The Ur-Quan Masters Discussion Forum

As you know, I'm undertaking a cost balancing project. Now, the method I used for this thing was taking the ships one by one and assigning each match-up of the current ship a value corresponding to how well it fares in the match-up.

I did so for a number of ships (I made a few mistakes in some places, but thanks to Cedric 6014 and myself, I fixed them) and I even used a _LAME_ algorithm for deriving new costs - simply adding the match-up how-well-it-fares-numbers.

Thanks to Elvish Pillager, however, I was able to compute a better algorithm: add the how-well-it-fares-numbers for each ship (to find out how hard it is to be killed, overall) and, so, find the X value, corresponding to how hard to be beaten is each ship.

So it will matter more to be able to destroy hard-to-destroy ships.

Anyway, this is the match-up value for each ship:
(WARNING! LONG LIST! TO SKIP PRESS HERE!) (http://uqm.stack.nl/forum/index.php?topic=3586.msg45603#msg45603)

VUX:
Ur-Quan -2
Chmmr -1
Kohr-Ah -2
Chenjesu -2
Orz -2
Yehat -0.5
Utwig -1
Mycon -2
Pkunk +1.5
Mmrnmhrm -2
Melnorme -1
Spathi -2
Druuge -1
Slylandro +2
Supox -2
Arilou +1
Androsynth -1
Syreen -2
Earthling +0.5
VUX 0
Ilwrath 0
Thraddash -1
Umgah +1.5
Zoq-Fot-Pik -2
Shofixti -1


Ilwrath:
Ur-Quan -2
Chmmr -2
Kohr-Ah -2
Chenjesu -2
Orz -2
Yehat -2
Utwig -2
Mycon +0.5
Pkunk 0
Mmrnmhrm 0
Melnorme -2
Spathi -1
Druuge -1
Slylandro +1
Supox -2
Arilou 0
Androsynth +0.5
Syreen -2
Earthling +1
VUX 0
Ilwrath 0
Thraddash 0
Umgah +0.5
Zoq-Fot-Pik -2
Shofixti -2


Umgah:
Ur-Quan -2
Chmmr -2
Kohr-Ah -2
Chenjesu -2
Orz -2
Yehat -2
Utwig -2
Mycon -2
Pkunk 0
Mmrnmhrm -2
Melnorme -2
Spathi -2
Druuge -2
Slylandro -1
Supox -1
Arilou -0.5
Androsynth -2
Syreen -2
Earthling -1
VUX -1.5
Ilwrath -0.5
Zoq-Fot-Pik -2
Shofixti 0
Umgah 0
Thraddash +2


Ur-Quan:
Ur-Quan 0
Chmmr 0
Kohr-Ah 0
Chenjesu 0
Orz -2
Yehat +1
Utwig -1.5
Mycon +1
Pkunk -0.5
Mmrnmhrm -1
Melnorme +1
Spathi -2
Druuge +1
Slylandro +1.5
Supox -2
Arilou +1
Androsynth +1.5
Syreen +1.5
Earthling +1
VUX +0.5
Ilwrath +2
Zoq-Fot-Pik +2
Shofixti -1
Umgah +2
Thraddash -2


Slylandro:
Ur-Quan -1
Chmmr -2
Kohr-Ah -1.5
Chenjesu -1.5
Orz +1
Yehat -0.5
Utwig -1
Mycon +2
Pkunk 0
Mmrnmhrm -2
Melnorme +1.5
Spathi 0
Druuge -1
Slylandro 0
Supox +1.5
Arilou +2
Androsynth +1
Syreen +1
Earthling +1
VUX +0.5
Ilwrath -1
Zoq-Fot-Pik +1
Shofixti 0
Umgah 0
Thraddash +2


Earthling:
Ur-Quan -1
Chmmr -2
Kohr-Ah -2
Chenjesu -1
Orz -2
Yehat -2
Utwig -2
Mycon -2
Pkunk -1.5
Mmrnmhrm -1
Melnorme -2
Spathi -1
Druuge +0.5
Slylandro -1
Supox -1.5
Arilou +0.5
Androsynth -2
Syreen +0.5
Earthling 0
VUX 0
Ilwrath -1
Zoq-Fot-Pik -0.5
Shofixti 0
Umgah +2
Thraddash +2


Shofixti:
Ur-Quan -1.5
Chmmr -1
Kohr-Ah -1.5
Chenjesu -1
Orz -2
Yehat -0.5
Utwig -1.5
Mycon +2
Pkunk -1
Mmrnmhrm -0.5
Melnorme -0.5
Spathi -1
Druuge -0.5
Slylandro 0
Supox -0.5
Arilou -1.5
Androsynth -1
Syreen -0.5
Earthling 0
VUX +2
Ilwrath +2
Zoq-Fot-Pik -1
Shofixti 0
Umgah 0
Thraddash +1


Kohr-Ah:
Ur-Quan 0
Chmmr 0
Kohr-Ah 0
Chenjesu -1.5
Orz +1.5
Yehat +1
Utwig -1
Mycon +2
Pkunk +1.5
Mmrnmhrm +2
Melnorme -1.5
Spathi -0.5
Druuge +2
Slylandro +2
Supox +2
Arilou +1.5
Androsynth +0.5
Syreen +1
Earthling +2
VUX +1.5
Ilwrath +2
Zoq-Fot-Pik +2
Shofixti +1.5
Umgah +2
Thraddash +2


Thraddash:
Ur-Quan -1
Chmmr -2
Kohr-Ah -2
Chenjesu -2
Orz -1.5
Yehat -2
Utwig -2
Mycon -2
Pkunk -1.5
Mmrnmhrm -1.5
Melnorme -2
Spathi -2
Druuge -1
Slylandro -2
Supox -2
Arilou -2
Androsynth +1.5
Syreen -1.5
Earthling -2
VUX +1
Ilwrath 0
Zoq-Fot-Pik -1.5
Shofixti -1
Umgah -2
Thraddash 0


