Recommend
60 
 Thumb up
 Hide
14 Posts

Roll for the Galaxy» Forums » Strategy

Subject: The Ruthless Math of Roll for the Galaxy rss

Your Tags: Add tags
Popular Tags: [View All]
Thomas Ting
msg tools
The Ruthless Math of Roll for the Galaxy

This is a quick strategy guide for "Roll for the Galaxy", which I'm writing after playing the game about a dozen times (nine of them in one day. Yes my group is a little crazy for this game). It seeks to reveal the core mechanics that determines victory or defeat in this game - the "Ruthless Math" if you will - an approach I've also taken with some videogame guides.

Unlike most strategy guides, which focus on the "How", I instead tend to focus on the "Why" in my "Ruthless Math" guides. The guides in fact seek to explain why certain players / strategies keep winning, while others don't.

The other reason why I'm posting this guide is because there still seems to be a debate between two particular strategies in this game - which is "building planets & developments" versus "shipping" - as these were the core strategies in the original RftG. However, I feel that these debates miss the larger point that the game is primarily about efficiency - meaning picking an approach that nets you the most VP - and in most cases you will need a mix of both in various proportions to win most games.

====

The first and most important strategic principle to embrace is that the amount of money you earn in the game ($) will be the prime determinant of your VP total at the end of the game. The game may be focused on the dice and its different faces, but players must realize that every dice must be "recharged" using money after every action - and every action in the game actually results only in two possibilities - VP generation (that is usually accompanied by some new abilities for your tableu) or more money generation which allows you to roll more dice to get more VP.

Dice, therefore, merely define the amount of $ you can "absorb" at any given time (and define what's the biggest Development or Planet you can build), but what's important is that you're constantly spending them so that you can use $ to make them available again for use in more actions.

In short, the player who moves the most dice from his Citizenry or the general supply to the cup will generally win. Every $ you spend will generally give you another VP at the end of the game. This is in many ways similar to RtfG, where the player who drew the most cards in the games is usually the victor.

As a corrollary to the above, this also means that games are generally decided by the abilities on each player's tableu and how often they are able to activate them. Everyone basically gets the same base income - $1 if you are efficient and have no money in the bank - so everything else must come from abilities that generate income. Even shipping for money requires having planets of the appropriate type and getting the produce/ship actions taken often.

====

The Ruthless Math of Scoring

So why is $1 = 1VP generally speaking, and how can I tweak its math to my advantage?

First of all, it's important to realize that there are only three sources of VP generation in the game, namely:

1) Settling Planets
2) Building Developments
3) Shipping

For both developments and planets, the math is seemingly obvious - you spend one dice per vp value of the planet to build it.
However, do note that you need to spend an extra Explore dice to actually get tiles beyond your initial two tiles - hence the cost of each planet/development is actually one more than its VP value.

What this means is that more expensive planets are actually more efficient at converting money into points. To expand:

1 VP Planet/Dev = $2, 0.50 VP/$
2 VP Planet/Dev = $3, 0.66 VP/$
3 VP Planet/Dev = $4, 0.75 VP/$
4 VP Planet/Dev = $5, 0.80 VP/$
5 VP Planet/Dev = $6, 0.83 VP/$
6+ VP Planet/Dev = $7, 0.85-1.0+ VP/$

That said, planets and devs are not just sources of VP; they also bring more dice and abilities to the table.

In particular, planets that immediately add to your cup (or as a good, which can be thrown into the cup too if you choose) basically give you a virtual $, since the first use of those dice was made free for you. This often offsets the additional dice used for explore. Planets which give you extra money outright are even more efficient.

Devs meanwhile are dependent on your build; albeit some are standouts especially if taken early - Merchant Guild for instance earns you $2 for every Produce, earning a profit the first time someone takes the Produce action and every time the phase is selected thereafter.

In short, while it is more efficient to build bigger planets and developments, you have to balance this against the fact that additional developments and planets can give you additional $ generation early on, which snowballs into being able to make MORE money later on and hence even more VP later on. If you rely entirely just on your starting dice and abilities your income generation will be very limited, which in turn leads to a lower final score.

