| 1 |
So in short:
|
1 |
So in short:
|
| 2 |
* "I'm worried that ZK loses competitiveness due to such a league system."
|
2 |
* "I'm worried that ZK loses competitiveness due to such a league system."
|
| 3 |
* "A game has competitiveness (which means it is competitive) if its goal is to maximize a competitive rating."
|
3 |
* "A game has competitiveness (which means it is competitive) if its goal is to maximize a competitive rating."
|
| 4 |
* "By competitive rating, I mean one that rates how good you play, basically exactly what basic WHR does."
|
4 |
* "By competitive rating, I mean one that rates how good you play, basically exactly what basic WHR does."
|
| 5 |
=> "I'm worried that ZK the focus on the goal of maximising WHR is reduced with such a league system."
|
5 |
=> "I'm worried that ZK the focus on the goal of maximising WHR is reduced with such a league system."
|
| 6 |
(Some rephrasing was required to translate the relative quantity grammar of "loses competitiveness" into a discrete goal")
|
6 |
(Some rephrasing was required to translate the relative quantity grammar of "loses competitiveness" into a discrete goal")
|
| 7 |
\n
|
7 |
\n
|
| 8 |
Yes, the focus on WHR is reduced by a league. That is part of the point. The only goal of the current ladder is maximising WHR, and this is a bad goal to have as the only goal.
|
8 |
Yes, the focus on WHR is reduced by a league. That is part of the point. The only goal of the current ladder is maximising WHR, and this is a bad goal to have as the only goal.
|
| 9 |
\n
|
9 |
\n
|
| 10 |
[q]Ultimately, all those goals can be tracked back to the more fundamental goal of having fun: If other factors than skill, like random or grinding, have influence on the rating, it can be frustrating. It can also be frustrating if games are poorly balanced.
|
10 |
[q]Ultimately, all those goals can be tracked back to the more fundamental goal of having fun: If other factors than skill, like random or grinding, have influence on the rating, it can be frustrating. It can also be frustrating if games are poorly balanced.
|
| 11 |
Apart from that, I think that a rating system is not the best source for fun or activity. Activity should rather come from fun in the game itself. [/q]
|
11 |
Apart from that, I think that a rating system is not the best source for fun or activity. Activity should rather come from fun in the game itself. [/q]
|
| 12 |
I agree with the last part, but look at it a different way. A good ladder should not try to be the source of fun or make up for fun that otherwise isn't there. Rather, a ladder should motivate and coordinate people to do what they already find fun, and to do the particular thing that they find the most fun. Psychology isn't so simple that people will just do what they find fun and somehow overcome coordination problems in order to do it.
|
12 |
I agree with the last part, but look at it a different way. A good ladder should not try to be the source of fun or make up for fun that otherwise isn't there. Rather, a ladder should motivate and coordinate people to do what they already find fun, and to do the particular thing that they find the most fun. Psychology isn't so simple that people will just do what they find fun and somehow overcome coordination problems in order to do it.
|
| 13 |
\n
|
13 |
\n
|
| 14 |
Queuing in the matchmaker is inherently risky. You might end up in an uneven game and not have that much fun. You might wait a long time for an opponent. You might lose. These factors affect players to different extents, and we can try to mitigate them, but the best we can do is make tradeoffs. We can't completely eliminate these problems. People have fun in the matchmaker, on average, but are discouraged from queuing due to these risks. The risks of queuing are exacerbated by lower ladder activity. Here is the coordination problem, everyone would have more fun if they all did the fun action, but taking the action alone is less enticing and repeatably being hit by the heightened risk could lead to burnout.
|
14 |
Queuing in the matchmaker is inherently risky. You might end up in an uneven game and not have that much fun. You might wait a long time for an opponent. You might lose. These factors affect players to different extents, and we can try to mitigate them, but the best we can do is make tradeoffs. We can't completely eliminate these problems. People have fun in the matchmaker, on average, but are discouraged from queuing due to these risks. The risks of queuing are exacerbated by lower ladder activity. Here is the coordination problem, everyone would have more fun if they all did the fun action, but taking the action alone is less enticing and repeatably being hit by the heightened risk could lead to burnout.
|
| 15 |
\n
|
15 |
\n
|
| 16 |
The problem with 'gain WHR' as the only goal for the ladder is that it does almost nothing to encourage activity. If you make a system that perfectly ranks player skill (ignoring the debate about whether WHR achieves this) then players are only rewarded by improving at the game. Essentially, we're handing players the monolithic goal 'Get Better At The Game' without giving them any guidance. If someone decides to get better by playing a lot of games, then that's great. Unfortunately, players could reasonably assume that the ladder is telling them to go and play the AI until they are better, or to just play a few times a month and hope that those above them fall off. The top 10 positions are prominently displayed on the site so seems to be a reward, but most of the players with this reward don't actually need to play more than few times a month to maintain it. People are risk-adverse so will generally be drawn to ways to satisfy the ladder without risk, that is, without playing the game, and thereby increase the risk of queuing for others. Avoiding risk is natural, even when taking it will, on average, lead the the most skill improvement and fun for themselves and others.