Utwig:
Ur-Quan +1.5
Chmmr -1
Kohr-Ah +1
Chenjesu 0
Orz -1.5
Yehat +0.5
Utwig 0
Mycon +2
Pkunk 0
Mmrnmhrm -2
Melnorme -1.5
Spathi +1.5
Druuge -1.5
Slylandro +1
Supox -2
Arilou +2
Androsynth +1.5
Syreen +2
Earthling +2
VUX +1
Ilwrath +2
Zoq-Fot-Pik +1
Shofixti +1.5
Umgah +2
Thraddash +2


Arilou:
Ur-Quan -1
Chmmr -2
Kohr-Ah -1.5
Chenjesu -1.5
Orz -1.5
Yehat -1.5
Utwig -2
Mycon -2
Pkunk 0
Mmrnmhrm -2
Melnorme -1
Spathi +1
Druuge +2
Slylandro -2
Supox -1
Arilou 0
Androsynth -1
Syreen -0.5
Earthling -0.5
VUX -1
Ilwrath 0
Zoq-Fot-Pik +1
Shofixti +1.5
Umgah +0.5
Thraddash +2


Zoq-Fot-Pik:
Ur-Quan -2
Chmmr -2
Kohr-Ah -2
Chenjesu -1.5
Orz -2
Yehat +1
Utwig -1
Mycon -2
Pkunk -1
Mmrnmhrm -1.5
Melnorme -2
Spathi -2
Druuge -0.5
Slylandro -1
Supox -1
Arilou -1
Androsynth -1
Syreen -2
Earthling +0.5
VUX +2
Ilwrath +2
Zoq-Fot-Pik 0
Shofixti +1
Umgah +2
Thraddash +1.5


Chmmr:
Ur-Quan 0
Chmmr 0
Kohr-Ah 0
Chenjesu +1
Orz -1.5
Yehat +2
Utwig +1
Mycon +1.5
Pkunk +1
Mmrnmhrm +1.5
Melnorme +1.5
Spathi +2
Druuge -1
Slylandro +2
Supox -0.5
Arilou +2
Androsynth +0.5
Syreen +2
Earthling +2
VUX +1
Ilwrath +2
Zoq-Fot-Pik +2
Shofixti +1
Umgah +2
Thraddash +2

I'm sure I confused you, but you'll understand when I finish.  ;D

Not only is that not what I suggested, but also, as far as I can tell, it's exactly the same thing as your original plan!

What I suggested was that, approximately, for each ship:
- you find all the ships that beat that ship
- you increase the cost of each of those ships by X divided by the total number of such ships (where X is some constant).
- thus, each ship's cost is the sum of the points other ships give it for beating them.

P.S.: the VUX and Umgah numbers against each other are inconsistent.

I said that nobody will understand my explanation! No, what I said is this:
(rather than -2,-1.5,-1,-0.5,0,0.5,1,1.5,2 I use 0, 0.25, 0.5, 0.75, 1, 1.(3), 2, 4, 8)
EXAMPLE:
A:
A 1
B 2
C 0.5

B:
A 0.5
B 1
C 0.75

C:
A 2
B 1.(3)
C 1

Then:
X(A)=1+2+0.5=3.5
X(B)=0.5+1+0.75=2.25
X(C)=2+1.(3)+1=4.(3)

Then:
Cost(A)=(1*3.5)+(2*2.25)+(0.5*4.(3))=10.15
Cost(B)=etc.
Cost(C)=etc.

About VUX-Umgah inconsistency - fixed (set to VUX b Umgah = +1.5)

After reading all this, any regret about dropping math at high school is gone. Not that I had any, but this just confirms it.

Anyhow, what does this mean for all us simple folk? Does it mean that some ships are harder or impossible to beat with any others? If so, you just came up with an incredible elaborate way of telling that?

Did you take into account player versus player or is this based on the AI? Just because you can beat the Mycon with an Arilou blindfolded and with 2 fingers in your nose at awesome AI, does not mean a player controlled Mycon can still put up a fight.

Sorry if I sound all dumb and all, but all these numbers and formulae doohickeys really do not matter to me when I play an action game. If I can make the suggestion anyhow, you might want to put all your accumulated data into a tabular format?

I have no idea either - I did geography instead. Anyway, presumably Step 2 is to somehow convert that data into a list of ships with brand new values. I'm not going to pick holes in Valagger's efforts until that is finished. Good luck to you!

You aren't dumb for not seeing anything in Valaggar's madness. If I were to test anything, I would do it by pitting Angus and Squisher against each other in every matchup. Why those two? Because they're very good and very close in skill level as far as I remember. If not those two, you need to find other similarly matched veterans to do the tests. He isn't doing this so his results will be wrong and the numbers he lists will be meaningless.

Imagine being tasked with the Spathi and Thraddash matchups - i nominate someone who isn't me

1. I said myself I'm mad (moreover, I like being named so! What a pervert I am...)
2. It's right, some values are wrong.
3. To test all 300 combinations with just two players... I don't think they'd accept.


Thank you.  :)

It's an exact, precise way of telling just how good each ship is, overall. It's a method to make ship choosing based on a maximum cost balanced. Cedric will appreciate this.

Ah... oops. Not the h-w-it-fares values AGAINST, not, just like the old values, only that I use 0, 0.25 etc.

Numbers are nice and pretty until your ship is a pile of slag!!!

But far be it from me to question your point balancing system, frankly I think the current set up is fine.

Nobody said that everybody must feel a need for cost balancing.
Nobody said that ANYBODY must feel it.
(anybody except me)

The current system is okay, but could be better. Why should we be content with mediocrity? some ships are over valued (Umgah, Uq-Quan) and others are under valued (Androsynth, Utwig).  Your ur-quan is more likely to become a pile of slag than my kohr-ah

Thraddash is overvalued too... argh!

Undervalued may be the Slylandro too, but you need Awesome AI-like reflexes.

Seeing the 200+ lines of numbers Valaggar has kindly prepared for us it would seem I'm quite a bit too late to post this, but still, IMHO the Ur-Quan and Chmmr ships shouldn't have their point values changed, because the fact the three better ships have the same value has a nice *flavor* interest I think shouldn't be gone.

