We need to talk about gems, attributes, and end-game builds.

I see two different types of endgame builds.

• 50 point end game builds
• Mid-game build that was being used when the game ended between 20-30 minutes in.

How often have you gotten to do your full 30 point build in a game? If the answer is rarely, then understanding deck curves is very important.

This is a concept borrowed from multiple collectible card games. It applies not just because we have cards and the same colors, but because our cards cost resources we accumulate over the game.

You start with 1000 gold. This can buy any one point or 1000 gold or less card.

Another minute or so goes by, and you have another thousand or more gold. As time goes by, you have more gold with which to buy attributes(and therefore cards) or to buy consumable/cultivate/combustible cards.

That is your available resource curve, mapped over time, your accumulation of resources.

If your deck doesn't have gold-costed cards, then it doesn't go down.

However, if you do said cards, your curve has a peak - right before you buy the combustible card.

This leads to the second type of curve - the cost of the things you can buy with your accumulated resources.

So think of your twelve cards, laid in piles of the same cost (1 pt cards go in the same pile for example), and draw a graph of your card costs by number of cards.

A lot of decks will gravitate around 3-5 cost cards being the most they have, and this is what you should focus your decks on. You have a couple of useful single point cards, then some that cost two points, more with three, even more with 4 and 5, then less sixes, sevens, eights and more.

The whole reason for this is, and why it is often good deck building strategy, is because you are trying to find the best cross-section of resources available and greatest impact by spending those resources. Every bunch of unused attribute points that could be going towards a cheaper card is inefficient.

(I say often, because there are times where breaking these guidelines is a valid strategy. )

So, quick summation: focus on having a few more cheap cards than expensive cards, because you get a bigger benefit from those, than not using the attribute points while you work towards a much more expensive card.

There's another pressure affecting this as well. We only have three slots, so being able to equip five cards simultaneously is just not possible. So when we're building decks, we're trying to build for that efficient cross-section above, while keeping in mind that the length the game will last is unknown, and that we have limited slots to put cards in.

It's a very hard optimization problem, and there's yet another factor to consider: gems.

The gems at point 25 are massive game-changers. They're all very powerful, and coupled with the attribute bonuses you gained getting there makes it a very important stage.

But how many of you have ever gotten to 25 points in an attribute in most of your PvP games?

By the time you've been able to equip your first end-game card, and are building to the next one, the game is likely over. I see a lot of games where people finish at between 20-30 attribute points purchased. And you can't purchase all in one stat. You need health/mana and regen for both fighters, rangers, and casters.

So, in most cases, equipping a high level gem (25 and sometimes 19 points) is a waste of a limited gem slot.

You want impact with your gems. Ideally, you want your end-game build to give you the stats to fill your role(s), as well as equip all of your gems.

If you can build a deck that uses only gems in the first three slots for Vitality and Intellect, then reaching your 26th attribute point, you get the benefit from all six gems.

That will make you stronger than the enemy where their gems are all powerful very late game slots, and they haven't equipped them all yet.

So like the cards you choose, its an optimization problem dealing with uncertainty and tight constraints.

So all of this is stuff to think of while you're building decks. I took a look at some Kallari builds before writing this, and then looked the player's success with Kallari. Of the fifteen games I looked at (5 builds), there was an average of two gems not equipped by the time the game finished.

SoloQ is unpredictable. You don't know if you're going to get steamrolled, or if its a tense fairly even match, or just something that's a massive grind to win.

But when you're competing in a tournament, you may have a pretty good idea how your opponent's decks work and how the game with them would go.

Therefore, there's ways of taking advantage of that knowledge, to use the extra information to make the optimization problems easier.

TLDR:

• Having more cheap cards than expensive late-game cards makes for a better deck which could lead to winning more
• Being able to slot more of your gems in the a game may help you win more.

(As always, skill plays a big role in winning a match, not just decks and gems. It's a very difficult challenge with very tight constraints. Always consider impact. You may find yourself winning more games with a deck that has a heavier impact earlier on in the game.)