost and taking the attack stat to 816. For anyone looking to specialise in 4* heroes he is certainly a good second project after Boldtusk (who whilst doesn’t quite benefit as much stat wise as Lancelot from the tree Boldtusk has better potential to make Revive skill more punishing). Whilst there are infographics floating around recommending which heroes to emblem and which seem based on the pre-Emblem performance of the hero and how suited the Class Talent is to that hero. What seems to have been missed is that certain class trees may have the potential to provide a more significant performance improvement to some heroes in the class than others in the same class. I hope to have more on this point in the coming weeks that may help people identify some heroes in their rosters that can be nicely polished up for certain use cases.
4 Likes
Ok … but how do you account for the fact that the three least desirable heroes in the Season 1 roster are the highest ranked in your analysis? How many people willingly use Quintus in their team for either attack or defense, yet he has the third highest stats power just behind Thorne and Elkanan, and well above the (Season 1) stalwarts of the game like Lianna, Sartana, Magni, joon …
Something here just doesn’t quite add up.
There’s less than a 1% spread between the stats of all S1 heroes. At that point, special strength/effectiveness, which is not factored at ALL into CP, totally dominates the equation.
1 Like
Thanks and no worries here is the analysis for the 4* heroes. Indeed this data really grew out of a project aimed looking how especially C2P and F2P players could maximise the potential of their 4* heroes with the class system by allowing them to see a comparison of all the combinations of stats that is possible for a specific hero to get with their class tree, rank each combination by a particular stat or compare them to other heroes etc etc. Essentially something that can help to plan rosters for War, Class Quests and Tournaments. 5* heroes were added as to the data for the project for a later release as they were anticipated they would be asked for but with the sheer number and rate of new releases sorting those would delay initial release much more and I’m yet to consider how I would keep up with the rate that 5* heroes are added.
Just before getting on to the tables one question I think that some will have is how close to 5* stats can you get with a fully emblemed 4* hero. So here goes:
The lowest ranked 4* in terms of Stat Power is: Hu Tao. If you decided you wished to emblem up Hu Tao and took a route through the tree that did the least to increase his Stat Power then he would end with a Stat Power score of 622.65. This is -1.29% of the Season 1 5* Hero Stat Power Average.
The highest ranked 4* in terms of Stat Power is: Danzaburo. If you decided to emblem Danza and chose the path that maximised Stat Power then Danza would finish with a Stat Power score of 640.99. This is +1.62% of the Season 1 5* Hero Stat Power Average.
So in terms of embleming 4* heroes you can get between -1.29% and +1.62% of the Stat Power of the Season 1 5* average. Though the attack power of their specials is lesser than the specials of the 5* the emblemed 4* won’t have cost you any 4* Ascension Materials to power up – and you can always claim back the emblems later on if your roster health changes. It is also worth remembering that the developers also test whether some specials they think may be OP on 5* heroes out on 4* heroes to gauge them – now you can power a few up to 5* level.
Anyway on to the Stat Power Analysis of 4* heroes that are maxed BUT NOT EMBLEMED:
Season One 4 Star Hero Stats Power
The Average Stats Power for Season 1 4* heroes is: 557.67. This value will be used later to compare the percentage improvement of season 2 and event 4* heroes vs the Average Season 1 4* hero.