*Proposal*
Ur-Quan: change blaster damage to 8 (so it can destroy small ships in one shot), proportional battery cost increase.
Kohr-Ah: reduce blade damage to 3, possibly increase FRIED from 2 (IIRC) to 3 to compensate.
Chmmr: slightly reduce laser damage.

They're not even half of the entire work to be done. In fact, I've done a bit more since then, there are only 9 ships remaining. I'll upload them shortly - I've switched to Excel, it's simpler than Notepad.

Well, my mod aims to balance the prices so that vanilla UQM is more balanced. But such a change as you propose would deviate from the original, so I don't think that anybody else would like it.
Anyway, if you want just that, you can simply change the values in the source code - I can do it for you if you haven't already installed the needed software.

May I beg you all to make some more criticism over the match-up values I've snatched yet?

Good point.

Nah, that was just a random idea after all. Thanks for the offer though.

The only way I can think of for testing those is to have human players play against each other with all the combinations... and, in all honesty, I think you're going to have to pay them for that.

No, I thought that veteran players can see at a glance some match-ups, and they can simply test one ship each or something.
I mean, IF they want.

I think, despite the occasional grumblings to the contrary, most people are fine with the way the ships are currently valued. Otherwise, this would have been done in earnest long ago.

Even if people are NOT fine with the current values, I don't think this is helping any. Your numbers are fairly consistently nonsensical.

Maybe nobody wanted to take the burden upon himself...
And why are my numbers nonsensical? I am very sure about most match-ups. Specify.

Anyway, I'll pollify this thread...  ;D

The current state of my project:
http://www.savefile.com/files/623661 (http://www.savefile.com/files/623661)
(link to an Excel sheet with the match-up values)

They're horrible and stupid.

...wait, you mean CONSTRUCTIVE criticism? :P

To make a judgement about the relative value of all ships, on your own, you need to be extremely proficient in all of the ships.
I'll take your values seriously when you can consistently beat Tiberian.

That said, maybe one day I'll add statistics gathering to UQM, so it can be seen objectively how much damage each type of ship does on average to each other type of ship. You could make different valuations for different groups of players depending on the data gathered (top players in tournaments, or newbies among eachother, for instance).

That said, to determine the true value of ships, you'd also need to take into account support ships. For instance, a ship capable of blowing up an Avatar satellite (like a Scout) can make the Avatar acessible for ships which otherwise wouldn't stand a chance. Or an intruder might manage to slow down a ship while being destroyed itself. Or a Broodhome which gets destroyed after draining the energy of a Jugger.

Yes! Statistics gathering must be the top priority for UQM 0.7.0!!! A higher priority even than a multiplayer lobby or anything else.

Now, just as a jokey thingy, here's my completed sheet:
http://www.savefile.com/files/625856 (http://www.savefile.com/files/625856)

Your sheet makes more sense now, but as above posters mentioned, it's still based on personal thoughts and formulaes. Some people are just better with specific ships than others, and you should find fairly consistent players who can play all match ups before you can make educated guesses on how to change point values.

Furthermore, except for the singular exception, I find the current point values sufficient as they are.

I would only change point values by the way and I strongly disagree with Vela's suggestion to change anything gameplay wise. If it's only the points in melee, I'm cool, but change the properties of a ship to accomodate point value, nope.

I think that the core team would say that any change of this nature, point values included, would be contrary to the purpose of UQM, and they'd be right. What is downloaded from Sourceforge should be as close as possible to what was on the 4 5.25 inch floppies I unpacked back in 1993.

I think perhaps, in the future, when we are racing around in jetpacks, that if there is ever a developed lobby/competiton system, that playing in it might require a mod to be be downloaded that incorporates slight value adjustments. I still personally believe that online Melee could have a broad appeal and raise awareness for  sequel. For the purposes of competition the values should be as correct as possible.

By the way Valagger, from a cursory glance of your spread sheet it looked ike you had Mycon beating Arilou.

The statistics issue along with the question of what/who the win/loss is against is going to be the overriding thing.  After all, most of what beats what is opinion, and we know there's quite a difference of opinion based on skill level of the person involved, whether they've learned that the AI is trash (awesome or *gasp* they're playing down) or not, whether it's a fluke victory or not (70% of my losses against the AI seem to be that way anymore - either involving an asteroid or a planet) and I'm sure many factors I haven't even thought about.  I don't know exactly what the numbers in the spreadsheet even mean for most part, but do know you need a much more representative sample before you consider acting on anything of this sort.

Keep your plasmoids inside the Podship and try killing the Arilou when he nears. Well, it's not really +2 for Mycon, maybe -1: sometimes the Mycon may hit.

Anyway, we'd need two Tiberians to really get the numbers right - someone so proficient in all of the ships, that he's no longer better at one ship than another.

Just for fun, here are my aberrant "balanced" costs using the Elvish Pillager-Valaggar method:
Ur-Quan......25
Chmmr........50
Kohr-Ah......50
Chenjesu.....35
Orz..........42
Yehat........20
Utwig........36
Mycon........17
Pkunk........16
Mmrnmhrm.....47
Melnorme.....37
Spathi.......20
Druuge.......28
Slylandro....24
Supox........28
Arilou.......13
Androsynth...35
Syreen.......41
Earthling....7
VUX..........11
Ilwrath......6
Zoq-Fot-Pik..10
Shofixti.....15
Umgah........3
Thraddash....2

Mod executable:
http://www.savefile.com/files/629984 (http://www.savefile.com/files/629984)

The more I think about this, it would be possible to make such an analysis for the AI alone (it plays the same for everybody, can't say that for any human player, even though it is trash at playing), assuming something was coded into a copy of UQM.  You'd have to make sure you use proper statistical analysis on the results, given the full data generated.  For example, you would have to track victory counts and crew, but recognize that beating a ship at 1/4 or 1/2 strength is different than beating it at full strength.  This is especially true, since there are a lot of matchups that are impossible for a ship against a full strength adversary, but very possible against a weakened one.

Basically for a statistical module, the game would need to track winners & losers, as well as losses by both in terms of percentage of total crew lost.  Yes the numbers for the Syreen would skew things a bit, since you'd have to base it off of 12 crew instead of 42.   But there's a lot of thought to be had still on a proper way to go with this.

