Recommend
5 
 Thumb up
 Hide
34 Posts
1 , 2  Next »   | 

Race for the Galaxy: The Brink of War» Forums » Strategy

Subject: Underpriced cards and the value of prestige rss

Your Tags: Add tags
Popular Tags: [View All]
Miika Oksanen
Finland
flag msg tools
mbmbmbmbmb
Even after many, many games (click on the game ids, if you want to see the tableaus and logs) I still think there is something wrong with either the cost of most cards that give a prestige when placed, or the prestige leader mechanic, or both. Thanks to the devastating effect of early prestige, the cheap prestige giving cards offer a huge value for their cost. Here are some of the cards I find most problematic:

-Pan-Galactic Mediator: Only military players might skip this one in the early game.
-Pan-Galactic Security Council: Almost as bad as PGM, sometimes even worse, thanks to the consume power.
-Alien Booby Trap: Almost always a very strong early play, if you can place it somehow.

Each of the three cards would be priced quite reasonably for what they do, if they did not give prestige. A cost or defence of 1 for a card that gives 1 VP and has some ability, that is quite useful in some situations, sounds about right. Just look at cards from base set and 1st expansion. But the prestige makes them almost auto-play cards. Personally I find that a bit boring, and adding unnecessary luck into the game.

Then there are some medium priced prestige giving cards, that are very powerful for their cost: Galactic Markets, Alien Burial Site, Galactic Power Brokers, Information Hub and Rebel Fuel Refinery.

What would be the right price for these cards? Here's what I think:

-Pan-Galactic Mediator: Two or three. Two would do, if it scored -1 points instead of 1, perhaps. It would still be very good play with Contact Specialist or some other similar power, or decent play just for the prestige, depending on the situation of course.
-Pan-Galactic Security Council: Two.
-Alien Booby Trap: 3 defence. It may sound harsh, but this is the only way to prevent it being layed without placing something else first.
-Information Hub: 4 defence. It would still be generally better than Malevolent Lifeforms, which has 4 defence, and isn't a bad card.
-Galactic Markets. 5. Now it's 1 cheaper than Consumer Markets, gives at least as many points, has roughly equally good or better powers (a bit worse for all-blue production, but a lot more versatile), and the additional value from the prestige. At 5, it would still be quite good.
-Alien Burial Site: 3. Taking Spice world, substracting one from the blue trade bonus, which is fairly niche anyway, and making it give prestige hardly seems like a balanced trade-off.

We could also ask how many VPs should these cards give, in order to be about as good on average, as they are now, if they did not give prestige. For PGM it would be 3-4, I believe. Obviously it would be better last turn play then, but in many cases worse 1st turn play. Pan-Galactic Security Council: 3-4. Alien Booby Trap: 3-4. Information Hub: 4-5. Galactic Markets: 5. Alien Burial Site: 3-4. Thus, it seems I would value prestige at somewhere between 2 and 3 VP on average, in the early game, 2 in mid-game and 1-2 in the last rounds. Any thoughts on this?

Of course, it remains a matter of debate, whether the cards even should be balanced in the first place. My personal opinion has been made clear. The great thing about RFTG to me is that it offers great variability like a trading card game, but also balance, unlike any TCG I have played. With the 3rd expansion, I think the game has fallen from exceptionally balanced to well balanced, which is still good, of course. I realize the sheer amount of cards and new mechanics makes the game extremely hard to balance.
8 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Tom Lehmann
United States
Palo Alto
California
flag msg tools
designer
mbmb
Race for the Galaxy is a game of risk/reward, not average returns.

The potential reward for playing a card depends a lot on the game situation and what card combinations you have in hand or draw.

The risk for playing a card depends on its "opportunity cost", what you give up by playing and paying for it, instead of something else.

Yes, some of the cards you mention have considerable upside in the right circumstances. However, I think your analysis focuses too much on the upside in one particular circumstance (namely, when playing that card gives you the prestige lead) and not enough on the overall costs and benefits.

