Choosing your race

G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

Hi,

I have added a chart to the start of this page:

www-furfur-demon

co.uk/geop/v4races-htm (please excuse anti-spam mangling)

which tries to compare all the races in one place, to help people pick
one for a given galaxy eg

"this is a mineral-rich, native-poor galaxy so Aczanny are a good
choice, but they don't tend to win very often".

The chart also has columns for "recommended player ability level" and
"preferred map type".

Obviously I don't know everything, and much of this is subjective, so
there are gaps in it and you might disagree with some of my judgements.
I'd be interested to hear peoples' views on the chart, so I can get it
more accurate.

And if anyone has a brilliant idea for another column, I'd be glad to
add it. For example, starting population might affect different races to
different degrees.
--
Paul Honigmann
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

I would recommend to set the UEA to:

Beginner +
(Simple and straightforward ship design, easy to use fighters, strong
economy in the beginning thanks to 200% city income and RedisCenter)

Large Maps
("Leave us alone. [...] Hooray! We've got the Pax - now we'll come TO
YOU!")

Medium to rich Natives
(tax bonus!)

Minerals to "any"

Comment Suggestion: "Fast raiders. Good vs. hyp-races and cloakers."

Suggestions for new columns: Starting Pop! How many players are in the
game? This race is good vs... This race is weak vs...

Maybe you can use some of the above.

Fabian
 

nameless

Distinguished
Jun 18, 2002
213
0
18,680
Archived from groups: alt.games.vgaplanets4 (More info?)

I wonder from whom you got the infos about the Cents.
As it is the Centaurs (unless of a few very exotic settings) do prefer
large maps.
As concerning to the natives: Currently only a large amorph population
does matter the other native types are neglateable.
Preferred Mineral setting is certainly not low, unless it is still easy
to obtain large quantities of Neutronium (that is even more important
if the Cents do not start with their normal starting ships and the 14 k
kt of fuel on the Shamia).
Strength: weak to very strong.
And as a comment I suggest: High Lerchin Price -> Rich Centaurs

As for the Dracs: I would change the strength to very weak to reflect
that they are the weakest vgap4 race currently available.
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

KlingonKommand wrote:
> Hi,
>
> I have added a chart to the start of this page:
>
> www-furfur-demon
>
> co.uk/geop/v4races-htm (please excuse anti-spam mangling)
>
> which tries to compare all the races in one place, to help people
pick
> one for a given galaxy eg
>
> "this is a mineral-rich, native-poor galaxy so Aczanny are a good
> choice, but they don't tend to win very often".
>
> The chart also has columns for "recommended player ability level" and

> "preferred map type".
>
> Obviously I don't know everything, and much of this is subjective, so

> there are gaps in it and you might disagree with some of my
judgements.
> I'd be interested to hear peoples' views on the chart, so I can get
it
> more accurate.
>
> And if anyone has a brilliant idea for another column, I'd be glad to

> add it. For example, starting population might affect different races
to
> different degrees.
> --
> Paul Honigmann
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

KlingonKommand wrote:
> Hi,
>
> I have added a chart to the start of this page:
>
> www-furfur-demon
>
> co.uk/geop/v4races-htm (please excuse anti-spam mangling)
>
> which tries to compare all the races in one place, to help people
pick
> one for a given galaxy eg
>
> "this is a mineral-rich, native-poor galaxy so Aczanny are a good
> choice, but they don't tend to win very often".
>
> The chart also has columns for "recommended player ability level" and

> "preferred map type".
>
> Obviously I don't know everything, and much of this is subjective, so

> there are gaps in it and you might disagree with some of my
judgements.
> I'd be interested to hear peoples' views on the chart, so I can get
it
> more accurate.
>
> And if anyone has a brilliant idea for another column, I'd be glad to

> add it. For example, starting population might affect different races
to
> different degrees.
> --
> Paul Honigmann

Which races are best is an interesting proposition, and one into which
I did some quick research. Scytale's approach is, as he admits, a bit
flawed because it does not adequately account for the number of times
that a race is played versus how often there is a win.

I attempted to solve this problem by applying a contingency analysis
using the chi-square statistic. The advantage to this approach is it
allows a determination of whether there is a significant difference in
the win/lose ratios for races and takes into account the number of
games for which each race is represented. Also by examining the
contributions of the diffences of observed to expected wins, you can
see which races contributed more wins than might be expected simply by
chance, and which races contributed fewer than you would expect by
chance. I did a quick run approach using Drewheads and RCWorlds data
on winners. What I found was there is a significant difference in the
number of wins that can not be accounted for by chance. By examining
the differences in the contingency tables I found that the Birds and
the Crystals won far more than might be expected and the Scavangers won
slightly more than expected. The Dracs and the Evil Empire won less
than expected, but their lack of wins was about the same as the
Scavangers increased number of wins. All other races won about as
often as might be statistically expected given the total number of wins
and how often each race won.

