Tomas Hejna
Czech Republic Prague Czech Republic

Is there any clue on how many of starting worlds will be included in this expansion? And on how they will be numbered (i.e. starting from the 5 / the 16)?
Also, will they be compatible with the basic game, or will they use any of the new mechanics (and therefore they will be playable only with this new expansion)?

Ross G.
United States Indianapolis Indiana
Wisdom begins in wonder.
Stale pastry is hollow succor to a man who is bereft of ostrich.

There is no information on specific cards at this time. Presumably they will start at 5, as this expansion is not meant to be mixed with the first three expansions.

Gareth Roberts
United Kingdom

Im going to speculate. I imagine: in the same way that Rebel Freedom fighters would be useless if you just played with the base set some of the new starting worlds will be useless with it as well. Also for reasons of balance you probably shouldn't incorporate them alone.
In terms of numbers I'd expect to see at least as many as included by the time you get to RvI as the starting world choice mechanic has been implemented. I would not expect to see as many start worlds as the full first arc because there isnt quite (almost) as large a deck to shuffle them into and for the most part start worlds are pretty cost ineffective to settle I think, so you don't want them coming up more than they do.

Tomas Hejna
Czech Republic Prague Czech Republic

ZiggyZambo wrote: There is no information on specific cards at this time. Presumably they will start at 5, as this expansion is not meant to be mixed with the first three expansions. In that case I hope there will be some preview, soon

Luke Stirling
Norway Trondheim

XehutL wrote: In that case I hope there will be some preview, soon Given that there's no word of the game showing at Essen, I'm not holding my breath for any major new info for the next couple of months.

Tomas Hejna
Czech Republic Prague Czech Republic

ilovedawkins wrote: Im going to speculate. I imagine: in the same way that Rebel Freedom fighters would be useless if you just played with the base set some of the new starting worlds will be useless with it as well. Also for reasons of balance you probably shouldn't incorporate them alone.
In terms of numbers I'd expect to see at least as many as included by the time you get to RvI as the starting world choice mechanic has been implemented. I would not expect to see as many start worlds as the full first arc because there isnt quite (almost) as large a deck to shuffle them into and for the most part start worlds are pretty cost ineffective to settle I think, so you don't want them coming up more than they do. Hmmm, maybe this could be a clue:
Quote: Everything gets used in the Alien Orb scenario, while players can also play the base game with just the new cards and start worlds. ... The only thing we use from the first expansion arc is the deal two and choose a start world after seeing your starting six cards mechanic introduced in Rebel versus Imperium. Nothing else. http://www.boardgamegeek.com/article/6266415#6266415 Personally  I will be very happy if the SW's numbering would continue from 16.

Rus
United States Mountain View California

Surely the new start world numbering will continue from 5, not 16. It has been made clear more than once that the new expansion and arc is to be totally disjoint from the first arc. That said, nothing prevents you from instituting a house rule to combine the expansions in any way you like, of course. But, I strongly suspect that any such house rule will not be officially endorsed.
Regarding the number of new start world cards, I believe it will be 5, in order to bring the total to 10. I remember seeing somewhere that the new expansion will add action cards for one more player, to make a 5player game possible. Therefore, there needs to be at least 10 start worlds total (given the "deal two and pick one start world" rule" which we know will be in this expansion). I also remember seeing a post about the components of Alien Orb somewhere that may actually explicitly give this number.

Tomas Hejna
Czech Republic Prague Czech Republic

rbelikov wrote: That is actually quite a lot of information  thank you for them



16 or 5 serves the same purpose, so I would hope they would pick 16.
An obvious starting world option would be nonmilitary and have alien synergy. I'm not sure if any of the starters would have orbspecific powers though, since that can create some significant imbalances with orb on/orb off.

Tomas Hejna
Czech Republic Prague Czech Republic

Stunna wrote: 16 or 5 serves the same purpose, so I would hope they would pick 16.
An obvious starting world option would be nonmilitary and have alien synergy. I'm not sure if any of the starters would have orbspecific powers though, since that can create some significant imbalances with orb on/orb off. It doesn't have to create significant dissbalances necessarily, but yes  it could create them, true.
And another good reason for continue in numbering from 16 is the possibility of adding those new worlds to a solitary robogamers: I think that there is still quite a lot of such players out there, who are enjoying AI plays.

