Analysis: Girls Versus Boys

Topic: How Game Theory works in the game 'Boys vs. Girls' - Triangle's version Boys VS Girls 4

Background
This game is a virtual relative of Tug of War. The teams peer groups is the traditional Battle of sex, girls versus boys, Men versus Women. The punishments represent the optional mud pool in the game. A simple post stands for a single pull on the tug. The score is the current balance.

Abstract level
The score starts with 50. One team, let's call them A, wins when it drops to 0 or lower, the other team, let's call them B, wins when it reaches 100 or higher. Members of one team can subtract 5 from the score, members of the other can add 2 to the score. The players do this via a simple post in the thread, but no player may post right after her- or himself.

Simulation
In the special case of just one member in B, A can only win because of forbidden double posts. In any other case, the winning team is determined by the posts of its members. A wins when it has 17 more posts than B. B has to perform 25 plus 2.5 times the posts of A to win.

We can simulate the game with a simple model based on the assumption that every player had a frequency to check and post, and furthermore on a second assumption that no one has more than double the frequency of every other player. In this model, we can sum up the number of posts per team and time, and therefore get the duration of the game as well as the winner. With more complexity, we could even assume that every player only plays for a certain time or even time windows, and hence get a more accurate prediction. We could observe the players in all these games. Then we could take any experienced player's stats to predict their most likely frequency and time windows, take the stats of all players in their first game for new players, and should, therefore, get a good prediction of the outcome of the game.

Reality check
The reality differs from these models in that the exact post time of all players is not predictable. Their frequency and time windows are just broad tendencies. Social psychology, peer pressure, the effects of visible competition as well as out of game events make the reality differ from any predictions.

Last 5 winners and losers
Since only the fast 5 posters of the winning team decide about the punishment of the last 5 posters of the losing team, there is an incentive to stop posting when you are not one of the last 5 posters of your team and it is close to losing. Except if you want to lose and get the punishment, you want to make a post just then and make sure you are one of the last 5.

Strategies
For the special case of one active player of A versus only one active player of B, A should post anytime they can, while B should not post. In any other case, any team should post as often as possible. Simple as that. How you get yourself or your teammates to do that, is a matter of psychology rather than Game theory, because you have no meaningful choice of game actions.

Game Design
Besides prediction of a game's outcome and analysis of dominant and poor strategies and choices, Game theory can also help to make alterations on the Game rules to change player behavior, strategies, game flow or outcome according to different goals. Much is possible depending on what you want from the game. Let's take some game details as examples.

Peer group: Girls versus Boys
While this makes up a lot of the atmosphere, it's one peer group against the other. So you could exchange those with Doms and Tops versus slaves and subs. In this game the peer group makes up the whole setting and atmosphere of the group dynamics, so we should take great care to match further details to it. For instance, unequal up to unfair point distribution might be very fit for Doms and Tops versus slaves and subs, whereas in Girls Versus Boys it just fits the practical goal of fairness, since girls playing tend to be less in numbers than boys.

Add/Subtract Score: The exact numbers of how much to add or subtract relatively to the winning scores make the whole dynamic of the game. If we wanted to accelerate the game towards the end or when getting close to a winning score to create more tension and thrill, we could increase the numbers to increase and subtract depending on the score range. For example, double increase when the score < 10 or score > 75, and double decrease when the score <25 or score > 90.

Posting rules:
The posting rules determine the choice if to post or not right now. Currently, posting whenever you can except for one special case is not interesting at all in the choice itself. If we wanted to make that choice more interesting, for instance, we could double the score change a post if the previous post has a timestamp at least 2 minutes lower or if the timestamp of the previous post from the opponent team has a timestamp at least 5 minutes lower. That would create a possible bet on when others are going to post, as well as the organization of your team in combination with that bet.

Different Adds and Decreases: This obvious handicap for one team regards the estimated relation of players of the peer groups. While this could add to the topic of peer groups like Doms and Tops Versus slaves and subs, it can quite differ from reality. If we'd base the added and subtracted amount on the number of announced opponent players, then the game would reflect that relation as well as adapt to a change of it during the game. Though, we also had to increase the winning scores if we didn't want to accelerate the game with increasing number of players (and practically the current score if we didn't want negative winning scores, or change the counting in total).

Punishment for all:
We ought to punish all members of the losing team. Via this commitment all player remain an interest to keep playing and in winning if they don't want a punishment. There is no more tricking out by holding back at the end of the game. The unfortunate couter-argument for this is, that the winners probably had to make all losing players aware of this, since they might not be up to date with all the posts and their conclusion.

Creative changes:
The current game is still just about the ability to post. The strength and potential technique of the Tug of War translate into the simple ability to spam posts into the thread. We could change that. For example, in Boys Versus Girls every post has to contain an activity or achievement the poster did, which a majority traditionally attributes to the opposite sex. Repetitions of the same player are forbidden. A girl could post it had courted a partner with flowers, gifts and inviting him, her or it to dinner and into a theater, having peed while standing, have repaired a car, mechanical or electronic device. A boy could post he had peed sitting, has made himself pretty for half an hour, acquired two pairs of shoes at once.
Or both teams would have to post a favor or something good for the opposite peer group, that they have done.

There are plenty of conditions we could demand on a post, even PM dares, and tasks, revealing a dirty secret, whatever your can think off. Or every post is allowed to state a riddle or question, and the next post only changes the score if it solves the riddle or answers the question. Another effort example: to make your post change the score, you have to edge more often than the previous post from your team, or you have to spank yourself 50 times to reset the amount of edges for your team's next post.

We could also make the game asymmetrical, by giving one group a certain amount of time to reach their score via posts under conditions. The opposite team could post rejections of such single posts with reasoning it according to rules. That could be an interpretation of doing a favor or activity counts or checking very restrictive rules like having to use at least 24 vowels.