Detailed Analysis

The usefulness of many cards also depends on how many players there are. For example, playing Pan-Galactic Mediator on round 1 is much more likely to earn you sole Prestige Lead in a 2-player game than in a 6-player game, where the odds are much higher that another player also places a prestige generating card (so that you get only 1 VP, not a VP and a card, from your play and may soon lose the Prestige lead to a Prestige "engine").

I have seen many players shoot themselves in the foot on turn one by playing PGM or Alien Burial Site or Pan-Galactic Security Council to end up only tied for the prestige lead. Had they played some more useful card -- for example, Investment Credits or a cheap windfall world to sell for cards, they would have been much, much better off.

Yes, *if* they do get sole Prestige lead, then these first round moves are reasonable, but there is a real risk that they won't get it (the odds of which rises in games with more players).

Suppose you are in the early mid-game, already have 1 prestige (so you could use the Prestige Opportunity card if you wanted to do so), and the Prestige Leader has 3 prestige. How many of those cards you cite are you now willing to play? Mostly, their prestige is now just +1 VP.

Cards that grant you ~2 VPs, but are useful only in fairly specialized situations cost 0-1. For example, Pilgrimage World in the base game is cost 0, a "gray" world, its Consume power can be interfere with your plans (especially if you have no other Consume powers to use instead), and it is worth 2 VPs. Sometimes, PW is exactly what you need (when you have lots of goods and no consumption) and placing it can yield 8 or more VPs. But, most of the time, it is just 2 VPs. Refugee World, another example, has a power (-1 Military) that can interfere, comes with a Novelty good, costs 0, and is worth 1 VP. Sometimes, RW can be quite useful (in Novelty produce-consume cycles), but often it is just worth 2 VPs (its good gets turned into a VP and never replaced). (The circumstances where RW is quite useful are more common than PW, which is why PW comes with a more interfering power.)

Are these cards "no-brainers" to place on round 1 if you have no other useful Settles? For some players they are -- why lose a tableau tempo and 2 VPs? These players feel lucky to have these cards in their starting hands. However, for many players, the opportunity cost of spending a card for a relatively small benefit and an interfering power is not worth it, unless it is part of a greater plan. Why not keep the card in hand to both avoid its interfering power, until you know better what path you're going to take, and as a card to pay for something that is both expensive and very useful that you draw?

With this perspective in mind, let's look at Pan-Galactic Mediator. It has an interfering power (-1 Military), a fairly specialized power that is worthless without a pay-for-military power (which only two of the 16 start worlds have, one of which only works on Alien military worlds), +1 Explore, comes with a Prestige, and is a development costing one. Played late in the game by a player who already has a prestige, no pay-for-military power, and is several prestige behind the leader, it is essentially just 2 VPs and, maybe, an extra Explore draw or two. Given that it costs 1, is PGM really out of line with cost 0 cards with interfering powers like PW or RW in the base set?

Sure, in the situation where you get it out early on and it gives you the Prestige lead for several turns, it is now worth a card and, say, +3 VPs. That's not bad, though arguably worse than what PW or RW can do in the right circumstances at 1 less cost. But, add in the circumstances where you do have a pay-for-military power and get several more prestige out of it, now it's pretty comparable, particularly if those several more prestige help you maintain the prestige lead and generate card draws.

I chose PW and RW for comparison because they also have interfering powers and are fairly specialized. Now, let's look a 1/1 cost dev from the base set, Expeditionary Force, which also has +1 Explore but has +1 Military. EF has no interfering power and +1 Military is far less specialized than PGM's prestige-when-using-pay-for-military power. There are many useful defense 1 worlds to get started, plus a player with Military 1 can use EF to conquer a valuable defense 2 world, such as the many Uplift 2s. So, EF has a more generally applicable power, no interfering power, but PGM comes with a prestige. Everything else is the same. PGM doesn't seem overpowered compared to EF. And, depending on circumstances, I might well take EF over PGM in my starting hand...

