# [TIGD] Analysis Paralysis (Common Problem #1)

19 replies [Last post]
Anonymous

Analysis Paralysis is this week's theme in TIGD, as we move from looking at themes to looking at specific common problems. On a broader point, definitions will probably be a lot less contentious, as we're not grouping games by content as much as side-effect. Therefore, I suspect we'll find some very varied games within each weeks parameters.

At its most basic, Analysis Paralysis (AP) is downtime caused by other players having to think about their move.

As to what causes it, it must be the case that it is somewhat individual: some people will find spatial problems paralysingly complex whilst others will only have such a problem with arithmetical functions. Tthere is the plain fact that some people are always very slow in taking a move in a game naturally, but this isn't something game design will account for unless there is a real-life time limit-- a mechanic/device I suspect we'll return to.

Good examples of AP abound, and I'll not worry too much about quoting them as the phenomenon itself is something we're probably all used to. Tikal and Tigris & Euphrates are two games I hear quoted very often for AP problems. Chess is a good example of a game with AP where it is traditionally regarded as a feature rather than a true flaw: it's a game where thinking ahead complex series of moves is an integral part of the game.

Interestingly, AP is usually an example of a surfeit of strategy (or an overdose of options, to continue this awfully annoying alliteration), when providing players with decisions is traditionally seen as the quintessence of good game design. In AP situations, the player cannot process the options and information he has regarding them. It is worth noting that for anybody, some games can be 'pure skill' on paper, but give so many functions that two humans will find it a game of luck, as noone can properly grasp what a semi-optimal option would be.

I suspect AP is mitigated most often in games where other players can at least consider their own future moves. If the game will have decisively changed (as is often the case after one move in chess, and most frequently in multi-player games), then this will rarely be fruitful.

There is a GeekList by Chuck Blahous on types of Analysis Paralysis that discusses the different sorts of games he's found to suffer from AP, although I'd suggest steering away from doctrinal disputes over which category of AP particular games are in! ;-)

As to how this affects game design, I'd suggest that "too many decisions/calculations" is a problem to be concerned about. Also, use of timers is one, rather drastic, way to address it. In one of my games, a partnership game, I had a rule banning a player or his partner from speaking during his turn. This meant that negotiations would be possible between opponents the longer you thought, and would prevent a "partner 1 tells partner 2 how to play" problem to a great extent (which is a slightly different issue, of course). Anyway, I suspect that for many people timers or silence rules will be seen as too harsh, meaning mechanical solutions are probably in order.

Cheers,

Richard.

GeminiWeb
Offline
Joined: 07/31/2008
[TIGD] Analysis Paralysis (Common Problem #1)

I wanted to comment about analysis paralysis, but there were just too many ways to approach it ... where could I possibly start?

...

But seriously, it's an interesting issue.

Richard_Huzzey wrote:

Quote:
Interestingly, AP is usually an example of a surfeit of strategy (or an overdose of options, to continue this awfully annoying alliteration), when providing players with decisions is traditionally seen as the quintessence of good game design. In AP situations, the player cannot process the options and information he has regarding them.

This is a nice way to approach it because, as game deisgners, we have the challenge of providing enough options to give the players control of their destiny, while not confusing them with too many options. Furthermore, we want there to be multiple ways of winning, rather than following a specific formula ... or else the game just comes down to following a formula.

Having said that, analysis paralysis can also be influenced by difficulties in trying to second guess other players, or feeling a need to have to plan too many moves ahead (where too many is generally just a few more than you can do).

So how do we mitigate this?

One way is that while one specific option might be the best in a given circumstance, the penalty function for choosing the second or third best isn't generally going to cost you the game ... although its fair that some options should be obviously suboptimal is specific circumstances.

Personally, I like games which encourage other players to 'suggest' strategies, often to their benefit and other players deficit. However, other people dislike that.

Also, it's nice if the best few options are generally intuitive, particularly if you've adopted a specific strategy. This is where a theme can add a lot of value by 'pointing you in the right direction'.

