Skip to Content

A game of time

12 replies [Last post]
Johan
Johan's picture
Offline
Joined: 10/05/2008

Hello

This discussion is about the time length of a game. The main question is: Where are the time limits?

I believe that time is one of the most complicated components when I designing a game. The length of the game has to feel right for that game. You will always end up with tests, but you don't want to come to the test group with a game that takes forever.
How do you calculate, how long time a game will take?
How do you see what effects a rule will have on the game?

Some month ago I tested a new game for the first time. I had two groups that tested the game at the same time and it was quick, fun and took around 1 hour to complete. After collecting there opinions and suggestions, I did some small changes. A new prototype was created and the game was up for test a second time. One of the small change in the rules made the game go from 1 hour to 2.5 hours and the game become both slow and boring.

One of the main problems with gaming time is the down time. I know that the question about down time between the actions come up several time, but we have also some other related situations about down time:
If you are removed/eliminated from the game, how long time can you wait? Can it be OK to wait 10 minute if you have been an active player for 50 minutes or and then eliminated. Is 20 minute wait and 40 minute active OK or…?

Can the components lower the time for a game? Usage of scoring board instead of paper and pencil will probably take away some time, but what about other things as game board to handle a card game...

Finale
Is there an upper and lower time limit for different type of games?

// Johan

Hedge-o-Matic
Hedge-o-Matic's picture
Offline
Joined: 07/30/2008
A game of time

Wow. these are good questions. I, too have seen the occasional small rule change completely change the game is major ways, including game length. In the case of the last rule change you made, by the way, I'd recommend going back to the "fun and quick' version, and ditch the rule change you made.

As for game length in general, it depends on the type of game. For abstract games, anywhere from fifteen minutes to an hour is about right. Longer games imply a particularly entangled situation, but just last night I had to add an additional victory conditon to one of my games, since my wife was killing me, but I was hanging on in a way that was just prolonging the agony.

For board games, I shoot for a playing time of fifteen minutes per player, plus twenty minutes. So if it's a four-player game, I try for a playing time of around 1 hour twenty minutes. But I also try to eliminate downtime as much as possible, keeping the players involved when it isn't their turn. This isn't always possible, of course, but if not the play of the person whose turn it is should be entertaining enough for everyone, rather than just watching someone move resources around. Interaction keeps the gameplay alive longer. Drier games need to head to a climax pretty quickly. As for downtime, if there's no way for players to affect the gameplay of the active player, I'd say each turn should take no more than five minutes, and two is better.

But generally, I'd advise against being too aggressive in changing rules during playtesting. After ever change, play the game a few times before making another, and don't hesitate to undo a change, if it has immediate negative consequences, like you last change did. In general, changes that make gameplay faster as always better, unless they eliminate much of the engaging complexity. But with four players in a game, their interactions will provide a lot of complexity you can't foresee with the rules, making the game longer than you foresee. Simpler is usually better, I say.

Rick-Holzgrafe
Rick-Holzgrafe's picture
Offline
Joined: 07/22/2008
A game of time

Quote:
How do you calculate, how long time a game will take?
How do you see what effects a rule will have on the game?

I start by playing solitaire games, just me pretending to be all the different players. This is not a good simulation of a game with real players, in a number of ways, but it can reveal gross problems without wasting anyone else's time.

Of course, solitaire play can't tell you everything. Play is always different when you don't know what the other players are thinking. Different people approach a game in different ways, and will surprise the designer (as well as each other) with their play. The time each turn takes will be different when players are wondering what the others are thinking, and so on. So I'm not arguing against play-testing, just saying that a few solitaire runs beforehand can be helpful in ensuring that a game is basically functional and playable.

Johan
Johan's picture
Offline
Joined: 10/05/2008
A game of time

Hello

Hedge-o-Matic wrote:
I, too have seen the occasional small rule change completely change the game is major ways, including game length. In the case of the last rule change you made, by the way, I'd recommend going back to the "fun and quick' version, and ditch the rule change you made.

Actually I did the other way around and let the rule become more dominate (I will see how it work next Saturday). The rule was about the number of cards you could play during your turn. In the first case, there was no limit, but if you had the right combination you will win that turn. That happened 20% of the time.

