In any game of Guild Ball, there will eventually come a point where the game has potential to end in the near future. This is often when a player reaches eight points, but it varies depending on the teams being played and the board state. As soon as this point is reached, both players’ priorities change dramatically. For any player with the potential to win, influence efficiency and trading up in resources fall in priority heavily, while sheer VP scoring potential becomes the most powerful tool available. Similarly for their opponent, if it prevents the end of the game then any investment of resources is worth giving up. If both players are at this point, the game becomes a frantic back and forth as players try to claw the last few VPs together from wherever they can be found.
If a player has potential to end the game immediately, then naturally high numbered game plans are the first priority. Influence efficiency matters a lot less then it did in the end game, which makes gameplans with low initiative numbers but bonus influence a lot less valuable, since there is a real possibility that the game won’t last long enough to spend that surplus influence anyway. Dig Deep is particularly awkward here since the effect does nothing once the opponent is at 8 VPs. The game plans with useful effects are still relevant, but usually initiative is the first priority. Lone Striker can be awkward even though it’s a +7, but is just fine if you aren’t worried about hitting an armoured or durable target, or you just want to guarantee first activation to get the ball away from the opponent. Grudge Match is exceptional in the late game, since it gives your captain +2 TAC to set up a first activation kill or tackle on a squaddie to end the game, and the downsides are irrelevant if you’re hoping to end the game before the opponent activates anyway.
If the game goes particularly late – to turn 6, or 7 with a minor guild – players pick their game plans back up and replay them, which means you know exactly what numbers they have available and can plan accordingly. This is important when it comes up but comes late enough into the game that it rarely matters. Be aware of what cards have already been played and what cards you discarded – if Seize has already been played and you have Lone Striker in your hand, you know your opponent can’t play a +7 card.
In the late game, your priorities remain somewhat similar – you’re still trying to get the most out of the points of influence you’re able to spend – but the payoffs and effects that actually matter change a lot. Sort of like in the first turn, each team is likely to have a small number of opportunities to actually interact and making the most of those opportunities is the important thing. In a reversal from the first turn, however, it’s the first few activations that matter the most, rather than the last ones. Because the game has high odds of ending before many models have activated, it’s often a good idea to stack up 1-2 models with as much influence as possible, if you know those models will be the ones that go first – allocating influence elsewhere has a high chance of not affecting the game state at all, since it might well not end up being spent. Influence that’s spent in an inefficient way is still more impactful than influence that never gets spent at all because it’s still on an unactivated model when the game ends.
If you have one obvious route to ending the game, then allocate everything you need to maximise your odds of success there, but the rest should go in places where you can attempt to salvage things if your primary plan doesn’t work. If your takeout fails and leaves your opponent on 1HP, you want to have someone else who can do some damage to finish the target off if they run away. Similarly, if your opponent has one obvious route to ending the game, your first priority should be stopping them from doing so if you have any relevant way to spend influence to do so. However, it’s usually a good idea to give yourself avenues to actually scoring VPs yourself rather then going all in on stopping your opponent from winning. This means that if you’re trying to stop the opponent getting takeouts on your model, you should probably also try to put a point of influence on someone in shot range, so that if you get the opportunity you can score a goal. If you’ve already allocated influence to someone to control an enemy VP threat and they don’t have any others available, you should be putting the rest of your influence somewhere it can be used to progress towards actually winning the game.
This idea goes further when your opponent has a way of ending the game available to them that is difficult to stop. For example, as soon as an opponent reaches 10 VPs, you should stop considering whether your model is at risk of dying when determining how much influence to allocate to them, because if they get taken out the game is over anyway, whether they had influence allocated or not. As such, if your only chance of winning the game is for the opponent to mess up or roll badly, allocate as though that is going to happen. If your model is likely going to be taken out and die first activation, set up your influence such that if the opponent fails to do so you can capitalise and try to win. If your opponent is on 8VPs and has the opportunity for a goal shot first activation, set up your influence such that if their shot misses or they whiff their tackle, you can control the ball and reclaim the advantage. This is generally much more likely to happen with goal runs than it is with takeouts, but there is absolutely no cost to planning like this even if the opponent has only very small odds of failing.
The First Activation
The most powerful thing you can do with your first activation in the late game is to win the game. This is far more impactful than generating a resources advantage in the midgame, since there is absolutely no chance of your opponent making a comeback afterwards. With this in mind, if you have an opportunity to win the game with reasonable odds you should pretty much always take it, unless the penalty for failing is particularly bad and you think you’ll be able to get a better opportunity if you are patient and wait until later. Because the game can end straight up on your first activation, doing as much as possible in that first activation is important. As such, captains – with their high max-inf and generally good output – are particularly valuable here. If you want to score a goal or take out a model with the absolute best odds possible, your captain is often the one to do so.
