Thoughts on points

From Armagetron
Revision as of 13:49, 20 May 2007 by 2020 (talk | contribs)

The discussion of the points system for Fortress is apparently pretty open, though the underlying mechanisms are surprisingly complex. Forum discussion on this point during may 2007 is a little too shallow, since there are too many factors. The objective of this page is

  • to determine what the factors are
  • prioritise these factors
  • suggest a points scoring system which works

At this point, I suspect there might be several alternative systems that could co-exist, but only a few (in the same way there are so many alternatives to defining servers with rubber, acceleration, turning etc etc and yet there are only a few 'ideal' forms, one of which is fortress, another capture the flag.

factors (rambling allowed)

Some starting points.

There are plenty of tactics and strategies in football (soccer), and yet there is only one way to get points... scoring a goal.

Redundancy. Eg holing is an expensive strategy for team in its own right since a team loses a player and most players don't want to throw their game away. Thus to penalise them by -1 one's team and +1 opponents is redundant... Come to think of it, the current system with giving 2 points to the opponents means a torpedo is only winning 8 points for the round. This leads to the argument proposed by luci amoungst others: just winning points for the team and no points for individual behaviour.

On the other hand, an individual player's contribution to a game keeps it 'human'; in the sense that we can put names to famous football players, who are normally the ones who score goals.

Another thing to keep in mind is that a kill is often not an active thing. A defenders high score at the end of a round is an accumulation of mistakes by attackers rather than an accumulation of consumate skill by a defender.

Regarding active and passive - The active thing is that a team defeats another team by killing all opponents or conquering their zone through an accumulation of a hundred different combinations of individual skill and combination and mazing and complexity... The passive thing is that individuals die through their lack of skill which may have been accelerated due to the consumate skill of opponents or bodged teammate communication


And a beautiful part of the game is that it is active everywhere, with focus at either end simultaneously, unlike most ball games. Hence the complexity and need for an alternative scoring system...

So, a balance between individual scores and team score seems to make sense, where we can celebrate the individual within a winning (or losing) team.


A useful nail to hang this on is a 'golden round'. The winning team conquers the opponents zone, dumps all opponent players and all teammates are still alive. A golden round to which perhaps we should award 100 points and immediate match win. Every other variation of winning and losing is relative to this. For example, what about the same but with a few deaths in one's own team...? And if we itemise each element, we are in danger of being blinkered. For example, as it stands, conquering the enemy zone should be awarded with 10 points. Dumping all opponents leads inevitably to the taking of the enemy zone, if play was allowed to continue, and should be awarded with the same 10 points, as is the current settings. But the style is different and perhaps psyko is right to emphasise this kind of play; indeed by killing all opponents, with current settings, this means potentially 16 points already in an 8v8 game.

One more general observation. The game is a little like yin-yang, at the team-level where there are two zones, and at the player-level where a person attacking can suddenly be the person defending and vice versa. This is partly why the game is so vital.

Oh, and I believe the reason why this has been forwarded as a moot point, is because a certain stalemate situation seems to be recurring in the higher-quality games. For example, in Tronic Ladle 10, there were numerous occassions when a round concluded with a 1v1 situation. When there is poor skill by both players, it ends usually quite quickly. When there is a mismatch of skill attack>defender, again this end quickly. However, when defender>attacker, the situation can go on for a while unless the defender gets bored and risks killing the attacker; not a nice sumo to watch. And even when both players are skillful, which can sometimes be very interesting to watch, the sumo can last for minutes at a time. Discussion began with ways to resolve this problem in minimal ways, for example, changing the scoring system.

The factors seem to be these:

  • taking the enemy base/zone (1)
  • successfully defending base/home/zone (2)
  • dumping all opponents (3)
  • actively killing opponents by trapping them (4)
  • enemies dying passively against your walls (5)
  • suiciding (6)
  • getting dumped by a teammate (7)
  • all positive points or positive/negative points combination (8)
  • numbers of players in teams (9)
  • number of points for dumping, etc, and winning zones, etc (10)

Looking at (4) (5) (6) and (7), it is almost impossible to determine whether a person's death is because of your skill or because they made a mistake miles back on your wall and you weren't even aware of it. In relation to the game engine, they just crashed into your wall. To attribute points at this event is never going to be an indicator of your skill. It is more like an indicator of their lack of it. But of course, we are all playing in such ways to increase the potential for opponents to make these mistakes, mazing etc. Even (6) suicides can be completely self-induced, which is rare, and is mostly a consequence of a response to another player's movement, even though no contact has been made.

What matters is which events should lead to an increase in teamscore. Once an opponent is dumped, the benefit to the team is immense because there is one less opponent. Does this need to be enumerated? And if so, do we merely consider the team-score incrementing or do we attribute the points to individuals?

(1) (2) and (3) are team-level, which is a collective result of skills demonstrated by individual players, from both sides, in ways inconceivable to tabulate and enumerate. We will start with the standard of 10 points for taking the enemy zone.

Factor 9 is also significant. With 8v8, there are potentially 16 points available for individual gpf (general protection faults) as well as a 10 point bonus for winning the zone, using current CVS Standard Fortress Scoring System. That is, the bonus is not as significant as it is with a 3v3 game, for example.

positive and negative points (factor 8)

A recent development has been the suggestion that a kill gets +1 and getting killed gets -1, suiciding is 0 (with the possibility of getting a point for the team at the end of the round for surviving).

Perhaps, if we are going to use positive and negative, we should have one score? That is, the blue team are negative and the gold team is positive. Dumping an opponent wins your team one point, conquering the enemy zone wins your team 10 points. At any point, a player can see the balance of the game by how far it deviates from zero. It means that individually, players have an accumulation of points, and there is no sense of losing points for the team.

This is a completely different way of scoring. I have never seen anything like this.

prioritising these factors (a bit more concise)

suggested scoring systems (simple solutions that everyone can understand)

cvs standard fortress scoring system

10 points for a team for a round win by conquering enemy zone or dumping all opponents, plus 2 points per opponent that crashes against player's walls (has been in contact recently with your wall before suiciding or crashing against teammates walls).

This system seems to work pretty well. It is very simple. Even new players get it easily enough.

The only