Of course, all bets are lost if the AI is changed in any way to "improve" it.

Ahoy, Captain, the AI is better at some ships (the quick-reflexed-based ships) and HORRIBLE at others (those for which you need to use your Brain). It's by no means a judge of the ships.

As to starting crew, you only need to test for both ships at full strength, since the ships play equally well, no matter how damaged they are.

Of course, the AI may be improved - Elvish and Novus rambled a bit about using a learning algorithm, but... it's only a dream.

It surprises me how decent those costs are.

Arilou and Shofixti are off by a lot.

Wow, I've been of the boards for a while (Everything going wrong at work these last couple weeks). Firstly, I would to thank Shiver for his complements.  Secondly I would like to make several comments in regards to peoples posts:

As far as the need to ballence goes, it is true that it is not nessessary.  However, I do believe that doing so does provide something positive for online supermelee.

People have mentioned that the costs of the ships are dependant upon the skill levels of the players.  While this is true, I do believe that the costs should reflect a player who is considered "good".  I came to this conclusion based on the concept of "optimal strategy".  Two poorer players seeing an imballence between two ships can easily be flipped due to a change of strategy, while two veteran players have analyzed the play enough to have the strategies evolved to a point where it is unlikely that a radical strategy will upset the ballence any longer.  Although new strategies can be determined, they generally dont swing from dominating of one side to the other.  To give an example, Valaggar give the Kohr-Ah -1.5 vs Chenjesu and +2 vs Earthling, when these numbers are actually reflective of poorer players, and ballence based on this will be virtually unplayable for good players.

As for the testing of all combinations; I dont think that this is nessessary.  I believe that if I drafted up the +/-'s of all the ship combos, and If Angus did the same, we would only have to play out the matches which had differences of opinion.

As to how the algorithm is currently being done, I believe it is flawed.  I believe that the ship +/- levels should be based upon the COST of the ship, and that the results of the algorithm would produce ADJUSTMENTS to the current costs.  I explain further:  Valaggar currently lists for the Chmrr the following values:

Umgah    +2
Spathi   +2
Shofixti +1

Now, it is clear that none of these ships can beat the Chmrr, but the spathi has no chance to damage, while the 4 Umgahs (7 food each) have a decent shot at taking out the chmrr.  My proposal is this:  Make the +/-s based on the current food level.  A shofixti is a great choice vs a Chmrr, and it should be reflected as such.  Based on current cost values, a Chmrr should be -2 vs a Shofixti.  Based on the way Valaggar is currently doing it, the Shofixti has 0 as the highest value it can acchieve, as it can not win without dying also (generally speaking).  Also, many people would agree that the current costs are close to what is appropriate, and a full overhaul is not nessessary; a simple balance will do.

Additionally, there are other factors which should be accounted for:  Ship Cost vs Ship vulnerability.  Consider a ship which simply pummells every ship, but has no chance of even damaging one certain ship.  By your system, this would be a high cost ship, which would destroy 1 enemy ship, then lose the next fight as the opponent will pick that ship to which it is weak.  Ships with low cost are inheirently stronger (relative to its cost) because even if they have a glaring weakness to a high cost ship, they can still be cost effective so long as they make your opponent choose that high cost ship, which will subsequently die quickly to your next choice. 

Anyone who has played tiberian will quickly recognize this.  So long as a fleet of low cost ships can collectivly destroy any high cost ship, they will always at the advantage because the high cost ship, despite being good against most ships, will still destroy 1 ship and lose to the next ship, something wich a high-cost ship fleet cannot afford to do.  Thus, it is a ships disadvantages which should be the largest impacting factor to ship cost.  Again, this can be easily seen by considering the effectivness of a ship which is strong vs all but weak vs one compared to a ship which is stong vs half, but weak against none.  Clearly the latter is the better ship.

Valaggar:
You are right about the AI not being a judge of the ships, but at the same time, it's something that is a constant if you wanted to try to figure out the relative strength between ships.  Of course, you get a skew in results given the poorer players versus the good players, but it's much less than if you consider two players.  Not to mention, the game mechanics are different between local and online games anyway (specifically movement timing, which is the most important factor in the game), and that plays a factor into things.

Squisherxxx:
There definitely would be a difference of opinion in most of these.  If I could understand the spreadsheet as it was done, I would probably draft much different numbers than Valaggar did, simply because my skills would be different, I'd be better at some ships than others, and a whole host of other things.  As far as your ship observations go:

1) A shofixti is indeed a great thing for a Chmmr, even if it dies, it wipes out a number of the zapsats (3 if you're really good), making the Chmmr vulnerable to much more.  Also a shofixti is much capable of taking out a few ships without using the glory device, so any observation based on the glory device alone is flawed.

2) I take it you did not see the video that was posted a few months ago regarding what a Spathi can do to a Chmmr. http://uqm.stack.nl/wiki/SuperMelee_demonstration_videos#Defeating_a_Chmmr_Avatar_with_a_Spathi_Eluder (http://uqm.stack.nl/wiki/SuperMelee_demonstration_videos#Defeating_a_Chmmr_Avatar_with_a_Spathi_Eluder)  So this actually reveals that there's another factor in there, preceived notions of impossibility.  One thing I've found personally is that my game got much better when I gave up the notion that any matchup is "impossible so I shouldn't even try it".

I'm lost i think - are you saying that a Syreen is superior to an Ur-Quan?

Well, his numbers for Ur-Quan...


It's better than he says against, at least, Chenjesu, Orz, Utwig, Pkunk, Spathi, Supox, Shofixti, and Thraddash.

(Especially Thraddash. What the hell is up with that number?!)

Ok, I think i understand. The method may work but so far the data is very bunk. We will need 1 experieinced player to go through Valagger's template, and for it to be subsequently peer reviewed by 2 or 3 others.

