General Color matters / devotion

CML

Contributor
Everyone agrees there shouldn't be Monopoly-level luck, but I'm sure there should be more than chess has. We should also yak about different notions of luck:
-Guy gets dealt AA three times in a row while we get KK. We are unlucky here but once both players have wagered everything we would be lucky to win.
-With 2 seconds left, the Pacers back off LeBron and he jacks up a 3 and connects, though he's only (say) 35% in that situation and they defended it well.
-We flood out in a game 3 where we feel we're heavily favored, and maybe we are. He's at 4 as we take one last draw step. It's Warleader's Helix. 'You got so lucky!' he yells.

Many game situations involve multiple kinds and they ought to. There's also the idea of a disconnect between intention and execution (throwing up a 3?) that interests me and that FSR didn't cover. This has always been a point of difficulty for sports video games to approximate, and I believe it exists for intellectual as well as physical attempts.

Anyway, 'My opponent's brain and intentions' are the essence of incomplete information and should likely form the backbone of each game. Luck should be on multiple levels, not "lol Park Place," but this is an obvious idea. The more interesting one I have is that game design is obligated to place this luck on multiple levels -- level 0 (dealt an unbeatable hand), level 1 (dealt an unbeatable hand that gets beat), and so on. This might help explain some of the stuff I typed in other posts.

A decision that doesn't matter isn't interesting. You could've done anything else.

Chess is a much reduced series of decisions, once you've memorised a bunch of openings, counter openings and counter counter openings. Top level chess is about doing that, although lower level chess is indeed a bunch of decisions. I sit with the ideal that games should be designed/balanced at the top level of play first, and for casual players second.


I played competitive chess in HS (I'm not very good) and so I understand it somewhat dimly, mainly metaphorically. But, damnit, I have played. I think you misapprehend higher-level chess, and I'm stoned, so I'm going to write some paragraphs:

You're right that it requires rote memorization to get to the top of chess, which is a game design flaw chess shares with Magic, but openings are not so important as you think -- calculation is, as is a general strategic understanding of planning and execution and so on. You don't get better by memorizing the Yugoslav Attack, though when you get good you should know what it is. This is OK, though, since at higher levels chess is, as far as I can tell, a good game. It has its flaws, such as the corrupt governing base, the tendency towards Grandmaster draws, the tournament structure, the lack of a community, and the superiority of engines to people. But there is still room for excitement and spectators and aesthetics. Hell, the most "beautiful" games of chess are played by computers.

When chess is decided by brilliancies and subtleties and calculation and foresight and bringing all its concepts together, it's pretty cool. When it is played by anyone who is not quite good, though, or without a long time control (say 90m / side), it will usually be decided by catastrophic mistakes -- mistakes that are "objectively" bad, that computers would reject in a millisecond, and so on. There's too much precision required by Chess. This is by far the most important skill you need to master to progress beyond beginner -- all the ideas about middlegame IQP chances with the c3 Sicilian and the pawn-storm and doubled pawns and pins are fascinating, but they do not matter, they are not of consequence, because chess prioritizes a different and more restrictive and less interesting skill set. There are other things in chess that indicate this truth -- if I walk by two people casually playing chess, I will instantly notice several absurdities in the position, and I can only assume that, if someone much better than me walked by me playing myself, the result would be the same. The better player wins too often in chess, and its ability to support prodigies also points to deficiencies -- chess is a more self-contained world than even the Magic nerd cult, which limits its metaphorical significance -- it's not a compelling or fun way of seeing the world, which I guess is why I majored in literature and not economics. There is little irony in chess, little that compels its players to revise their judgments, and irony is what lets humans tower above mere gods. The Luzhin Defence is a decent novel about this. Anyway, below an elite level chess is just boring, and above that level it's subject to diminishing returns, in my opinion. This may very well come down to a judgment of taste, but none of us have trouble bashing EDH or Grim Mongo Cubes.

The best board game of all time is Scrabble (another Nabokov thing) and it's not close. Computers are poor Scrabble players. Guess why?
 
I stand corrected.
Anyway, below an elite level chess is just boring, and above that level it's subject to diminishing returns, in my opinion.
This is kind of my overwhelming view of chess from a game-design/balance perspective, although I wasn't aware of the nuances.

