1234567890123456789012345678901234567890123456789012345678901234567890123456789
Gunbound Avatar Optimization and Strategy Guide
v0.2
Copyright Chez Maradine 2004
+=================+
| I. Introduction |
+=================+
I've spent a reasonable amount of time in GunBound watching the way people
play. Sometimes, there doesn't seem to be a whole lot of logic involved.
Other times, I see bizarre flame wars and friendly fire over silly things
such as "Kill Stealing". I hope to show that all of that is irrelevant.
This document is meant to address two related issues.
1. What is your goal?
2. How are you planning on getting there?
1. is addressed through a few simple logic problems. 2. is addressed with
some speculative math and empirical data collected by myself and my friends,
as well as anyone who would like to contribute.
But first, lets deal with the red tape!
+==============+
| II. Legalese |
+==============+
As mentioned above, below, and here, this work is Copyright Chez Maradine,
2004. GunBound is a registered Trademark of softnyx, Inc. All game content
belongs to them. All analysis in this FAQ belongs to me.
You are permitted to utilize the content of this work under the Creative
Commons Attribution License 1.0, available at:
1. Display, copy, and re-distribute this work.
2. Make derivative works based on this work.
3. Make commercial use of this work or your derivatives.
Under the condtions that you:
1. Give prominent credit to the original author (me).
Or, in really short terms:
I hope you find this useful, do whatever you want with it.
I recognize that people who want to use this document are going to do so
regardless of what I say here, which is why you are licensed to do as you
please. If you are planning on ignoring the one condition set out in this
license, I won't stop you. Its not worth a law suit. But I will certainly
question the honor of anyone who can't give credit where credit is due.
Now, all that being said, I'm a big fan of community-based learning. If you
have some helpful information or would like to contribute, please let me know.
You are welcome to do so, and will be credited appropriately.
The most recent version of this document will always be posted to GameFAQs.com.
If you are not reading this at GameFAQs, then you are probably behind.
DISCLAIMER
----
THIS INFORMATION IS PROVIDED AS-IS WITH NO WARRANTY, EXPRESS OR IMPLIED. USE
AT YOUR OWN RISK. AUTHOR AND GAMEFAQS ARE NOT RESPONSIBLE FOR THE EFFECTS OF
THE USE OF THIS INFORMATION UNDER ANY CIRCUMSTANCES.
+======================+
| III. Version History |
+======================+
V0.1 - 02/08/2004
----
Intial version with inital dataset. Based on my avatar.xls 0.5. Still
need to finish some arguments and get all the girl gear stats done (though
they are fundamentally identical).
V0.2 - 02/08/2004
----
Updated to use avatar.xls 0.6, which includes newly released items.
+=======================+
| IV. Table of Contents |
+=======================+
I. Introduction -- What I'm trying to do.
II. Legalese -- How I'm protecting myself.
III. Version History -- Where I've been.
IV. Table of Contents -- What you see.
V. GunBound Goals -- Why are *you* here?
VI. Stat Metrics -- How the numbers work.
VII. Avatar Optimization -- How to use the numbers.
VIII. Data Collection -- How you can help me.
IX. MetaFAQ -- What you wanted to know.
X. Credits -- Who I couldn't do it without.
XI. Appendix A: Scoring -- What were they thinking?
XII. Appendix B: Stat Tables -- Where am I?
+===================+
| V. GunBound Goals |
+===================+
GunBound is a great deal of fun, and a fantastic time waster to boot. But why
are you playing? Are you playing to have fun, or are you competing?
If its the former, more power to you. I hope I meet you in a jewel match some
time. But the rest of this guide really isn't geared towards you. Still, you
may find parts of it useful and entertaining.
If its the latter, then we need to have a talk.
Logic Demonstration I
----
| P1: GunBound is a scored game.
| P2: Score is based solely on accumulation of "GP".
|----
| C1: To "win", you must gain GP faster than anyone else.
Obvious, I know. But we need a solid foundation. For the next parts, refer to
the Scoring List in Appendix A.
| C1: To "win", you must gain GP faster than anyone else.
| P3: The largest single consistant award of GP comes from winning a 4v4 game.
| P4: In Score and Jewel modes, 4v4 games don't take much more time than 1v1.
|----
| C2: To gain GP quickly, play 4v4 Score or Jewel games.