The math for shipping on the other hand always requires spending $2 per ship - since you need to spend $1 for both the goods dice and the shipping dice itself. This leads to three basic possibilities:

1) Ship with no matching dice - spend $2 for 1 VP (0.5 VP per $)
2) Ship with one matching dice - spend $2 for 2 VP (1 VP per $)
3) Ship with two matching dice - spend $2 for 3 VP (1.5 VP per $)

What this means is that having two matching dice is actually an incredibly efficient way of earning VP - better than almost anything save 6 Devs. Having one matching dice is also better than straight-up converting dice into Devs/Planets for VPs alone. However, the problem with shipping is that you do NOT get additional abilities or income-generation unlike Devs/Planets; hence shipping too much too early can lead to stunted income generation overall and an inability to snowball your economy. Shipping with no matching dice is not even worth considering - it's as inefficient as making a 1 Dev/Planet with no abilities or additional dice!

As a side note, it is also worth noting that the $2 cost per ship also applies to selling goods. Hence, if you take this cost into account, selling a Novelty good (blue) at $3 will in fact only net you $1 in additional profit, rare materials (brown) at $4 will net you $2, and so on. Not taking this into account is very often why shipping-only strategies fail; shipping for money is not as profitable as many abilities that simply give you money outright and is reliant on two phases (produce + ship) being chosen consistently.

This is one of the biggest differences compared to RtfG math-wise: In RtfG "produce" actions did not take a card from the player's hand - it was taken from the deck. In Roll for the Galaxy the goods dice must come from the player's pool; hence you simply cannot apply RtfG shipping math (and their resultant strategies) to Roll.

====

Hence, based on this math, players can draw the following conclusions:

Core Principle: The amount of money you earn in the game ($) will be the prime determinant of your VP total at the end of the game.

1) Buying low-cost developments and planets is less efficient than more "expensive" ones because of the extra dice you pay to "explore". However, Developments provide you with additional abilities while Planets provide additional dice. Given that the core of the game revolves around money generation leading to VP generation, it is very important to create additional sources of $ generation from the outset regardless of your strategy.

Putting out cheaper planets and development earlier so that their abilities get used more can lead to a "snowball" effect. This means that you early income generation allows you to buy more income-generating tiles sooner; which in turn leads to an even bigger VP total by the end.

2) Planets that add dice directly to the cup or come with a good also "discounts" the cost of playing them. Keep this in mind when building. 6 VP developments are also incredibly efficient because they produce additional VP beyond their base VP value.

3) It is very inefficient to ship without matching dice. You should try not to do it at all unless it earns you additional VP or money by other means.

4) Shipping with one matching dice is generally more efficient than building most developments or planets from a VP perspective, but it doesn't provide you with abilities or more dice. In general, you should avoid shipping with one matching dice for VP early in the game because you can stunt your ability to snowball into more income and VP generation later

5) Shipping with two matching dice is very efficient and is only beaten by high-scoring 6 Devs in sheer efficiency. You should consider shipping with two matching dice early in the game if you can still maintain production of your planned development and planets. This also points to the great value of the purple dice - which counts as any color for shipping.

6) It still costs $2 to sell a good via shipping - $1 for the good, and $1 for the ship dice. Keep this in mind when selling goods via shipping as it really makes for very thin profit margins without some abilities to support you.

Finally, I have to note - once you get the hang of these concepts games can get extremely tight; very often with players actively "racing" to end the game while they have a VP lead whereas others are still trying to convert a superior economy into a higher VP total.

Happy Rolling!
  • [+] Dice rolls
Nathaniel Chambers
United States
Austin
Texas
flag msg tools
mbmbmbmbmb
wow, really interesting analysis.
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Bruno Wolff
msg tools
This also shows why buying big worlds can be very good, even without discounts. For example, you need 7 dice (counting the scout) to build Alien Sentinals, but you get $3 back and a die in your cup. So the net cost for 6 victory points is only $3. That's a better ratio than shipping with 2 matching dice. With discounts the ratio gets ridiculus.
5 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Thomas Ting
msg tools
brunowolff wrote:
This also shows why buying big worlds can be very good, even without discounts. For example, you need 7 dice (counting the scout) to build Alien Sentinals, but you get $3 back and a die in your cup. So the net cost for 6 victory points is only $3. That's a better ratio than shipping with 2 matching dice. With discounts the ratio gets ridiculus.
Yes, worlds that give you money back as part of construction are effectively giving you a discount and hence improve the money-to-VP conversion. Developments that confer discounts to production have a similar effect.

That said, I have to note that there are also developments that improve the efficiency of shipping as well. Having +1 VP per ship for instance improves efficiency to 2VP per $ with two matching dice.

My sense though is that - at least in the base game - the discounts to developments and planets are stronger because you can complete multiple planets and developments in a single turn. With a lot of low-cost developments and planets, the Planet/Dev rusher can thus dictate the pace of the game by triggering the 12 tile end game condition while they still have the advantage. They are, in a sense, the "clock" of the game that other players must race against.

