Tuesday, October 24, 2006

A Good Scoring Game

What makes a good score for a game? This is almost the same as asking: what makes a good game?

Is a game good when one player simply wins and all others don't? When all players score very closely? When players score proportional to their strengths?

Let's take a look at a few case and see what we can tell about the games from the final score. Some scores leave fingerprints with which we can tell not only the game but a bit about how it was played.

In the following game ending scores patterns, I examine games that produce different types of scores. But note: my method of scoring might appear to be different from the scores the game assigns. For instance, in Blokus you score the number of points remaining on your pieces, but you could just as easily score the number of pieces that you have placed on the board. The difference between these two methods is either subtly psychological or a matter of convenience.

ScoresNotes
0-0-...-1This is a game that has no gradual scoring element. It is not a race game, because the progress of the loser in a race can be recorded. One person simply wins.

"First to guess" is an example of this type of game. The game ends when one person has completed a task that is not progressive in nature. The second player receives no partial progress score because there is no way to tell when he or she would have finished the task.

This is unlike a race game, where you could say that the second player ended with "the legs and arms done, and only the face remaining." You can add granularity to this game by allowing each player to continue until they have completed the task.

For instance, the first to guess writes down the answer, verifies its correctness, and then records the time it took him or her to complete the task. Each player continues to do so until everyone is finished or a time limit has been reached. The resulting scores will then not be 1-0 but the times taken to complete the task.

Other types of games may appear to score this way, but don't. For instance, a game of combat where each player controls a single unit may appear to be a 1-0 game. But in all likelihood, the losing player can estimate how far he or she was in completing the task before his or her opponent won, e.g. number of rounds remaining or damage inflicted.

Does a score of 0-+-1 make for a good game? It makes for a quick moving party game, where the movement of each round is more important than the quality of the results.
0-2
0-1-2
0-2-2-2
2-2-3
0-0-1-4
Here we have an extremely low scoring game. Every success is a huge advantage. The game won't end in a tie if the game is designed to end immediately when one person acquires the game ending score or condition.

Generally speaking, what distinguishes this scoring type from the following one is that there is only one way to score. There may be multiple paths to gain each objective, but only one objective yields points, and each objective is one point each. Also generally speaking, final scores are in the low single digits.

An example of this type of game is Yinsh. The first player to remove three rings wins. Another example is Cosmic Encounter.

Does this scoring type make for a good game? Yes, assuming the following: 1) That ties are not decided by some arbitrary mechanism. If they are, then the scoring rules should be rewritten properly. For instance, that X's are 100 points each, and Y's (such as cash remaining) are 1 point each. That doesn't really change the rules, but it gives the scoring a semblance of sense. 2) That the length of the game is appropriate to the scoring. 3) That the skill necessary to achieve the results is also appropriate.

In a game of this type, the scores are so close as to eliminate the idea that a score of 5-3 is somehow better than 5-4. Players will look only at the end results, and gladly trade progress with their opponents once they have a lead in the scoring. Unlike other games where a larger differential in scoring can be achieved, differentials in scores at the end of these games have no significance as to the strength of the players.

As score of 0-2-2-2 indicates a poorly designed game, unless the game is a negotiation game, in which case the tied results indicate an alliance.

A score such as 2-2-3 indicates either a) that the winner gave up what was needed in order to win, b) that the player's strengths are very close, or c) that the results were as much a matter of luck as of skill, while a score such as 0-0-1-4 was probably an unsatisfying game, either due to unbalanced skill levels or disproportionate luck.
3-4-7
4-7-10
8-9-10-13
4-11
This is a low scoring game, similar to the previous scoring type, but usually indicating that points can be scored in multiple ways.

4-7-10 is instantly recognizable in my group as the end score for a game of Settlers of Catan, 8-9-10-13 for Cities and Knights. In the latter case, I would be willing to guess that the Longest Road was stolen and a metropolis built in a single turn. Not only that, but players 3 and 4 made the mistake of forgetting that the ultimate winner may have been behind a few points but was well within reach of winning. They probably overlooked the point from the Merchant, too.

The example of 4-11 is a typical short table tennis game, and probably serves as well for other dexterity games such as Crokinole and so on.

