forums.marpirc.net

About the scoring system, I made two simulations.
<p>

The first score is the sum of the rounded percentages
of highest score of each game. ( Deca2000 scoring )

<p>

The second score is just weighted by the ranking, just as our current leaderboard : each score on a game is :
100*score/(best score) X 0,85^(rank - 1) , then rounded.
So, here are the rankings for the 20 players who qualify, playing all the 10 games :

<p>

Percentage scoring Percentage and ranking scoring
1 sutton 668 1 sutton 522
2 lamat 625 2 lamat 455
3 tiihonen 584 3 tiihonen 423
4 bjw 535 4 bjw 347
5 blost 461 5 boxster 270
6 boxster 452 6 cicca 245
7 bbh 429 7 blost 237
8 cicca 405 8 bbh 196
9 sandman 367 9 iur 181
10 axe 342 10 sandman 168
11 iur 306 11 axe 114
12 donut 253 12 lagavulin 106
13 lagavulin 251 13 donut 84
14 guru 188 14 guru 51
15 rodewald 176 15 rodewald 38
16 millardet 148 16 millardet 18
17 julie 132 17 julie 14
18 gameboy 122 18 gameboy 14
19 queenmary 64 19 queenmary 3
20 jump 28 20 jump 1

<p>

So, the question is :
Why not adopt one of these methods
( According to me, the first one is better, less frustrating for weak or occasional players )instead of the current one, quite curious and giving no benefit to players who leave others far behind them : when the best score is 100000points, it's not fair to have the same points if you make 101000 or 700000 !

<p>

Lagavulin.

--
darre@club-internet.fr

Percentage scoring is by far the most unfair and unbalanced scoring
method for any tournament. You can check Deca2000 and see how it
really doesn't work that well.

<p>

I can see two main reasons for this:

<p>

1. When a game has a maximum possible score (example Track and Field)
then all the top scores are bunched together and a vast majority of
people score 95% or over - what's the point of this? You gain no
possible advantage if you're in 2nd place compared to someone who may
be in 9th place if the scores are so close.

<p>

2. Only the 1st place score really has any effect on the tournament
once several scores are in place. You gain no real advantage over a
fellow competitor in 5th place in you score 4th place, especially of
the 1st place score is way out in front - see Deca2000's Burgertime
for a classic example of this. In order to make any difference to the
overall scores you have to be 1st.

<p>

This shows that percentages don't work - and that in order to beat a
person you just have to play overall better in total scoring, and if
you lose just one game by a huge margin it may not be possible to
catch up. This means you could be slightly better on 9 games and
still lose - does this mean you are not the better player? No.

<p>

Whereas MARPs tournament scoring method rewards overall placings, and
you beat a person in 10 games if you place better in 6 games, which
is fair.

<p>

There is no method that can please everyone and which would prove to
work best in all circumstances, but I'm sure that the percentage
system is by far the worst.

<p>

Crash.

--
crash@tcp.co.uk

We've been discussed so long on which scoring system to adopt for the
tournaments, and we decided for this one....it's not the best, and
it's not the worse.<br>As the stats maintainer, I could include
theese two different LBs, and consider them in the "overall" LB.


Cicca

--
cicca@writeme.com

I'm in agreement with both Lagavulin and Cicca on this one. You have
to understand that each scoring system has its drawbacks. There will
probably never be a system that ALL players will be happy with. As
for MARP's system, Cicca said it best, "It's not the best, and
it's not the worse.". Same can be said for the TG (D2K) system. I
personally like it the way it is with both tourneys using different
methods.

<p>

Dean, I fully understand your objections in regards to Track & Field
in the Deca2000 tourney. Most of us in this tourney saw this
situation coming a mile away, so we chalked this game up to a "non-
factor" status. That is fine with me as it is a controversial game in
the first place and shouldn't have been in the tourney. It's not
Ron's fault, the votes came in and the democratic process left us
with this game...like I said, no big deal because it's not a real
factor anyway.

<p>

As for example 2, I can only partially agree with you. In a
tournament like D2K, it is imperative that you can "hang with the big
dogs" on at least a couple of the games. And that is basically the
point of D2K...if you want to be a top player, you have to be able to
keep up with the other players on several games. Yes, there are games
such as Burgertime and Tapper that have experienced somewhat of an
early blow-out making it futile for some players to pursue. But, that
doesn't mean you still can't gain something by improving your own
score. You just have to concentrate on games where you can gain
significant points in return for an improved score. There are still
those types of games at D2K. This is the example that has me agreeing
with Lagavulin. As he stated, "It's not fair to have the same points
if you make 101000 or 700000!". I don't think it is unreasonable to
reward the dominant player of any given game in a tournament. This
reward doesn't guarantee that the dominant player will prevail...it
only gives him a foot up on the competition and justifiably so.

<p>

By no means am I saying you're wrong in your thinking, Dean, I'm just
saying that each system has its own merits. And for that reason, it's
great that you can choose which you prefer.

<p>

JoustGod

--
pinballwiz1@msn.com

I guess it's just a shame that after the Deca2000 votes were in,
they didn't have the same debate system in use at MARP - it's
basically become a tournament with 9 games imo.

<p>

oh yeah, I'm not playing in Deca2000 as all the top scores are just
too good for me, and I just don't feel like playing Frogger THAT much
to improve my score Although I guess I could get 85 or so % in
T&F.....heh.

<p>

I'm not overly critical of either method - they both have their
advantages and disadvantages, but I just feel percentages are more
uneven when compared across a range of games between the same group
of players. I do think that 1 point awarded is better than 0.03% as
well

<p>

Crash.

--
crash@tcp.co.uk