Hedge-o-Matic wrote:
For board games, I shoot for a playing time of fifteen minutes per player, plus twenty minutes. So if it's a four-player game, I try for a playing time of around 1 hour twenty minutes. But I also try to eliminate downtime as much as possible, keeping the players involved when it isn't their turn. This isn't always possible, of course, but if not the play of the person whose turn it is should be entertaining enough for everyone, rather than just watching someone move resources around. Interaction keeps the gameplay alive longer. Drier games need to head to a climax pretty quickly. As for downtime, if there's no way for players to affect the gameplay of the active player, I'd say each turn should take no more than five minutes, and two is better.

Every game is unique, but I think that could be a good base rule.

Rick-Holzgrafe wrote:

I start by playing solitaire games, just me pretending to be all the different players. This is not a good simulation of a game with real players, in a number of ways, but it can reveal gross problems without wasting anyone else's time.

Of course, solitaire play can't tell you everything. Play is always different when you don't know what the other players are thinking. Different people approach a game in different ways, and will surprise the designer (as well as each other) with their play. The time each turn takes will be different when players are wondering what the others are thinking, and so on. So I'm not arguing against play-testing, just saying that a few solitaire runs beforehand can be helpful in ensuring that a game is basically functional and playable.

I have never managed to get the feeling of how long time a game would take out of the solitary test. I have a hard time to predict what the players would come up with.

One game was build around that player should try to get the cards on the table. Each time a card was taken, a new card where put in that place. The game was ended when the last card was placed on the table (you received points for the cards you had in the end).
I calculated the time to be around 1 hour and the solitary study did show the same. In the first test, the players tried to rob each other instead and after 2 hours I stop the test. Only half of the deck was used.

// Johan

Scurra
Scurra's picture
Offline
Joined: 09/11/2008
A game of time

Hedge-o-Matic wrote:
For board games, I shoot for a playing time of fifteen minutes per player, plus twenty minutes. So if it's a four-player game, I try for a playing time of around 1 hour twenty minutes.

Yes, that tends to be the rule I use too. Of course, all parts of the equation are subject to circumstance: if you've got a fixed-turn game, then it's easier to aim at, say, "fifteen minutes per turn" rather than base it around number of players.

I'm not necessarily convinced that downtime has to be reduced to a minimum though (although it's usually better that way.) And I certainly don't think that "drier" games have to head to a climax more quickly: but hey, everybody likes different things! (Then again, I have been to *watch* Chess tournaments... :-)

I do agree that watching someone else moving resources around can be dull, but if those actions change what you can or cannot do on your turn, then that can be significant. (It can lead to different issues - one of the reasons I hate Tikal! - but that's a different discussion!)

I do think there is also an interesting "first game" factor: there are some games that will just take much longer on their first (or even second!) outing as a "typical" session. Puerto Rico is probably a good example of this - I'm sure most people's first games of this took well over two hours, but I'd be surprised if that continued. During early testing, it's sometimes difficult to assess whether the length is due to unfamiliarity or game issues.

larienna
larienna's picture
Offline
Joined: 07/28/2008
There is no rule, it is a matter of taste

Hello

Long games can be as fun as short game. It depend on the type of people you are marketting at. Still even in long games, if you could make some speed optimisation/simplification, it would be appreciated.

The main important thing with long games is to make sure that the players are always in the game. If the players must wait after others to play, it can be boring for them so you should avoid this. You can also ask other players to plan their turn in advance if they have no choice. Or make sure that there is some elements in the game that will always keep the player active.

I remember playing a WW2 game call "bells of Wars" which is a much more detailled "axis & allies" game. We played the game for more than a month and it was interesting. Each saturday night we gathered together and played the game for around 5 hours. Then we covered the game with a plastic board and leave it there until next week. What was cool, is that during the week, I could look at the board and start planing my strategy for next saturday. I even realised that it could be more fun just to have a free day of thinking between each turn. Like some sort of long term strategy game.

sayonara(^_^)

Anonymous
A game of time

Well, if you go with the german model of gaming then in the most broad view your game probably shouldn't go past a hour and a half, no players should ever be eliminated, and downtime should be absolutely minimal.

There are exceptions to these guidelines that come out of germany, but most games will do all of these things.