Stats Power Table for all Season 1 4* Heroes
| Hero | Collection | Stats Power |
|---|---|---|
| Kiril | Season 1 | 562.8 |
| Rigard | Season 1 | 559.93 |
| Wu Kong | Season 1 | 559.93 |
| Boldtusk | Season 1 | 559.86 |
| Skittleskull | Season 1 | 559.86 |
| Kelile | Season 1 | 559.23 |
| Sonya | Season 1 | 558.67 |
| Caedmon | Season 1 | 558.53 |
| Colen | Season 1 | 558.32 |
| Boril | Season 1 | 558.25 |
| Kashhrek | Season 1 | 558.18 |
| Grimm | Season 1 | 557.76 |
| Gormek | Season 1 | 557.55 |
| Chao | Season 1 | 557.55 |
| Cyprian | Season 1 | 557.13 |
| Little John | Season 1 | 556.99 |
| Scarlett | Season 1 | 555.87 |
| Li Xiu | Season 1 | 555.52 |
| Sabina | Season 1 | 554.89 |
| Tiburtus | Season 1 | 554.54 |
| Melendor | Season 1 | 553.98 |
| Hu Tao | Season 1 | 553.35 |
Current Event 4 Star Hero Stats Power
Without the upward stat pressure brought by introducing a new 4* to the game each month the Stat Power gain has been much more modest with a number of these heroes having Stat Powers less than the better Stat Powers of the higher ranking Season 1 heroes. The Average Stat Power increase for these heroes is 562.53 which is just a +0.871% increase on Season 1 4* hero average.
Stats Power Table for all Event 4* Heroes
| Hero | Collection | Stats Power | Comparison to Season 1 4* Heroes |
|---|---|---|---|
| Jack of Diamonds | Wonderland | 565.6 | +1.422% |
| Cheshire Cat | Wonderland | 564.48 | +1.222% |
| Gaffar | Summer | 564.41 | +1.209% |
| Buddy | Christmas | 564.27 | +1.184% |
| Jabbar | Summer | 564.27 | +1.184% |
| Valeria | Halloween | 564.27 | +1.184% |
| Jack O’Hare | Easter | 563.29 | +1.008% |
| Falcon | Teltoc | 561.96 | +0.770% |
| Hansel | Grimforest | 561.89 | +0.757% |
| Gretel | Grimforest | 561.61 | +0.706% |
| Jackal | Teltoc | 560.91 | +0.581% |
| Lancelot | Avalon | 560.42 | +0.481% |
| Peters | Pirates | 560.35 | +0.481% |
| Boomer | Pirates | 560.21 | +0.456% |
| Merlin | Avalon | 560 | +0.418% |
Season Two 4 Star Hero Stats Power
Similar situation as for the 5* Season 2 heroes the Season 2 4* heroes have the best Stat Totals for 4* heroes currently. The Average Stats Power for Season 2 4* heroes is: 565.36 representing a +1.377% increase on the Stat Power Average compared to Season 1’s 4* heroes.
In terms of stats Season 2 heroes are CURRENTLY the most powerful 5* in the game with the best Stats Power scores in the game. The Average Stats Power for Season 2 5* heroes is: 648.75 representing a +2.845% improvement in stats over the Average Stats Power for Season One 5* heroes. Whilst the heroes were released incrementally there exists artwork released before Season 2 was first released in which obscured versions of many cards can be seen (if not fully read at the time) indicating they were likely originally all devised around the same time with just beta tweaks months before they would be drip fed into the game so a comparison by family wouldn’t reveal anything but here is the table for individual heroes.
Stats Power Table for Season Two 4 Star Heroes
| Hero | Family | Stats Power | Compared to Season 1 Average |
|---|---|---|---|
| Danzaburo | Sakura | 567.42 | +1.749% |
| Triton | Atlantis | 567.14 | +1.700% |
| Gobbler | Lagoon | 566.23 | +1.535% |
| Sumitomo | Sakura | 565.81 | +1.460% |
| Proteus | Atlantis | 565.11 | +1.335% |
| Wilbur | Lagoon | 564.97 | +1.309% |
| Ameonna | Sakura | 564.97 | +1.309% |
| Agwe | Lagoon | 563.64 | +1.071% |
| Gadeirus | Atlantis | 562.94 | +0.945% |
All 4 Star Hero Stats Power
As there is a good deal of overlap between them and fewer of them here is another table of all the 4* heroes in the same table to compare and contrast across the groupings.