The problem with my approach is that it is based on the number of wins.
Comparing the number of wins against the total number of games played
would be more valid. It would also be instructive to compare the
number of wins in a points game, which is usually based on points
awarded for population, and therefor favors certain races, versus the
number of wins in a military victory game, which favors other races.
For example it would be very difficult to win a victory point game
based on population with the Evil Empire or the Centaurs, however, in a
military game both of these races would be much more likely to win.

If Scytle, or anyone else, has data on the number of wins and which
races were in each game, I would be glad to run this same analysis and
share the results. This approach will tell us which races win more or
less often than expected, but not why they win.
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

The whole strength rating thing cannot work.

Cents are a good example.
Dracs too, because in a *very* setup they excel with the extra income from
every government center and are very hard to beat .... if played well.


Lordfire
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

That should be: very *poor* setup

Lordfire

"Lordfire" <a@b.com> wrote in message news:38mutjF5odft3U1@uni-berlin.de...
> The whole strength rating thing cannot work.
>
> Cents are a good example.
> Dracs too, because in a *very* setup they excel with the extra income from
> every government center and are very hard to beat .... if played well.
>
>
> Lordfire
>
 

Sparrow

Distinguished
Mar 16, 2002
239
0
18,680
Archived from groups: alt.games.vgaplanets4 (More info?)

Wining or loosing is mainly not based on the race, but on the player.
An example:
Aczanny are considered to be weak by many people. I'm currently hosting
a game where an excellent Aczanny player is beating up races which are
considered to be much stronger (e.g. Liz), but are played by less
experianced players. It may be that Aczanny are underestimated, but I'm
sure that there are not very many players who could repeat this!

It also depends on the setup. E.g. a Borg on a small map without
natives is not the same than one on a very big map with abundant
natives.

Furthermore you would have to exclude private wars (same races playing
against each other) and team games.

Last but not least alliances are difficult to evaluate, too as they are
stronger than the sum of each race individually.

To make a long story short: The help can / should not be to find out
which race is the strongest, but rather to find the race which suits
one best (for a specific setup).

Just my 0,02€
 

nameless

Distinguished
Jun 18, 2002
213
0
18,680
Archived from groups: alt.games.vgaplanets4 (More info?)

Lordfire wrote:
> Cents are a good example.
> Dracs too, because in a *very* setup they excel with the extra income
from
> every government center and are very hard to beat .... if played
well.
>
>
> Lordfire

Regardless of settings, with the Dracs you will only have a real chance
to win, with a race pack which does not yet exist, at least if the rest
of the players do know how to play.
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

> Which races are best is an interesting proposition, and one into
which
> I did some quick research. Scytale's approach is, as he admits, a
bit
> flawed because it does not adequately account for the number of
times
> that a race is played versus how often there is a win.

Correction: I _DO_ take into account how often races get played. In
fact, that is a key part of my approach. Since I already gather how
often each race gets played, I normalize the points given to each race
by the percentage of playtime they get. So a win for the Borg or the
Bots counts for significantly less than a win for the Cents or Aczanny,
since they get played a lot more often.

The flaw in my approach is that I don't take into account player skill.
Your approach doesn't take that into account either, and I'm not
convinced any method ever could.

Question: How did you determine "winners" in the games you looked at?
If you went by VPs along you'll be way off, sadly. Very few games are
actually won on VPs, at least at Drewhead's. I'm forced to scour
banter boards to try to figure out who's winning games.

Scytale
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

disagreements listed:

Borg : Intermediate+ skill level, any mineral setting [Biocide for
alchemy], natives [medium to rich], strenght (med-strong) Newbies
playing the Borg artifically drop their strength on Scythale's analysis
Priv: Expert skill level, strength = med. Non-popularity with newbies
inflats their strength on Scythale's analysis

Dracs: Expert skill level if your wanting to win with them.

Peeps: Expert skill level, strength = med. (Beginners build the wrong
ships)

UEA: Looks like a good Beginner's race to me.

Aczanny: Intermediate+

Solarians: Expert level. Different fuel rules, need to monitor it like
a hawk.
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

" [..] at least if the rest of the players do know how to play."

Don't have to teach you anything about this sentance, do I ? :)

Lordfire
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

In answer to your question I did something similar to you, although I
wasn't as thorough (I didn't have the time and was only interested in
getting a feel for which races I might want to play). Part of the
reason I never brought this up before is I didn' t have as much data as
I would like.

The part of your approach that I don't like is your weighting system,
particularly giving points based on the number of winners against the
number of races. I think your assumption here is that it is more
difficult to win a game with larger numbers of players, but I am not
certain that is true. Certainly it is something that should be tested,
rather than assumed.