AAAAAAAAAAAAAAAAAAAAAAARGH!!! THAT WAS IT! I KNEW! I KNEW!
I wrote -2 for Ur-Quan vs Thraddash, then I tested again and I discovered that I was wrong... but I forgot to change the value!
:P :P :P :P :P :P...
Plus, I suck at Ur-Quan, not because I lack skill with it, but because I like shooting like mad, even if I don't have any chance of hitting.
Ur-Quan will be reviewed immediately.

Yes. Do it. Please!...

As for the suggestions related to taking into account HOW MANY ships are needed to destroy the <current> ship... yes, in fact I did kinda like that - I converted:
+2 to   8
+1.5 to   4
+1 to   2
+0.5 to   1.(3)
0 to   1
-0.5 to   0.75
-1 to   0.5
-1.5 to   0.25
-2 to   0
(the numbers on the right represent how many <current enemy> ships must be lost to defeat the <current> ship)
More precision is unattainable since the match-ups depend on skill and luck too.

As to the SuperMelee demonstration video with the Eluder defeating an Avatar - it's against Mr. Awesome!

Fixed Ur-Quan. Ur-Quan vs Utwig pending, I will test this on #uqm-arena.

Workbook:
http://www.savefile.com/files/632108 (http://www.savefile.com/files/632108)

Actually, any match vs AI has absolutly no credibility.  Show me a human player losing that matchup then we'll talk.

Playing single battles and calculating the outcomes is a very wrong way to make adjustments to ship values.

The only proper way to re-think the values would be a real-time discussion with lots of people who know everything there is to know about melee. That is never going to happen.

Trying to fix something that is not broken will often lead to breaking it.

Typical arrogance.  It's unfortunate how this board has gotten in the last few months.  I'm thankful I haven't gone into #uqm-arena as much as I might have had.

Firstly, you do not understand the meaning of arrogance.   Secondly, you are arrogant to the people on this forum.

Rather than flaming the boards, better test the match-up on #uqm-arena.

No, a real-time discussion is not necessary. We can simply have the melee experts (you and the others on #uqm-arena) make a few tests and judge the match-ups.
Also, cost balancing would be a good thing for tournaments.

All the "tests" have been done already past these 15 years. Melee experts (me and the others on #uqm-arena) don't have to play single matches to determine which ship is better than the other.

Also, "cost balancing" is not a good thing for tournaments, since the ship values are already balanced. Tweaking it would most likely result in a worse outcome.

Someone could now call out "but hey, ship A clearly has a too low value and ship B has too high value!" That is true, there are ships like A and B, but no matter what the values are, there will always be lots of players who feel the values are not correct. Since we, the "experts at #uqm-arena", and the developer team don't see a problem, I'm pretty sure there will be no changes.

When PR3 and FF made SC, they did not have the time and resources to play a bazillion melee games to properly test the values they attributed to the ships. Whilst I think what they did was pretty spot on, it's subsequently evident by my experiences in PvP that some ships ARE overvalued and some ARE undervalued. And most good players, Tiberian included, would agree on which ships these are. The issue is to what degree are they over or underrated. Its not by much but it is enough to ensure that every player (if they want to win) picks an androsynth and no player picks an ilwrath. Despite my normal socialst tendencies, I say we let the market decide. Would people pick an ur-quan if we dropped it a couple of points?

I think online melee would be enhanced considerably if the full array of ships were used. I get sick of constantly battling kohr-ahs, orz and earthlings. Bring on the umgahs and chenjesus! Also, for some reason all the sucky ships are red, green or purple.

Anyway, like a said before, UQM shoudn't be changed as such but perhaps a mod could be used for competitive situations.

If you want to play the underused ships some more, I'd suggest letting each player build the opponent's team.
The number of ships of a type should be limited, and the agreed ship cost would be a lower bound, instead of an upper bound.
Maybe you'd even want to make it a tournament. :D

Hey that's a pretty fun idea.
Don't know if it would fly for competition purposes though

Of course not. Everybody would choose the overvalued ships for the opponent etc. Then, the underused ships will become the overused ships. No change.

Very well said. That's why we need a complete rewrite to fix those values. We'll then decide what to do with the other values.

Right on the spot. And they're my favourite colours... *sigh*

Workbook updated:
http://www.savefile.com/files/635012 (http://www.savefile.com/files/635012)

How is "underused ships becoming the overused ships" no change?
And you'd obviously use this play method in addition to the traditional gameplay. Noone is expecting you to stop playing the game the traditional way.

Well meep eep brings up an interesting idea.  Consider the following:

Perhaps we have been looking at the cost ballencing from an incorrect light.  What if instead of looking at which ship beats which, we look strictly at ship selection.  As each player has their own views and playing styles of ships, in the end they end up choosing ships which they feel are undervalued or ships which they are more proficient in.  If it were possible to do statistical analysis on which ships are chosen, and simply re-ballence them due to supply and demand, it could produce something interesting.  You would not need to take into account variables such as Avatar zap sats, pkunk respawning or VUX limpeting.  Players would make their own analysis of this based on their ship selection. 

Imagine a mod which lowers the cost of the 3 least used ships by 1 each month, and increases the cost of the 3 most used ships by 1 each month.

Because what today is underused will then be overused. And what today is overused will then be underused. Of course, it IS something good, but it's still not the best.

I’ve played with these rules several times. It can be quite fun for a match or two, but the teams always seem to boil down to the Rejects of the Hierarchy + Zoq-Fot & Supox and whatever else is deemed an easy kill.

An interesting novelty for sure, but I wouldn’t want it for tournament play.

Yes this is the kind of thing I was getting it with 'letting the free market decide'. The terms 'undervalued' and 'overvalued' are very subjective - as evidenced by this heated debate. In my view, the aim of this discourse is to result in a situation where each ship is selected about is often as every other ship. I dont even think a statistical analysis is necessary. I think the better players have a fair idea of what adjustments to make.

So I think the best way forward is to make a mod with value adjustments and just experiment with it for a while in PvP. It will take a while for it to actually occir to players that they can pick a mycon over a jugger. Actually, I think it will take a loooong time for the magic of market forces to make their impact, now i come to think of it.

Hmm...I like the idea of making a mod to test the adjusted prices, and then encourage people to play it. I'd try it.

Thank you for supporting our cause, Angus!