I haven't played the expansion yet, but my impression is that it does try to balance this out and give shippers a better chance to be the "clock".
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Justin Dugger

Corvallis
Oregon
msg tools
mbmbmbmbmb
I posted a similar analysis to reddit a while back. A few additional thoughts:

1) The average world sets you back 2-3 credits to settle, regardless of settle cost -- a six cost world may give you 1 die in cup and 3 dollars. The end result is that it's much closer to 1 VP per $ for settling.

2) Matching color dice for shipping is hard -- colored planets yield only one die. Best you can expect is to 3 point match on half your production worlds and trade the other half for money. Getting a purple die helps, since you can't produce on their worlds, but since you could have had a production world with a colored die, the primary value is matching flexibility.

3) As you state the victor is whoever spends the most money. Since both players have the same number of turns, the victor is then whoever can send dice from the cup back to the citizenry (thus completing the economic engine / cycle) the fastest. Reassign powers grant you that ability.

====
New analysis: Reassign

Reassign is a tricky thing to evaluate. When moving dice, you steal from one phase to fuel another. Using Galactic Influence (reassign as the baseline, you're forgoing Blaster Gem Mines, or more importantly returning in subsequent explore actions for a different planet (which usually gives a 2 dollar rebate). In order to come out ahead vs building a world, you need to successfully use it three times, meaning, stealing from a phase that wasn't called to power a phase that was called and you benefit from.

Factors that affect you pulling this off include the number of dice you're rolling (you already get one free reassign as phase selection), the number of assign powers you already have (you start with dictate), the phases you can't benefit from (you can't ship goods you don't have), the distribution of dice rolls, and the number of phases activating each turn (3 in a 2-3 player game, potentially more in higher player counts). That's a lot of factors! It gets far more complicated with say Mad Scientists or Executive Power.

There's also a benefit in the form of faster ROI. Galactic Exchange is expensive to build, and would likely take you two turns to hammer out that 6cost dev. But with reassign powers, you might be able to get there in one turn without locking up too many dice.
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Andrew W
Australia
flag msg tools
mb
My play group is finding that military dice are brutally effective, doubly so if combined with any sort of cost-optimisation development.

Background: Unlike Race, all actions in Roll are directly proportional in power to the number of dice you commit. In Race, even a no-card produce-consume engine could run at full power every turn. Add some way of generating cards, and you could leech off other players' build actions. Contrariwise, if they had spare production and consume powers then they could generate VP (and possibly cards) for free off your actions.

In the completely optimal Roll strategy, you roll all your dice every turn. In practice, this means that about half your dice are devoted to income, and the other half to action.

The problem our group finds is that a military Develop / Settle + Explore strategy keeps the dice cycling, plus each completed tile increases the number of dice in play. You're also less dependent on phase selection.

In contrast, a Produce + Consume strategy is hampered by four issues:

(1) the need for both phases to trigger to keep the engine rolling. If you produce and no-one ships, then half your dice sit for a turn and the other half must roll ship to free up your engine. If you ship and no-one produces, then you get all your dice back but have effectively wasted the round.

(2) the dice required to set up your VP engine are different to the dice required to execute it. You need building dice to complete a sufficient number of worlds, and then you need a balanced number of colour-matched produce and consume dice. As such, your dice are either sub-optimal to set up your VP engine or they are sub-optimal to execute it.

(3) Colours matter. The builder just needs to coerce his dice to Explore, Develop or Settle, and the exact balance isn't usually critical. The consumer needs worlds, Produce dice and Consume dice to line up or he loses any VP advantage. Yes, you can be choosy about what worlds you settle, but that is another limiter.

(4) When and how to shift. The builder just does more of the same, only faster. Each incremental advantage in building improves his primary strategy. In contrast, the shipper needs to finesse a point to switch strategies. Too soon, and his engine is too weak. Too late, and his engine doesn't get enough time to run. And unlike Race, an engine can't "run in the background"; you can't participate if you don't commit dice, and if you commit dice and they don't operate efficiently then you fall behind.

My questions:

Is there a counter to building (especially military building) except by ALL other players only going produce and ship (thus denying the builder either exploration income or completion)?

