Recommend
1 
 Thumb up
 Hide
22 Posts

BoardGameGeek» Forums » Board Game Design » Board Game Design

Subject: Is ability as a designer helped by skill as a player? rss

Your Tags: Add tags
Popular Tags: [View All]
Andy Haigh
United Kingdom
flag msg tools
Avatar
mbmbmbmbmb
A flippant question for you: do you need to be able to play well to design well?

And the thought process that led to it: in a game where multiple strategies emerge from (more or less) complex synergies between game elements, to what extent has that all been planned out intentionally by the designer?

Is it that the designer has foreseen all possible interactions, and adjusted the game accordingly, or is it more a case of finding out through playtesting what breaks the game? If it is (as I suspect) a little from column A and a little from column B, how far along the line and in which direction does it lie?

Does a designer need to be adept at spotting all the interactions between game elements (one of the hallmarks of a good player), or just good at adjusting the design when problems crop up?
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Steve Hope
United States
Woodside
California
flag msg tools
badge
Likes: Mountains, Tundra Turn-offs: Serpents, Marsh
Avatar
mbmbmbmbmb
I think good designers have to be good (but not excellent) players, but good or excellent players are not necessarily good designers.
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Rob Harper
United Kingdom
Wantage
Oxfordshire
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
flibertygibbet wrote:
A flippant question for you: do you need to be able to play well to design well?


I don't think so. Analysing what works in a game to get a strong play experience is not necessarily the same thing as analysing a game to spot the strongest moves or strategies.

On the flippant side, since I got more seriously into game design it's possible I have become a worse player, as I am now analysing games in different ways.

Quote:
And the thought process that led to it: in a game where multiple strategies emerge from (more or less) complex synergies between game elements, to what extent has that all been planned out intentionally by the designer?

Is it that the designer has foreseen all possible interactions, and adjusted the game accordingly, or is it more a case of finding out through playtesting what breaks the game? If it is (as I suspect) a little from column A and a little from column B, how far along the line and in which direction does it lie?

Does a designer need to be adept at spotting all the interactions between game elements (one of the hallmarks of a good player), or just good at adjusting the design when problems crop up?



I think that this is a matter of designer style. Some will plan everything out in advance, while others will evolve the game during playtesting. Experience helps you make initial choices that are closer to the final version, and also to spot how things need to be revised later.

As far as designing or developing those synergies between game elements, this is a skill that develops (again, whether it is during initial design or later development is a matter of personal style). I certainly look to include elements that I think will interact in an interesting way from the beginning, but I very often miss the mark on the first attempt and iterating over the design helps focus in on the necessary improvements. I think that as I gain experience I am getting better at this and my first shots are usually (but not always) better, but I have a long way yet to go in building my skills.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Bojan Prakljacic
msg tools
Avatar
Eric Lang lost playing 'Others' during the showcase. So...
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
David Thomas
msg tools
designer
mbmbmbmb
I think being a good player initially helps in guiding you towards what kinds of games and designs you might like, but the more time you spend designing the more your playstyle changes as you do start deconstructing games when you play them. I find myself looking at games and wondering how I can break them, or what kinds of edge cases I can cause simply because I want to see how the designer worked around them and if that information could be useful in my own designs.

As for when I do my own designs I typically come up with a few combos or interesting interactions, but many tend to be a result of watching my play testers and seeing what they try to do and if it seems interesting tweaking things to facilitate the kinds of things they are trying.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Robert Wolkey
United States
Spokane Valley
Washington
flag msg tools
Avatar
mbmbmbmbmb
Does a programmer need to be good at Chess to create a program that will beat a grand master?
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Gil Hova
United States
Jersey City
New Jersey
flag msg tools
designer
publisher
badge
Avatar
mbmbmbmbmb
flibertygibbet wrote:
A flippant question for you: do you need to be able to play well to design well?


No, it's not necessary.

What is necessary is to have playtesters who can break a game in 30 seconds or less. They're the ones who are going to find ways to win that a designer never even contemplated.

This is completely speculation - I have no hard evidence - but designers tend to be so close to our designs that we're the worst people to find degenerate strategies, since we're so closely attuned to how people "should" be playing the game.

I have noticed, anecdotally, that ambitious new designers who go in obsessed about balanced wind up making a game that is perfectly balanced... and completely boring to play. Good game design is just as much about player experience and strong incentives as it is about mechanical balance.
13 
 Thumb up
0.25
 tip
 Hide
  • [+] Dice rolls
Rob Harper
United Kingdom
Wantage
Oxfordshire
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
IngredientX wrote:
I have noticed, anecdotally, that ambitious new designers who go in obsessed about balanced wind up making a game that is perfectly balanced... and completely boring to play.