Well, it seems that my method, while good, can't be implemented so easily due to the lack of veteran players support (save for Angus). We'll try, however, to implement it.
Unfortunately, we can't get prices that balanced by simply stretching them, as the earlier part of the Best value ships Thread shows.

I must extend once again the call for expert help on my workbook (or you can make one from scratch, if you want).

OK, let's see what's wrong with the Syreen (it's got a way too high value in my mod, you know):
Syreen
Ur-Quan -1.5
Chmmr -2
Kohr-Ah -1
Chenjesu -2
Orz 2 ? 0
Yehat 1.5 ?
Utwig -2
Mycon 2
Pkunk 1.5
Mmrnmhrm 1.5
Melnorme -2
Spathi 2
Druuge -0.5
Slylandro -1
Supox 2
Arilou 0.5
Androsynth 1
Syreen 0
Earthling -0.5
VUX 2
Ilwrath 2
Zoq-Fot-Pik 2
Shofixti 0.5
Umgah 2
Thraddash 1.5

The values marked with "?" are unsure; others may be unsure too, but I don't see them right now.
The proposed new value is after the "?", with yellow.

On the high end, these values also look suspicious:

Mycon 2
Pkunk 1.5
Supox 2
Androsynth 1
Ilwrath 2

Hey hey hey... let's see these new values for those problem-ships...
Here they are:
Mycon -1
Pkunk 0.5
Supox -0.5
Androsynth -0.5
Ilwrath -1
Orz 0
Yehat 0

Next on my list (well, I didn't make a list.. anyway:)

Mmrnmhrm:

Ur-Quan 1
Chmmr -1.5 ? -2
Kohr-Ah -2
Chenjesu 1.5 ? 0.5 or 1
Orz -1
Yehat -1.5
Utwig 2
Mycon -1.5 ? 0.5 or even more?
Pkunk 2 ? 1 (don't forget resurrection)
Mmrnmhrm 0
Melnorme 2
Spathi 2
Druuge 0.5 ? -0.5 or -1
Slylandro 2 ? 1.5
Supox 2 ? 1.5 or 1
Arilou 2 ? 1.5
Androsynth 0
Syreen -1.5
Earthling 1
VUX 2 ? 1.5
Ilwrath 0
Zoq-Fot-Pik 1.5
Shofixti 0.5
Umgah 2
Thraddash 1.5

Alright, I've fixed Mmrnmhrm without help.
I've returned at working at this project.

For Mmrnmhrm, these values need to get less extreme:
Kohr-Ah -2
Yehat -1.5
Melnorme 2
Spathi 2
Syreen -1.5
Umgah 2

And this one needs to get more extreme:
Thraddash 1.5

Thank you. I had overlooked those.

Another aberrant ship:
The Orz Nemesis.

Ur-Quan 0.5
Chmmr 1.5
Kohr-Ah -1.5
Chenjesu 1.5
Orz 0
Yehat 0.5
Utwig 1.5
Mycon 2
Pkunk -1.5
Mmrnmhrm 1
Melnorme 1
Spathi 1
Druuge 1
Slylandro -1
Supox 1
Arilou 1.5
Androsynth -2
Syreen 0
Earthling 2
VUX 2
Ilwrath 2
Zoq-Fot-Pik 2
Shofixti 2
Umgah 2
Thraddash 1.5

These values seem off:

Chmmr 1.5
Pkunk -1.5
Melnorme 1
Arilou 1.5
Earthling 2
Thraddash 1.5

Orz vs Arilou can launch a dozen of Marines and storm the Skiff with rocket launchers from afar.
The other values are now changed.
(Orz vs Chmmr is now +2, I hope this is right)

Chmmr 1.5
should go DOWN. The chmmr can planet-camp, and the Orz only has to make one mistake to die. I'd put this somewhere near 0.

Pkunk -1.5
Arilou 1.5
What can you do to the Arilou that you apparently _can't_ do to the Pkunk? Really, launching lots of marines against Arilou is awfully stupid, and you're likely to just miss and die.

Ah, got it. Chmmr is now -0.5, Pkunk and Arilou 0.5.

A difficult match: Arilou vs Melnorme. I gave the Arilou a +0.5, but I'm not sure.

And now, the Pkunk Fury: (note that you need to take into account resurrection; currently, I'm doing this in this way: If the Pkunk's rating is negative - let's say that in 1 life, the Fury gets a -0.5 rating. Then I calculate: [(RatingForOneResurrection*2)+RatingForTwoResurrections+RatingForThreeResurrections]/4. If it is positive, no change.)
Ur-Quan -1
Chmmr -1
Kohr-Ah -1.5
Chenjesu -1
Orz -0.5
Yehat -1.5
Utwig 0
Mycon 1
Pkunk 0
Mmrnmhrm -1
Melnorme -1
Spathi 1.5
Druuge -1.5
Slylandro 0
Supox -1
Arilou -0.5
Androsynth -1
Syreen -1.5
Earthling 1.5
VUX -1.5
Ilwrath 0
Zoq-Fot-Pik 1
Shofixti 1
Umgah 0
Thraddash 1.5

I don't have much experience with Pkunk, I can't say whether any of those values are off.

Anyway, Pkunk is not that hard to test. The problem is that I haven't applied that same algorithm everywhere, but I will redo it soon.

Shofixti Scout:
Ur-Quan -1.5
Chmmr -1
Kohr-Ah -1.5
Chenjesu -1.5
Orz -2
Yehat -0.5
Utwig -1.5
Mycon 2
Pkunk -1
Mmrnmhrm -0.5
Melnorme -0.5
Spathi -1
Druuge -0.5
Slylandro 0
Supox -0.5
Arilou -1.5
Androsynth -1
Syreen -0.5
Earthling 0
VUX 2
Ilwrath 2
Zoq-Fot-Pik -1
Shofixti 0
Umgah 0
Thraddash 1

With a gravity whip, the Shofixti pretty much has to detonate against the Mycon - you can't really steer the globs into a speeding Mycon, he'll easily regenerate from peashooter damage, and one mistake or an unlucky asteroid is all it takes for a plasmoid to kill you. The same is true to a lesser extent for the Ilwrath and VUX if they get a gravity whip. And these ships usually will get to the planet, because it takes a long time for a Scout to kill something with those energy darts.