Still not shocking stuff. I'll get around to demonstrating the supporting
facts on P4 at some point.
| C2: To gain GP quickly, play 4v4 Score or Jewel games.
| P5: In Jewel play, many useful gold and GP bonuses are disabled.
|----
| C3: When trying to gain GP rapidly, Score is better than Jewel.
So at this point, we're playing 4v4 Score games, preferably with people you
know, as they are less likely to lag out or quit. Now we get to the offensive
part.
| P6: A "kill" is worth 1 GP.
| P7: In a Score game, there can only be 5 kills before the game ends.
|----
| C4: Getting every kill in the match (sweeping it) nets you 5 GP.
Still with me?
| C4: Getting every kill in the match (sweeping it) nets you 5 GP.
| P8: Winning the match is worth 12 GP.
|----
| C5: THERE IS NO POSSIBLE CIRCUMSTANCE WHERE INDIVIDUAL KILLS ARE
WORTH RISKING LOSING THE GAME.
So what am I trying to say?
There is no "kill stealing". While the intial anger at losing a hard-fought
kill is annoying, you'll get over it. Why? Because the enemy that just
dropped doesn't get to act anymore. He's gone until he spawns again. He can't
hurt you or your teammates.
So if the choice is between letting your buddy finish the kill when the kill
still has an action left before that happens, or swooping in and ending it,
end it. Fretting over kills is never worth risking the match. Never.
That being said, understand that I am not advocating being a prick. If your
team is up by 4 kills, you are in little risk of losing. Let your teammates
finish what they started, and expect them to do the same. Play nice.
Note that this logic applies in all game sizes, not just 4v4. In 3v3, the
ratio becomes 9 to 5. In 2v2, 6 to 5. In 1v1, well, no one will be stealing
your kills in 1v1.
+==================+
| VI. Stat Metrics |
+==================+
Now that we have our goal firmly in our mind (to win, right?), we have a
fundamental choice to make. Do we play avatar-on, or avatar-off?
Avatar-off is very tempting. Everyone, no matter how long they've played, and
no matter how much money they can throw at the problem, is equal in an avatar-
off server. It is simply a question of skill and asthetics. If our goal is
to win, then this is where the even odds are.
However, right now, there is a bit of a misunderstanding within the Gunbound
community that makes avatar-on servers interesting. That is, namely, this:
Everyone thinks attack is better than defense. They are wrong.
A high attack stat is certainly good for picking up damage-based bonuses,
mostly Gold. However, our goal is winning the match, right? Let's see why
this is important.
The Argument of Shields versus Swords
----
Two Grubs are next to each other, siumltaneously trading shots. Grub One is
doing 20% more damage than Grub Two. Observe their health totals over time.
Health Totals
----
| T T+1 T+2 T+3 T+4 T+5 T+6 T+7 T+8 T+9 T+10
G1| 1000 900 800 700 600 500 400 300 200 100 dead
G2| 1000 880 760 640 520 400 280 160 40 dead
This time, Grub One is taking 20% less damage than Grub Two.
For equal magnitudes of enhancement, defense allows us to finish an engagement
in significantly better shape than attack. This scales with increased damage
potential weapons.
Enough theory. Let's observe the damage over time of two actual Grubs from a
test match. Grub one is at +12 Swords and +6 Shields. Grub Two is naked.
Those of you who have been keeping up should notice something odd at this
point: Defense scales one and a half times as fast as attack power. Each
point of attack was buying us roughly 2 points (or 2%) of damage. Each point
of defense was buying us about 3 points of damage reduction.
The Argument of Shields versus Hearts
----
[finish text brick. upshot: shields are better]
The Argument of Speed
----
[why having less power over more decisions is better than the inverse]
+==========================+
| VII. Avatar Optimization |
+==========================+
So how do we capitalize on these findings?
The Argument of Cash versus Gold (strictly Maradine's opinion)
----
Though the following tables and information are useful to anyone, there is a
certain benefit to realizing the power of Cash. There are things you can only
buy with Gold. It therefore stands to reason that you should save your Gold
for them. Some people get very offended at the premise of paying money for
GunBound. Some threaten to quit if it ever becomes a pay-for-play game.
