gametechmods

Robot Arena => Discussion => Topic started by: RpJk on October 16, 2011, 04:07:33 AM

Title: A table for bot type efficiency
Post by: RpJk on October 16, 2011, 04:07:33 AM
Ok This is an idea to help people understand what bot types are normally good and what they are weak against.

I've got an idea where it is like Pokemon types, Fire> Grass> Water> Fire

This is kinda like this when it comes to Robot Combat.

 VS > Rammer > HS >VS

So I thought this would help alot for people who are new to competitive building in RA2.

However I will need suggestions when it comes to hybrids.

Is this a good idea?
Title: Re: A table for bot type efficiency
Post by: ty4er on October 16, 2011, 04:18:59 AM
Would be useful if you actually knew what types are good against others.

HS isn't always better than VS
Title: Re: A table for bot type efficiency
Post by: Pwnator on October 16, 2011, 04:25:28 AM
Would be useful if you actually knew what types are good against others.

HS isn't always better than VS

And Scizor can still take on Flareon with the right moveset. Specializing something will always be an exception.

So yeah, that's pretty much it.
Title: Re: A table for bot type efficiency
Post by: Naryar on October 16, 2011, 04:45:03 AM
What they said.
Title: Re: A table for bot type efficiency
Post by: RpJk on October 16, 2011, 05:09:32 AM
Ok so give me suggestions.
Title: Re: A table for bot type efficiency
Post by: Pwnator on October 16, 2011, 05:22:12 AM
That's pretty much it, really. List down the other types you would want to include to make this easier.
Title: Re: A table for bot type efficiency
Post by: Naryar on October 16, 2011, 05:45:25 AM
The only clear one is Popup > Rammer. Other than that...
Title: Re: A table for bot type efficiency
Post by: ty4er on October 16, 2011, 05:58:19 AM
pfft my rammer beat a popup [/brag]

Hows about hammer > SS then?
Title: Re: A table for bot type efficiency
Post by: Fracture on October 16, 2011, 08:34:04 AM
The only clear one is Popup > Rammer. Other than that...
Warmachine
Title: Re: A table for bot type efficiency
Post by: Pwnator on October 16, 2011, 08:36:34 AM
Warmachine

Specializing something will always be an exception.

:P
Title: Re: A table for bot type efficiency
Post by: Fracture on October 16, 2011, 08:42:54 AM
Longbow would be a specialized rammer. Warmachine is pretty typical.

Anyway RA2 isn't exactly comparable to Pokemon in that there are concrete matchups of certain types beating other types, so I'll say good luck with making this idea work.
Title: Re: A table for bot type efficiency
Post by: System32 on October 16, 2011, 08:52:27 AM
SnS > VS
SnS > Popups
SnS < Rammers
SnS = HS
 
 
HS > Rammers
HS < VS
HS = SnS
HS = Popups
 
VS > HS
VS < Popups
VS > Rammers
VS < SnS
 
Popups > Rammers
Popups > VS
Popups = HS
Popups < SnS
 
 
Title: Re: A table for bot type efficiency
Post by: Mr. AS on October 16, 2011, 09:14:07 AM
Stock or DSL?
Because saw-bots are a lot better in DSL than stock...
Title: Re: A table for bot type efficiency
Post by: Naryar on October 16, 2011, 09:17:03 AM
The only clear one is Popup > Rammer. Other than that...
Warmachine

That's because the popup had no defense. 90% of popups, at the HW level, have defense.

And one rammer win on a popup doesn't count, cause 95% of the time when a classic rammer faces a classic popup the popup wins anyway. One expection doesn't makes the statement "most of the time popups destroy rammers" false.

Edit : And yes, stock or DSL changes the odds. So we can't really make a simple table for bot efficiency ala Pokémon. Besides, RA2 is more complicated. Bot type isn't the same thing as a type in Pokémon.