Stats Power Table for all 4* Heroes
| Hero | Collection | Stats Power |
|---|---|---|
| Danzaburo | Season 2 | 567.42 |
| Triton | Season 2 | 567.14 |
| Gobbler | Season 2 | 566.23 |
| Sumitomo | Season 2 | 565.81 |
| Jack of Diamonds | Wonderland | 565.6 |
| Proteus | Season 2 | 565.11 |
| Wilbur | Season 2 | 564.97 |
| Ameonna | Season 2 | 564.97 |
| Cheshire Cat | Wonderland | 564.48 |
| Gaffar | Summer | 564.41 |
| Buddy | Christmas | 564.27 |
| Jabbar | Summer | 564.27 |
| Valeria | Halloween | 564.27 |
| Agwe | Season 2 | 563.64 |
| Jack O’Hare | Easter | 563.29 |
| Gadeirus | Season 2 | 562.94 |
| Kiril | Original | 562.8 |
| Falcon | Teltoc | 561.96 |
| Hansel | Grimforest | 561.89 |
| Gretel | Grimforest | 561.61 |
| Jackal | Teltoc | 560.91 |
| Lancelot | Avalon | 560.42 |
| Peters | Pirates | 560.35 |
| Boomer | Pirates | 560.21 |
| Merlin | Avalon | 560 |
| Rigard | Original | 559.93 |
| Wu Kong | Original | 559.93 |
| Boldtusk | Original | 559.86 |
| Skittleskull | Original | 559.86 |
| Kelile | Original | 559.23 |
| Sonya | Original | 558.67 |
| Caedmon | Original | 558.53 |
| Colen | Original | 558.32 |
| Boril | Original | 558.25 |
| Kashhrek | Original | 558.18 |
| Grimm | Original | 557.76 |
| Gormek | Original | 557.55 |
| Chao | Original | 557.55 |
| Cyprian | Original | 557.13 |
| Little John | Original | 556.99 |
| Scarlett | Original | 555.87 |
| Li Xiu | Original | 555.52 |
| Sabina | Original | 554.89 |
| Tiburtus | Original | 554.54 |
| Melendor | Original | 553.98 |
| Hu Tao | Original | 553.35 |
5 Likes
Please note that this is an analysis of POWER CREEP OF HERO STATS and not an analysis of which hero is the best. The purpose is to show at what rate the stats of heroes introduced into the game have been increasing. As mentioned before the quality of a hero is comprised of its stats and its special skill of which this analysis is looking at just the increase of stats alone.
If you want to try and calculate a way of numerically analysing the best heroes please feel free to expand upon this by quantifying each hero’s skill to factor into this. The challenge you face is how to quantify a skill like Wilbur’s which is average that does no damage or any healing. You can’t simply assess it by the speed of the skill and the damage or healing of the skill as once again Elkanen would beat Wilbur but most would agree Wilbur has the better skill.
Now as this isn’t about which hero is the best but simply the growth of hero stats. This does mean Elkanen comes top of Season 1 heroes because he has the highest Stat Power total - Elkanen’s problem isn’t that he hasn’t got enough stats it is that the distribution of his stats is heavily towards his defensive stats but his special skill depends on having a high attack stat. If Elkanen were to receive a Thorne Version 20 style makeover to have some of his defensive stats moved into attack (an example would be 759 Attack, 702 Defence, 1220 Health - incidentally giving the same Stat Power Elkanen currently has) he would definitely be more highly desired for the increase in both his fast attack and the healing he would receive in return.
Elkanen is a good example that when it comes to the Class Trees for heroes it is less the Total Stat Power that matters but more HOW those stats are distributed… Unfortunately for Elkanen he’s in the class that probably has the worst possible tree for him to improve on his shortcomings.
5 Likes
One would probably be better served by creating some axes to assess the specials along, and then giving ratings for each special on each axis.
2 Likes
I like the idea and given the intellectual challenge of it would probably give it a shot after I’ve finished my to do list on the current projects, the only thing that might put me off though would be the inevitable “why did you rate this special more than that one,” posts I’d be neck deep in… Not that isn’t worth trying, I like a challenge!
1 Like
Just remember: complaints about a free rating system need to be submitted by registered mail 
3 Likes
With a stamped addressed envelope enclosed if they want a response…
2 Likes