That's enough for now ... need to recover from my random thoughts ...

zaiga
Offline
Joined: 12/31/1969
[TIGD] Analysis Paralysis (Common Problem #1)

I remember Wolfgang Kramer saying in an online interview that analysis paralysis is mostly a player problem. Is it a coincidence then that a lot of Kramer's games suffer from AP?

I think Kramer is partly right. There will always be people who overanalyze a problem or who simply take a lot of time coming to a decision. On the other hand, I think there is a lot a designer can do to mitigate this problem.

I think it is a common misconception of designers to think that presenting the player with a lot of options is a good thing. This is not necessarily so. Sure, you want to give the player some options to chose from, but it is the quality of these options that matter, not the quantity.

In the ideal game the player would be presented with 2 options. Both options would seem more-or-less equally good (or bad), but would steer the player down two different paths. The player knows that one option will lead to victory and the other won't, but at this point in the game it isn't completely clear which option will lead to success.

I think these kind of strategic decisions are the most interesting. They don't lead to analysis paralysis because you simply cannot compute them all the way through, you have to rely on gut instinct and perhaps experience to make a good decision. However, they may cause turn angst which is, in my opinion, a more tolerable form of downtime. It does show that not every decision in a game needs to be a teeth gnashing, nerve wrecking one. Some decisions can be fairly easy, to keep the game's pace brisk.

The type of decisions that I don't like are the tactical ones, where you can calculate through many permutations of actions and after some minutes of number crunching reach an optimal choice. There is no real tension, except for the fear that you may have missed a permutation or made a mistake in a calculation. I haven't played "Tikal", but from what I've read it seems that it suffers from exactly this problem.

In a strategic game, the increments by which the game progresses should be small, so that you can plan your turn ahead. If the game's configuration changes drastically between your turns, you may have to rethink your whole plan and this costs time. Of course, a game that is completely predictable leads to a different sort of analysis paralysis, so you have to strike a balance here. What I found to be a good trick is to force players to perform a "non-action" every now and then. For example, take an action to draw cards, get money or another action that is necessary for a player to advance in the game, but that doesn't alter the game state in a drastic way.

Analysis paralysis also often occurs right before the end of the game or right before a scoring round. In those cases a player may want to spend some time to find the best way to maximize their score. This may be mitigated by having a "random" ending (ala "Union Pacific", "E&T" or "Ra") or by making a portion of the scoring hidden ("Taj Mahal", "Ticket to Ride").

Simultaneous action selection (SAS) doesn't solve analysis paralysis, but it does solve the downtime problem. In essence, with SAS every player suffers from analysis paralysis at the same time. This allows for a game to have a fairly complex action selection mechanic and get away with it. Games that use this mechanic include "Wallenstein" and "Maharaja".

In a certain way, "Puerto Rico" makes use of "simultaneous action execution" to speed up the game. One player selects a role and then every player more-or-less simultaneously executes the action associated with that role. This speeds up the game tremedously and is the prime reason why such a complex economic building game can be compressed into a reasonable timeframe. This doesn't really have to do with analysis paralysis, but it does have to do with downtime and I wanted to mention it anyway.

Fos
Offline
Joined: 12/31/1969
[TIGD] Analysis Paralysis (Common Problem #1)

I think simultaneous action selection can go a ways to solve analysis paralysis if there actually aren't any turns at all.

This was done with some success in the card game Falling in which players played cards (modifying the game, how cards were dealt, etc.) whenever they felt like it. Granted, this particular system left only frantic tactical decisions, but I think with the right design a game could be produced where if you spent too long between actions the optimal strategy for you would change under your feet, thus creating a sort of time-based mechanic where players have to come up with the right balance between fast tactics and more methodical strategy.

Anonymous
[TIGD] Analysis Paralysis (Common Problem #1)

Fos wrote:
I think simultaneous action selection can go a ways to solve analysis paralysis if there actually aren't any turns at all.

An interesting example; I suppose this does run the risk of hurrying players unnecessarily and bringing in questions of reaction time that most people choose to keep out of their games.