Perfectly describing my first 20 or so game designs. I'm a slow learner.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
james napoli
United States
Westwood
New Jersey
flag msg tools
badge
Avatar
mbmbmbmbmb
i think there is something to be said about being a thorough gamer, and you may say they are good.
since good is a bit subjective.
i've had the fortunate to game with a lot of folks in our industry and a lot of the well known designers do tend to be good gamers, gamers who often more easily identify a winning strategy and work towards executing it.

i think in some ways a very thoughtful designer would likely get in their own way, my wife and i tend to balance this way as she's very creative and often thinks of interesting ideas/concepts and i often (and quickly) tend to denote the issues/exploits and loopholes. We havent landed on a working design yet, but someday!

for me, i think a good designer is likely a good player but as Gil mentioned having varying playtesters ultimately what will make a good game.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
J C Lawrence
United States
Campbell
California
flag msg tools
designer
badge
Avatar
IngredientX wrote:
This is completely speculation - I have no hard evidence - but designers tend to be so close to our designs that we're the worst people to find degenerate strategies, since we're so closely attuned to how people "should" be playing the game.


It is hard to play the game on the table as versus the one in your head: you're ever so much more familiar with the one in your head.
3 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jeremy Lennert
United States
California
flag msg tools
designer
Avatar
mbmbmbmbmb
I think being a skilled player helps with design. It allows you to do more of the iteration in your head and makes you less dependent on others to playtest your game. Also, it's easier to see what things will or won't appeal to a particular group when you're part of that group.

I don't think it's the primary skill that a designer needs, though.
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
patrick mullen
United States
Washington
flag msg tools
mbmbmbmbmb
My flippant answer: no.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jeremy Lennert
United States
California
flag msg tools
designer
Avatar
mbmbmbmbmb
Wolkster wrote:
Does a programmer need to be good at Chess to create a program that will beat a grand master?

I don't think that's the same question, though this question has some interesting nuances in its own right.

First of all, you clearly don't need to be a good Chess player to duplicate an algorithm that someone else has already perfected for playing chess. But that's not saying much.

What would you have needed if you were a pioneer in the field?

A typical Chess AI has two main parts. The first part is a tree search of all possible moves: the computer lists all the moves it could make right now, then for each of those moves it lists every possible counter-move the opponent could make, and then for each of those counter-moves it lists every possible counter-counter-move it could make in response, and so on.

To do a tree search, the only thing you need to know about Chess is how to determine the legal moves--no strategic insight whatsoever is necessary.

If you have enough computational resources, you can continue this search all the way to the end of the game, and then you know for sure what sequences of moves will win or lose. That's called "solving" the game. It's easy for some games (e.g. Tic-Tac-Toe) and harder for other games (e.g. Go), but theoretically you can do it for any strategy game if you just have a big enough computer.

So with a big enough computer, you can make a perfect AI player for any game (including Chess) without the programmer knowing anything at all about the strategy of that game.

But we don't have computers big enough to do that with Chess, which is where the second major part of the typical Chess AI comes in. This part is known as the heuristic and it's a math formula used to estimate how favorable or unfavorable a given position is. For instance, a position where you've taken the enemy queen is probably more favorable than one where you've lost your queen.

When you can't carry the tree search all the way to the end of the game--which, in Chess, is almost all the time--you carry it as far as practical, and then you apply your heuristic to decide which is the best outcome you can achieve (keeping in mind that your opponent is working against you).

Unlike the tree search, the heuristic does rely heavily on the programmer's knowledge of Chess strategy. A lot of the numerical parameters can be tweaked empirically (e.g. if you're not sure whether a queen should be worth 9 pawns or 10 pawns, you can try both and see which performs better), but figuring out which traits you should check for in the first place is mostly on the programmer; you have to at least realize that it matters whether you've got doubled pawns, whether your rook is on an open file, etc. or else you won't even create a parameter to tweak.

Even when you're using empirical tests for the values of your parameters, you probably have to get into the right ballpark before you start, or you won't make much progress. (If you think your rook is only worth a pawn, you'll trade it for a pawn, and then it was only worth a pawn to you in that game...) Your AI is likely to have some weaknesses that you can only see when you have it play against an outside expert--which means you need an outside expert for it to play against.

So if you wanted to create a Chess AI "from scratch" (rather than building on the work of others), and you follow the traditional path for building one, then yes, you need to have some skill at Chess yourself.


Of course, there's more than one way to make an AI. Tree search + heuristic has proven to be a very powerful technique for Chess; not so much for some other games.

World-class computer AIs for Backgammon are a bit different. They still do some tree search, but since Backgammon relies heavily on dice-rolling, there's a high branching factor (i.e. many possible outcomes per turn) and not much ability to "prune" the tree (you can ignore some possible moves because they are stupid; you can't ignore any possible dice rolls). A great Chess AI might look dozens of moves in advance; a great Backgammon AI might look, say, two or three moves in advance.

The backbone of a typical Backgammon AI is a neural network that evolves a strategy by playing millions of games against itself and trying to predict which side is going to win. Then, it makes the move that it predicts has the best chance of winning.

In this case, the programmer doesn't need to know any Backgammon strategy, because he doesn't tell the neural net how to win--he just tells it how to learn.

(Of course, to get a truly world-class AI, you probably mix in a few subtle tricks to handle some special cases, rather than relying 100% on the neural net. Those tricks require some additional insight into the game being played.)