Johan Haglert
Sweden Örebro

XehutL wrote: Hmmm, maybe this could be a clue: Personally  I will be very happy if the SW's numbering would continue from 16. Me to. Also "with only the new cards + start worlds"? Why not atleast make those possible to mix in with arc 1? Though maybe they wouldn't contribute anything of value.

Dan Moore
United States San Francisco California

aliquis wrote: "with only the new cards + start worlds"? Why not at least make those possible to mix in with arc 1? Though maybe they wouldn't contribute anything of value.
I think the answer is Because! . . . Mr L felt he'd taken the 'arc' to the breaking point?

Tom Lehmann
United States Palo Alto California

aliquis wrote: Why not at least make those possible to mix in with arc 1 catmando wrote: I think [..] Mr L felt he'd taken the 'arc' to the breaking point? There are lots of reasons why RFTG:AA is not compatible with RFTG:Arc1:
A) Look at the BoW comments: a common refrain among those who really like BoW is that RFTG:Arc1 is "now too hard to teach to new players". That's a clear sign that a reboot is needed. Allowing the next expansion to interact with Arc1 would only make this problem worse, not better.
B) Adding more cards to an already large deck makes RFTG too luck dependent: where whether one player's combo hits and another doesn't outweighs playing skillfully the cards drawn. BoW came awfully close to doing this  without Search to help players out when none of their early cards fit together, I believe BoW would have had too high a luck factor.
C) Dilution, Dilution, Dilution. Between the BoW cards and the Prestige symbols on a few RvI cards, roughly 25% of Arc1 combined involves Prestige in some way. Even with variance, most players will see a few cards involving Prestige during a game. Drop this below 20% due to AA and Prestige becomes too variable. Similarly, adding in lots of AA cards without any takeover powers makes Takeovers too variable.
This applies the other way. Some AA cards interact with the Orb scenario. I think it's important that card play can affect the Orb and Orb play can affect card play. However, I don't want lots of cards to have Orbspecific powers (since players won't always be playing the Orb scenario and then these powers are just clutter). I can't make these proportions work simultaneously with just the base set (where AA represents ~30% of total cards) and also all of Arc1 (where AA would represent just ~17% of the total).
D) Bad/messy interactions. Arc1 Start World design was restricted since I had to ensure that +Military start worlds couldn't be taken over right away by other players (which would be incredibly frustrating). This is why so few Arc1 +Military start worlds are actually military worlds.
Since AA Start Worlds don't work with Arc1, I can now design low defense, +1 Military Start Worlds with an extra interesting power. This has two useful properties: these worlds now count towards most Military 6devs and these worlds get played more often when they *aren't* Start Worlds. Now, a player with low Military who draws one of them can conquer it to boost their Military, plus gain a useful power. If I had to make AA Start Worlds compatible with Arc1, I couldn't do this.
Similarly, I can include new ways to Settle multiple worlds in a single phase in AA if I don't have to worry about compatibility with TGS and Improved Logistics. Or, I can include a 6dev rewarding Imperium in a different way in AA if I don't have to worry about a player drawing it along with too many Arc1 Imperium 6devs.
E) Finally, let's be honest. There was some "power creep" in Arc1. A reboot lets me both scale back complexity and reduce power creep. If AA cards had to work with Arc1, then they would have to be more powerful than they need to if they just have to work with RFTG base.
For all these reasons (and more), AA is a reboot. Yes, I know some players will combine AA with Arc1 and then complain about weird rules interactions, too high variance, Orb powers not showing up often enough, or their Start World got taken over on turn 1, and so on. I can't help that except by trying to draw a sharp line between Arc1 and AA.

Serge Levert
Canada Vancouver British Columbia

Tom Lehmann wrote: B) Adding more cards to an already large deck makes RFTG too luck dependent: where whether one player's combo hits and another doesn't outweighs playing skillfully the cards drawn. BoW came awfully close to doing this  without Search to help players out when none of their early cards fit together, I believe BoW would have had too high a luck factor. This issue has been itching me for a while. The idea that a larger deck results in more luck and less skill. I'd never heard of such a thing before coming to bgg. And yet here i hear the complaint time and again.
Can anyone link me to some hard facts on this issue? Something evidencebased, a formula, a dissertation, anything. Alernately just the scientific name of the phenomenon, so i can dig on my own.
Thanks!

Pete Martyn
United States Guilford VT
EXCELSIOR!!!
ZOMGALOMES!!!

entranced wrote: Tom Lehmann wrote: B) Adding more cards to an already large deck makes RFTG too luck dependent: where whether one player's combo hits and another doesn't outweighs playing skillfully the cards drawn. BoW came awfully close to doing this  without Search to help players out when none of their early cards fit together, I believe BoW would have had too high a luck factor. This issue has been itching me for a while. The idea that a larger deck results in more luck and less skill. I'd never heard of such a thing before coming to bgg. And yet here i hear the complaint time and again. Can anyone link me to some hard facts on this issue? Something evidencebased, a formula, a dissertation, anything. Alernately just the scientific name of the phenomenon, so i can dig on my own.
Someone probably can  not me, though!
But think of it this way: you've managed to get several Uplift worlds into play, and it looks like you're in contention to win the game if you can maintain some sort of card advantage. "What I really need," you think, "is to draw and play Uplift Code! Then I'll get extra points and extra cards in Production phases!"
In this scenario, would you rather be drawing from a deck with:
A) Forty cards, one of which is Uplift Code
or
B) One hundred twenty cards, one of which is Uplift Code?
In this case, clearly you'd prefer scenario A. The more cards in a deck, the less likely it is you'll get the one card you really need (see also: Magic: The Gathering, Dominion, and A Few Acres of Snow.) In RftG, the game still manages to feel balanced, because no one card is necessarily going to win the game. In the example above, several other cards would also be helpful, and this is mirrored in other strategies  you've got a good chance, with all the expansions, of using military, Rebel, Imperium, or Alien strategies.
But consider this: if a new expansion added all of those kinds of cards except Alien ones, then the alien strategy is diluted. The ratio of Alien cards to total cards is less favorable, and winning with that strategy requires more luck as a result (probably to a point where experienced players give up on the strategy as too risky, thus devaluing every Alien card in the game. By starting anew, we get new challenges without forcing the designer into trying to balance a dozen different strategies...well, to a degree, anyway.
Does that make sense?



the pete wrote: entranced wrote: Tom Lehmann wrote: B) Adding more cards to an already large deck makes RFTG too luck dependent: where whether one player's combo hits and another doesn't outweighs playing skillfully the cards drawn. BoW came awfully close to doing this  without Search to help players out when none of their early cards fit together, I believe BoW would have had too high a luck factor. This issue has been itching me for a while. The idea that a larger deck results in more luck and less skill. I'd never heard of such a thing before coming to bgg. And yet here i hear the complaint time and again. Can anyone link me to some hard facts on this issue? Something evidencebased, a formula, a dissertation, anything. Alernately just the scientific name of the phenomenon, so i can dig on my own. In this scenario, would you rather be drawing from a deck with: A) Forty cards, one of which is Uplift Code or B) One hundred twenty cards, one of which is Uplift Code? This is a clear choice, but what about this:
A) Forty cards, 2 of which would win you the game.
or
B) One hundred twenty cards, 6 of which would win you the game.
Let's say you are seeing 20 cards for the rest of the game: The probability that you find (at least) one of your wanted cards is equal to
A: 1  (20 ncr 2)/(40 ncr 2) = 0.756 B: 1  (100 ncr 6)/(120 ncr 6) = 0.674
The clue here is "assuming you see the same amount of cards". When the deck size increases, the chances that all your wanted cards are clumped together at the end of the deck also increases.



..and this becomes more intuitive if you think of this scenario:
C) 20 cards, 1 of which would win you the game.
Seeing 20 cards, the probability is obviously 100%.
For a visualization, see: http://www.wolframalpha.com/input/?i=plot+1++%28%28x20%29+...