E: the tendency to grandmaster draws is, I think, what I was subconsciously alluding to, as I knew that was a thing. The reasons for it are clearly not what I thought though!

Also, you're wrong, the best board game is probably (and unironically) Dungeon Petz. It is, at least, a combination worker placement and resource management game, that is both simpler than agricola, and not dull as ditchwater.
 

Laz

Developer
The best board game of all time is Scrabble (another Nabokov thing) and it's not close. Computers are poor Scrabble players. Guess why?

Hmm... I wonder how hard it is to design an algorithm for making Scrabble moves...
Once you have a huge dictionary (probably a complete dictionary), score maximisation is the easy part.
Planning the effect of making certain letters available to be used based upon all public information (number of letters, complete dictionary, letters used on board, tiles in hand, etc) is solvable, but not simple.

The former is probably enough to beat most human players, while the latter is one step towards 'optimal' play.
 

Onderzeeboot

Ecstatic Orb
That was a very entertaining and insightful post CML! Let me add then that a game that is inherently flawed can still be fun. I like chess, even though you would probably spot a billion inconsistencies in my games, it is still fun to play against someone of my own level. The question then is, can an inherently flawed game be a good game, and I think it can. As long as a game is fun and worthwhile, it does not need to be perfect. Chess is such a game.

I also agree Scrabble is beyond awesome. Sadly I don't get to play it too often, and I'm not about to join one of those senior play groups that get all grumpy when you try to lift the serious mood with a lighthearted jest. Dixit is also a game I enjoy a lot, which a computer would suck at.
 
Scrabble is a good game, but like most games it has flaws.

In particular, I think the scoring system doesn't properly reward long words enough and it overvalues specific places on the board and high value letters. To fix that, I added this rule which plays a little on the already existing 50 bonus point rule for using 7 letters.

You gain additional points for using 4 or more letters. These bonus points are in place of any bonus points you would receive for adding an "S" to extend another word (you may choose to take the points for extending the other word with an "S" or you may take the bonus points for using more than 4 letters - whichever is higher).

4 letters = 5 points
5 letters = 10 points
6 letters = 25 points
7 letters = 50 points

Simple rule addition and it opens up a lot of options since you are really rewarded for coming up with long words (which both make the game play better by opening up more of the board, and it gives another tactic besides high value letters on high value board spots - both of which are highly random in nature).
 

FlowerSunRain

Contributor
I find that as a general rule, random elements are a weakness in games and as such exploring how to best layer that randomness is a design-wise dead end: game design should be seeking to minimize random elements.

Still, I enjoy plenty of games with “luck” and I think games should use randomness when it enables them to create a unique gameplay experience that wouldn’t be possible without it. While the random elements are not desirable, the game remains fun in spite of them because it allows me to gain access to a different experience. In short, I think in any game design, if given a choice between a random and a non-random that will serve a particular mechanical function, the non-random element is always superior. The main reason I feel this way is because when a loss is directly traceable to a random element, it is never interesting play because the loss was without any agency on the part of the players (though it might be interesting as a story, like a TV show, being excited to “see what happens next”).

The idea that games need to have randomness strikes me as false. Particularly in multiplayer games, multiplayer chaos gives the game all the variance you need. I think the best multiplayer games are those with zero randomness, zero hidden information and systems that encourage complex, malleable and opaque incentive structures, where winning is a product of being able to both shape the gamestate in a way that the other players actions benefit you as well as exploit the gamestate in a way that harms the other players. The game never gets in the way of the players and is purely a language of expression that rewards the players.

That said, the shorter a game is, the more randomness a game can have for two main reasons. First, it’s a lot less frustrating to lose 20 minutes to a bad draw/roll then 2 hours. Second, if it’s a short game, you can play it more than once where the luck might favor someone else next time. In other words, I’m ok with getting a bad beat in Magic or Money, but no in Imperial because I can play 12 games of Magic or Money in the time it takes for 1 game of Imperial. It’s also easier to appreciate the “see what happens next” aspect of a short game.

