Recommend
10 
 Thumb up
 Hide
24 Posts

Space Alert» Forums » General

Subject: An Analysis of Resources in Space Alert rss

Your Tags: Add tags
Popular Tags: [View All]
Dylan Gould
United States
San Francisco
California
flag msg tools
badge
Avatar
mbmbmbmbmb
* Is it better to just destroy this threat myself, or should I ask for help?
* How many actions should I take to save an energy cube?
* Is it worth interrupting what everyone is working on to ask who jiggled the mouse?

I want to look at these questions by using the example of a very simple enemy setup in a partial game; in fact, I’d say this is the simplest possible setup. Also, I’ll focus on only half the game – 2 players and 2 zones – in a game with 4 crew members on the Sitting Duck (the standard game board).

The setup is the following: A Fighter (E1-07) appears in the red zone on turn 2 (t+2: threat, zone red). The Fighter has 3 speed, 4 health, 2 shield, and attacks on X, Y, and Z. I try to generalize my analysis to any trajectory.

To Collaborate or Not to Collaborate?

That is the question I’d like to focus on: whether ‘tis more efficient to get the job done yourself without involving the other players, or whether it’s worth saving a little energy by bringing down the Fighter together with teamwork.

Thus I will consider 2 distinct strategies: Strategy A and Strategy B. In Strat A, a single player (Yellow) plays actions Left (L), A Button (A), A Button (A). In Strat B, Yellow still moves left and presses the A button, but in this case he only presses A once, because he waits for a second player (Purple) to move into position and press A at the same time. Yellow plays L, Wait (blank), A; Purple plays L, Deck-change (D), A.

Strategy A 1 2 3
Yellow: L A A

Strategy B 1 2 3
Yellow: L A
Purple: L D A

In both cases, the team does 4 damage to the Fighter by turn 3, destroying it so that it doesn’t get to execute its second movement. In Strat A, they use 2 energy cubes, but in Strat B, they use only 1 energy cube. So Strat B is more energy-efficient.

Does that make Strat B better than Strat A? Energy is an important resource in Space Alert. There are some ways you can make a difference against external threats without spending energy – the lower deck laser cannons, rockets, and interceptors – but these only do 2 or 3 damage. Against most threats’ shields, this barely makes a dent. Those weapons all seem to be supplemental to the upper deck laser cannons, which need to deplete the ship’s energy in order to have a substantial impact on destroying threats. And of course, energy also fuels the shields to protect you from damage. Damage makes you lose the game as well as severely reducing your score if you win; avoiding damage is of the utmost importance. So you should always choose an energy-efficient strategy, like Strat B, right?

In Space Alert there’s another resource we should try to use efficiently: actions. It may help to think of them as action points; each player gets just 12 action points per game, and if you’re spending a disproportionate number of them on one easy threat, then you might find you just don’t have enough action points later to take care of something more serious.

Looking back at Strat A and Strat B, it’s obvious which one is more action-efficient. Strat A uses 3 actions to kill the Fighter on turn 3, while Strat B uses 6 actions to kill the same threat by the same turn. (The Wait counts, since you can’t take it with you.) If you employ Strat A, then Purple can go and deal with some other threat from some other part of the ship. So Strat A is more action-efficient.

You make a tradeoff when you choose one strategy over the other, Strat A instead of Strat B, for example. I’ll put it more explicitly: If you choose Strat A, you spend 2 energy and 3 actions. If you choose Strat B, you spend only 1 energy but 6 actions. So if I compare the strategies, Strat A saves 3 actions at the expense of 1 energy; conversely Strat B saves 1 energy at the expense of 3 actions.

Let’s take it one step further and consider whether this tradeoff is a fair trade. Which is worth more, actions or energy? I don’t know at this point, but the question suggests a way to choose between the strategies: Find an independent method of valuing energy in actions. If 1 energy turns out to be worth less than 3 actions, then Strat A is better. But if 1 energy turns out to be worth more than 3 actions, then Strat B is better.

One factor I’d consider is the total available amount of each resource. The assumption here is that the designer, Vlaada Chvatil would have given us about the right amount of both resources, so that if we overuse one in part of our plan, we should conserve that amount in other parts by overusing the other resource. The initial amount of energy is always 10, with 3 fuel capsules that can be converted to a maximum of 15 energy, ignoring heroic/special actions. This gives us a total of 25 energy, but it assumes superhuman efficiency in draining and refilling the reactors, so I’d suggest about 22 energy (4 per capsule) as a realistic upper limit. In a 4-player game, the players have a total of 48 actions, but in a 5-player game, they have 60. This suggests that the valuation ratio between energy and actions changes depending on the number of crew members, and that makes intuitive sense as well – a 5-player game only adds 1 regular threat, or 14% more threat value, while it adds 25% more actions, when compared to a 4-player game. The extra actions should give the players more room to waste them and still be able to win the game. As for the ratios, we have 22 energy : 48 actions in a 4-player, or 1 : 2.2. And in a 5-player, we have 22 : 60, or 1 : 2.7. In both cases, an energy is worth less than 3 actions, so this line of thought suggests again that Strat A is better.