Many war games probably fit in here, as well. A few units versus a few units, or a few units versus number of rounds left before a mission would have been accomplished.

This scoring range is not good for games where there are fewer strategies for scoring. As an example, some of Knizia's shorter games have end scores like these, and I am generally left unsatisfied as a result. The scores don't really reflect either the true nature of the skill of the players or the luck involved in the game. In fact, they seem rather arbitrary, most likely the result of a single leap forward that could have happened to anyone. Colossal Arena is an example, as is Quo Vadis.

Maharaja is also a game in this scoring range that would have been better off in the previous range of 0-3. There is only one way to score points in Maharaja, and the last three points seem to have been added just to extend the scoring range.

If the scores are very close, such as 10-9-9-8, the game was probably exciting and fun. If too far apart, such as 13-7-2, the game was probably lackluster. If 10-3-3-2, as was the scores in a game of Settlers that I once won, the game wasn't fun.
10-12-76
35-67-98
5-9-105
98-99-100
Here are a great many Eurogames, such as Puerto Rico, Princes of Florence, El Grande, and Taj Mahal, to name only a few.

With little exception, scores in these games progress through several stages at the end of various rounds, actions, or events. This is true even if you don't keep score on a scoring track, and also true if some of the scoring is kept hidden, so long as the bulk of it isn't.

There are different progression types for this type of scoring. In some games, the initial scoring is low, and near the end of the game there are sudden leaps. This represents a build up of infrastructure cashed in before or at the game's end. Players that are losing may only know that they are losing at the very end of the game. The lack of point differential early on hides this fact. Players who didn't know they were losing may feel like the scoring came out of nowhere; in bad games, it may well have.

Another progression is a simple and steady progression. Unless the game is auto-balancing, such as Traumfabrik or Power Grid, where movement forward represents an expenditure of resources that other haven't expended, the game will develop clear winners and losers early on and probably be boring for the losing players. This is especially true if the same items used to score are those that produce the resources necessary to score, such as in St Petersburg.

Simple progression may work fine for games where the scores in each part of the game are NOT based on infrastructure build-up, such as trivia games or sports. That opens up the possibility for a late game comeback.

A score like 35-46-63 may be a good game in the former type of progression, but bad in the latter type. A score of 10-12-72 is probably bad either way.

Scores of 98-99-100 also may be good or bad. My feeling is that it is probably bad, although I'm aware that most people probably feel differently. When the differences in the end scores are so low in proportion to the scores themselves, it may represent a fiercely close game that was exciting all around. But it may also represent a win by a simple margin of turn order or lucky draw, or a game where no particular strategy or tactics made any real difference to the score, other than basic common sense.
155-267-415-425
403-403-403-404
The next order of magnitude's scoring illustrates my problems with close ending scores. 403-403-404? Is that really a victory? What about 40,000-40,000-40,001? Of course it depends on the length of the games and the granularity of how points were collected. But unless we're talking about a ten year game, an ending like this is meaningless.

Balanced yet diverse ending scores such as 155-267-415 are probably the result of someone failing to accomplish a major challenge in the game. The result of this failure knocked the player out of serious contention. An example of this is Modern Art, where an entire series of paintings failed to score any points.

In a sense, therefore, this is a little like the second scoring type, where only players who have achieved the basic objectives will reach the final stretch, and the scoring then comes down to their secondary achievements.

Regardless of the final scores in a game, the end result of just about any game is the same: 0-...-1, i.e. one winner and several losers. Aside from negotiation games and team games, where a few of those 0's may also be 1's, there are very few exceptions to this. Can you name any?

The answer to that question, and alternative solutions to simply winning or losing, will be the subject of next week's post.

Yehuda

3 comments:

Andy said...

I play a game a lot that ends with scores like:

30:20:-15:-35
150:100:-50:-200

OzGamer said...

How would you record the scores for a game like Niagara, where the winner is the first to reach one of a number of possible winning conditions?

Fraser said...

How would you record the scores for a game like Niagara, where the winner is the first to reach one of a number of possible winning conditions?

I would assume, using the same or similar nomenclature, that it would be 1-0-0-0, or first past the post. One person wins and the rest do not, although some of them may have been closer to winning themselves than others.