Re: Continuous scoring

[Tom Holub <tom@CSUA.Berkeley.EDU>]

On Mon, May 08, 2000 at 01:48:37PM -0400, Dave Ahn wrote:
> On Sat, May 06, 2000 at 06:41:19PM -0700, Tom Holub wrote:
> > In article <8evoac$hdg@northshore.shore.net>,
> > Ken. Hanson <mirror@shore.net> wrote:
> > )
> > )
> > )   The way that this would be done is as follows:  Each team would
> > )begin the game with zero points. The server (or the bot) would, at 
> > )the end of every minute of the game, check the current planet count. 
> > )For each planet that a team holds at that point in time one point 
> > )would be added to that team's score.  Therefore, for a 90 minute 
> > )game the total points possible would be 1800 (i.e. 20 planets times 
> > )90 minutes); if neither team took a planet the entire game, the end 
> > )score would be 900-900. 
> > 
> > Dave Ahn mentioned a similar, more complicated proposal on the
> > vanilla-l and inlcouncil mailing lists recently.  I think both proposals
> > are well-intentioned but based on flawed assumptions.
> 
> Well, my assumptions are that 1) we keep the 90+30 minute game; and 2)
> do nothing to change the game mechanics; 3) want to make the first
> half of the game more meaningful for end-game; and 4) still use end-game
> planet results as part of the winning condition.

You also are assuming that we don't care about the effect on the 
predictability of games.  Your system makes game results more predictable;
the existing predictability is already too high.

> > If you look at successful sports and games, they all have obvious
> > and fairly simple scoring systems, and they all have the ability for
> > people who are behind to come back and win.
> 
> Yes, but up to a point.  A team that lags far behind for most of the
> game will be behind by so many points that even a perfectly executed
> strategy wouldn't allow them to win in the end.
> 
> In netrek, it's entirely possible for one team to be completely dominated
> 5-15 for 80 minutes of the game to end up winning 11-8-1 simply because
> they successfully delivered 25 stashed armies for the win.

You have a fallacy in your thinking here.  A team that is behind 15-5 in
planets but has 25 armies is not behind by 5 planets in the same way that a
team with no armies is.  A football team that's down by two touchdowns
with 2:00 left, but has three timeouts and is first and goal on the 
opponent's 1-yard-line, is in a much different situation than a team
that's down by two touchdowns with 2:00 left, with no timeouts and the
ball on their own 1.

Trading control of territory for consolidation of resources is not just
a netrek strategy, it's a war strategy, and often an effective one.

> > I think Ken's proposal
> > is reasonably simple but doesn't give people a chance to come back;
> > if you get ahead 13-7 and hold it for 5 minutes, the game's basically
> > over.  I think Ahn's proposal is too complicated to work in practice;
> > it would not be possible to make informed decisions near the end of 
> > close games because of the complexity of the system.
> 
> I agree that making informed decisions with my proposal is harder.
> Assuming that there are no other problems with the idea, we can use the
> point system to change the winning condition at end-game so that the
> system is translated to the familiar planet-count method.
> 
> Let's divide the game into opening, mid-game and end-game.  At the 30
> minute mark, the team point count can be tallied and the winner can get
> a handicap towards the final score.  Same deal at the 60 minute mark.
> This means that if one team completely dominates the other team for 2/3
> of the game, then the lagging team must come back to win by, say, 12-7-1.
> The size of the handicap would require fine tuning, of course.

Again, with this kind of system the game is over 30 minutes before the
end.  Emprical evidence shows that 11-8-1 or closer is the expected
score between two reasonably equal teams in a 90-minute game.  If behind
behind 11-8-1 is now a winning condition, you've now given the team that's
AHEAD a great incentive to play passively and store armies; how are you
going to get a 12-7-1 advantage against a team that just controls their
front and never tries to drop armies in your space?
 -Tom