If a production engine can compete, what is the optimal point to switch?
5 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
So far in our 2-player meta we've found that fast rush loses to big developments, which loses to big developments + consume (using the great P/C powers of those devs), which is vulnerable to fast rush.
4 
 Thumb up
0.05
 tip
 Hide
  • [+] Dice rolls
Bruno Wolff
msg tools
pwnguin wrote:

2) Matching color dice for shipping is hard -- colored planets yield only one die. Best you can expect is to 3 point match on half your production worlds and trade the other half for money. Getting a purple die helps, since you can't produce on their worlds, but since you could have had a production world with a colored die, the primary value is matching flexibility.
I don't think "hard" is the correct description. The matching isn't really hard, the issue is that only some of your dice can be used to match. What you want to count is the number of worlds you expect to consume from plus the number of dice that will be able to score points on those worlds. (In rare cases you might have more than two matching dice per world and should just cap that at two per world.)
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Bruno Wolff
msg tools
Nomau wrote:

(1) the need for both phases to trigger to keep the engine rolling. If you produce and no-one ships, then half your dice sit for a turn and the other half must roll ship to free up your engine. If you ship and no-one produces, then you get all your dice back but have effectively wasted the round.
When choosing produce, you can still try to leech with some of your dice. In two player there is at least a 1/3 chance of explore happening, which you can use to raise money. If you get credits for producing, then leeching builds is also practical.

Nomau wrote:

(2) the dice required to set up your VP engine are different to the dice required to execute it. You need building dice to complete a sufficient number of worlds, and then you need a balanced number of colour-matched produce and consume dice. As such, your dice are either sub-optimal to set up your VP engine or they are sub-optimal to execute it.

(3) Colours matter. The builder just needs to coerce his dice to Explore, Develop or Settle, and the exact balance isn't usually critical. The consumer needs worlds, Produce dice and Consume dice to line up or he loses any VP advantage. Yes, you can be choosy about what worlds you settle, but that is another limiter.
I think the solution to the above problems is to get Galactic Mandate or Executive power. As each costs 3, making them relatively painless to build early and each guarantees being able to get 4 dice in the phase you select when rolling at least 5 dice.

Nomau wrote:

(4) When and how to shift. The builder just does more of the same, only faster. Each incremental advantage in building improves his primary strategy. In contrast, the shipper needs to finesse a point to switch strategies. Too soon, and his engine is too weak. Too late, and his engine doesn't get enough time to run. And unlike Race, an engine can't "run in the background"; you can't participate if you don't commit dice, and if you commit dice and they don't operate efficiently then you fall behind.
That is what I think makes produce / consume strategies hard to execute well enough to win.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Justin Dugger

Corvallis
Oregon
msg tools
mbmbmbmbmb
brunowolff wrote:

I don't think "hard" is the correct description. The matching isn't really hard, the issue is that only some of your dice can be used to match. What you want to count is the number of worlds you expect to consume from plus the number of dice that will be able to score points on those worlds. (In rare cases you might have more than two matching dice per world and should just cap that at two per world.)
I didn't think I'd need to mathematically prove this, but here we go.

To get three points per planet, you need twice as many dice as production worlds. A ratio of 2:1.

There is exactly one development that gives dice that can color match production worlds. Every production world at most yields one color matching die, thus adding more brings the average down towards 1:1. Worse, you can only color match in pairs, so the first colored world of each type pushes you towards 0:1. Purple is the rarest citizen, and only comes in via nonproduction worlds. These are pretty damn important to high efficiency produce-ship, as they push towards a ratio of 1:0. So, The universes in which you exceed 2:1 at scale is going to be extremely rare, and not worth considering.

Conversely, shipping one alien good is worth 6 money. Assuming you can convert money into points at 1 money:1 point, you don't even need to color match; you're much better off just taking the future 6 points than 3 points now. Applying that logic, even novelty goods are break even.

1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Tom Lehmann
United States
Palo Alto
California
flag msg tools
designer
mbmb
pwnguin wrote:
To get three points per planet, you need twice as many dice as production worlds. A ratio of 2:1.

There is exactly one development that gives dice that can color match production worlds.
That depends on how you count things. For example, Galactic Renaissance effectively turns one non-matching die for a given production world into a matching die. Given you start with 5+ dice, adding 5+ production worlds to leverage your 2:1 ratio into 3 VPs per world is fairly trivial with this dev.

Two cheaper, more specialized versions exist (Minor Research Labs and Space Refineries). Info Tech gives you two Novelty dice and Organic Shipyards gives you two virtual shippers, one of which can color match.

So, by my count, there are 5 devs that counter your average from sinking to 1:1.

The Ambition expansion throws this further off since everyone starts with a color-matching die (the black Leader die) and the objectives produce color-matching Talent Counters. During testing, I saw some huge shipping for VPs moves (15-20 VPs in one shot) by players who exploited their Talent Counters.

