Know when to move and when to stay

Most of the strategy of Warring States' predecessor Cosmoclasm can be applied to the Chinese struggle as well. The two strategic dimensions of the card management and the board management exist in both games. However, there are some important differences to take into account.

The board management

The numerical facts of Warring States are exactly the same as in Cosmoclasm:

The board consists of 12 states; 3 central states bordering 6 other states, 3 semi-central states bordering 4 other states and 6 peripheral planets bordering 3 other states.
Around those states there are 37 strongholds; 13 triple strongholds bordering 3 unclaimed states, 9 double strongholds bordering 2 unclaimed states and 15 single strongholds bordering 1 unclaimed state.
The more unclaimed states a stronghold borders, the more potential majorities can it contribute to and the more valuable it is.

If the placed armies had been permanent as in Cosmoclasm, an army placed in a double stronghold (B) had been twice as strong as an army placed in a single stronghold (C) and an army placed in a triple stronghold (A) had been three times as strong. However, the presence of strategies obscures this view.

Fire is more likely to be used against strongholds with two armies and triple strongholds are more likely to attract armies. Hence, claiming triple strongholds is begging for fire attacks.
Wind is more effective in the single strongholds, from which armies usually get to choose between two double strongholds to move to. Armies in double strongholds are quite flexible in the early game, when the adjacent triple strongholds still have room for a second army, but as the board fills up, their only paths lead outwards to the single strongholds. Armies in triple strongholds will quickly get hemmed in.
Mountain is a double-edged sword. Used in central strongholds, they can be powerful for defence but if your two armies get hemmed in, you have effectively paid two armies for a stronghold. Used in peripheral strongholds, they seldom serve any defensive purpose but they can serve as a strategic reserve ready to send when and where they are best needed.
Forest is the one strategy whose power isn't related to the board position. Whether a state is central or peripheral doesn't matter, it's whether it runs a risk/opportunity of a tie that counts. An army placed in a future tied state is a gained army while an army placed in a non-tied state is usually a lost army.

The card management

Let's now move on to the card management. To win a fight on the board, you must first win a fight with the cards. The outcome of this fight determines two things: when do you get to place armies (the timing) and how many do you get to place (the quantity). However, you cannot have both - if you fall back early, you typically get to place early but less armies, and if you fall back late, you typically get to place late but more armies. So far no difference compared to Cosmoclasm - you have to decide what's most important for you at the specific state.

The tactical considerations are also similar to Cosmoclasm, with the reservation from above that the differences between peripheral states and central states is less important in Warring States:

Central state (6 triple strongholds): All card winners get a triple stronghold so timing is less important than quantity.
Semi-central states (3 triple strongholds): Three of four card winners get a triple stronghold so both timing and quantity is important.
Peripheral states (2 triple strongholds): Two of four card winners get a triple stronghold so timing is more important than quantity.

What is different is when to play for the different strategies.

In the early game, Mountain and Forest are strong, since they bring extra armies to the board. Fire accomplishes less, since removing opponents may simply leave too few armies around the state for it to be resolved. Wind also accomplishes less until you have enough armies on the board to move around.
In the middle game, Mountain and Wind can be a powerful combination for adding and moving armies to the right place at the right time. Fire can also be powerful to break ties in your favor, particularly if used to move two opponent armies from a stronghold at once. Paradoxically, the tie-breaking capability of Wind may be less powerful, since the other strategies will be powerful for avoiding ties at all.
In the end game, there may be too little time to take advantage of Mountain's extra army. Fire and Wind can still be powerful for breaking ties but if the board is clogged by armies, there may simply not be any open roads left for moving armies. Instead, the end game may be Forest's great moment for breaking that decisive tie.

In spite of the more flexible gameplay of Warring States compared to Cosmoclasm, it's still a good sign of a well executed strategy if you enter a middle game where your opponents have claimed one state each but where your total value of strongholds is the greatest, giving you an advantage when the final states are fought over. Remember that unlike Cosmoclasm, states are not permanently controlled in Warring States but can change owners several times. However, this strategy also requires a well managed hand of cards and the means for this differ significantly from Cosmoclasm.

In Cosmoclasm, the card rules limits you to play one suit at the time. In Warring States on the other hand, you can play any cards as long as the symbols are "linked" (similar to previously played cards).

The options this gives you are similar to those of Cosmoclasm but the value of them differs:

Spread your symbols so that you have flexibility to play the symbols that the other players don't play.
Balance your hand between 3-4 symbols and focus on 1-2 of them when you play.
Focus on few symbols and empty your hand each round. Then pass 1 round completely to build up a new strong hand again.

I still prefer option 2, since this gives you the possibility to participate each round and contribute to tied states you're currently not interested in. However, it's also risky if other players play the same symbols, since it's more difficult to switch symbols. Option 1 is more flexible and should give you at least 1 battlefield card each round but there is a risk that you fail to link cards and seldom get more than 1 battlefield card. Option 3 is probably weaker compared to Cosmoclasm, since there are more symbols to fight for and you shouldn't waste all your cards on just a few of them.

As in Cosmoclasm, a balanced hand should also aim for a balance between double symbol cards and single symbol cards (advisors). It may seem that a player with only double cards will spend less cards to win battlefield cards but by adding single cards, you may be able to beat their double cards and fall back before they have the time to respond.

In the example to the right, player 1 has played a double and a single crossbow card and unless any other player has the same combination, player 1 will be able to fall back with the most crossbows, no matter how many the others have left on their hands.

How many cards to spend per round is dependent on the group psychology. If all players are aggressive, you can afford to be that as well, and if all players save their cards, you should as well. The important thing is to maintain the balance in states, armies and cards so that not a single player benefits from the other players' depleted card hands (unless, of course, you're the benefitted player).

How about the spy cards then? They do not allow any armies but rather let the player determine the next state and starting player. This is not very important in the game opening but as your armiess spread across the Empire, you will find states where you have more adjacent strongholds than other players and hence an advantage in upcoming superiority fights. Spies may let you advance the resolution of those states before the other players catch up. You may also choose to delay the resolution of those states, if your card hand is too weak to take advantage of your armies and you want to save them for later. Depending on your hand, you may also want to go first and fall back first (if you focus on few strong symbols) or last (if you focus on many weaker symbols).

A majority of spy cards also gives you an extra card. Not only does this compensate you for the spent card but it also gives you more options for the cards you intend to play the next round. Perhaps you have a strong hand in one symbol that you know another player is also strong in? Save them for later and take three cards with another symbol instead.

Spy cards can be particularly powerful in the end game. At this stage, you have typically decided which state that is necessary to win and started to build up a strong hand to get it. A spy majority will not only let you advance the resolution of this state but also let you get the extra cards necessary to make your hand invincible.

But each spy symbol on your hand also means one less of another symbol, doesn't it? Yes, it does, so when taking a spy symbol, make sure you do it on the expense of a symbol that you don't need anyway. Say that you collect crossbows one turn and are given the opportunity to take a cavalry card. If you don't have any other cavalries, you're much better off taking a spy-crossbow card instead.

Conclusion

To conclude our discussion, you must formulate a strategy in terms of both board management and card management to be able to play both tactically (place armies) and strategically (manipulate armies) where they make the most difference, either by maintaining a balance between the other players or by giving you the edge.