As such, I’d go with Imperial, Clippers, The Great Zimbabwe, Stephenson’s Rocket and some 18XX titles as being the best designed multiplayer games because they manage to craft the most interesting gamestates that give the players total agency over the outcome of the game. It’s not that other games aren’t fun/good/interesting/engaging or anything, it’s just not the “ideal”. I’ll never turn down a great game like Quo Vadis, Mascarade, Container, Majaraja, Princes of Florence, Amun-Re, Traders of Genoa, Bohnanza, Samurai, Contract Bridge or a good Magic Cube just because it has some luck in it. Having that luck lets them play differently from zero luck games, which is nice for variety’s sake.

For two player games, I can see wanting some luck for variance, but I still think Caylus is the best two player game around and the only variance is pregame and public, so even their I don’t agree.

Lastly, I only really care about how a game makes me feel, so the fact that a lot of the games I listed aren’t and will never be remotely popular is not a concern of mine. Games like Poker will always be x100,000 times more popular then games Container, but I don’t think quality of game design is the reason for that.
 

CML

Contributor
I really think poker is the perfect design. A minute to learn, a lifetime to master. I hear the same things about bridge but I've never played.

You guys are right about Scrabble and I am dumb. It is pretty good though.

Keep iteratin
 
FSR,

I agree with most of what you wrote. The one additional variable that impacts Magic in particular is balance (or in the case of Magic, the lack of). Games with no random component (purely skill based) only work if they are completely balanced. If there are inherent imbalances in the game, it falls apart. Players will find and exploit them and the game will break down.

Chess is perfectly balanced. Both players have the same board position, pieces and options. Magic can never have that kind of balance because of all the cards in the game (each deck is unique). Therefore, a totally skill based version of Magic (where you don't draw random cards, you just grab what you need when you need it) would be an awful game that would devolve into figuring out which cards are the most powerful and running just those. You might not even have a meta outside one or two decks. Either it would be pure rock paper scissors at that point (all about matchup), or it would be even more broken and simply be a question of who goes first and does either guy make the wrong move. Regardless, it would be terribad.

Long story short, Magic only works because it has randomness. Most games boil down to which guy drew what he needed. The trick (and what makes it fun) is making games not "feel" like that is happening. Where you have a bad matchup and you feel like you overcame that matchup with good play choices and not just because your opponent didn't draw what he needed (even if that is actually the case). Make no mistake about it. A lot of magic games are over before they start (most people just don't realize it).

And this is where the subject of consistency comes into play. Too much consistency and the game becomes more a matchup question (assuming similar player skill levels). Not enough consistency and you get games that end with screw/flood/silver bullets where the loser feels like they had no control over the outcome.

Because the power level of all the cards in cube is so high, you avoid a great deal of the second scenario (especially if you have strong mana base support). I think the trend here on Riptide of removing many of the OP cards (i.e. GRBS) contributes to this goal and it's why I would not go back to a power max design. I wouldn't want a completely flat power curve, but exponential on either side makes the meta worse IMO.
 

FlowerSunRain

Contributor
I really think poker is the perfect design. A minute to learn, a lifetime to master. I hear the same things about bridge but I've never played.

You guys are right about Scrabble and I am dumb. It is pretty good though.

Keep iteratin

I don't think the strength of poker is its game design. Rather, I think the simplistic design combined with the monetary stakes creates an epic environment to enjoy the "metagame" parts of gaming.

Which, in itself, is great game design in its own way I suppose, though in a completely counter-intuitive fashion. While I prefer games that form the means of interaction between the players and generally otherwise stay out of their way, poker does the exact opposite, constantly dictating outcomes and giving players little means to interact, but by doing so forces play solidly into the realm of stakes, intentions, metagame connections and perceptions which is certainly an interesting place to be, even if my gamesnob side tells me its not "much of a game". It certainly has substance and it certainly is a game, so saying poker isn't a strong design isn't much of a claim.

Its probably silly for me to be waxing poetic about zero luck incentive web games on a forum about Magic (which is much closer to Poker then Imperial), but I'm sorry I'm eclectic and even though Magic contains basically none of the things that I look for in a good game, I like it anyway!
 