The question that I was attempting to address is whether any races win
more (or less) often than can be predicted by chance. It has very
few assumptions. I simply count how many times a race was played, and
how often the race won. I did not include games in which only one race
played (like Rebel versus Rebel) . I did include team wins (which I
think now may be a mistake). The chi-sqaure statistic is based on
contigency tables that automatically take into account the number of
times a race is played, there is no need to normalize the data. There
were some cells with low N (numbers below 5 in any cells can affect the
calculation of the statistic). Several races, the Centaurs being one,
did have cells with low Ns. In the end the chi-sqaure statistic did
reject the null hypothesis that number of wins for each race was based
on chance (which takes into account how often a race was played). The
races I gave either won more or less frequently than could be accounted
for by chance.

The reasons why is a different question. It might be player skill. It
might be starting positions, the number of races faced or other factors
as well. What I can say (with 95% confidence) is that certain races
are winning more or less often that can be accounted for by chance
alone.

I also think a flaw in both of our approaches was to include victory
point game winners with winners of other types of games. There are
clear differences between the two types of games, and I strongly
suspect that different races are better suited for one or the other,
with a few perhaps suited for both (I am thinking the Rebels may be one
of the latter).

I do agree that player skill is a factor. I could probably figure out
a statistical method that would allow us to address that question, but
I don't think the required data is available anywhere to apply such an
approach.

Another problem is that Tim has been changing the rules. Thus some
races that were very powerful in the past (for example the CoM) are
less so now, while other races (the Bots) are more so because of the
rule changes. Because the games take so long to be resolved, the
results of these changes may take several years to become apparent.

Scytale - you do lots of research that is invaluable to other players,
including me, and you often share it. This is a great service to other
players in the game, so I hope you don't take my comments personally.
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

Take it personally? Gosh no, I value your input. I'd love to see your
data and analysis.

The question of whether or not to give more points to winners of big
games vs small ones is interesting. I definitely agree that it's not
as simple as "he beat more players", since it's much more likely those
other players beat each other up first. On the other hand, consider a
race that only seems to get played in games with 15+ people (at the
moment at least, the UA seems to fit that). If all the races and
players were equal, they'd only win a fraction of the games. On the
other hand, races that are popular choices in 1v1 or 2v2 games (like
the Borg or Privs) will, all things being equal, win half the time.

It would be nice to test my approach to point awarding vs yours, but
there are some major problems with that. The obvious is that it takes
so long to gather any meaningful data. There's also the problem of
determining who's right. If the two methods get different answers,
which one is a better measure of race strength? How can we tell?

I don't think including victory point winners in the tallies is a
mistake, since VPs is a valid win condition used by some hosts.
However, a better analysis would divide the data by the different win
types, so we could see which ones are strong in VP games and which ones
are best in last-man-standing, etc. I perhaps should have been doing
that from the beginning, but then, I never actually meant to post my
results when I started collecting data.

As to Tim changing rules (and race power levels), yes, it definitely
takes a long time to see results. This is probably made worse by my
constantly-increasing point system. For example, just because the COM
is in 5th place on my list doesn't mean they're 5th strongest overall.
They've been slowly dropping since their cutbacks a while back. The
other races need to log wins to overtake them, and since I only get a
couple win results a month, that takes time.

Like you, I definitely feel that some races win more than they should
by chance. Crystals are my prime example (their point value is crazy
high). They win more than they get played. I'm not convinced they are
overpowered though... A lot of their wins come from beginner VP games.
I think they may have more of the "opponents don't know how to fight
them effectively" edge.

The Solorians I feel may be too strong. They've been shooting up the
ranks constantly for a while now. I believe the race is well-built and
strong enough to win if they had an average growth rate. I think their
crazy high growth is just too good for them.

Scytale
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

Sparrow wrote:
> Aczanny are considered to be weak by many people. I'm currently
hosting
> a game where an excellent Aczanny player is beating up races which
are
> considered to be much stronger (e.g. Liz), but are played by less
> experianced players. It may be that Aczanny are underestimated, but
I'm
> sure that there are not very many players who could repeat this!

A comment from one of the inexperienced players (me - 3 1/2 games)
whose Lizards are indeed getting beat up badly by the excellent Aczanny
player (Lord Owl) in Sparrow's excellently hosted game:

I agree with everything Sparrow says, but want to point out that the
Azcanny, like many interesting races, have a glaring weakness -- their
key swarming hulls, the ones with the annoyingly high evasiveness --
are easily torn up by mines, and they have no countermining ability.
They can sweep mines, but they remain extremely vulnerable to cloaked
barbitics. In this game, Lord Owl has cleverly overcome this weakness
by making great use of an ally and ally hull. So, to take nothing away
from Lord Owl, who is obviously an excellent player, and who perhaps
would have had the same success without his ally (but perhaps a bit
slower), you can't necessarily view this as evidence that the Aczanny
race alone -- considered as a standalone race -- could have the same
success. It is, at least in some part, an example of the synergies
that come from good alliances.