Your mileage may vary, but it seems to me that you are over-valuing prestige and under-valuing the potential impact of interfering powers and generally applicable versus very specialized powers.
28 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Andrew
Japan
Tokyo
flag msg tools
badge
Avatar
mbmbmbmbmb
Tom also posted a comparison (which I can't seem to find) using PGSC as an example. Basically it costs you a bunch of cards that midgame would have funded a 6-cost, losing you points comparable to maintaining uncontested prestige lead.

Of course, if your opponents are doing the "multiplayer solitaire" thing and dawdling, this might not be apparent.

Edit: Here's one post (not the one I was thinking of) by Tom on prestige.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Belgium
Brussel
flag msg tools
badge
Avatar
mbmbmbmbmb
We've been playing without the prestige leader rule for a while now, and prestige cards are still at least worth playing and usually even ahead of the curve.

Conclude from that what you will about the prestige leader rule - we don't miss it, in any case. The game is a lot more interesting without it.
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Miika Oksanen
Finland
flag msg tools
mbmbmbmbmb
Thanks for your replies, everyone, especially Mr Lehmann.

Quote:
Race for the Galaxy is a game of risk/reward, not average returns.

This is true, but it still might be useful to look at expectation values, to get more information on whether the risks are balanced with the rewards. I forgot to mention I assumed "correct play" in my analysis, meaning no choices, that can be fairly easily to be proven to be suboptimal. This would mean for example ruling out PGM in 95% of early game situations, if starting with military power, or having a good reason to go for military power. When we rule out these quite obvious situations, I would say the prestige point on PGM would be roughly worth a random number from 1 to n, where n is 8 minus the number of the current round in 2 player advanced, or something like 7 minus the number of the current round, divided by 2, with 3 players. So when I said I would value the prestige point for 3vp in early game on average, I meant this.

My problem number 1. is, that the expectation value for say, PGM seems so high in the first turns, ruling out the forementioned situations, that it's very often (almost always, I believe) an obvious choice. Of course, these obvious card choices are not a new thing in BoW. Examples include Interstellar Bank and many of the military worlds, when you can place them. But I think these obvious card selections occurred less often before. This may be partly explained by myself being more experienced now. I would like having a hard time more often with choosing placed cards and discards. I would also like to see more varying kinds of tableaus.

I also think prestige adds more luck into the game. However, I don't think this is much of a problem by itself, unless I were to participate in some sort of RftG competition; it might be troublesome to go through a large number of games, to decide who plays better. The game is still very hard to figure completely, so I would say small skill differences will probably become evident even with the best players, given enough data, say 50 games.

I own all the expansions and I still think the game is one of the best out there, but I also think it could be better! wow

EDIT: I think the comparison of Spice World and the prestige giving blue world (forgot the name), is a bit off. It only applies assuming you are planning to live off trading the blue goods, and do not draw cards in produce phase for example.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Holger Hannemann
United Kingdom
Upper Heyford
Oxfordshire
flag msg tools
badge
Avatar
mbmbmbmbmb
My wife and I have about 200 plays of RtfG, this is the game I introduced my wife to gaming with, and we were excited like 4-year-olds on Christmas upon opening every expansion box.
Yes, nowadays this game is a luckfest for us because our skills are on really equal levels, and I think we've seen everything you can do in this game. We still highly enjoy it, and despite lots of other games we tried and own RtfG is still the number one game for us. We know some cards are really out of balance, but as Tom said, some of the cards you mentioned are really worthless in some setups.

If I wanted to play a perfectly balanced game I'd play chess.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Troy Adlington
United States
Dallas
Texas
flag msg tools
badge
Avatar
mbmbmbmbmb
Stormparkiet wrote:
We've been playing without the prestige leader rule for a while now, and prestige cards are still at least worth playing and usually even ahead of the curve.