So for the more general question of "do you need to be good at some particular game to make an AI that's good at that particular game?", the current state-of-the-art seems to be: "sometimes no, sometimes yes...and sometimes you can't make a good AI in either case." It depends on what technique you use and how amenable that particular game is to that particular technique.
6 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Christopher Dearlove
United Kingdom
Chelmsford
Essex
flag msg tools
SoRCon 11 23-25 Feb 2018 Basildon UK http://www.sorcon.co.uk
badge
Avatar
mbmbmbmbmb
Reiner Knizia is a strong game player. Martin Wallace is not as strong. Make ofvthat what you will.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Boss Beau Blasterfire
United States
Berrien Springs
Michigan
flag msg tools
badge
Avatar
mbmbmbmbmb
I imagine it should help while play testing.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matt Green
England
Horsham
West Sussex
flag msg tools
designer
badge
Avatar
mbmbmbmbmb
IngredientX wrote:
I have noticed, anecdotally, that ambitious new designers who go in obsessed about balanced wind up making a game that is perfectly balanced... and completely boring to play.


Agreed. "Designed by ex-Pro Tour Magic player..." usually means it's going to be a statistical value-management exercise.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Bojan Prakljacic
msg tools
Avatar
IngredientX wrote:


I have noticed, anecdotally, that ambitious new designers who go in obsessed about balanced wind up making a game that is perfectly balanced... and completely boring to play. Good game design is just as much about player experience and strong incentives as it is about mechanical balance.


Eldritch Horror is completely unbalanced. There are few mythos cards in there (and items / spells / tasks) that can brake the game toward the quick defeat or quick win. With all that said, it is still my favorite game and I play it almost every week. Game experience > Balance.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Jeremy Monts

Louisiana
msg tools
I think a designer should be good at their games, lol. This does not mean that they should win, just that they should have all the information.

I think a designer does NOT have to be able to see all the interactions. One reason we playtest is to have people with distance from the game assist us in determining the quality of the game, and help us find broken bits.

My team and I have this discussion quite frequently BTW. And then we have to pull out our spreadsheet of who has won what game to analyze who is the best.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
David Web
msg tools
flibertygibbet wrote:
A flippant question for you: do you need to be able to play well to design well?

Does a designer need to be adept at spotting all the interactions between game elements (one of the hallmarks of a good player), or just good at adjusting the design when problems crop up?


Answer is 'depends on your definition of play well'. Using the general definition of 'play well' as in 'winning the most', no you don't need to play well at all.

But using the definition of 'play well' being 'discovering more things, making cool tricks, knowing the game and taking it to its limits', then yes, a good designer needs to enjoy exploring the possibilities of a game. Here is why:

Given player A called TryHard who plays to win, and player B called FlashyPlays who plays to discover more stuff. Player A may be the best at a game, finding the most efficient path to victory in the shortest number of rounds, but player B is able to think and discover more ways of playing which, ultimately makes him know all the tricks of a game. That is crucial to game development, because there is always one of these super smart people who just wants to break the game and will find a way to break it through the poorly written rules that the designer designed. Being FlashyPlays, you can prevent that.

Moreover, being FlashyPlays, the game becomes more fun because he or she will aim to have a game with more possibilities while TryHard will just make the game harder.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Andrés Santiago Pérez-Bergquist
United States
Mountain View
California
flag msg tools
badge
Avatar
mbmbmbmbmb
mgreen02 wrote:
IngredientX wrote:
I have noticed, anecdotally, that ambitious new designers who go in obsessed about balanced wind up making a game that is perfectly balanced... and completely boring to play.


Agreed. "Designed by ex-Pro Tour Magic player..." usually means it's going to be a statistical value-management exercise.


Top competitive players tend to be good developers, who can take a rough design that's already got a bunch of neat ideas in it, then prune, refine, and balance it into a great experience.

Good designers, on the other hand, have a different skill set. They should be good enough at games to be able to see how systems work very quickly, be able to predict the gameplay consequences of rules, and have an understanding of how different sorts of players will interact with rules systems. If you're not a skillful player in general, all that will be very hard, but designers don't need to be the best players; they need to be "good enough", while also possessing the ability to invent novel, enticing rules systems that draw players into engrossing experiences.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
JT Schiavo
United States
Frederick
MD
flag msg tools
mbmbmbmb
I think it's entirely possible to accidentally design an intelligent system. It may not be fully functional in the first playtest, but there can be enough of a spark to be iterated upon.

Sometimes you just throw things against the wall to see what sticks. Sometimes an unexpected afterthought can become the heart of a design, or a future design. Sometimes something just works and you don't know why.

I'd say it's more important for a designer to understand how a game should feel, and how to extract that feeling, rather than predetermining how every possible interaction and strategy can play out.

But having a good analytical mind, and knowing a lot of games, surely doesn't hurt.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Posthumous Jones
United States
Milwaukee
Wisconsin
flag msg tools
badge
Avatar
mbmbmbmbmb
They seem like different skills to me. Good design is primarily creative (this varies widely, depending on the game) while good play is more of an analytical endeavor. You obviously have to be good at games to understand how they work (or don't) but your primary job is creating something where previously there was nothing.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.