If you have a lot of players, say 5+, then normally the game is designed so that people are capable of acting in every phase of the game. Such as with the Settler's 5-6 player expansion, they add a rule that anyone can build after the main player builds, that way you don't have to have a static position as you watch five other players do your turn before you can build again.

Lots of other games are built so that players can act all the time, such as with Puerto Rico, or Cosmic Encounter. It's the old school tradition of american game design where you saw lots of downtime in games.

Likewise, player elimination is almost completely absent in german game design. Rather than creating a zero-sum situation, such as with risk, where the player who eliminates everyone else wins, instead german gaming is typically about getting as many victory points as possible before the game ends. I'm actually straining to think of a german game that does have player elimination.

I'm not saying that you should aim to make a german styled game, but it is the dominant model at the moment in the boardgaming industry and thus if you make something that either takes a long time to play, elminates players, or has lots of downtime then it probably won't get good reviews.

jwarrend
Offline
Joined: 08/03/2008
A game of time

I would like to gently contradict the idea that downtime is a bad thing; I would say that a more accurate statement is that a game needs exactly the right amount of downtime, and no more. I can give a simple example from a game I recently tested: I have whittled the turn sequence down to very short, punchy actions. The only problem is that the actions are so punchy, that people weren't prepared when their turns came up -- "It's my turn already?" The result is that the game took much longer to play than it should have, because there was effective downtime associated with people planning their moves.

This founders an important point, that a little bit of downtime is a good thing, as it gives players time to plan their moves. In that sense, what I think David is objecting to with games like Tikal is that if the board situation changes dramatically between turns, then this downtime is useless: you can't be planning your moves between turns, because things change too much for you to plan effectively.

Addressing Johan's question more specifically, though, it is indeed a tough thing to gauge how long a game will take. As David notes, it's sometimes tough to separate "first time playing" length from actual game length. (And as a general note, it's often hard to separate player effects like poor play from actual game flaws -- I've had some games pronounced "broken" by new players, and I think it's common for players to assume that any flaws in a prototype are due to the game itself, and not their own sub-optimal play).

I try to assign a reasonable length of time for each decision, and a useful guideline is perhaps 1 minute. Unfortunately, even if an action is very quick -- "draw a card", eg, if the player must choose it among several alternative actions, it will still take a minute for them to reach that decision. For deciding whether to make a bid or not, perhaps 20-30 seconds; for deal-making, perhaps 1-2 minutes. These are very crude, and highly variable, but I think that they roughly give the effect that deal-making games are time consuming, and phases like auctions can be time-consuming, particularly if the auctions goes around the table multiple times, because it involves many decisions being made in serial.

I generally try to think in terms of how quickly experienced players will be able to play, and so I think I tend to underestimate game lengths, but I think that actual lengths tend to converge to my estimates as players get more experienced.

In my experience, solo testing is an abysmal gauge for game length. It's just too hard to keep track of four different positions and the strategies that you're adopting for each position, and to be able to go back and forth between them and take actions as quickly as someone who was worrying exclusively about that position. (Of course, solo tests are great for other things).

I realize I haven't shared anything definitive, but I think the best strategy is to form a reasonable guesstimate of the length each action/decision will take, and then just multiply. Obviously, factor in peculiarities of your own group -- "Joe the human turtle" will take longer to make decisions, "Frank the human CPU" will take less time. But to be safe, you shouldn't really factor unusually fast players into your analysis, as the estimates you'll derive won't be as universal.

Hope this helps,

Jeff

zaiga
Offline
Joined: 12/31/1969
A game of time

It might be useful to compare your design to existing games to get a grip on how long it will take. Things that take time are making decisions, moving components around, or counting scores. The time it takes to make a decision can vary in time a lot, depending on the players, the complexity of the decision, whether a player can plan for this decision and whether the decisions are being taken in serial or parallel order (simultaneously).

Example. El Grande has nine rounds. During each round each player makes, basically, two decisions:
- bidding (est. half a minute per player)
- taking an action card (est. 1 minute per player)

Three times during the game there's a scoring round which takes about 5 minutes in total. Let's assume about 5 minutes extra for setting up the game, shuffling cards, etc.