One issue with Produce-Ship in Roll is that other players can often see when you are going to call Produce with lots of workers and can leech off of this by placing one die under Produce and calling Ship to Trade a high value good for big $ (fueling their empires), while all your dice are mostly stuck on Produce setting up a big Ship for next turn.

With a bunch of Talent Counters, a shipper setting up a big Produce-Ship sequence can safely put most of their dice on Produce while placing all their Talent Counters on Ship. If someone calls Ship to leech off your Produce, then this "conditional Ship" fires for big VPs. If not, your Talent Counters are unused, no opponent leeching occurs, and next turn you call Ship with them and get your big VPs.

For players who feel that Shipping is too weak in base Roll, I would encourage them to explore Ambition's objectives. Not only are many of them geared towards Produce-Ship, but Talent Counters can offer one shot color-matching in a huge way if you can get enough of them.
9 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Bruno Wolff
msg tools
pwnguin wrote:
brunowolff wrote:

I don't think "hard" is the correct description. The matching isn't really hard, the issue is that only some of your dice can be used to match. What you want to count is the number of worlds you expect to consume from plus the number of dice that will be able to score points on those worlds. (In rare cases you might have more than two matching dice per world and should just cap that at two per world.)
I didn't think I'd need to mathematically prove this, but here we go.

To get three points per planet, you need twice as many dice as production worlds. A ratio of 2:1.
That doesn't make "matching" "hard". What it means is that shipping victory points are limited by the number of dice that will match the worlds you consume from. It will normally be the case that you don't have enough of the right dice to score 3 VPs from every world you ship from.
The effect of this is that you should count worlds plus usable dice rather than count 3 per world, when estimating what you will likely score using produce ship.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Bruno Wolff
msg tools
Tom Lehmann wrote:

One issue with Produce-Ship in Roll is that other players can often see when you are going to call Produce with lots of workers and can leech off of this by placing one die under Produce and calling Ship to Trade a high value good for big $ (fueling their empires), while all your dice are mostly stuck on Produce setting up a big Ship for next turn.
We weren't playing with goals, but this kind of situation did come up in a game I was playing. I was mostly shipping for VPs and my opponent was doing some shipping for VPs. There were very few VP chips left and I needed to produce and then ship. I thought my opponent was likely to settle, but if I did produce and speculate on settle (I got income for producing, so I would get the dice back.) and my opponent called ship, I was screwed. They had a number of blue dice, so a roll that gave them the ability to produce and ship wasn't that unlikely. So I speculated on ship as a safety play. As it turned out I would have done about 8 points better by speculating on settle, but I still got enough points to win when I shipped the following turn.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Joel Horo
msg tools
Avatar
mbmbmbmb
I love your analysis, it's helped me get a much better mental model of the game, which I struggled, even though I am Race FTG fanatic.

A few things I would add that perhaps you didn't emphasize so much
[A] even though 1$ is ~ 1VP, in practice you only get to turn one onto the other if it comes to the right phase. Every dice that goes back in your cup at the end of the round was a wasted opportunity. I see 3 ways to improve that
(1) getting re-assignment powers
(2) always having a stock of devs / planets in construction and planets which can produce, or are ready to ship (easier said than done)
(3) favoring rolling more dice rather than less

What I mean for 3 for instance, is that, if I have 3 dice in my cup, and 4 in the citizenry (and no money to buy them), no matter where the 3 come out, I'll try to assign them to explore, in order to get cash to buy the 4 other ones. Even though this wastes a turn, it prevents from being at a stage where you just have 1 dice, which you assign to explore to get another 2, etc... which is really slow and tedious

which leads me to
[B] I now try to always have as much cash as I want. When I do abandon/stock, I try to order my planets in such a way that each settled planet will either give cash, or a good that I can trade to get money to settle the next planet.

[C] developments can of course be great (replicant robots that give a 1 discount to planets), but there are some more subtle cases where a dev can make an enormous difference when they allow you to get the money right when you need it. An example is the one that gives you 2$ per green worker sitting on a green planet. This makes it that you can do produce, then ship in 2 rounds (sometimes 1 if you can piggy back on opponents), instead of having an intermediary round where you get more cash to get dices back.

Overall, Roll has that in common with Race (and many games in fact) that, after a while, you start 'chaining' things, i.e. you ensure that each action leaves you in a good state to perform the next one (instead of having to start from scratch, i.e. explore).
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls