The problem with competitive IRL in RA2

[Reier]

This will undoubtedly be controversial but I needed to say it cause I feel like no one else has, or at least in the way that I feel about it. I know I’m not the best at explaining what I think or feel so please try to understand the spirit of what I say and not the letter.

------

In recent years GTM has had a huge shift from the formerly popular ‘DSL-S’/’Standard” style of building that reigned supreme from like 2005-2014 to a more ‘IRL’ style. Now at risk of sounding like some disillusioned geezer who hates change, let me preemptively say that properly built and designed IRL bots are some of the coolest looking designs that I’ve seen ever built in the game.

That’s the issue. They look cool. They suck at fighting.

Little history. Back when RA2 was released the line between ‘stock unrealistic’, ‘standard’ and ‘IRL’ was much more blurred and the terms hadn’t really been invented yet. Now I wasn't an active member until after the official forums died but I’ve done my fair share of looking at as many of the earliest RA2 posts as I could find.
The original RA2 bots were obviously the stock AI, which despite being built with the buggy RA2 engine which allowed parts to intersect with each other, were generally built to look as realistic (ie mimicking actual robots) as possible. But I’m hard pressed to find any showcased bot from ‘04-05 on the official forum which wasn’t stock unrealistic. They were just flat out better at fighting than IRL style because RA2’s engine is not designed in a way that building a robot that would work in real life will be the best in RA2. Why add one spike in the front when you can add 16? A bot with 16 spikes would suck in real life, but RA2 isnt real life. Stock was dominated by these glitch-exploiting, unrealistic, but extremely well engineered bots as a result. That’s basically what stock has remained since then. Extracting every ounce of efficiency possible out of the game engine as you can to make the best robot you can. That was the goal of builders. To make a good robot, and to make it cool if you could on top of that.

So then DSL 1.0 comes by in ‘05 I believe. Can’t say I played much of 1.0 but I remember it being released with all the replicas but none of the stock bot remakes that Click+FB did for 2.0. BBEANS2 and DSL 2.0 were released around the same time, meaning that the generally accepted “standard ruleset” had been around before that for DSL since both the BBEANS2 bots and the 2.0 remakes used the system. It’s a system that works, works well, and has stuck around for like 13 years for a reason. And one that I believe is the best system to build in for RA2 specifically because it has enough of the freedom of stock unrealistic but very little of the silly and abundant exploits that makes stock inaccessible and tedious. I think this is where IRL falls flat on its face.


‘RA2 IRL’ as is popular now is designed for making things look cool, not making things that can fight well. As I said, RA2 as a game is not designed in a way that bots that would work in real life would work well in the game. In RA2 there is a sliding scale of building from ‘efficiency’-to-‘realism’ with stock unrealistic on one end and DSL-IRL on the other. If you make a bot that uses every mechanic in the game to build the absolute most efficient fighting machine possible, it would look nothing like a DSL-IRL bot. (For those who don’t believe me, try adding 4x more spikes to your IRL bot and tell me with a straight face that it isn’t better at winning fights) And on the flip side, making the most realistic looking IRL bot by design isn’t gonna be as good at fighting as the first, since it doesn’t take advantage of mechanics in RA2... it takes advantage of mechanics from real world physics. RA2 doesn’t have real world physics. There is no fall damage. There is no structural integrity, internal or external. Damage and force is not divided by surface area. So on. DSL-S standard is just objectively a more efficient way to build if your focus is to win battles. Period. A DSL-S bot would annihilate an IRL bot every time without question if they are both built with the same level of quality.

It’s the whole root of the problem with building IRL bots “efficiently” (aka- to win) because it inherently is not efficient. To make it more efficient (eg weapon spam) you have to make it less IRL. So people try to make some subjective standard to how many weapons you can use, or how many weapon supports you need, or how many wedges you are allowed to have, or blah blah blahhhh. No no no! It is way too subjective, and if you think it isn’t, I just don’t know how to tell you the sky is actually blue.
 
This is only a problem with IRL tournaments. If you're building IRL to make a cool robot and nothing more, using RA2 as a simulator of sorts just so you can drive it, I’ve got no problem whatsoever with that. What I do have a problem with is people trying to play IRL to win tournaments in which I ask, what is the point? An IRL tournament is like the special olympics. You can watch two disabled bots duke it out (IRL), but why not just watch two professionals do it better (standard)? If you try to build an IRL bot that is good at fighting, it no longer is strict IRL, people get mad, and you’d be better off just building efficiently in Standard. If you build the most realistic thing you can, it’ll be total garbage at fighting due to the nature of RA2. So again, what is the point?   