Dan Bradshaw
United States Farmington Connecticut

entranced wrote: Can anyone link me to some hard facts on this issue? Something evidencebased, a formula, a dissertation, anything. Alernately just the scientific name of the phenomenon, so i can dig on my own.
Thanks!
It's hard to analyze that for a game like this, because defining something like "good cards" and "bad cards" in an objective sense is questionable at best.
That said, in general with statistics, large sample sizes introduce larger degrees of variance, and that provides at least some parallel to this situation. For example, it's much easier to predict how many times you'll get heads on 10 flips of a coin than, say, 1000 flips of a coin. For this situation, the standard deviation (which means, essentially, the degree of uncertainty in the result) depends on the square root of n, the number of "trials" that you conduct. As n increases, the standard deviation increases, and more uncertainty is introduced into the outcome.
For drawing cards from a deck, what you're essentially looking at is the number of permutations; that is, out of the large deck you'll only draw a certain subset of cards, and the order in which you draw them is important (getting New Vinland in your start hand is awesome, not so much later in the game). The number of permutations depends on the factorial of the number of possible outcomes, which is a function which increases unbelievably quickly. As the deck grows, therefore, the number of possible "outcomes" (that is to say, sets of cards you'll draw) increases dramatically. The number of "good" outcomes (again noting the difficulty in defining that) doesn't increase nearly as rapidly.
The whole thing is muddled by the fact that you have some degree of control over how many cards you will see (through choosing Explore actions and powers, leeching with things like Interstellar Bank, etc.) and that interactions between cards are so important (Galactic Federation is awesome, unless you don't have useful developments). Mathematically exploring this would, as a result, be tricky at best. Hopefully, though, this gave you a better picture.
In regards to the more luck/less skill idea, I don't think I buy that. A game like RFTG is very much about managing luck. Sure, sometimes you'll get a perfect storm of cards which synergize perfectly (a bunch of cheap blue production worlds and Free Trade Association) and you can win blindfolded. Sometimes you'll get nothing but stinkers and get steamrolled. These are extreme situations. Most of the time you get a hand that is useable, but needs management. If the number of options goes up, you need to be even more flexible and have a better idea of strategy in order to make it work.

Mark Delano
United States Stamford Connecticut

entranced wrote: This issue has been itching me for a while. The idea that a larger deck results in more luck and less skill. I'd never heard of such a thing before coming to bgg. And yet here i hear the complaint time and again.
Can anyone link me to some hard facts on this issue? Something evidencebased, a formula, a dissertation, anything. Alernately just the scientific name of the phenomenon, so i can dig on my own.
Thanks!
(Edit: I messed up the numbers by not correctly subtracting the removed cards from the deck. The corrected numbers are in parentheses)
Let's say you have a 10 card deck, 5 are developments and 5 are worlds. If you draw 7 cards you are guaranteed to draw a world. Now change that to a 20 card deck still 50% developments and worlds. You have a .05% (corrected .15%) chance of not drawing a world on those 7 cards. If you increase the deck again to 100 cards now it is .5% (.62%), or 10 (4) times more likely. Raising it to 200 cards increases it to .63% (.7%). The effect levels off for a given number of card draws as you increase the deck size, but it becomes more dramatic if you are looking for rarer cards.
Let's look at trying to draw a 6 cost Development. Let's say they constitute 10% of the deck, which is close enough for our purposes. Now even with our 10 card deck there's a small chance we won't draw one on a 7 card draw, 1.81% (30%). 20 cards it jumps to 12.53% (41.05%), 100 cards it's up to 37.66% (46.67%) and 200 it's at 42.5% (47.26%).
With a deck when you draw a card it reduces the chance of drawing a similar card since you can't draw that exact card again. Compare that to dice where it doesn't matter how many times you roll a 6 the chances of rolling it again remain the same. The larger the deck, the closer a simulation of a die it becomes.

Lacombe Louisiana
Suddenly a shot rang out! A door slammed. The maid screamed. Suddenly a pirate ship appeared on the horizon! While millions of people were starving, the king lived in luxury. Meanwhile, on a small farm in Kansas, a boy was growing up.

GoodOmens wrote: That said, in general with statistics, large sample sizes introduce larger degrees of variance, and that provides at least some parallel to this situation. For example, it's much easier to predict how many times you'll get heads on 10 flips of a coin than, say, 1000 flips of a coin. For this situation, the standard deviation (which means, essentially, the degree of uncertainty in the result) depends on the square root of n, the number of "trials" that you conduct. As n increases, the standard deviation increases, and more uncertainty is introduced into the outcome.
That is, in fact, the exact opposite of what happens to the standard error as the sample size increases... and it's disanalogous anyway to the question being asked.
The previous example concerning looking for 1 card from an 80card deck or 1 card from a 120card deck was much better.
Quote: For drawing cards from a deck, what you're essentially looking at is the number of permutations; that is, out of the large deck you'll only draw a certain subset of cards, and the order in which you draw them is important (getting New Vinland in your start hand is awesome, not so much later in the game). The number of permutations depends on the factorial of the number of possible outcomes, which is a function which increases unbelievably quickly. As the deck grows, therefore, the number of possible "outcomes" (that is to say, sets of cards you'll draw) increases dramatically. The number of "good" outcomes (again noting the difficulty in defining that) doesn't increase nearly as rapidly.
Yes, but this muddles the issue. If only a small subset of cards are "useful", it's harder to find them in a large deck, regardless of factorial growth.