Richard.

zaiga
Offline
Joined: 12/31/1969
[TIGD] Analysis Paralysis (Common Problem #1)

I'm a bit surprised that this topic hasn't attracted more responses. The other topics in this forum were pretty busy. Are people not interested in how to deal with analysis paralysis, or do they not perceive it as a problem in their designs?

Just thinking out loud...

Scurra
Offline
Joined: 09/11/2008
[TIGD] Analysis Paralysis (Common Problem #1)

I have to confess that I think AP is much more a symptom of players rather than the game, especially if it is a game in which the board position changes sufficiently that preplanning is very difficult.
I don't know of good solutions to this: the one player in my regular group who is prone to AP doesn't necessarily make the game agonising for the rest of us, and imposing some sort of external time-limit might well make him feel unwelcome. Of course, if there were two regular players with this problem then we'd probably have to do something about it... :-)

At root, any game must involve analysis to some extent (otherwise it would be Snakes & Ladders) - the key is in making the decision tree a manageable size without compromising on depth. As this seems to be particular to the game in question, I'm not sure you can generalize about it very easily.

Anonymous
[TIGD] Analysis Paralysis (Common Problem #1)

Rene- I too was surprised this one proved so unpopular, although David probably makes good points about the difficulty of discussing it. As to whether I agree with his (and Wolfgang Kramer's) point that AP is mainly a player problem... I think the "decision tree" point is the key, and that games with too broad-branched a tree will have the unfortunate effect of creating such difficulties.

Richard.

GamesOnTheBrain
Offline
Joined: 07/24/2008
[TIGD] Analysis Paralysis (Common Problem #1)

While I agree that too many decisions can cripple a game and that AP is a real problem, I think it often comes down to the perceptions of the players.

While I've occasionally said...

"Hurry up! You're taking too long!"

... I've never said ...

"Turns take too long. This game stinks!"

Anonymous
[TIGD] Analysis Paralysis (Common Problem #1)

I have noticed some degree of AP in games that involve deduction and bluffing. With the possible second and third guessing going on, a player can take a while to make a move that they feel comfortable with. I have noticed this to a small degree in playtest sessions of my recent GDW offering, Good Cat Bad Cat. It is possible to try and outsmart others to such a degree that it could go on forever:

"I think they're trying to do this, but are they really or do they just want me to think they are because they're really doing the opposite but are they really or do they just want me to think that..."

This tends to occur even though the decision tree is actually very bare.

It can get interesting to see people's reactions to bluffing situations.

What to do about it? That's a good question. I have noticed that in more complex games, certain actions/decision can be streamlined or made more intuitive. That offers either simpler (or fewer) choices, or the benefit of having choices that are intuitive to the players and don't need to be mulled over at length. In a game I'm working on, I'm now going through and streamlining complex decision trees to make gameplay more smooth.

It's amazing that I never even considered streamlining gameplay until I playtested with other people. I have extensively solo playtested with generally positive results, but I also understand all the rules and how they work. When others become involved, it becomes very clear which parts of the game are overly complex or unclear, causing players to bog down as they try to decipher their options. After the last playtest, I went back and extensively reworded card text and game components to simplify the decision making aspect. Now I have to playtest again to see if I'm headed in the right direction.

Dralius
Offline
Joined: 07/26/2008
[TIGD] Analysis Paralysis (Common Problem #1)

Other than Joining GAAP (Gamers Against Analysis Paralysis) there seems little you can do to affect a person that must think out every possible permutation of outcomes in a game. If you limit the options to such a degree that they can think through it that easily you are also risking taking out the experience of accomplishment you get by discovering a move that is not obvious; discovery is one off my favorite aspects in game play. Putting timers in games that should otherwise not need a timer just to keep the AP people playing at a pace that we can live with seems a poor solution. The AP player will feel rushed and not enjoy the game since they only had time to explore a small number of play options.

When people say that a game is prone to AP what they are saying is that it is a complex game with many options. This is not a flaw just an attribute of the game, no different than the number of players or suggested age range printed on a box.