I think that the assumption I made that Chvatil gave us close to the just-right amount of each resource is questionable. To more accurately value energy against actions, I would measure the energy and actions used in a large number of games won by the players. This should result in a more accurate valuation for both energy and actions.

Pulling Focus

I want to introduce another resource that is a less obvious, but no less important, part of Space Alert. I call it focus. Focus is defined as the attention of a given number of players for a given amount of time. I think of it in terms of units that I call focus units, creatively. 1 focus unit is equal to 1 player paying attention to something for about 12 seconds. So a full, single-action, 4-player game gives the players 200 focus units.

There are a couple of different types of tasks that use up the players’ focus units in Space Alert. These are, generally, calculation tasks and communication tasks. Whenever you do the mental arithmetic needed to determine how much damage a threat will take from an attack, how far away it is, or what actions you need to take before you can hit it, you’re doing a calculation task. Each mission requires the players to perform a large number of calculation tasks, but each task can be performed by just 1 player. Communication tasks happen when a player needs to tell any number of the other players what to do to coordinate an attack or avoid duplicating work and getting in each other’s way. There are generally fewer communication tasks than calculation tasks, but in contrast, communication tasks use up multiple focus units at the same time, since by definition, multiple players must pay attention to communication, even the ones who are merely listening.

Here’s how I break down the tasks required to follow Strat A:
1. Yellow calculates the L card.
2. Yellow calculates the (first) A card’s effect on the reactor.
3. Yellow calculates the (first) A card’s effect on the threat.
4. Yellow calculates the (second) A card’s effect on the reactor.
5. Yellow calculates the (second) A card’s effect on the threat.

Strat A therefore uses up about 5 focus units of the team’s focus resource pool. (Note that it may not actually take a player 12 seconds to do these calculations; 12 seconds is really just an example – I’ve never measured it, so I don’t know how many focus units are spent on a successful mission, and I don’t have an average length for a focus unit. I’ll put it on my to-do list.) The important thing is how this compares with Strat B’s cost in focus units, which I broke down as follows:
1. Yellow calculates his L card.
2. Purple calculates her L card.
3. Yellow and Purple communicate that they will fire at the same time. (2 focus units)
4. Purple calculates the D card.
5. Yellow and Purple communicate that Purple cannot be in position to help until t+3, so Yellow will have to spend an action waiting. (2 focus units)
6. Yellow calculates the A card’s effect on the reactor.
7. Yellow and Purple calculate/communicate their A cards’ effect on the threat. (2 focus units)

With Strat B, the team is spending 10 focus units, so again, Strat B comes out costlier than Strat A. Note that tasks 3 and 5 each use up 2 focus units, because 2 players are paying attention to those tasks. Also note that task 7 is a bit hand-wavy, since it’s possible each player first calculates the damage and then they communicate about it afterwards, spending a total of 4 focus units. However I’m pretty sure that would overstate the resource usage of these moves, and that it’s probably best to allow some gray area about whether a task must fall distinctly under either category, calculation or communication.

Putting it all together, Strat A spends 2 energy, 3 actions, and 5 focus units, while Strat B spends 1 energy, 6 actions, and 10 focus units. The winner seems clear – your crew should go with Strat A as much as possible, right?

Well, I feel confident that’s true in this highly simplified example. And my experience playing the game tells me it’s also true in many missions, but not in every mission. I feel equally confident that experienced Space Alert players will agree with me when I say that in some situations, Strat B is better. I also believe that there are some things you can do to improve Strat B without drastically increasing its cost; the same cannot be said for Strat A.

Here are a couple of options:
Strategy B’ 1 2 3
Yellow: C L A
Purple: L D A

Strat B’ provides for the team handling the red zone threat to also jiggle the mouse, which frees up the other crew members to go work on other threats in other parts of the ship.

Strategy B’’ 1 2 3
Yellow: L B A
Purple: L D A

Strat B’’ makes use of the energy cube saved by Strat B and also fills in the wasted Wait action with a more useful B button (B) action, boosting the red zone shields so that they can stand up to more damage.

When is Strat B (or especially its descendants) better? It’s incrementally better in a 5-player game, where the extra player gives your crew both extra actions and extra focus units to use towards winning the game, as mentioned above. It’s also a little better when the threat comes in on trajectory T2 or T5, since the Fighter gets an attack before being destroyed; Strat A leaves 0 energy in the red zone shields in this case, and 0 is a lot less than 1. It’s substantially better when there’s another red zone threat incoming right on the Fighter’s tail. And it’s particularly better when that threat is a Destroyer or Plasmatic Fighter, threats that reward the use of shields.