|
16 |
The problem with 'gain WHR' as the only goal for the ladder is that it does almost nothing to encourage activity. If you make a system that perfectly ranks player skill (ignoring the debate about whether WHR achieves this) then players are only rewarded by improving at the game. Essentially, we're handing players the monolithic goal 'Get Better At The Game' without giving them any guidance. If someone decides to get better by playing a lot of games, then that's great. Unfortunately, players could reasonably assume that the ladder is telling them to go and play the AI until they are better, or to just play a few times a month and hope that those above them fall off. The top 10 positions are prominently displayed on the site so seems to be a reward, but most of the players with this reward don't actually need to play more than few times a month to maintain it. People are risk-adverse so will generally be drawn to ways to satisfy the ladder without risk, that is, without playing the game, and thereby increase the risk of queuing for others. Avoiding risk is natural, even when taking it will, on average, lead the the most skill improvement and fun for themselves and others.
|
| 17 |
\n
|
17 |
\n
|
| 18 |
Augmenting the 'gain WHR' goal with 'play matchmaker games' ([i]augment[/i], not replace) encourages people to behaves in a way that gains them WHR and that makes the ladder more fun for everyone. Being told to get better is daunting, whereas playing games is easy. We're essentially breaking down the goal into smaller parts. Also, gaining points at the end of each game, no matter the quality or outcome, means that the perceived worthiness of having played a game that you're a bit sour about is bounded below by how much you value the points you received. It dampens losses and sets a minimum reward, mitigating some of the risk of queuing.
|
18 |
Augmenting the 'gain WHR' goal with 'play matchmaker games' ([i]augment[/i], not replace) encourages people to behaves in a way that gains them WHR and that makes the ladder more fun for everyone. Being told to get better is daunting, whereas playing games is easy. We're essentially breaking down the goal into smaller parts. Also, gaining points at the end of each game, no matter the quality or outcome, means that the perceived worthiness of having played a game that you're a bit sour about is bounded below by how much you value the points you received. It dampens losses and sets a minimum reward, mitigating some of the risk of queuing.
|
| 19 |
\n
|
19 |
\n
|
| 20 |
(
As
an
aside,
I'm
not
entirely
sure
that
every
player
should
receive
league
points.
At
first
I
thought
it
could
be
a
system
to
solve
inactivity
in
the
top
10,
and
to
enable
us
to
crown
a
winner
of
the
league
that
didn't
get
there
by
being
good
and
then
barely
playing
for
three
months.
@malric
made
some
decent
arguments
about
how
league
point
could
be
good
for
everyone.
If
we
also
display
the
current
ladder
rating
so
that
non-top
players
can
track
their
progress,
then
I
think
having
league
points
for
everyone
would
be
a
reasonable
thing
to
try.
)
|
20 |
(
As
an
aside,
I'm
not
entirely
sure
that
every
player
should
receive
league
points.
At
first
I
thought
it
could
be
a
system
to
solve
inactivity
in
the
top
10,
and
to
enable
us
to
crown
a
winner
of
the
league
that
didn't
get
there
simply
by
being
good
and
then
barely
playing
for
three
months.
Being
unseat
the
top
player
because
they
aren't
playing
is
frustrating,
and
if
there
is
no
benefit
to
activity
then
the
top
player
will
be
motivated
to
not
play.
@malric
made
some
decent
arguments
about
how
league
point
could
be
good
for
everyone.
If
we
also
display
the
current
ladder
rating
so
that
non-top
players
can
track
their
progress,
then
I
think
having
league
points
for
everyone
would
be
a
reasonable
thing
to
try.
)
|
| 21 |
\n
|
21 |
\n
|
| 22 |
The league points in a league aren't the only part that attempts to overcome the coordination problem and make queuing less risky. The league also coordinates players through time. With the current system players wander in or wander back to Zero-K, play the ladder for a bit, and then go do something else for a while. A league with a set start and end date creates events that tell people to come along and play at the same time. More players at once makes for a better matchmaker and the effect could persist past the event. A league is also less daunting to join and generates less complacency. Players who are new to the current ladder have to climb past all the inactive players. On the other hand, existing players are free to wander away for a while and come back to their old rating.
|
22 |
The league points in a league aren't the only part that attempts to overcome the coordination problem and make queuing less risky. The league also coordinates players through time. With the current system players wander in or wander back to Zero-K, play the ladder for a bit, and then go do something else for a while. A league with a set start and end date creates events that tell people to come along and play at the same time. More players at once makes for a better matchmaker and the effect could persist past the event. A league is also less daunting to join and generates less complacency. Players who are new to the current ladder have to climb past all the inactive players. On the other hand, existing players are free to wander away for a while and come back to their old rating.
|
League Ideas