Fos
Offline
Joined: 12/31/1969
[TIGD] Analysis Paralysis (Common Problem #1)

Unless the game is entirely reaction/speed based, there will always be players who take too long to make a play. I don't think this is analysis paralysis exactly. Rather, I think of AP as a problem arising when there are no obvious actions a player can take to help him reach his tactical or strategic goals, which is a design problem, not a player problem. Players taking too long might be a symptom of AP, but I don't consider it AP itself.

Richard_Huzzey wrote:
An interesting example; I suppose this does run the risk of hurrying players unnecessarily and bringing in questions of reaction time that most people choose to keep out of their games.

I've been mulling over this gameplay idea for a few days now, and I've come up with a fairly comprehensive list of design restrictions that absolutely must be implemented to avoid confusion, keep the game fun, etc. I still think it could work beyond a card game, but it feels like there are a million ways to do it wrong.

One of these restrictions is a forced mix of tactical and strategic gameplay. Too much emphasis on quick, exploitive moves and the game would be all reaction/speed. On the flip side, too much focus on grand schemes planned 20 moves ahead, and I think the game would have a show-stopping analysis paralysis problem.

I'm going to do a design with this system because I see it as an interesting challenge that should help me control analysis paralysis better. Players who take too much time even if there isn't AP are going to be heavily penalized within this system, however, if true AP exists and there really aren't obvious actions to reach a goal, the game will shift to pure reactionary tactics. While I don't believe a turnless game is an instant fix to AP, I do think it will make AP problems readily apparent.

jwarrend
Offline
Joined: 08/03/2008
[TIGD] Analysis Paralysis (Common Problem #1)

While I would agree that slow, analysis-prone players tend to slow things down, I would say that there's such a thing as analysis paralysis that goes over and above something that can be blamed away onto the players.

When people are "analyzing", what exactly is it they're thinking about? I think there are a couple of sources.

One aspect would be the "too many options" problem; maybe I have 6 Action Points (interesting that "Action Points" and "Analysis Paralysis" have the same acronym; coincidence?), and the number of permutations for these is huge. Or perhaps there are lots of different places I could place a tile, and considering the benefits of each takes a long time.

Another aspect would be the "prognostication" problem; "If I do this, then he'll do that, which means I'll want to do that, to which he'll respond with..." and so on. In a multiplayer game, this gets out of control quickly.

Related to this is the "multiple outcomes" problem; "If I roll the die and hit, then I'll do this, but if I miss, I'll be forced to do that".

Now, I would say that if you have several of these levels of thinking going on in a game at the same time, baby, you're in for a looong ride. For that reason, I think forcing players into one mode of analysis only is optimal, if it can be done.

For the "many options" problem, the obvious solution is restriction. However, as someone else noted, sometimes a more open system can actually require less thought because it's more forgiving. I think that's valid, but I still feel that restriction of actions is the very soul of interesting decisions. Making the actions short and punchy can also help a lot. Puerto Rico is a masterwork in this regard -- when someone else selects "Builder", you have only one very specific and limited decision to make, so there's not too much to think about.

For the "prognostication problem", making the future a little cloudy can help with this. Of course, that can lead directly to the "multiple outcomes" problem, since people will try to plan for every eventuality. Adding chaos to the system via event cards or the like can mitigate that, but it's not always terribly satisfying way to handle things. Giving players secret information can also help; if I can fully evaluate Joe's position, I'm more likely to try to do so than if Joe has a secret hand of cards whose contents I have no idea about.

Anyway, bottom line is that what makes games fun (for many of us) is interesting decisions, and so our job as designers is to motivate those decisions. In that sense, "analysis" need not be seen as a bad thing, and one need not remove any possible avenues for analysis from the players. Rather, one should be careful to restrict the levels on which a player can analyze the game, making only one or two likely to be rewarded.

There's my take on the subject, anyway...

-Jeff

Anonymous
[TIGD] Analysis Paralysis (Common Problem #1)

All I have to say about AP is that games either have it or they don't. If you have a game where a ton of strategy is involved there will always be AP. If you have a trivia game the only AP should be the person figuring out the answer, although in Trivial Pursuit we found the smarter someone is the more AP exists ("Oh man.... I know this...(Scratches head)..."). I've always been a fan of long board games but not because they are long, they usually have a good reason other than AP, such as Runebound. Runebound rarely has AP but it's a long game. So I guess what I want to say is that there is a huge difference between a long game and a game that takes long due to AP.