A Word about Cards as a Resource

Another resource players may wish to consider is cards, but I don’t actually count them as a resource. Each player starts the (4-player or 5-player) game with only 5 or 6 cards, depending on which expansion components you include. These are all you get for phase 1, and you should expect to use 3 of them. Statistics tells us that when you have a little sample size, you’re more likely to get a disproportionately large or small number of a given result than you would be if you had a big sample size. This means that you might be constrained by the fact that you have too few of a symbol you need in order to complete your task, so these constraints may seem like a resource squeeze. However the group as a whole, as well as each individual in it, has well above the number and variety of cards they need in order to win the game, every single game. In addition, phase 1 always has at least 1 incoming data and at least 1 data transfer, so you should always be able to get the cards you need, as long as you’re communicating and planning sufficiently. I think that the purpose of the cards is to provide one more need for calculation and communication tasks, and hence acts as more of a focus unit-sink.
7 
 Thumb up
1.00
 tip
 Hide
  • [+] Dice rolls
Jan Bazynski
Poland
Warsaw
Mazowieckie
flag msg tools
publisher
badge
Avatar
mbmbmb
fishboy wrote:
As for the ratios, we have 22 energy : 48 actions in a 4-player, or 1 : 2.2. And in a 5-player, we have 22 : 60, or 1 : 2.7. In both cases, an energy is worth less than 3 actions, so this line of thought suggests again that Strat A is better.


Yes, but as you added later in your post, wasting an action on waiting in this situation is just bad play. If the player did something useful instead then the team spends only 2 actions more to save one energy, which is better than the ratios above.

fishboy wrote:
1. Yellow calculates the L card.
2. Yellow calculates the (first) A card’s effect on the reactor.
3. Yellow calculates the (first) A card’s effect on the threat.
4. Yellow calculates the (second) A card’s effect on the reactor.
5. Yellow calculates the (second) A card’s effect on the threat.

Strat A therefore uses up about 5 focus units of the team’s focus resource pool.


I don't agree that this is that simple. After these actions there is no more energy left in the red reactor. I would not be happy if someone with whom I play simply used all the energy and left it that way. So now Yellow must communicate that there is a lack of energy in the reactor and probably think of the best solution to re-supply (do it by himself, ask someone to do it, check if there is any energy left in the central reactor etc.) From my experience it is re-supplying reactors that costs the most focus.

With no other data I would rather go with strategy B'. Not knowing what will happen later on I don't want to leave the red reactor empty by either shooting twice or powering up the shields which I am not sure will be useful at all. The costs in actions and, as you call them, focus units of re-supplying the reactor afterwards is in my opinion too big to go for strategy A.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jack Spirio
Austria
flag msg tools
Avatar
mbmbmbmbmb
I have to disagree with you.

1. If there is no Energy left (like in Strat A) you have to fill it up which costs more Actions.

2. Mostly at the start there are not so much threats, so it is good if player Purple uses his actions to go to another station. So you can combine that with yhooting at the fighter. The next turn he can take the battlebots, shot a rocket or load up the Energy.

1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Cameron Chien
United States
Rancho Cucamonga
California
flag msg tools
mbmbmbmbmb
I find this in-depth analysis fascinating, but in my group I highly doubt any of this would be remembered and used, "in the heat of battle".

Usually the problem is making sure the right people have cards to take care of a threat to begin with, let alone try and optimize.

Cameron
3 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Grant Johnson
United States
Cedar Park
Texas
flag msg tools
Well me known for eating cookie when me don't they shout
badge
Look! He try to throw loyal fans a curve!
Avatar
mbmbmbmbmb
I agree with Cameron. We had a mission we failed the other night that would have easily been doable had someone been holding a "Press A, then move blue card". But since it wasn't there and we had limited time we weren't able to work things out.

On a side note, I'm back to loathing the Psionic Satellite.
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Sean McCarthy
United States
Seattle
Washington
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
This is an interesting thing to try to analyze. Certainly if you're completely new to the game, it's probably going to work better to take out the fighter by yourself. That's because of the focus and communication stuff, which you mentioned.

However, having two players take out the fighter is probably going to be MORE action-efficient. Look: you probably wanted someone down there in the lower red zone anyway, if for no other reason than to refill the energy. The cost of plan A is 2 actions (shoot-shoot), the cost of plan B is also 2 actions (2x shoot). Plan B also has an extra energy left though.

Here is how I would calculate the tradeoff:

If we do plan B, we are committing a player to lower red in exchange for one energy.

Is this good? Very likely yes. For it to be bad you'd need to have a use for that fourth or fifth player worth more than 1 energy which is rare in the first phase. And you probably wanted someone in lower red to be able to grab the battlebots, anyway.
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Cameron Chien
United States
Rancho Cucamonga
California
flag msg tools
mbmbmbmbmb
tallgrant wrote:
On a side note, I'm back to loathing the Psionic Satellite.