Serge Levert
Canada Vancouver British Columbia

Nice, thanks again everyone.
frunkee wrote: Let's look at trying to draw a 6 cost Development. Let's say they constitute 10% of the deck, which is close enough for our purposes. Now even with our 10 card deck there's a small chance we won't draw one on a 7 card draw, 1.81%. 20 cards it jumps to 12.53%, 100 cards it's up to 37.66% and 200 it's at 42.5%. The 7 card draw being 1.81%, that can't be right can it? You have 1 6dev in a 10 card deck. You draw 7 cards. How is it only 1.81% you whiff? Not intuitive if it's right.

Tom Lehmann
United States Palo Alto California

The core issues are:
A) When average proportions of various attributes are held constant across an increasing deck size, each given player still sees only a *subset* of the deck as it is gradually drawn. Thus, the *sample* variation increases (given that players are drawing about the same number of cards per game), even though the underlying *population* variation (frequency) actually remains the same.
B) When a game involves synergies and combos (as RFTG does), drawing a clump of cards (on a sale or mix and match Explore) that work very well together is powerful, whereas drawing a clump of cards that don't work together is not. As sample variation grows with increasing deck size, this luckofthedraw factor starts to loom larger.
(I used a lot of tricks in the expansions to reduce this luck factor including: making more cards multipurpose, so that the actual deck proportions for many characteristics rose to compensate for the increased sample variation; adding Goals to add more scoring "dimensions" to the game, so that the game is less about drawing a single effective combo; adding more "draw1" explore powers (plus some "draw then discard" powers in BoW), so players would see more cards (effectively increasing the sample size to compensate for increased sample variation); and, in BoW, adding Search to help players get started (the first three categories), reduce the risk for some highvariance strategies (the Uplift, Alien, Military 5+, and Takeover categories), ensure that players could find at least two 6devs (since 6devs are a major source of VPs); and ensure that players could find some multigood Consumption (since Produce/Consume is another major source of VPs).)
C) As a deck draws larger, some previously oneofakind "reward a given strategy" 6devs get duplicated (to maintain the deck proportions). Before, when there was only one "Imperium 6dev", say, players couldn't draw two of them, as two simply didn't exist. Add a second one in the expansions (to maintain average deck proportions relative to strategies) and occasionally a player will now get lucky and draw both of them. Playing both of them will score better than placing one of them, plus another 6dev that didn't combo so well.
Both high and average scores rose across Arc1 in part due to A) increasing "power creep" from introducing more multipurpose cards (here's where solving one problem  increased sample variation  leads to another problem); B) the expansion of maximum world VPs from 7 to 9 (this compensates for more reinforcing 6devs) and C) the increased chance of drawing "reinforcing" 6devs due to there being more of them (absolutely) in the deck.
However, high scores rose a lot more than average scores did. That's a indication of increasing luck factor (although high scores will always rise more than average scores over time as players simply experience more games within the "long tail" of possible score distributions).
Does this make sense?

Mark Delano
United States Stamford Connecticut

entranced wrote: Nice, thanks again everyone.
The 7 card draw being 1.81%, that can't be right can it? You have 1 6dev in a 10 card deck. You draw 7 cards. How is it only 1.81% you whiff? Not intuitive if it's right.
And of course I screwed up the numbers. I knew I shouldn't of posted that early in the morning. I'll fix them now.

Serge Levert
Canada Vancouver British Columbia

Tom Lehmann wrote: Does this make sense? Yup, great post, fascinating. The way you dealt with the problem is genius, i have to give you props!

Brendon Russell
New Zealand Auckland

entranced wrote: Tom Lehmann wrote: Does this make sense? Yup, great post, fascinating. The way you dealt with the problem is genius, i have to give you props!
And if this approach was continued indefinitely, you'd eventually end up with an expansion where the new mechanics and card powers consisted solely of solutions to the variance problem, having no room for anything else  so props also for "rebooting" before reaching that point of absurdity!