Conclude from that what you will about the prestige leader rule - we don't miss it, in any case. The game is a lot more interesting without it.


I play a lot online. And I RARELY EVER focus on prestige as a way to win a game. (Unless my starting hard pointed to it as my engine)

I win a respectable amount so I kind of smile at these prestige are over-powered arguments.

If anything I find produce/consume strategies to be weaker than before.

My personal weakness? I don't focus on the goals enough. However this does not lead me to think that the goals are broken or to prefer to play in games without them.

Just my 2c


4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Michael Brough
msg tools
mbmb
Something to bear in mind when doing this kind of evaluation: if you're playing Produce/Consume*2, any production world is generating you 1VP each round (2VP in 2pa). And this is reliable, unlike the 1VP a round you might get for being Prestige Leader - it's not unlikely for someone to be able to get ahead of you in Prestige, but (apart from in very specific circumstances involving takeovers) they can't take your production worlds from you!
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Chris Linneman
Canada
Vancouver
BC
flag msg tools
Avatar
mbmbmbmbmb
I think Prestige is clearly better balanced for games with many players (4+) than few (2-3). The risk/reward that Tom talks about is very real in games with high player counts, whereas in a 2p game an early prestige point can very easily return many VPs and cards before your opponent catches up. Even if you lose the prestige lead, you prevented your opponent from getting it 1 prestige earlier.

In 2p, I would venture to say Uplift Gene Breeders is broken if you draw it in your opening hand.

But, I don't think there is any way to perfectly balance a card game like RftG across multiple player counts. I don't think it's necessary for every card to be equal, as long as there are no cards that are utterly useless. Well, Outlaw World notwithstanding.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Chris Linneman
Canada
Vancouver
BC
flag msg tools
Avatar
mbmbmbmbmb
UlyZed wrote:

Is Outlaw World worse than Expanding Colony?


Yes. The produce on blue windfall power is actually useful once in a while.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jason T
msg tools
Troymk1 wrote:
I play a lot online. And I RARELY EVER focus on prestige as a way to win a game. (...)

I win a respectable amount so I kind of smile at these prestige are over-powered arguments.


I find this puzzling. I've been playing a lot against the RFTG AI and I find that whoever gets Prestige Leader first almost always wins by a pretty wide margin, even between the two AI opponents. If there are goals in play that reward you for being the first to get to 3 Prestige or 2 Prestige/5 VP, that widens the margin even further. It's possible that the AI has some flaws in its strategy, but I'm commonly getting games where two of us score 40-50 and the third scores 70-80. I've only played four games of Brink with friends, and the point differentials weren't as dramatic, but the early-game Prestige Leader did win 3 out of 4 games (and lost by a couple points due to a small mistake at the end in the fourth game).

I never saw this as an issue with card point costs so much as an issue with the Prestige Leader Mechanic itself. I have been wondering whether it would work better if it only gave a card OR a VP to the Leader, and being tied means only a card to each, but I'm considering playing without Prestige Leader entirely to see how playing with Prestige might still work.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Serge
Canada
Burnaby
British Columbia
flag msg tools
badge
Avatar
mbmbmbmbmb
coinop25 wrote:
I've been playing a lot against the RFTG AI and I find that whoever gets Prestige Leader first almost always wins by a pretty wide margin, even between the two AI opponents.

The AI strongly overvalues prestige. Therefore it will often seem to be a large factor if it beats you.

coinop25 wrote:
I've only played four games of Brink with friends, and the point differentials weren't as dramatic, but the early-game Prestige Leader did win 3 out of 4 games (and lost by a couple points due to a small mistake at the end in the fourth game).