Oh, my friend HATES all of the Satellites, especially Psionic Satellite.

There's always much cursing and gnashing of teeth from him when it gets drawn.

I kind of like it, since it let us get the Ghost Ship achievement once

Cameron
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Hung Nguyen
United States
Berkeley
California
flag msg tools
badge
Avatar
mbmbmbmbmb
So I'm the OP's primary gunner/external threats person, and my personal opinion is that it is better to have the extra battery and have a person positioned in the red zone. That one extra battery may be enough to tackle a regular threat in the same zone with use of the light laser combined with a rocket or interceptor. Not having to pull additional batteries into the red zone is a HUGE benefit for me.

Dylan: One thing you failed to take into account is how often you have blank spaces. I know in our games, we're never 100% utilized in with respect to actions. We use maybe 80% of our actual actions (10 out of 12)on a mission? We usually play 4P mixing in new players all the time, so it's usually all whites, and occasionally scrambled white/yellow.

Also, you have to weigh what the player would be doing otherwise if he was not going to help out and fire. It is nice to have someone already down there that can pull additional batteries and/or grab the battle bots.

The situation changes a little if it was the blue zone, because in my experience, you will almost always have a person down there at some point firing rockets, so you'll have someone there that can recharge easily. On the red side, if you don't have someone easily positioned, they'd have to use 1-2 movement points in addition to refueling the red reactor.

In summary, my opinion is that having a battery ready for the next threat and positioning a person in the lower left is the superior choice.

-Hung
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Dylan Gould
United States
San Francisco
California
flag msg tools
badge
Avatar
mbmbmbmbmb
I appreciate all the feedback so far!

Writing this article, my purposes were primarily to spark a discussion about the relative value between the resources, and to establish focus units as a framework for discussing, as one of those resources, player attention over a duration of time.

I admit that the article isn't well-informed about double action games -- my group has only played single action games so far. I'm anxious to try double action, but we just haven't found the time yet. However, it's a really good point that in a double action game, you are far less likely to be able to get the exact card somebody needs to whoever needs it, because there are more distinct cards and therefore fewer copies of each distinct card.

Some of you have focused on the question of whether Strat A or Strat B (or a descendant) is better. I think that each one has strengths and weaknesses, so that one or the other is better in a given situation. Strat A is better if the Fighter is the only red-zone threat, for example. You conserve actions and focus units, which are better spent on other threats, and you burn through energy, which is not needed in the red zone later. You don't know by the end of phase 1 whether there will be a second red-zone threat through the whole game, but you DO know by the end of phase 1 whether a more serious threat appears in the white or blue zone or internally and whether a second red zone threat appears right on the tail of the first. I've definitely lost games before by over committing resources in a zone that didn't end up getting attacked very much.

But back to my primary purpose: Do you find the concept of focus units useful? And how many actions do you think is too many to spend in order to save 1 green cube?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Dylan Gould
United States
San Francisco
California
flag msg tools
badge
Avatar
mbmbmbmbmb
arod324 wrote:

Dylan: One thing you failed to take into account is how often you have blank spaces.

True, I only addressed the cost of a Wait action -- 1 action point and 0 (or a negligible amount) of focus units -- but I didn't talk about their likelihood / frequency of occurrence.

Could you please explain more about how to address that topic? I'm not sure I see your point yet.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Hung Nguyen
United States
Berkeley
California
flag msg tools
badge
Avatar
mbmbmbmbmb
fishboy wrote:
arod324 wrote:

Dylan: One thing you failed to take into account is how often you have blank spaces.

True, I only addressed the cost of a Wait action -- 1 action point and 0 (or a negligible amount) of focus units -- but I didn't talk about their likelihood / frequency of occurrence.

Could you please explain more about how to address that topic? I'm not sure I see your point yet.


You need to take into account that in a 4P game, that each player may only actually be using 10 out of the 12 possible actions. So instead of 48 possible actions in a 4P game, you really end up with 40. Also, you might want to also take out actions that are directly related to Visual Confirmation, as they really aren't required to complete the mission. So perhaps you only use 9 out of the 10 actions?

One other major assumption in the analysis you made was the fact that all of the energy gets used. So far, in our games, we have never come even close to using all of the energy in the game. A more rigorous analysis would look at the average amount of energy used in the game vs. the average amount of actions taken in a game to then compare the energy/action ratio. This would give you a more accurate insight into the relative value of an energy cube.

I don't have the numbers, but my gut reaction is that we use far less energy than the 22 you assume to calculate the ratios.

Additional research: I'm a bit tired at the moment, but an interesting, and simple analysis would be to look at the damage per action to external threats. This would be based purely on how much gross damage (pre-shield) is done for each weapon (heavy/light lasers, pulse gun, rockets). In this, I think you have to assume that everyone already is in the correct zone (it will always require movement to get to the right threat) and just see how much damage per action each weapon can do.

For example, firing the central gun requires an additional 1/5 of an action (assume the central reactor is filled from 0 to 5), whereas firing the red/blue gun requires an additional 6/15th action (1/5 for refilling the central reactor, and 1/3 for filling the side reactors). Obviously these are best case scenarios, so we'll have to make some assumptions on what a realistic level of energy in each one is and do the math.

In the end, I think the results will make a very compelling argument to always have a guy spamming the interceptors no matter the scenario because it doesn't require additional actions.

I'll post an analysis sometime tomorrow :-).
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Sean McCarthy
United States
Seattle
Washington
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
Looking at the ratio of actions used/available to energy used/available doesn't even make any sense. Why would that ratio matter?

Take the ratio of available energy to available window-look scorings (3). According to your logic I should conclude that 1 window-look-scoring-opportunity is worth about 7 energy. Or you could compare it to the number of players who can be in the interceptors at once. Since only one interceptors use is available on any given turn, whereas in a four-player game four player actions are available every turn, by analogous logic I conclude that the interceptors are four times as valuable as a normal player action.

Not to go on too much but here's my favorite: the game gives you 12 captain-actions and 36 non-captain-actions. I therefore conclude that an action by the captain is worth three times as much as an action by a lesser crew member.

A more productive way to calculate a ratio of value, as in any game, is to look for situations where you have a tough choice between X of one type of thing or Y of another type of thing, and you're not sure which one's better. By this measure we conclude that a captain-action is actually worth about the same as a non-captain-action after all. Interceptor actions are typically worth less than non-interceptor actions because a large portion of the time when both possibilities are available (i.e. someone with battlebots has been in the upper red deck) you opt not to use them.

So, action vs energy. There's at least one very clear case where you have the choice between one action and one energy. That is when there is one energy left in the central reactor after charging the main shield at the beginning of the game, and you have the opportunity to waste an action to do something with that energy before cracking a can. Maybe someone can pull the energy to a side reactor before the can gets cracked, or maybe the person cracking the can could pass for a turn instead of moving into position in a side station turn 3, to allow that energy to be fired at a turn 3 white zone threat. Either way this opportunity tends to come up in quite a few games, and you take it some of the time. How much? I would prefer the action in most games. I'd pretty much only take the energy if we weren't going to use that action anyway. That does happen a lot though: the first phase is often a struggle to use as many actions as possible for at least some benefit. Only in the second phase are there suddenly not enough players to do everything that needs to get done.

I think useful actions are often much better than one energy, unless there is a specific circumstance making that particular point of energy necessary (because more energy cannot be produced and shifted in time). However often it's hard to use certain actions offensively; in those cases being able to trade those nothings for an energy is a fantastic deal.
1 
 Thumb up
1.00
 tip
 Hide
  • [+] Dice rolls
Benjamin Grey
United States
Minnesota
flag msg tools
Avatar
mbmbmbmbmb
Fun article, I'm enjoying the discussion.

One factor that was missed in the original analysis was "deviation from the Plan". In our games the first couple turns follow a standard pattern. One player heads to lower Red and grabs the bots, one heads to Blue forward to do the same, one stays in White to calibrate the computer and throw logs in the reactor, spare crew spreading out to the non-bot wings in preparation. We're all on "auto pilot" at this point, communication tends to be minimal and basic.

In scenario "B" above, purple is likely already heading to lower Red. The energy savings are a natural by-product of "the Plan". In this case, having Yellow fire the main Red cannon twice is unnecessary and wasteful.

On the other hand, lets suppose that something else came up. An intruder internal threat popped up in White with 2 health requiring both bot teams to go deal with it. Lower Red is now empty when the fighter shows up in Red sector. Is it worth it to pull whoever is handling White sector down to lower Red just to save an energy cube? No, this deviates from "the Plan". Now somebody else needs to handle "White's" duties, somebody else needs to handle that person's station, and the chances of confusion and mistakes rise.

(You touch on this with the "focus" resource but I'm not sure you go far enough. Pulling someone from their assigned duty costs everyone focus, because now everyone needs to pay attention to a station that one person had covered before.)

I realise that Space Alert is a game about managed chaos and "the Plan" is probably going to be abandoned around turn 5 out of necessity. Deviation from "the Plan" should be limited to threats that can't be defeated any other way, and a simple fighter doesn't qualify. Keeping things as simple as possible is the way to go if you can. If firing the main cannon twice will keep someone else from having to abandon their position, do it. If saving the energy cube will prevent someone from having to waste time leaving their post to refill the reactor, do it. If it's more complicated, then the crew have to make a judgement call. Lowering group confusion is more important than resource management any day.
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Guy Srinivasan
United States
Kirkland
Washington
flag msg tools
Avatar
mbmbmbmbmb
I just read a book on queueing theory as applied to product development, so I'm seeing queues everywhere. Here are rambling thoughts. The biggest gain from optimizing from a queue perspective is reducing the size of the largest queues, because they're often much larger than anyone's noticed and the cost of delay of things in the queue is also large. Often the cause of large queue size is the combination of a) working a resource near its capacity in the name of efficiency, and b) incoming job variance.

In Space Alert, there are two queues, actions and focus, which seem to fit these criteria. As SevenSpirits noted, we can't simply trade off actions for energy because there's some concept of "commitment" - actions 3 and 6 are not fungible. Instead let's think of our available action slots as our processing power per time (turn), and the jobs we've taken on within the mission as our queue. Similarly we have a certain amount of processing power per time (focus) and things to think about during the game are jobs in queue.


Action Queue: If this queue size becomes large, a threat usually does lots of damage to the ship or destroys it. The cost of delay of jobs in the action queue is huge, because the longer certain jobs (threats) sit in this queue the further down the track they get, hitting X, Y, or Z and doing damage to the ship.

What causes this queue size to become large? Usually it's when we work a resource to capacity, aka tie up a player for an unusual number of turns. Internal threats that move around and phasing threats tend to cause this problem. But in itself that causes no problems. The difficulty comes in the variance of incoming jobs - sometimes the only person in Blue needs to get the battlebots and defeat the slime, but everyone else thinks "oh Alice is in Blue, if a threat comes that's not so bad" and then a threat comes that requires two people to deal with and Bob says "Alice come defeat this threat with me" and Alice says "I can help starting turn 9" and then everyone begins scrambling because turn 9 is far too late.

How can we mitigate this? When you find a threat that ties you up for an inordinate amount of time, ask how much damage the threat will actually do if unchecked. If you're in upper red and Commandos are appearing in lower red, you could go grab the bots, chase the Commandos and hit them once, go recharge your bots, chase the Commandos (now in upper White), and hit them a second time. Or you could take under half of the time and just hit them once and get hit for 4 damage. Maybe that's worth it! Tying yourself up for essentially the whole game to save 4 damage might not be a good idea. If your team has already eliminated an early White threat and you've got threats coming down the wings, keeping yourself utilized at 40% instead of 100% means you can jump on possibly costlier threats.

Or you could break the job into smaller chunks and farm them out. Maybe letting someone already in Blue take out the second half of the Commandos is far cheaper than doing it yourself. In that case, examine their board for holes, figure out where they should put their C to grab the bots during a free turn (they should already have done this actually), and when it looks like their Focus queue is not being worked to capacity slip in a new job: "hey go to White and hit battlebots in upper White on turn 7, 8, or 9 and while you're there maybe you should wiggle the mouse". It's often easy to break down external threats into smaller chunks, and we do it all the time: "someone come help fire the lasers in red, and we need a rocket shot on turn 8".

Variance in jobs is smaller when chunk size is smaller, and lower job variance gives smaller queue sizes.

Of course you can also pass the job around. If you're overworked and a new threat comes in, put it in someone else's queue. Tell your teammates that you're not going to be able to got wiggle the mouse after all. This gives me an idea: we usually move our minis to the appropriate location during the game so others have an idea where we are. Next time I'm super-booked, I will take my mini off the board.

What you really do not want to do is have a new threat enter your queue, keep dealing with your current threat and think "okay once I've dealt with this I'll deal with the next", and later realize you've incurred huge delay costs (it shoots you for 2 damage twice before you start to deal with it!) for no good reason.


Focus Queue: Again, this becomes large when players are thinking at capacity and variable jobs come in. If you're still trying to puzzle out how to break those Phasing Troopers and someone demands to know where you'll be on turn 8, they're going to have to wait. Delays in this queue are not as costly unless the queue size gets quite large. The main concerns are a) not getting to think about a threat that requires certain actions in slots 6-7 before phase 3 begins; b) delaying thinking about a threat that will turn out to require help for long enough that when you do think about it, you invalidate someone else's planning, introducing massive variance into their incoming Focus jobs; and c) not getting much time to think about a later threat and find the action-efficient way of dealing with it, thus incurring action-queue size costs.

How do we mitigate this? By far the most important thing is triage. Be vaguely aware of all the threats so that if your Focus-queue is empty you can take a job from the longest queue (whoever is busy thinking but somehow got tasked with 2.5 threats) and thinking about that. Sometimes it's better to think about the next threat near you, that you can defeat, but first look to see if there are neglected (aka at the end of long queues) threats. These are the most important.

But dealing with large queues is better done before they become large. Instead of reducing job variance we can reduce the capacity utilization of our most efficient Focusers. Whoever is best at quickly thinking holistically about all the threats all over the ship should probably be assigned the fewest tasks - it's more important that you deal with an influx of jobs quickly than that you deal most efficiently with each job, because the gains from efficiency will be more than wiped out by the losses from cost of delay as later jobs sit in the queue untouched. Are you playing with a new player? Put her on a task immediately. You want the least versatile Focus tied up first, even though the most experienced player could figure out how to defeat the threat more quickly.

(This is assuming you're likely to win unless something unexpected happens - in these cases you want to sacrifice a bit of value for a much greater chance of dealing with the unexpected. If you're unlikely to win unless you deal with everything very well, then it may be better to assign your most efficient Focus to tasks immediately, since you're banking on nothing unexpected happening.)

I hope that made some sense. I don't have a lot of experience thinking in these terms yet and far less communicating in these terms.
5 
 Thumb up
5.00
 tip
 Hide
  • [+] Dice rolls
Hung Nguyen
United States
Berkeley
California
flag msg tools
badge
Avatar
mbmbmbmbmb
SevenSpirits wrote:
Looking at the ratio of actions used/available to energy used/available doesn't even make any sense. Why would that ratio matter?

A more productive way to calculate a ratio of value, as in any game, is to look for situations where you have a tough choice between X of one type of thing or Y of another type of thing, and you're not sure which one's better.


What Dylan originally posted was an effort to quantity something that is very difficult to quantity, because different situations call for different measures, but he's trying to get an overall "feel" of the value of each resource to prep people before going into a game. I agree that it is better to look at specific situations and determine what the best outcome is. But the purpose of this thought exercise is to determine if, in general, saving that energy cube is worth the extra action.

The reason why the ratio of actions to energy matters (and is more important than the trivial ratio of captain's actions vs. other crew member's actions) is because it gives you an idea of the relative value between two resources that are limited and critical to winning the game. As you pointed out, a captain's action is essentially identical to any other person's action. In Dylan's analysis, he's trying to put a value, in terms of actions, on that energy cube to see if, in general, it's worth saving.

What's the right answer? There is probably not a "right" answer from a simple analysis, because each game has so many different factors and the cost/benefit analysis can't be done until you're in game. However this is just an attempt to try to quantify that to dictate overall strategy. For example, if the energy cube is typically worth more than an action, you could have it in your mind that the "default" move is to spend the extra action to save the cube. But while playing it, if you see that there is a severe negative effect by not using the cube, then go ahead and use it.

As a follow up to my previous post, I created another post which details the "true cost" of doing damage to external threats for each weapon and it takes into account the "action cost" of creating the energy, and pulling the energy to the side reactors. I think this is a more rigorous analysis, as it ties doing damage directly to the actions required to create/pull the energy.

I kept it in a separate post so I wouldn't derail the topic of this post.

http://boardgamegeek.com/thread/744065/analysis-of-damage-vs...
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Sean McCarthy
United States
Seattle
Washington
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
arod324 wrote:
What Dylan originally posted was an effort to quantity something that is very difficult to quantity, because different situations call for different measures, but he's trying to get an overall "feel" of the value of each resource to prep people before going into a game. I agree that it is better to look at specific situations and determine what the best outcome is. But the purpose of this thought exercise is to determine if, in general, saving that energy cube is worth the extra action.

The reason why the ratio of actions to energy matters (and is more important than the trivial ratio of captain's actions vs. other crew member's actions) is because it gives you an idea of the relative value between two resources that are limited and critical to winning the game.


But that's just it - it doesn't at all. There's no reason to think that that ratio is a meaningful number. It's based on the assumption that the sum value of the player actions in the game is equal to the sum value of the energy in the game, which is utter nonsense. It happens to produce a plausible result, which is why I tried applying the exact same methodology to some other types of resources as a demonstration (where, it turns out, you get an obviously wrong result).
1 
 Thumb up
1.00
 tip
 Hide
  • [+] Dice rolls
Dylan Gould
United States
San Francisco
California
flag msg tools
badge
Avatar
mbmbmbmbmb
GreyArcher wrote:
Fun article, I'm enjoying the discussion.

One factor that was missed in the original analysis was "deviation from the Plan". In our games the first couple turns follow a standard pattern. One player heads to lower Red and grabs the bots, one heads to Blue forward to do the same, one stays in White to calibrate the computer and throw logs in the reactor, spare crew spreading out to the non-bot wings in preparation.

Thanks.

I have a similar "The Plan" when I play solo missions, but not as much when I play a standard 4-/5-player game. In my "The Plan", Captain Mustard goes down to refuel the reactor and generally stays down to help refuel and provide supporting fire on whichever side is more heavily attacked. Miss Scarlett wiggles and then goes redward to deal with the red threats, Ms. Peacock B's up the shields and then goes blueward to deal with blue threats, and Professor Plum goes blueward to grab the bots.

Now I've seen more comments than just yours indicating that people love picking up the red zone battlebots, but I don't get that. I like the blue bots because they don't require any risky gravolift trips in many cases. Do you prefer the red ones, and if so, why?

I might be misunderstanding you, because in your "The Plan", you actually have 2 people picking up battlebots, so maybe you don't have a preference, and you just love battlebots in general.
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Cameron Chien
United States
Rancho Cucamonga
California
flag msg tools
mbmbmbmbmb
Internal threats sometimes only give you a turn or two window to engage them before the poop hits the fan. Having both battlebots ready is a good thing...unless you get Battlebot Uprising, of course

I also don't consider grav lifts to be "risky". You ARE communicating with your crew mates, aren't you?

One move we like to do is B, down B. Charges the white zone shields to full and refills the reactor.

Cameron
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Dylan Gould
United States
San Francisco
California
flag msg tools
badge
Avatar
mbmbmbmbmb
fishboy wrote:

One factor I’d consider is the total available amount of each resource. The assumption here is that the designer, Vlaada Chvatil would have given us about the right amount of both resources, so that if we overuse one in part of our plan, we should conserve that amount in other parts by overusing the other resource. The initial amount of energy is always 10, with 3 fuel capsules that can be converted to a maximum of 15 energy, ignoring heroic/special actions. This gives us a total of 25 energy, but it assumes superhuman efficiency in draining and refilling the reactors, so I’d suggest about 22 energy (4 per capsule) as a realistic upper limit. In a 4-player game, the players have a total of 48 actions, but in a 5-player game, they have 60. This suggests that the valuation ratio between energy and actions changes depending on the number of crew members, and that makes intuitive sense as well – a 5-player game only adds 1 regular threat, or 14% more threat value, while it adds 25% more actions, when compared to a 4-player game. The extra actions should give the players more room to waste them and still be able to win the game. As for the ratios, we have 22 energy : 48 actions in a 4-player, or 1 : 2.2. And in a 5-player, we have 22 : 60, or 1 : 2.7. In both cases, an energy is worth less than 3 actions, so this line of thought suggests again that Strat A is better.

I think that the assumption I made that Chvatil gave us close to the just-right amount of each resource is questionable. To more accurately value energy against actions, I would measure the energy and actions used in a large number of games won by the players. This should result in a more accurate valuation for both energy and actions.


I think the first paragraph that I quoted above is what Sean is primarily objecting to, and I agree. I said as much in the next paragraph. The assumption that the whole energy pool is worth as much as the whole action pool is quite questionable. But I included it because I wanted some example of an independent method of valuing one in terms of the other. I could also do it by recording results from games that were won, but I didn't have time to do it within the timeframe that I wanted to post this article.

On the other hand, there's another way we could independently value energy in terms of actions: We could poll BGG users who read articles about Space Alert. So here we go.

Poll: What's the largest number of actions (including movement and wait actions) that you'd be willing to spend in order to save one cube of energy in Space Alert?
1
2
3
4
5
6
7
8
9
10
11
12
      4 answers
Poll created by fishboy
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Cameron Chien
United States
Rancho Cucamonga
California
flag msg tools
mbmbmbmbmb
I don't process Space Alert this way. Energy is a tool used mostly to kill threats. I never try and "save energy".

Efficiency in Space Alert is about trying to have the right people in the right places with the right cards.

I suppose if you only played Space Alert solo then it might matter, but I don't, so it doesn't.

Cameron
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Benjamin Grey
United States
Minnesota
flag msg tools
Avatar
mbmbmbmbmb
fishboy wrote:
I might be misunderstanding you, because in your "The Plan", you actually have 2 people picking up battlebots, so maybe you don't have a preference, and you just love battlebots in general.


We like being prepared. Having someone with battlebots on both sides of the ship makes it more likely that someone with battlebots is on the side of the ship that needs battlebots. Nothing screws up "the Plan" like having someone in red sector chase down a blue internal threat, or having someone who wasn't expecting to have to grab bots becoming internal security. Also, if a threat has two health and each attack knocks out a bot team, it makes things a lot easier to have 2 teams to take it out instead of one that has to repair after the first hit.

I do have to admit that I'm partial to the red bots. I tend to be the guy on late mission cleanup duty in the interceptors taking care of any threats that weren't quite killed when they were supposed to. Between me and the guy in the pulse cannon hopefully we can take out the ships with one hit left (always seems like the cannon we need is the one that gets damaged) before they hit Z.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Cameron Chien
United States
Rancho Cucamonga
California
flag msg tools
mbmbmbmbmb
There's also the nasty red zone airlock internal threat.

Cameron
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Alec
United States
Houston
Texas
flag msg tools
badge
Avatar
mbmbmbmbmb
This is easily the most thought-provoking and statistical analysis for this game I've seen. People can pick and choose what the agree/disagree with, and come to their own conclusions, but this was awesome and interesting.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Dylan Gould
United States
San Francisco
California
flag msg tools
badge
Avatar
mbmbmbmbmb
Oops! Made a couple edits, because I noticed that a part of the article got cut off near the end.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.