To them I say, grow up. Where do you think your data feed comes from? How
much do you pay *per month* to keep it? If you spend $20 (roughly) for the
ideal cash kit, and use it over 6 months of play, how is that worse than going
to the theatre at $15 a pop?
If you can't afford to pay cash or can't get a PayPal (credit card?) account, I
feel for you. You'll have to do this the hard way. But I have a section for
you as well below. I hope I can help.
A Word on Metrics
----
When generating the stat tables for the conclusions I'm about to come to, I
use a binary-weighted metric to make final decisions. In Appendix B: Stat
Tables, you can see what I came up with. Some highlights:
A-Metric
----
The A-Metric is a simple additive sum of the stats of an item. This is a
"pure", unweighted comparison of items.
B-Metric
----
A Binary weighted metric: Shields get four points, swords and white hour-
glasses get two, and hearts, blue hourglasses, and glowshields get one. This
metric shows us what items satisfy our criteria from Section VI. When choosing
items, you want the best item you can afford on this scale. Every one else
is running around with attack-centric avatars. All other things being equal,
a maxed avatar in this metric will win. That is, until everyone figures it out.
C-Metric
----
When torn between two items of equal or very close B-Metric value, use the
C-Metric to decide. This metric indicates the relative cost effectiveness of
the item versus its B-Metric. Think of it as efficiency.
Metrics aside, I present to you Maradine's Custom Kits, three collections
of items to perform three different task.
Legend
----
WH = White Hourglass, speed stat
SW = Sword, Attack stat
HT = Heart, Health stat
SV = Shovel, "Bunge" stat
ST = Star, Money Stat
SH = Shield, Defense stat
BH = Blue Hourglass, item speed stat
GL = Glow Shield, shield regen stat
----
AM = A-Metric
BM = B-Metric
CM = C-Metric
----
Maradine Custom Kit #1
Cash Special
----
This kit is the best you can do on Cash alone. Optimized for winning matches
on Avatar-on servers, it should serve you well while you build GP.
----
Maradine Custom Kit #2
No Holds Barred
----
Provided my data holds up, this is as good as it gets. Expensive, but worth
every penny. Addidionally, impressive looking as it gets. Note that this
outfit must be bought in Gold.
----
Maradine Custom Kit #3
Gold Digger
----
If what you need is Gold, Gold, Gold, then this is your kit. Built for
maximum gold potential with minimum expense, this really helps with saving
up for that big purchase. Like, for instance, any of the above. There are
better items for this category if you are concerned with other stats, but none
more cost effective.
Now, with the popularity rating beginning to apply in Avatar-off servers
relatively soon, this may change. Also, new items are on the way. Stay tuned.
+=======================+
| VIII. Data Collection |
+=======================+
I can always use assistance to help refine my metrics.
There is a lot of data that I haven't collected yet. If you want to help me
help you, please consider setting up one of the following scenarios and report
the results. Please send all results to gbfaq@speakeasy.net. Use plain .txt
files. If you send anything else, it will be discarded.
These are the most useful things you can send me as of V0.1:
----
1. Attack vs. Defense Independant Confirmation
----
Scenario: 1v1 Score, Two Grubs, Avatar-off, then Avatar-on.
Grab a friend, a pair of grubs, and three health packs each. Go to a map where
you can close to near point-blank range within few turns of movement. Proceed
to lay into each other with Shot 1, recording the damage for direct hits only.
Use health as needed to prolong the match and get more numbers. Eventually,
someone will have to win. That's OK. The other guy can win the next one.
Repeat the experiment, this time in an avatar-on server. One player needs to
be naked. The other needs to have gear that increases attack, defense, or
even better, both. Do the whole thing again, using Shot 1 only and recording
both players' damage. Send me the results of both runs, with all the
individual shots, and what the one player was wearing suring the second run.
Expected Results:
During the vanilla match, grub Shot 1 should average 100 damage. Gunbound-HQ
clocks its max damage at 150, but that's not what we're interested in. We want
a reliable estimate of damage over multiple shots, sans equipment.