Title: Re: A table for bot type efficiency
Post by: RpJk on October 16, 2011, 11:55:02 AM
The only clear one is Popup > Rammer. Other than that...
Warmachine

Can I see a pic of Warmachine?
Title: Re: A table for bot type efficiency
Post by: Sage on October 16, 2011, 12:04:30 PM
It depends on the freaking skill level of the bot builder. Generally HS rape everything and Popups rape everything else.
Title: Re: A table for bot type efficiency
Post by: Mecha on October 16, 2011, 12:21:05 PM
Can I see a pic of Warmachine?
Here you go.
(https://gametechmods.com/uploads/images/67149war machine.png)
he's the guy in the sky.
Title: Re: A table for bot type efficiency
Post by: Virus Bomb on October 16, 2011, 01:29:48 PM
funny thing about that pic is that the rammer actually isn't warmachine, but just a pretty damn accurate replica of it that i made.
Title: Re: A table for bot type efficiency
Post by: Badnik96 on October 16, 2011, 01:36:17 PM
how about we make a table. Like for a round-robin tourney or PKMN type matchups.

We pit a REALLY GOOD example of two different bot classes against each other (like Seism 13 vs AC for popup vs HS) and do it 10 times.

The robot that wins more times is better than the other. If they both win 5 times, then they are equal in terms of strength.

I'll volunteer to do this over Christmas break if this is a cool idea. (of course we'd have to do both stock and DSL ones)

also we need good examples of each bot type for each weight class (AW, BW, LW, MW, HW, and SHW). Might need some people to send some examples IN THE NAME OF SCIENCE
Title: Re: A table for bot type efficiency
Post by: Naryar on October 16, 2011, 02:04:18 PM
how about we make a table. Like for a round-robin tourney or PKMN type matchups.

We pit a REALLY GOOD example of two different bot classes against each other (like Seism 13 vs AC for popup vs HS) and do it 10 times.

The robot that wins more times is better than the other. If they both win 5 times, then they are equal in terms of strength.

I'll volunteer to do this over Christmas break if this is a cool idea. (of course we'd have to do both stock and DSL ones)

also we need good examples of each bot type for each weight class (AW, BW, LW, MW, HW, and SHW). Might need some people to send some examples IN THE NAME OF SCIENCE

No, one example of a bot is not enough. To be precise, we should just go and amuse ourselves to take stats from every single non-SFTW (cause a SFTW pretty much loses to any bot type) entry into every single tournament and challenge that has ever been done on Gametechmods, see how it won or lost against what bot type, and calculate probabilities out of this.

It'll take a heck of a long time, but it will be much more accurate than just one example of a tournament.
Title: Re: A table for bot type efficiency
Post by: That Robot is a Spy! on October 16, 2011, 03:42:55 PM
Or we could just get 10 of each bot type, each slightly different and pit them against each other, that would take less time.
All would have to be good though.
Title: Re: A table for bot type efficiency
Post by: Naryar on October 16, 2011, 04:09:25 PM
Actually I think my idea would take less time... we just have to observe old tournaments rather than making a new tournament just for stats.

And if we are a sufficient number and we're doing our job well, it will take less time.
Title: Re: A table for bot type efficiency
Post by: Squirrel_Monkey on October 16, 2011, 04:14:04 PM
FG>Everything
Title: Re: A table for bot type efficiency
Post by: Naryar on October 16, 2011, 04:19:03 PM
FG>Everything

Most certainly, my dear sir. Hammer ? VS ? Yes, of course.
Title: Re: A table for bot type efficiency
Post by: Squirrel_Monkey on October 16, 2011, 04:25:42 PM
FG>Everything

Most certainly, my dear sir. Hammer ? VS ? Yes, of course.
If build stable enough. I reckon one could be made at HW level.
Title: Re: A table for bot type efficiency
Post by: Somebody on October 16, 2011, 04:50:42 PM
What's FG?
Title: Re: A table for bot type efficiency
Post by: cephalopod on October 16, 2011, 05:01:46 PM
^Thought the same until I remembered...
Flying Guillotine.
Title: Re: A table for bot type efficiency
Post by: Badnik96 on October 16, 2011, 06:48:35 PM
what about a TRFBD?
Title: Re: A table for bot type efficiency
Post by: Urjak on October 16, 2011, 07:20:47 PM
FG>Everything