I've played 400 f2f games of BoW and prestige/prestige leader are ignored half the time (i don't mean we play without it, i mean it's not as big a factor as it initially seems). You still just make the best play with what you draw, prestige doesn't change all that much in the grand scheme of things. It certainly doesn't seem unbalanced, which isn't surprising given how dedicated the creators are about playtesting the game to death.

coinop25 wrote:
I have been wondering whether it would work better if it only gave a card OR a VP to the Leader, and being tied means only a card to each

As has been discussed before, that makes prestige stronger not weaker, because cards are more valuable than VP early game.

coinop25 wrote:
but I'm considering playing without Prestige Leader entirely to see how playing with Prestige might still work.

Did you take out Galactic Trendsetters when you were playing the base game?
Did you take out Terraforming Guild, Alien Toy Shop and Improved Logistics when playing TGS?
Did you take out Pan-Galactic Research when playing RvI?
Then you don't need to take out prestige leader when playing BoW. ;)

Those aren't even great comparisons, because BoW does a lot to balance all strategies equally. In Base/TGS engines are overpowered; in RvI military is OP; while in BoW, nothing is OP.
3 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Wei-Hwa Huang
United States
San Jose
California
flag msg tools
designer
Avatar
mbmbmbmbmb
The MatrixCube wrote:
If I wanted to play a perfectly balanced game I'd play chess.


Chess's balance is much worse than RFTG. Half of your pieces rarely have more than two possible moves, and all the others can basically run all around the board. Two of those are limited to what squares they can reach. And while the cost of losing your pieces is somewhat commensurate with their power, there's one glaring exception -- a mid-level strength, short-range piece, but you have to keep it alive or you lose the game.

Chess may be mostly a fair game (there's still the first-player advantage), but it ain't balanced.
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Sean McCarthy
United States
Seattle
Washington
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
onigame wrote:
The MatrixCube wrote:
If I wanted to play a perfectly balanced game I'd play chess.


Chess's balance is much worse than RFTG. Half of your pieces rarely have more than two possible moves, and all the others can basically run all around the board. Two of those are limited to what squares they can reach. And while the cost of losing your pieces is somewhat commensurate with their power, there's one glaring exception -- a mid-level strength, short-range piece, but you have to keep it alive or you lose the game.

Chess may be mostly a fair game (there's still the first-player advantage), but it ain't balanced.


I'm not sure what you are taking "balanced" to mean. TheMatrixCube (I believe) meant that in Chess, the positions dealt to each player - which is the bit out of their control - are evenly matched (which is not quite true because white goes first, but close enough).
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Miika Oksanen
Finland
flag msg tools
mbmbmbmbmb
Balance can mean many things, but I think the interesting question here is: to what extent and how often is the game decided by luck?

A game's balance function f could be defined like this: For each pair of skill levels (x, y), f(x, y) = p(x, y), where p(x, y) is the likelihood of the higher level player beating the lower level player. We could use this information to assess the game's depth, as in how early does the game's skill level plateau: for example, what is the smallest skill level c, so that f(x, c) < 0.55 with all x higher than or equal to c?

The interesting question then would be, how easy and fast it is to reach level c? We may not see anyone or anything do it with chess, before there are AI players that can take every possible later board situation into account at every situation, and play optimally. I haven't reached c yet in RftG, but I'm guessing it will be easier now with BoW, than it was with RVI, partly because of prestige and cheap prestige giving cards.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Rob Neuhaus
United States
New York
NY
flag msg tools
Avatar
mbmbmbmbmb
UlyZed wrote:
Lofski wrote:

A game's balance function f could be defined like this: For each pair of skill levels (x, y), f(x, y) = p(x, y), where p(x, y) is the likelihood of the higher level player beating the lower level player. We could use this information to assess the game's depth, as in how early does the game's skill level plateau: for example, what is the smallest skill level c, so that f(x, c) < 0.55 with all x higher than c?


How many people understand this? Surely there is a better way of communicating your point?



I'd bet at least four at sometime regular posters will understand the OP within 3 minutes of reading it.

Imagine you have some rating system (possibly Elo or TrueSkill). Call C the level or rating of an "expert" player. This expert is good enough to beat everyone (including the best player) at least 45% of the time. The OP wants to measure the depth of the game by considering the level/rating of the lowest rated expert.