During the second match, an avatar with attack gear should do somewhere between
2 points and 2% more damage per Sword stat per shot. That same avatar with
defense gear should take about 3.1 points less per Shield stat.
Important:
Do not forget that someone needs to be naked during the second round!
----
2. Do Attack and Defense Stats Operate Linearly?
----
Scenario: as per Scenario 1, but with a pair of A. Sate instead of grubs.
Perform the two tests as above, but with A. Sate, using Shot 1 only. Record
only direct hits. Send in the results.
Expected Results:
Damage metrics from Sword and Shield icons will scale linearly. A Sate's Shot
1 should average 120 points naked. This should be enough deviance to check for
linear or percentage multiples.
----
3. Just How Fast is Fast?
----
Scenario: 1v1, Avatar-off, then Avatar-on
Take bot out with a friend. Set your fire action to dragshot. What I'm
looking for is the fastest shot you can pull on your turn with Shot 1. Record
each shot's delay and send me the lowest for your mobile (don't forget to tell
me which mobile!).
Then, repeat on an Avatar-on server. Wear any gear you can get a hold of with
White Hourglasses. Perform the experiment, and send me the results.
Expected Results:
I have only done this for A. Sate. Results:
Fastest Naked Shot 1: 740
Fastest 24x WHG Shot 1: 680
Fastest 39x WHG Shot 1: 643
This indicates a scale of between 2.2 and 2.5 delay points per WHG. Useful,
but not shocking. I need more data.
+=============+
| IX. MetaFAQ |
+=============+
And now, the requisite FAQ FAQ.
1. Why are you doing this?
Because I can. Half the time, I walk out of a server thinking to myself,
"y'all are whack." Age groups and maturity levels notwithstanding, I think
everyone will have more fun if they're working for something and know how to
get there. Ultimately, if everyone understands an optimal path to a goal,
then the competition becomes a matter of skill of execution alone. And I
like that.
2. Wiggedy Whack?
No, just regular type.
3. I have this great hack for GunBound!
No, you don't. I'm an Information Security Operative by day. I know the
difference. Go scam someone younger. If you ask me for my password backwards
again, I'll kick you.
+============+
| X. Credits |
+============+
People who I appreciate and why:
softnyx, Inc. -- splendid job. Carry on.
DoomStalk, SeekerGio, AG1ES -- thanks for putting up with boring stat runs.
HomTheRobot -- continue to laugh at my jokes, and I will keep cracking them.
If I forgot to mention you, drop me a line at gbfaq@speakeasy.net and yell
at me till I put you in.
+=========================+
| XI. Appendix A: Scoring |
+=========================+
Paraphrased from http://gunbound.net/game/score.asp
Task Gold GP
-------------------------------------------------------------------
Scoring a Double-Kill +500 +4
Dropping an opponent off the map +100 +1
Doing the final damage to an opponent +100 +1
Doing over 500 damage to an opponent in a turn +50 0
Doing over 250 damage to an opponent in a turn +20 0
Doing over 150 damage to an opponent in a turn +10 0
Scoring 50 damage or more through a hurricane +20 0
Shot >70 degrees, flys >2.5 seconds, does >50 damage +15 0
Shot >70 degrees, flys >4 seconds, does >50 damage +30 0
Shot flight path changes direction in mid-air +20 0
Shot hits target behind you +25 0
Do over 3000 damage over entire match +100 +4
Do over 2000 damage over entire match +100 +2
Do over 1000 damage over entire match +100 +1
Falling suicide penalty -50 -1
Do 50 damage worth of friendly fire -25 -1
Kill a teammate -50 -2
Victory Award - 1v1 game +100 +3
Victory Award - 2v2 game +150 +6
Victory Award - 3v3 game +200 +9
Victory Award - 4v4 game +300 +12
+=============================+
| XI. Appendix B: Stat Tables |
+=============================+
Legend
----
WH = White Hourglass, speed stat
SW = Sword, Attack stat
HT = Heart, Health stat
SV = Shovel, "Bunge" stat
ST = Star, Money Stat
SH = Shield, Defense stat
BH = Blue Hourglass, item speed stat
GL = Glow Shield, shield regen stat
----
AM = A-Metric
BM = B-Metric
CM = C-Metric
Note: Misspellings in item names are theirs, not mine.