Shofixti can still usually kill all three of these ships and can maybe ping them to death sometimes, but needs the glory device to do it consistently. This also raises the question of how crippling an enemy ship or taking it down with you should be rated.

Corrected.

Normally. No matter if you used the Glory Device, Fusion Blasts, MX Missiles or anything else, the outcome will still, for example, that you lost your ship and halved the enemy crew. (=-1)
Using the Glory Device to damage the enemy is loss, not victory, since he survives. Using it to annihilate both ships is a draw.

Upon looking at this thread again, I am more firm in my belief that adjustments based upon ship selection is the proper way forward.

There can be no denying that a free market is the way to perfect balance.   

Consider a system which either 1) attempts to create a static ballence at a fixed date or 2) is a full free market   ,  using the following guideline:

Every (lets say 2) weeks, the 3 ships which are selected the most for battle have their ship costs increased by 1.  The 3 ships which are selected the least will have their ship costs decreased by 1.

After enough iterations, the system will be ballenced.

Lets look at the first iteration.  Those whom I have selected as the most/least used are purly subjective, but are used to show the concept behind it.

Lets say the 3 most used ships are:
Androsynth -> $16
Kohr-Ah -> $31
Chmrr -> $31

and the lest 3 used ships are:
Umgah -> $6
Illwrath -> $9
Sylandro -> $16

What this does is promote the lesser used ships to be used more, and the most used ships to be used less.  As the cost adjustment per interval is a mere $1, there wont be large swings each iteration.  Now consider a ship which is overvalued.  We shall use the androsynth for this example.  A cost increase to 16 will not stop me from using it as often, as, in my opinion it is still a bargain for 16.  There will be a group of users/instances in which the cost of 16 doesnt justify selecting it over annother ship.  Nonetheless it may still be the most selected ship.  But there will be a saturation point in which, at a cost of $18 lets say, that it wont be the most selected ship.   Now it has reached a point of 'ballance' and annother ship which is the most selected will now start to become ballenced, along with the other 2 most selected ships.

I deny that the free market is the way to perfect balance. I majorly deny it.

<some_guy> Why r u using so many Slylandro Elvish?
<Elvish_Pillager> To pump up the cost so Tiberian doesn't get as many ships on saturday!

:P

Even besides silliness like that, though, you could have most of the ship picks being done by mediocre players, in which case the balance would be not perfect but mediocre. ::)

As I see it, free market will be the best way to achieve balance. Without intentional manipulation, an equilibrium is mathematically inevitable.
The biggest problem will be intentional manipulation, but that can be mitigated by only looking at tournaments to determine the prices, and perhaps even by weighing the number of wins a played has. There could even be different markets for beginners and advanced players (after all, some ships are more valuable to an expert than to a newbie).

I too think a cost/demand approach to ensure ships are as popular as each other will work better than a complicated quantative analysis. The method being devised at the present (whilst very interesting) does not take into account factors such as strategic ship selection (thinking 1 or more moves ahead). There will be other more intuitive reasons that ships are selected, that will be almost impossible to account for in this approach.

We’ll see what the results are and maybe new costs can be used as a starting point. Ultimately though, I believe that the value of ships is determined by how popular they are amongst the best players.

So, I do not think Meep-Eep’s idea of having two separate ‘economies’ for new and experienced players is sensible – this is not how a free market works. Experimentation should be carried out using experienced/good players only.  If new players aren’t proficient with a particular ship then they will have to learn and adapt to fit in with the true ‘market’. Players must adapt to fit the market – not the other way around.

There'll be an equilibrium, yes, but not a balanced one.

...

Also, some people pick the same ship a lot because they're trying to practice and get good with it, not because it's a good ship in its own right. By the time they did become good with the ship, their actions would have resulted in it becoming more overcosted.

#define balanced

And practice games are irrelevant if only ships used in tournaments are used to adjust the market value.

Huh, does that cause an error, or does it define "balanced" to ""? Or something else entirely?

I practice in tournaments.

The second choice; any reference to the "balanced" (as a word in the code) is removed by the preprocessor. So it doesn't balanced matter how balanced I balanced use that word anymore, because it balanced gets ignored. I can still use "balanced" in quotation marks balanced balanced balanced balanced balanced. ;D

Right, and if we have a piano, it is ours. And that piano is a piano because if it wasn't a piano, there would be no piano. I'm sure you all understand.

The more I think about this, the more I think this effort will ultimately be futile. There are just too many opinions on how the costs should be adjusted, which ships, etc.

I think the best solution is to have matches/tournaments where ships, not points, are the limiting factor. So you’d have the rules set up something like this:

• 12 ships max
• No duplicates
• 1 or 2 heavy ships max*

*Optional if we get tired of every fleet having a Chmmr, Khor-Ah, Ur-Quan, and Chenjesu.

Of course, the number of ships is variable. But with a higher number, I think you’d start to see some of the misfit ships, like the Mycon, creep back into play.

And a Drone is worth the same as an X-Form?

Well, if the values differ only slightly, we can choose the average. If they differ more, only one of them will be allowed to select a final value (the best one, Tiberian, for example).

How about allowing something like "7 types of ships on your team, up to 30 points worth of each ship type"

It'd have to be fixed up some, but how's the basic idea?

It's the same as having a point limit: since the ship values are not balanced, the result will in no way be balanced.

Like Angus's idea, it is not meant to try to make the current system balanced, but to encourage play with some of the less well used ships.

Agreed.  We should therefore begin collecting all ship selection data from each tournament.  Perhaps wait until data is collected from 3 -5 tournaments before making our first adjustment, so that most manipulative anomolies can be limited

Well, I’ve been thinking about this a bit lately and came up with a list of what I feel the ship point values should be. This is based purely on my own experience playing net melee and how effective I feel each ship is. I’ve also been playing a lot of ‘reject’ ships lately to see just how useful they are against some of the better players. (Had I made this list a year ago, before net melee was available, I’m sure it would have been pretty different.)