In a somewhat similar vein, others have proposed using the difference between the minimum and maximum rated "serious" player for a fixed rating system to be a good measure of the depth of the game.
3 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Miika Oksanen
Finland
flag msg tools
mbmbmbmbmb
UlyZed wrote:
Lofski wrote:

A game's balance function f could be defined like this: For each pair of skill levels (x, y), f(x, y) = p(x, y), where p(x, y) is the likelihood of the higher level player beating the lower level player. We could use this information to assess the game's depth, as in how early does the game's skill level plateau: for example, what is the smallest skill level c, so that f(x, c) < 0.55 with all x higher than c?


How many people understand this? Surely there is a better way of communicating your point?



Ok, to make it more simple, I think balance in these types of games, is about how often and to what extent the outcome is decided by luck, as in not skill. One interesting question would be, how easy it is to get to a skill level c, where you can beat anyone or anything in 1v1 at least 45% (or some other percentage, from say, 40 to 49) of times, or any two players in a 3p game at least 29%-32% of times, and so on. (To make the comparison fair for shorter games, we could consider either a single longer game, or some sort of 2-4 hour match, consisting of a series of games). The skill level of the game could be said to plateau at this point c.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Børge N
Norway
Porsgrunn
flag msg tools
badge
Avatar
mbmbmbmbmb
UlyZed wrote:
Lofski wrote:

A game's balance function f could be defined like this: For each pair of skill levels (x, y), f(x, y) = p(x, y), where p(x, y) is the likelihood of the higher level player beating the lower level player. We could use this information to assess the game's depth, as in how early does the game's skill level plateau: for example, what is the smallest skill level c, so that f(x, c) < 0.55 with all x higher than c?


How many people understand this? Surely there is a better way of communicating your point?


As Rob pointed out, he wants to measure a game's luck/skill ratio by finding what skill level you need to achieve before skill doesn't matter "in practice" anymore.

Like: Does the top 10% players win about half of the games amongst each other? Or at what other skill level percentage?

See the geek list Levels of Expertise for some good discussions about skill/luck in various games.

Edit: Ninja'd by Miika, and some added clarifications.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Rob Neuhaus
United States
New York
NY
flag msg tools
Avatar
mbmbmbmbmb
45% is somewhat arbitrary. It's a value not too much less than 50%.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Rob Neuhaus
United States
New York
NY
flag msg tools
Avatar
mbmbmbmbmb
It is arbitrary in the sense that any value from say, c = .3 to c = .49 is unlikely to matter much.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Edward
United States
New York City
flag msg tools
badge
Avatar
mbmbmbmbmb
So when you said "no one understands this", what you really meant was that you don't understand it.

Let's put it another way. The card game War is not "deep" because everyone, even someone with a very low Elo rating, can beat the best player in the world, let's say, 40% of the time. [This is an arbitrary number. It doesn't matter. You'll see why.]

Chess, however, is very deep. There's maybe only 10 people who can beat the best player in the world 40% of the time. They have extremely high Elo ratings.

Miika's and Rob's point is therefore to measure the depth of a game by the lowest skill level someone can be and still beat the best player 40% of the time. War is not deep because that number is extremely low (you can be very bad and still beat the best player 40% of the time); Chess is deep because that number is very high (you have to be very good to beat the best player 40% of the time). You could replace 40% with 30% or 20%, but as long as you're applying the same %age to everyone it doesn't matter that much in the final analysis, since you're just comparing games' depth against each other.
3 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Edward
United States
New York City
flag msg tools
badge
Avatar
mbmbmbmbmb
Suppose I have five cars. (This is how you know this is a thought experiment.) I want to see which of them has the best acceleration, and which has the worst. So I plan to test each of their 0-60mph times.

But then you say, no, 60mph is arbitrary! Why not 40mph? Why not 80mph? All the rest of your acceleration test is an assertion stemming from such a large assumption!

