I’ve written about each cards’ points value before, but our latest update to the game is the first time we’ve changed the points of existing cards. So I thought I’d run through the what and why of those changes.
We decided to focus our first round of points balancing on the 13 Warlords already in the game, so that we can ensure each is appropriately costed against the eight new Warlords we’re adding. So let’s look at the point changes we just made:
Captain Acheran – remained the same points (32)
Lieutenant Tolmeron – now 20 points (was 27)
Grand Master Voldus – now 57 points (was 62)
Ghazghkull Thraka – now 40 (was 38)
Boss Zagstruk – now 30 (was 32)
The Corpse Lord – now 25 (was 32)
Commander Shadowsun – now 35 (was 29)
Zephyrblade – now 20 (was 24)
Ghosar – now 34 (was 28)
Servants of the Emperor
Canoness Veridyan – now 29 (was 19) – the largest points increase
Magos Dominus Ivasnophon – now 23 (was 21)
Gloguthrox the Foul – now 17 (was 29) – the largest points reduction
Abaddon the Despoiler – now 33 (was 40)
So why did we make these changes? First up, as I mentioned way back, we plan to use point changes to balance cards which are proving too powerful (make them more expensive) or too weak (make them cheaper). We’ll only adjust a cards’ stats if we’ve made a fundamental mistake with it, as we feel that’s more disruptive than a points change.
We wanted to make the points change now, because although we have thousands of players, that’s still a lot fewer people being exposed to the point changes than in a month or so, after worldwide launch. Also, we’ve found that our current player base is great at sending us feedback, so we’re hoping they’ll let us know what they think of this whole process, the new points, etc.
With so many cards in the game already, and more being added each week, we have to use an automated system to suggest point changes (though we still make the final decision by hand).
Our previous system ‘Dream Crusher’ (don’t ask – one day I’ll get the coders to run through all the ridiculous system names we have) had cards fighting each other hundreds of times. This works well normally, but struggles with ‘utility’ traits and Warlord special rules. For example, the Distract trait has no value in a one-on-one battle, but can be very powerful on the battlefield.
We’re still using that system, but also have an algorithm (no ridiculous name yet) which suggests what each card should cost based on the total value of its’ stats (with each stat points’ value worked out with data from the thousands of battles players have fought).
By combining these systems, we’re hoping to come up with an accurate model for what each card should cost.
This is our first round of point changes, but we intend to work our way through every card at some point, ensuring each is correctly costed for its’ effectiveness. However, the next round of point changes won’t happen for a few months, as we want to use plenty of battles won / lost data to look at the changes we’ve just made.
This is important for two reasons:
- First, it allows us to make further (hopefully smaller) adjustments to cost of the 21 Warlords.
- Second, it’s important that we refine our point balancing system to ensure that what it predicts matches reality. For example, it may tell us that (purely as an example) a cards’ Wounds are its most important stat by far. But from looking at win / loss results for each card we may see that in reality that’s not the case, so we’ll adjust the algorithm to put less emphasis on a cards’ Wounds. The goal is to get the algorithm to match the reality of how the community plays the game as closely as possible, so we know the point changes we’re making are as accurate as they can be.