So here is my list, presented in descending order of what my values are:

SHIP Chmmr      Kohr-Ah      Ur-Quan  Chenjesu    Utwig    Orz            Yehat    Melnorme      Mmrnmhrm Androsynth    Pkunk        Slylandro  Druuge Mycon        Supox            Arilou Spathi        Syreen      Earthling    VUX      Ilwrath        Thraddash    Umgah Zoq-Fot-Pik  Shofixti

ORIGINAL VALUE 30 30 30    28 22        23    23    18 19 15  20 17 17 21 16 16 18 13 11 12    10 10 7 6 5

ANGUS’ VALUE 34 33 28 27 24 23 20 20 20 18 18 17 17 17 16 15  15 13 12 12 9 9 6 6 6

NET CHANGE +4 +3 -2 -1 +2  0 -3 +2 +1 +3 -2  0  0 -4  0 -1 -3  0 +1  0 -1 -1 -1  0 +1

So why do this? Because I’m afraid that sooner or later any competitive online matches are going to be between two homogenous fleets, with only 1 or 2 different ships.

How boring. I miss playing with and against Mycons, Spathis, Yehats, etc. I think they’re interesting ships and they’re consistently MIA unless I’m in a “I don’t give a hoot if I lose” kinda mood and pick them anyways.

As you can see, there are some pretty drastic changes there. But then again, I think there are some pretty drastic discrepancies.

I’ve stated above that we’ll probably never get everyone to agree on a modified point system, and that’s still probably true. Nevertheless, this is my list (and I’m sure they’ll be some disagreement). Maybe I’ll get around to modding it one day to test it out…

I've recently tried a new approach, with Cedric6014 and Elvish as victims volunteers.
Each player got a number of points, and every ship was auctioned off to the person bidding the most points.

We did this manually, which takes a lot of time, and is a lot of work for the auction master, but it could be built into the game.

It adds a whole new level of strategy; each player must make sure he/she has appropriate counters available, and a player may choose to buy a ship he/she doesn't want just to make sure the opponent doesn't get it.

The advantage is that the value is determined automatically; there are no artificial predefined values. It also gives you original matches, as every ship (if it isn't excluded from the auction) is used once, and the usual counters may not be available. You also won't have any mirror-matches, which usually are little fun.

The prices that the ships go for could be collected, and the results could be used to determine a new list price for the ships.

Sounds very cool. I always though the Supermelee netplay has endless gamemode/ruleset gamedevelopment possibilities.

--Eino

Angus, your values are very similar to the ones I suggested a while back on this thread
http://uqm.stack.nl/forum/index.php?topic=3355.15

And you reasons for change are the same too. So I like your thinking!

The only thing is Thraddash - it should probably be increased by 5 points if only to discourage people choosing it. Currently there's a gentlemen's agreement between Elvish and any one who cares to play him that it doesn't get picked.

Meep-Eep's auction thing is great fun. Every ship gets picked in the end. The labourious thing will be actually getting folks to report back auction results. I will do so for any that I am involved with..

After some more thought, I think that the auction result will not be very useful to determine a new price list. Because in the auction the price is not just determined by how much you want to use the ship, but also by how much you don't want the opponent to use it. For matches where having a ship doesn't prevent the opponent from having it too, some ships will be worth less.

Still, the auction in itself can be used to avoid the fixed prices altogether.

I did an auction with Shiver, I'll just say what the prices came out to:
Androsynth: 20
Arilou: 13
Chenjesu: 24
Chmmr: 36
Druuge: 19
Earthling: 12
Ilwrath: 7
Kohr-Ah: 36
Mmrnmhrm: 24
Mycon: 15
Orz: 24
Pkunk: 12
Shofixti: 6
Spathi: 8
Supox: 10
Syreen: 13
Ur-Quan: 26
Utwig, VUX, Yehat, ZFP: No reading (they were at the end of the auction where it gets screwy.)

The stock market  type approach (adjusting the prices for the most and least used ships) some have mentioned would be the best way to balance the values.  Assuming of course you have good and experienced players doing it, you wouldn't want me doing it for instance since I'm terrible at multiplayer.   

The idea of using the AI to test it, doesn't work at all.  I actually did that a couple of years ago.  Using the AI, I had each ship fight one each of all the other ships in as big of groups as the ship could handle.   I then calculated what the ship "should be" valued at.  (Repeat experiments proved the results to be more consientant than I was expecting.)   

For instance the shofixti ship is a 5, which if it were perfectly valued for the setup I used, would mean that it would take 437/5 = 87.4 shofixti ships to defeat one ship of every type.   Here's the results I got (which actually tells you about the AI more than the ships):     

ship - real value - calculated value - change

Androsynth   15    18.3   +22%
Arilou           16    10.1    -37%
Chenjesi       28    25.1    -10%
Chmmr         30    48.5    +62%
Druuge         17    11.3    -34%
Ilwrath         10    10.6    +6%
Kohr-Ah       30    30.0     ----
Melnorme    18     9.9     -45%
Mmrnmhrm 19     8.8     -54%
Mycon         21     9.3     -56%
Orz              23    10.1     -56%
Shofixti        5       7.7    +54%
Ur-Quan      30    26.6    -11%
Utwig          22    55.3    +151%
Vux             12    10.0    -17%
Yehat          23    32.6    +42%


These ships couldn't be valued because they had at least one ship (which is listed after it) which the AI couldn't beat even with 14-1 ship numbers against it.   


Earthling - Chmmr
Pkunk - Syreen
Slylandro - Chmmr
Spathi - Chmmr
Supox - Chmmr
Syreen - Chmmr
Thraddash - Chmmr
Umgah - Spathi
Zoq-Fot-Pik - Chmmr

Also, from what I read I guess everyone is in agreement with the drop in value the Mycon had from SC1 to SC2.

What about the change for the Earthling, Chenjesu, and Syreen ships. 

Mycon and Syreen were more expensive in SC1 because of their powerful strategic mode abilities. Also, the Mycon is decent to good against most SC1 ships and bad against most new SC2 ships. THe Mycon's cost should be reduced another couple points, though.