If the opponent is threatening to end the game and you don’t have the option to answer them by ending it yourself first, you need to stop the opponent’s plan. This often involves tackling the ball and booting it to somewhere they can’t use it to score a goal, even if by doing so you sacrifice a model. Control effects are good for this plan, but so is damage – by this stage of the game models are often damaged, and taking out a model that’s threatening to score VPs itself is a great way to get back into a game that’s looking worrying. Be aware, though, that raw denial isn’t enough to end a game on its own (barring clock related situations).
Similarly, if your only chance of winning a game is to roll a spike on an attack and oneround someone you are unlikely to get the takeout on, then you should go for it. The same applies to a one die shot on goal or similar – if your odds aren’t great, but not rolling the dice only makes them worse later on, you should roll those dice. Recognising when your odds are only going to be getting worse is an important thing to consider, usually depending on what ball control options each team has and what health everyone’s players are on. This applies particularly strongly if you have a way of ending the game your opponent may not have considered. Common examples would be AOE based double takeouts, pushing models off the pitch, and scoring goals with non strikers. If you have a possible VP source that is not being considered by your opponent, they are far more likely to give you the time to deal with their plan or win yourself, if they think they have spare time with which to maximise their odds you’re sometimes able to snatch victory from under their nose.
Reasonably frequently, the player that’s currently activating can’t immediately end the game, but the opponent’s next activation is going to do so. They then spend their activation stopping the opponent from winning, and setting up their own team to win next activation instead. Then the opposing player has to do the same thing back again, and so on. The most common examples are when each team is at 8 points but the team without the ball is activating first. The first player then moves a model up, tackles the ball, and passes is back towards their teammates. Then the second player tackles back and passes back to their team, and so on for a number of activations as the ball bounces back and forth and nobody gets anywhere. At the same time nobody can afford to do anything else, since letting the opponent start an activation with possession of the ball already probably results in a loss.
These scenarios can be weird to deal with – especially if both teams are primarily fighting teams, since neither has the long range goal threats to both tackle the ball and immediately score. The things to consider in these scenarios are which team has the most retrieval options – if one player frontloaded all their inf onto three models, and the other spread it around, the player with the more evenly distributed influence is likely to end up with the ball. This scenario happens a lot less with takeouts because models’ ‘takeout prevention’ options are generally a lot worse than a model’s ability to prevent a goal – you can only heal a teammate once per turn, and so if a player puts all their influence into a model in a turn it’s likely to get taken out no matter how much you heal it. The ball, on the other hand, can be tackled any number of times in a single turn. A such, if your team is better at fighting than the opponent, rather than pass the ball back to your team to get tackled again, consider booting it into space somewhere that neither team can reach it. This pushes both teams towards a brawl instead. Be aware that slowing down your own route to 12 isn’t necessarily a mistake as long as the impact on your opponent is greater – the important thing is getting to 12 before your opponent, not getting to 12 quickly – they just overlap often.
Other deadlock scenarios can happen as well, especially if both players are on enough VPs that a single takeout ends the game. If lines have been broken apart again – similar to on turn one – committing a model exposes it to getting ganged up on by multiple opponents. This means that whoever commits first is likely to lose their model, unless you can immediately take a model out in the same activation as engaging. Like on turn one, the team that’s advantaged here is usually the one with the momentum advantage – whether through the ball, or by going in last activation and generating momentum. As above, this means that if neither player has any good engagement opportunities, but your opponent is going to end up with a momentum advantage or the last activation, it can be a good idea to go for engagement beforehand – especially if a spike can get you that takeout, or you can potentially commit with a model that is not likely to get one rounded in response, giving you time to follow up with other models.
The Game Ender
So you or your opponent are actually going for the activation that ends the game. The thing to be aware of is that if this is going to end the game, there’s very little cost to putting every resource you have available into whatever you can. That means bonus timing every attack or kick, and using your counterattacks and defensive stances if possible. Any resources you happen to still have available might as well be spent at some point during the activation. Just make sure you have enough for everything you’re going to be doing – so keep some momentum back to actually pay for your shot. If the play your opponent is making isn’t guaranteed, it’s also work keeping back a point of momentum if possible, if you’re expecting to need it immediately afterwards.
If you’re not close to clocked out, try to take the last activation carefully and make sure you’re taking everything into account. Proxy base out your movement before you take it and work out whether you’ll need to do things like Glide or position to avoid a Rush Keeper. Look over your players’ cards and your opponent’s cards and make sure you are taking everything into account and aren’t going to get caught out by Reanimate or similar effects. Clock time is a resource just like momentum and influence, and if you want to maximise your odds of success then it’s a good use of that resource to make sure you aren’t missing anything which will cause you problems. If you are low on clock, then this naturally goes completely out the window, but it’s still a good idea to be as prepared as possible. That means being as aware as you can of what you need to do to actually win, and making sure you have enough time to actually pull it off. Think about what dice pools you’re going to be rolling and what actions you’re going to be taking while your opponent allocates their influence and takes their activations, which saves valuable moments when you’re actually resolving things yourself.