But honest, it doesn't matter. What's important is how the cars perform in this test relative to each other, and as long as you pick a reasonable constant and stick to it, you get what you're looking for. You'll get basically the same ranking if you test the cars' 0-55mph instead, so complaining that the number is arbitrary is neither here nor there.
9 
 Thumb up
0.02
 tip
 Hide
  • [+] Dice rolls
nishadham
United Kingdom
Bristol
Avon
flag msg tools
Avatar
mbmbmbmbmb
UlyZed wrote:
Plainly put, you guys are putting a lot of weight on a 'let's say 40%' that then becomes the basis for all further analysis

rrenaud wrote:
It is arbitrary in the sense that any value from say, c = .3 to c = .49 is unlikely to matter much.

Alex - I think what Rob said here is that you could do this Game-Depth test several times using a range of assumption-values, e.g. from 30% to 49%, and the results would be roughly the same. I.e. the "Depth Scores" the games came out with would be roughly the same relative to each other. Therefore, he predicts that the actual number that you use can be arbitrary. If you didn't think so, you could do the test several times and find out. The test itself sounds pretty reliable.

(+The car analogy is inspired.)
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Miika Oksanen
Finland
flag msg tools
mbmbmbmbmb
theory wrote:
So when you said "no one understands this", what you really meant was that you don't understand it.

Let's put it another way. The card game War is not "deep" because everyone, even someone with a very low Elo rating, can beat the best player in the world, let's say, 40% of the time. [This is an arbitrary number. It doesn't matter. You'll see why.]

Chess, however, is very deep. There's maybe only 10 people who can beat the best player in the world 40% of the time. They have extremely high Elo ratings.

Miika's and Rob's point is therefore to measure the depth of a game by the lowest skill level someone can be and still beat the best player 40% of the time. War is not deep because that number is extremely low (you can be very bad and still beat the best player 40% of the time); Chess is deep because that number is very high (you have to be very good to beat the best player 40% of the time). You could replace 40% with 30% or 20%, but as long as you're applying the same %age to everyone it doesn't matter that much in the final analysis, since you're just comparing games' depth against each other.


Actually, 20% or 30% could be a bit problematic for games that involve a lot of randomness, but where skill clearly shows in the long run. Comparisons between deterministic games, where all the information is visible to both, and games like RftG, might not be fair. To make the test work for many different kinds of games, I would go as far as using 45%.

Furthermore, I think we should rather consider the amount of time it takes for skill differences to show, instead of the number of games, since lengths of games vary a great deal. Some games (and poker sessions) take a very long time, so some adjustments for them might be in place to prevent them seeming deeper than they should. Same thing for very short games, except different kind of adjustments. For games taking about half an hour, we could use 45%, 47% for games taking only 15 minutes, 40% for games taking about an hour and 20-39% for long to extremely long games. These adjustments are completely arbitrary, but something along those lines should work, and the fairness of the adjustments could easily be evaluated analytically.

To get meaningful results, we could look at the percentage of players, who can beat anyone or anything with at least the aforementioned probability, for each game, like Børge suggested. If 80% of grown-up people with normal intelligence and who have played the game in question at least a few times, have reached this skill plateau, the game would obviously not be deep at all. On the other hand, something like 1% would indicate a very deep game. This approach should work, if all the players are humans. In case they aren't, we could rather try to evaluate how hard it is to reach the skill level in question, but this might be hard to do objectively.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Nolan Lichti
United States
Indianapolis
Indiana
flag msg tools
mbmbmbmbmb
theory wrote:
So when you said "no one understands this", what you really meant was that you don't understand it.


Whoa. To be fair, I didn't think it was very well communicated either. I didn't take the time to parse it; my eyes mostly glazed over.

On the other hand, your explanation was easy to parse, and made it easy to go back to what Miika said and parse that as well.

Interesting read, all. Thanks.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
1 , 2  Next »   | 
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.