Anonymous
[TIGD] Analysis Paralysis (Common Problem #1)

Analysis Paralysis typically rears it's head in games with a broad decision tree, particularly when you can make multiple moves, and the order in which you make those moves makes a difference.

The obvious way to decrease it is to see whether you can sensibly reduce the decision tree. Particularly, it is good to try to make the different actions independant of each other, at least in this move. One typical example is that when a counter has performed an action, they can't perform another.

The other typical place is where it is hard to visualise the concequence of a particular move. This depends crucially on the form of the game, and beyond trying to make simple rules, there are few band aids one can directly apply.

Offline
Joined: 12/31/1969
[TIGD] Analysis Paralysis (Common Problem #1)

Glenn Rahman, designer of Divine Right, is a gentleman and a scholar, but he is a really slow player. While playtesting the Right Stuf edition of Divine Right several years ago, I found that Glenn was one of those people who doesn't anticipate his turn at all, he stares at board in unfamiliar wonder at the start of each of his turns.

To speed up the pace of the game, we moved the Diplomacy phase to the end of the turn. That single diplomacy dieroll had a potentially large impact on the decision-making process, so putting it at the end of the turn kept players focused on resolving the current turn. Then came the diplomacy roll, and the player could ponder the implications of that roll during the next player's turn, instead of bringing the game to a halt.

Based on that experience, I feel that analysis paralysis could be reduced by simply delaying important outcomes to the end of the game turn, and maybe also by reducing player knowledge of the relevant factors. Call it the fog of war, and allow for as much secrecy as the game theme can handle, whether it's cards in the hand, or face down counters on the board.

FastLearner
Offline
Joined: 12/31/1969
[TIGD] Analysis Paralysis (Common Problem #1)

Indeed, it's a great trick. Even a simple game like Lost Cities goes much faster because you draw your new card at the end of your turn, giving you the whole other player's turn to consider how it impacts you without slowing the game down. Though I can't think of them right now, I have noticed this trick on a number of other games, too, where new information to consider is pushed to the end of each player's turn. Great solution when it's not too weird (as you note).

-- Matthew

Gogolski
Offline
Joined: 07/28/2008
[TIGD] Analysis Paralysis (Common Problem #1)

Tile laying games like carcassonne, fjord or metro speed up considerably when you draw your new tile at the end of your turn...

Cheese!

FastLearner
Offline
Joined: 12/31/1969
[TIGD] Analysis Paralysis (Common Problem #1)

Excellent examples. Not in the rules, mind you. With Carc it eliminates friendly kibbitzing (or at least distorts it), so I suppose that's the concern there. With Fjords... well, I guess it might discourage me from finding a place for my opponent's tile after he gave up, if I knew what I was about to play.

I definitely prefer Carc that way, by far. We don't even consider playing it any other way: it's both faster and, imo, more fun, as you're (a) not pressured to decide instantly (you've got everyone else's turn to consider your move), and (b) you get that "aw, man, you took my spot" feeling a lot more often as you watch your ideal little places get destroyed as the turn moves around. :)

-- Matthew

clearclaw
Offline
Joined: 07/21/2008
Perception rules the day

Its not the game, its not the players, its the perception. AP is only present if the players consider it to be. Consider: I once played a two player game of Blokus that lasted a little over two and a half hours. What that AP? Blokus games normally last 20-45 minutes. No, it wasn't AP for the simple reason that neither of us considered it to be AP. We were enjoying sweating the decision tree on every turn and so it was fine. On the other hand if either of us has been complaining at the play rate it would have been AP (for them).

For me accusations of AP are a sign that the players playing the game are mismatched in their expectations, nothing else. In particular it says nothing about the game.

I've recently started playing Splotter Spellen's Bus and have found that there are two core camps of fans: those that consider it fun little 30-45 minute romp and those that see a two to three hour careful analysis challenge. The best I can say is that they are both right, the game is both those things, and those people shouldn't play Bus with each other.