I don't think the strength of poker is its game design. Rather, I think the simplistic design combined with the monetary stakes creates an epic environment to enjoy the "metagame" parts of gaming.

Yup. Poker without the "stakes" component is a pointless and awful game. The only real skill in Poker is knowing when to bluff and when to fold and that dynamic does not exist if there isn't something at stake (worth losing).
 

FlowerSunRain

Contributor
ahadabans: I'm absolutely not saying Magic should try to be Imperial. Different games do different things, that's what makes them different. While I would say Magic isn't the best designed game, it more then makes up for its faults by having an off the charts "See What Happens Next!" quality to it that can only come from having a huge pool of cards that are assembled freeform, then shuffled and played out with tons of hidden information. A lot of the fun comes from the satisfaction of seeing your creation work, the joy of pulling off a blowout and the surprise of how two decks will end up interacting. Even if the actual gameplay ends up being pretty simplistic, the experience of playing through the games can still be a lot of fun.

I guess what I'm trying to get at here is that in terms of fun, the actual quality of the game design (in terms of the gameplay it creates) isn't nearly as important as we may think it is. There's a good reason why most of you probably have never even heard of most of the games I listed above even though this is a gaming forum (where one might expect people to have heard of games).
 
I guess what I'm trying to get at here is that in terms of fun, the actual quality of the game design (in terms of the gameplay it creates) isn't nearly as important as we may think it is.

I'll drink to that. :)

I'm often pretty critical of the game, but take that with a grain of salt. If I didn't really think this was a fun game, I wouldn't still be playing it or be on a cube forum.

Your point in particular about "what happens next" is a great one. It's probably my favorite aspect of the game. It's why I like cube in particular because there are so many interactions and you are always finding new ones each time you play (since you build different decks instead of playing with pre-built ones). Its also why I tend to favor multi-player games just because of all the crazy scenarios that tend to play out.
 

FlowerSunRain

Contributor
Nope, white goes first. I remember reading that this does something like give it a 50% win rate to black's 49.5% or something, but forget exactly what.

Hilariously enough, there are some people who believe that white has no advantage/black has an advantage in chess and the white advantage is due to lack of skill or "mass psychosis".

With any abstract strategy game, by definition ideal play can only lead to one possible outcome (first player wins, second player wins, game is a tie). The reason these games work is because ideal play is so opaque that we (and even our most powerful computers) can't find it so that the game is effectively balanced by our limitations.

Although some simpler games like Tic-Tac-Toe (game is always a tie) and Connect 4 (first play always wins) have been solved.
 
playing a single hand of poker is pointless random and dumb

That the money management side of things can turn thirty hands of poker into something interesting is all kinds of amazing.

Also, I feel like I should bring up cosmic encounter because magic was Garfield's first attempt at Cosmic Poker, his platonic ideal game. Cosmic Poker sounds pretty rad.
 

Chris Taylor

Contributor
playing a single hand of poker is pointless random and dumb

That the money management side of things can turn thirty hands of poker into something interesting is all kinds of amazing.

Also, I feel like I should bring up cosmic encounter because magic was Garfield's first attempt at Cosmic Poker, his platonic ideal game. Cosmic Poker sounds pretty rad.

I think a few of my friends may have played that while high. Then again maybe they just had a really good time playing poker.
 

Jason Waddell

Administrator
Staff member
Even poker has some variance reduction techniques that people use. Some games use "run it twice", where, once the players are all-in, you actually deal out two sets of remaining cards. Then you can win all, half, or none of the pot. Theoretically you could just assign each player their statistical share of the pot, but I imagine this would be less satisfying from a gambling perspective.
 

CML

Contributor
Even poker has some variance reduction techniques that people use. Some games use "run it twice", where, once the players are all-in, you actually deal out two sets of remaining cards. Then you can win all, half, or none of the pot. Theoretically you could just assign each player their statistical share of the pot, but I imagine this would be less satisfying from a gambling perspective.


the best reducing-variance technique is A LARGE SAMPLE SIZE, which is not really possible in Magic and is also very difficult to attain in live poker or even online tournament poker. this is one of many reasons MTG theory these days reads like "Super System" or that generation of poker books, groundbreaking but primitive and of little but anthropological and historical interest
 
Top