Except well built SnS and VS.
Title: Re: A table for bot type efficiency
Post by: russian roulette on October 16, 2011, 09:32:52 PM
I could see this working for amount of wheels on a certain bot type but other than that there are way to many variables for this to be completely accurate, good idea though.
Title: Re: A table for bot type efficiency
Post by: Badnik96 on October 16, 2011, 09:59:49 PM
agreed but I like Nary's idea.
Title: Re: A table for bot type efficiency
Post by: Clickbeetle on October 16, 2011, 11:32:32 PM
If we're making analogies to Pokemon, then popups are the Psychic type in gen 1.  Nothing resists them except other popups and there are no supereffective attacks against them.
 
HS are the Fighting type (gen 2 onwards).  Good coverage, supereffective against a lot of stuff, but weak to Psychic (popups).
 
Flying Guillotines and TRFBD's would be Dark and Steel types, respectively, since they resist popups and are more recent developments.  Also they are weak to HS (or at least weaker than they are to popups).
Title: Re: A table for bot type efficiency
Post by: Squirrel_Monkey on October 17, 2011, 01:01:48 AM
FG>Everything


Except well built SnS and VS.
Yeah, the problem with this discussion is the "well built" clause.
Title: Re: A table for bot type efficiency
Post by: RpJk on October 17, 2011, 01:12:47 AM
Actually I think my idea would take less time... we just have to observe old tournaments rather than making a new tournament just for stats.

And if we are a sufficient number and we're doing our job well, it will take less time.


Like the idea
Title: Re: A table for bot type efficiency
Post by: Virus Bomb on October 17, 2011, 10:29:01 AM
I'm surprised no one mentioned crawlers yet
Title: Re: A table for bot type efficiency
Post by: Pwnator on October 17, 2011, 10:32:21 AM
I'm surprised no one mentioned crawlers yet

I don't think there's a Pokemon relatable to that. :P
Title: Re: A table for bot type efficiency
Post by: Trovaner on October 17, 2011, 09:37:08 PM
Unlike Pokemon, the types in RA2 are not set in stone. The finite number of types in Pokemon are hard coded into the games and there are no exceptions to the strengths/weaknesses chart (unless looking at some glitches).

RA2 has nothing like this built-in. There are also going to be problems when sub-dividing bot's into specific types. We can't be too vague because of things like spinners (FS, HS, VS, FBS, drills, etc.) but we also can't be too specific (due to sheer numbers and a disproportionate scale). 

Builders usually judge bots by several different and varying factors including but not limited to type, efficiency, quality, and weight class. This combines both categorical and numerical sets of data that people weigh differently on a case-by-case. We would need to quantify the categorical data in order to even consider comparing the bots.

I usually look at these four points:
Offense (based on speed, type of force (linear or angular), direction (vertical or horizontal), concussion, and pierce)
Defense (based on HP and Fracture)
Mobility (based on number of wheels, motors, weight, and invertibility)
Design (based on wedge, combat style, opponent, and build quality) (highly opinionated)

Indiscriminately judging bots in a chart or even a formula would undoubtedly provide faulty information. I've attempted numerous times to do this but, without the human element, the results are never accurate.
Title: Re: A table for bot type efficiency
Post by: 123savethewhales on October 18, 2011, 02:34:00 AM
Lol this whole post makes no sense.

First of all, what arena are we even talking about?  Open?  Close?  Or one of those Joeblo ones?  Obviously in close arena Crawler is King.

Next it is weight class.  The balance changes greatly from LW MW to HW SHW, especially in DSL.

Third, histories of past tournaments bear little relevance.  Whatever any specific bot type used to be weak again in the long past doesn't matter if the design has since change to deal with the weakness.  So to say Flail Sheck is weak vs popup, or popup is weak vs Flail SnS, is utterly pointless by today's standard.  Historically they are true, but shell nowadays has way more than enough bottom armor to tank popups, and wide popup more or less render Flail SnS obsolete HW and up.