If your opponent is going for the game end, be aware of what options you have to interact. If your counterattack is only live if the opponent misses their first KD (and you don’t need the MP elsewhere) then you should probably counterattack on the off chance they fail to put you on the floor. Timing your counterattack well is a useful part of dealing with game ending activations – in particular, be aware of what your opponent actually needs to do to win the game. If they are constrained on influence, something as small as a 1″ push can put them out of range of a goal shot, or prevent them from avoiding a countercharge, while even if you can’t get away from engagement a 1″ dodge can get you into cover or away from a ganging up bonus.
If both players are better at denying the opponent VPs than they are at actually getting them themselves, the game often stalls out quite a bit. This is particularly relevant if both players have frontloaded all of their influence into maximising their VP scoring or control abilities, because the important thing in the endgame is to get as much work done per activation as possible, rather than per influence. This means you often have a load of dead activations at the end of the turn where players might move around a bit and spend any leftover influence but don’t get much done.
An important thing to consider in these aftermath situations is that they are extremely punishing for a clocked out player. If you’re clocked and mess up a game ending activation, then there is a very real chance that your opponent is going to get a free 6VP out of the deal as you activate each of your models – it’s not like you can choose not to activate them. Incidentally, this means that killing a player before it activates is actually worth less VPs than killing them afterwards, if they’re clocked out, though it gives them a chance to get away. If you’re close to clocking out and have a single payoff activation that isn’t likely to end the game but will put you at an advantage or able to end the game next time round, it can be a good idea to quickly take all your filler activations first (if the opponent can’t interact with you) to get them out of the way before you go below the clock out point and start giving up VPs. This lets you take two big payoff activations in a row across the turn line while only giving up 1 VP on the clock, where if you took the big activation and clocked out straight away you’d have given your opponent up to 5 additional VPs.
As soon as your opponent has ended up back in a position where they aren’t going to end the game immediately, you can focus on setting up to end the game yourself or put your opponent in a position where they are reacting to you rather than the other way round. The big thing to consider is the momentum race for next turn – maximising momentum generation is very important if the first activation has chances to end the game, and if it was at risk once there’s a high chance it will come around again next turn. If a model activated early in a turn and got close to getting VPs but failed, if it gets to activate first again next turn then you’ll need to do something to stop them from just finishing the job. Just like in the mid game, good ways of doing so include taking the threatening model out, going first and applying control effects, and leaving it Knocked Down which makes its ability to go first and be useful much reduced. If neither player has much left to be doing, keep in mind what each player has to do stuff without influence, like Furious, Heroic Plays, and passive traits – getting a bit of free damage off or an extra couple attacks for a momentum advantage can help a lot with that initiative roll, and denying the opponent those opportunities by engaging Furious players can also come in handy.
In scenarios where the end of the game in sight, the important thing to consider is having a path to 12 VPs. Spending your entire turn scrambling to prevent the enemy from winning has some big issues, mainly that if you aren’t progressing towards your own victory condition things are probably going to turn out badly. In Guild Ball it is generally easier to score VPs than it is to deny VPs – it’s difficult to break out of engagement once your opponent has closed on you, a team generally does more damage in a turn than healing, and positional control effects cost a lot more influence than just moving your own models. This means the game is pretty much always moving towards someone winning it somehow. There’s no way to remove VPs once they have been scored, after all. Because it’s harder to deny the opponent VPs than it is to score them, the best answer to an opponent with the potential to win the game is to get to 12 yourself first. Sometimes that means denying them the opportunity to interact with you – like if you’re 4-10 down, you probably want to try and set up a last activation goal before trying to get another at the start of the next turn, giving your opponent as few opportunities to connect with your models as possible – but the important thing is that you’re still thinking about how you are actually going to end the game yourself. If you only have one takeout threat in a turn, you probably shouldn’t use that model up trying to stop an enemy goal run if they have backup options – you slow the opponent down but give up your own ability to progress in exchange. In scenarios like this, obviously the odds are against you, but the higher odds play is probably to get into a position to actually reach 12 points yourself and hope your opponent fails to end the game in some way.
The end game plays out very strangely because of how quickly it shifts between huge urgency (if someone is threatening to win) and much slower more calculated play if neither player is immediately winning, as set up for next turn is applied. Being aware of which scenario you are in helps a lot, as does being able to make your opponent think the game is in one state when it’s actually in another. While it’s a great idea to have multiple paths to 12 VPs – and a lot harder to disrupt – the most important thing is to have at least one, since the first ‘path to 12’ is the most important one.
This concludes the current series of articles on the different stages of the game. Each stage plays out very differently and needs to be looked at in a different way, as priorities change and different options and tools become available. Understanding the game state is definitely important in Guild Ball, and I don’t think there are that many resources that really go into the actual decisions we make within a game, so hopefully this series has been useful in helping people think about their options. I’ve updated the Article Index with links to all the pages in this series, and hopefully soon there will be some more updates to the Guild pages as new releases come out.
Thanks for reading,