Assuming five players, that brings the total amount of time needed for the game time to: ((0.5 + 1) * 5 * 9) + (3 * 5) + 5 = 87.5 minutes, which sounds about right for experienced players.

Now, suppose the game is played with slower players, or players who or new to the game, who take a bit longer taking their decisions. Assuming each decision point takes half a minute longer it would lengthen the game by 45 minutes!

Another example. Ticket to Ride. How many turns does this game have on average, approximately? The game ends when someone has played 43, 44 or 45 of his trains (44 on average). That means he has played 44 cards, which means he has drawn 40 cards. That's at least 20 turns of card drawing. Probably a bit more assuming he has drawn a joker card a few times, and perhaps he ends the game with a few cards in hand as well. So, let's assume 24 turns of card drawing.

In my experience players draw extra tickets during the game once or twice on average, so let's assume 1.5 turns spent drawing tickets per players.

Tracks are between 1 and 6 train spaces long, which means the average track length is 3.5. That means that playing 44 trains costs approximately 12.5 turns.

To sum it up:
- 24x draw card
- 1.5x draw tickets
- 12.5x claim track

The action of drawing tickets is definately a time sink, and costs approximately 2 minutes. Drawing cards, and claiming track cost less time, I'd say around 15 seconds per player. That makes: (24 * 0.25) + (12.5 * 0.25) + (1.5 * 2) is about 12 minutes per player. A five player game would then take about 60 minutes, not including initial setup and choice of starting tickets, which mirrors my experience with the game.

The trick is finding out how the decisions in your game compare with the decisions in other games. There are far more decision points in Ticket to Ride than El Grande. However, the decisions in Ticket to Ride are a lot "snappier", resulting in a shorter playing time. In my experience designers tend to underestimate the time it takes to make the decisions in their own games, and they have too many of them resulting in a playing time that is too long.

Hedge-o-Matic
Hedge-o-Matic's picture
Offline
Joined: 07/30/2008
A game of time

One technique I've been using lately, and failed to mention, is playing "solo" playtest games and keeping notation of what I do in each turn. This way, I get a good feel not of how long each turn takes, necessarily (since I'm writing notation for the turns, as well as learning the gameplay details), but to come to grips with how long the game plays in turns. This number will remain somewhat constant, even after players speed up gameplay by learning the rules. Then it becomes a matter of how long each individual turn takes. EDIT: Some of this was discusses in the post above mine... need to read those other posts after a week of absence! /EDIT

This is an easy way to do pre-playtest breaking of your games. First, try to break your game in the simpler way: have one of your "player personas" dedicate all their efforts to hindering advancement of anyone, in order to keep the game prepetual. If your game lasts a hundred turns without noticable gains in the player's situation, it doesn't matter how short the turns are. The game can be brought to a screetching, boring halt by an obstructioninst player.

The second break-test is trying to find perfect strategies, and thus outside the play time concern. But keeping notation may help you discover these as well.

Johan
Johan's picture
Offline
Joined: 10/05/2008
A game of time

Thanks for the input. There were several good points how to make my time calculations better

The main reason I asked the question was that I normally estimate the game time on experience (just took a look at the rules, the components and number of players and guest the gaming time). Most of the times there has OK but I have also got some really disasters. When the time to test games are limited, I don’t want to be the one that make some prototypes be dropped and they have to wait for a month to get it tested.

// Johan

Pt314
Offline
Joined: 12/31/1969
A game of time

I once designed a starship combat card game, but it was a disaster. When I had the rules for regenerating shields, the game never got anywhere, it could have gone on forever. When I got rid of the regenerating shields it completely changed the game (And it still took over 2 hours to play my game).

Since then I have worked on other games, trying to avoid this mistake as much as possible.

Anonymous
A game of time

I'm currently working on a game, where players are in charge of controlling planetary defense satellites. The SATs are composed of tiles, everz tile represents one part of the SAT (thruster, mass driver, solar panel etc.)

Currently, the player has to manage power by physically transfering energy tokens from the solar panels to other systems (much like in Illuminati)

Cool mechanism, has a lot of possibilites, but It's very likely that I'll have to drop it because of the extra game lenght it generates

Another cool mech has been already dropped for the same reason.

Syndicate content


forum | by Dr. Radut