Given how much I like strategy games, it's no surprise that I have formalized opinions on how they should be designed - and consder them objectively superior to yours :P Here are (what should be) the design goals of any strategy game in a loose order of priority:
- Maximize the skill ceiling.
- Minimize RNG.
- Balance the game.
- Avoid feedback loops, especially negative ones.
- Avoid tug-of-war gameplay.
- Maximize interaction between game elements.
- Maximize reactivity.
- Maximize variety.
- Minimize match duration.
- Minimize complexity of the rules.
- Maximize design space for game elements.
This is the most important one because once you reach the skill ceiling, the game is completely ruined. No one would want to play chess if they knew the perfect move in every situation. And the worst part is, to be really into a strategy game, you have to be working toward that end, at least passively. That's why it's absolutely essential to raise the skill ceiling as high as possible when designing a strategy game.
This is the second most important objective because RNG is the antithesis of skill. While it can be kinda funny sometimes to see the tide of a match completely turn due to RNG, it's really frustrating when you're trying to actually have a battle of intellect with someone, and that's the real point of strategy games. Games without RNG are much more satisfying and arguably healthier for the human spirit because not only do they make your victories feel better, they force you to admit that you're responsible for your losses, which can help foster a sense of personal responsibility for your failures in real life.
Note that balance between first/second player in a turn-based game falls under RNG, since usually the person who goes first is determined randomly in automatched play, so you just want to make it not matter.
Obviously, it's important to have game elements (ie. cards, units, characters) all be at an appropriate power level. If any element is too powerful, high-level play will revolve around it to the exclusion of all others, and if it's not powerful enough, it just doesn't see play. Either way, it's bad. You want every game element to see play in a reasonable variety of situations and not overshadow any others.
Note that this doesn't necessarily mean everything should be equally powerful. In a CCG, it probably means that. But in a game like Prismata, it's okay for a unit like Centurion to be much more powerful than other units, because since it's only role is absorber - which makes up only a small part of your strategy - it doesn't overshadow other units by being so powerful (except other pure absorbers, but those tend to be pretty powerful as well).
A positive "feedback loop" is a game dynamic where being in the lead actually makes it easier to get even farther in the lead (and by extension, being behind makes it harder not to fall even further behind). This is present in a lot of games where you control units and attempt to destroy or capture enemy units, such as Chess: the more pieces you lose, the harder it is to protect the pieces you still have. Even Prismata has this problem. It's a problem because it makes it so that a small early mistake can lose you the match even if you play much better than your opponent for the remainder of the game. This can only be upsetting to the player. Still, it's not the end of the world for a game to have this issue since you can just resign once you get behind and go into the next game.
A negative feedback loop, on the other hand, is a far worse problem. This is a dynamic where the closer you get to winning, the harder it gets to stop your opponent from catching up. This is bad because it makes good players feel that they're being punished for playing well, which leads to a lot of salt. It becomes a complete deal-breaker when combined with tug-of-war gameplay (see below).
Tug-of-war gameplay is when the two sides' goals are strict opposites, so that making progress toward winning is the same thing as undoing your opponent's progress. This is bad because it means the game can never be close: if one side is close to being eliminated, the other side is necessarily not close. You can never have a climactic moment in a tug-of-war game. It also leads to a lot of very long games, since if the players are equally skilled they'll just keep undoing each other's progress, making it extremely difficult for either side to win.
By interaction, I mean that game elements should not perform the same in all situations. Each unit should have things it's good against, things its weak against, things it combos with, etc. Prismata is a game full of emergent interactions between its units. I could write an entire article just listing them.
This is the biggest flaw with Dominion: you don't need to look at what your opponents are doing at all to play the game well, most of the time. But human interaction is fun. So (all other things the same) we'd rather play games where the best move depends heavily on what your opponent is doing, like Prismata and Go.
This is the biggest flaw with Chess: every game starts the same. The same pieces and openings are available. And as a result, the optimal strategy is the same every game. Dominion and Prismata show a fun alternative: different units/cards are available for purchase in each game, so you get a lot of different types of games. Prismata has games where the optimal strategy is a rush, games where the optimal strategy involves a huge economy, games where the optimal strategy involves making a lot of sacrifices to cheese out a powerful legendary unit early on, games where the optimal strategy is timing a concentrated burst of damage to crush your opponent on the spot, you name it. I am convinced that Dominion and Prismata have the perfect format for a strategy game. It doesn't seem to have any drawbacks.
The longer a game takes, the less opportunities people will find to play it, and the more you feel like you've lost when you lose. There really is just no drawback to shorter match times. You might argue that excessively short matches don't allow for many of these other objectives to be reached, but that's only a result of the matches being too short. It's not the shortness itself that's the problem.
A game that achieves deep strategy by using complicated rules can be fun, but it's better if it achieves deep strategy through simple rules, like Go. Go has basically one rule and is probably the deepest strategy experience ever created by humans. It's so elegant. I'm just in awe of the game's design every time I play it. It also makes it easier for newcomers to pick up the game.
It really helps the longevity of a game to be able to add new units (or cards, characters, etc), and design space is basically how long you can do that for without running out of possibilities. Prismata has a small design space because each unit can only have one click ability (due to the way the UI is designed), isn't allowed to break Shalev's rule, etc. In fairness, the restriction on multiple click abilities could be lifted fairly easily if the devs were willing to complicate their UI a little more, and Shalev's rule is pretty much a self-imposed limitation (though I do agree it's for the better). Most CCGs, on the other hand, have a huge design space.