Also, if alliance wins are being excluded from these "race-win" charts,
I think wins from games with hull or hull plan trading allowed should
also be excluded, as hull or hull plan trading allows many, if not
most, of the synergistic effects of an actual alliance, even if the win
is not an "alliance" win per se.

-- Karnak Prime

P.S. Verdammte transphased world crusher missles!&$(#^$&@%#@!@! ;)
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

I think if alliance wins were eliminated from the charts, there would
be very few games with winners. I can count exactly one game in Planets
4 I've been involved with that has had an indivudal winner that wasn't
in an alliance with another major power.

In the first one it was clear the Birds would evenually have won but
they were allied with the EE for much of the gmae.

In the second one a Borg player was allied with the Peeps [this is when
Peeps were overpowered] and they agreed to a joint victory.

In the third one, it was set up as an alliance game.

In the fourth one, the CoM winner had some help from multiple players.
(Sure he was dominant)

In the fifth one, Lordfire playing the EE managed to get into a winning
position without alliances, with only Non Agresive Treaties.

In the six one, there was no clear winner when the game ended at all.
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

Thanks guys! Yes, the original stuff was my own biased opinions picked
up from amorphous impressions over the years... I have thought about
what you've all said, and ended up copying it all down (heck, I've never
played half these races).

If anyone has any more comments, bring 'em on!

The page has been updated and no doubt will be re-updated by the end of
the weekend...
--
Paul Honigmann
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

KlingonKommand wrote:
> Thanks guys! Yes, the original stuff was my own biased opinions
picked
> up from amorphous impressions over the years... I have thought about
> what you've all said, and ended up copying it all down (heck, I've
never
> played half these races).
>
> If anyone has any more comments, bring 'em on!
>
> The page has been updated and no doubt will be re-updated by the end
of
> the weekend...
> --
> Paul Honigmann

Put a disclaimer on your page.
Something like:

Remember an excellent player using the weakest race in your game setup
can beat the very poor player using the strongest race in that setup.
With that in mind play the race you will have the most fun with.

A player that starts a game with a hard to play race and then quits
by turn 20 or so upsets the entire game and all strategy breaks
down as someone else profits from your quiting.
This fact alone plus that of player skill and allied wins will
skew any practical results that have come from this kind of sample.
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

Comments:

Rebels : Long range fighter wings!

CoM: Start with one ship of each type (included a Virgo on fire) in
lieu of a base.

Peeps: Complex ship list; many of the ships competely useless.

Aczanny: Very fast light ships.

Native settings:

UA : Any

RCS: Any

Solarians: Low

Minerals:

Peeps: Any

Solarian: Low. (Specially low fuel: They generate their own from stars)

Map Size:

Solarians: Any map size, but perfers high density of stars within it
for refueling.

Robots: In addition to a small map size, they also prefer high denisty
of stars for overlapping Gun Zeros and Bug Hive ranges.
 

nameless

Distinguished
Jun 18, 2002
213
0
18,680
Archived from groups: alt.games.vgaplanets4 (More info?)

Lordfire wrote:
> " [..] at least if the rest of the players do know how to play."
>
> Don't have to teach you anything about this sentance, do I ? :)
>
> Lordfire


Wall I guass you do not hava to (ops I must have used the same keyboard
as you did).

I guess I should be more exact. There are absolutly no normal host
settings, under which the Dracs would
not belong to the three weakest races.
Moreover in a one-on-one regardless of normal host settings, meaning
map size (we can even include the snowflake map), normal starting ships
(yes/no),native setting, mineral setting, map size, and regardless of
starting position (no prepicked settings), I am pretty sure that I
would for more than half of the races need no more than the shareware
version regardless of who is playing the Dracs to beat them (it might
take long, but eventually there is only one possible outcome) -
moreover I would exactly state which strategies and tactics I would use
to win before the game started (I might forget a few standard tactics
but no more). Sometimes the reason would just be outbreeding the Dracs,
while making sure that the Dracs would get no big amount of prisoners
from me. In some cases it would be the higher mobility. In some cases
the Drac owned Gov Bases would be assaulted. In some cases the money
making gimmicks
of the other race are just more profitable.
Heck in many cases even the Dracs starting in the middle of the epsilon
galaxy (and any other race on the fringes) and the Dracs would still
have no chance at all.
Oh and then I am pretty sure that you, so far, did not play the Dracs
yourself.

Of course many a player is not really playing vgap4 , they are merely
manipulating the client.
Of course one is able to win as an expert against intermediate skill
level, especially if you are a good Diplomat.
Or you might be lucky as the Dracs and somehow end up in the winning
alliance, but otherwise no chance.
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

More fantastic feedback there. Included it on the page. Thanks everyone!
--
Paul Honigmann