Then you get reaaal nitpicky and subjective when you try to ban ‘problem’ things in tournaments. “don’t use weapon spam”. “limit of 2 bursts”. “weapons need to have believable supports”. At what point do we just stop using RA2 to build in if we aren’t really gonna build to RA2’s strengths? Just design stuff in CAD at that point. Wanna know what the rules are in Standard?
      1. Don’t make stuff intersect that couldn’t have a slot cut through it in real life (extremely objective)
      2. Don’t use external programs to edit stuff (totally objective)
      3. That’s it
Rules in Stock?
      1. Don’t use external programs to edit stuff (totally objective)
      2. literally again thats it
Rules for IRL?
      1. no one agrees on anything because you really can’t (subjective)
      2. make it look realistic (extremely subjective)
           2a. make sure it has a believable number and design of weapon supports (which are unnecessary in RA2, tying up weight in a pointless thing, and is subjective to the host if they are “realistic”)
           2b. Use a very specific but vague amount of weapons where it is just enough to be competitive but not enough to be “unrealistic” or “tryhard”, and pray that whoever is hosting will agree with you and not see you as leaning to heavily towards the latter (lean too heavily to the latter and you get kicked out, too heavily to the former and you just lose)
          2c. Don’t use too many motors to be unrealistic but enough to be competitive (same as above)
          2d. make the chassis look however you want including making it unrealistically small and unhittable as long as it looks realistic with tons of extenders (and totally not just to make other parts sacrificial and make the actual chassis hard to hit, trust me…)
      3. external programs are okay to use (CF, RO7B, BFE) despite the fact that they count as cheating in other mods due to their ease of exploitation
      4. don’t use cheat parts except sometimes (flatmotors, judge burst), even though using them in other rulesets is cheating due to their power

it’s all so silly.

It’s impossible to make a universally agreed ruleset for IRL because it is just a subjective way to build in the first place. And if you balance components buffing them to make them “““competitive””” for IRL then it make them horribly unbalanced for actual competitive modes like dsl-s where your only goal is to build the best bot you can with the limits of a single rule. and it doesn’t even fix the problem, cause you could just use two of the buffed weapon instead of one, again a subjective and arbitrary condition.


RA2 sucks as a game engine but its the best we got. Stock and standard rulesets both play to its strengths, as unrealistic as they may be. When we start adding all these silly arbitrary and VERY subjective rules like IRL wants to do it just gets real dumb real fast. What joy is there in winning an IRL tourney when you realize that you didn't build the best bot you could, you just built the bot with juuust enough weapons to win (meaning more weapons than anyone else) but not too many that you got kicked out by the host for being “tryhard”. That is sooo lame.
Why is it so taboo to be tryhard? To just build the best bot you can build with a universally agreed ruleset (eg stock or dsl-s)? Just build the best bot you can with stock or standard and no one can tell you that your bot is too ‘X’ because the rules are actually objective.

Man. I just want to be able to fight again. I want to build efficiently without people complaining about being a ‘tryhard’. It’s almost the entire point of the game. Why build if you don’t fight. And if you fight, why limit yourself to some bizarre subjective rules that no one agrees on but still all limit you to building something inefficient? ugh.

I get why people who were sick of the difficult stock rules and the efficiency-mentality of dsl-s and stock wanted a reprieve to just 'build what they want.' I do get it. But the pendulum swung way, way too far as a result.
 
I got no problem with people that build to be efficient. That’s what I do. And I got no problem with people that build to make something cool even if it sucks at fighting (the only thing IRL can do WELL). I DO have a problem with people that build something cool that can’t fight and expect the standards of every other bot to be lowered so they have a chance of winning. And that is competitive IRL.



Reply if you want, or dont. I’m not necessarily trying to change anyone’s mind. Just wanted to say what I think and why I thinking building IRL competitively is broken from the start.

[Badger]

I agree with a lot of what you've said, but I think you're missing the point of why IRL has dominated the community and why DSL-S and stock have basically died. To be blunt, nobody gives a sh** about RA2 as a game in iteself anymore, or at least nobody that could be vaguely described as 'new' or even 'moderately experienced'. People are coming into the community and playing RA2 not because they want to play RA2, but because they want to make/fight bots like they see on TV/YouTube and RA2 is the best way to do that right now. IRL is the dominant mode because people want to experience robot combat without actually having to make a robot. They want an official RW/BB game that's actually functional. Stock and DSL-S are about skill at the game, IRL is essentially a spectator sport. If you're into cars, it's like the difference between rally and gymhkana. One is about skill, and one is about spectacle.

If you want DSL-S to return (and I do as well), you need to make people excited by RA2 as a game, instead of RA2 as a means to an end. That's my view at least. Again, I agree with basically everything you said, but I think you're barking up the wrong tree insofar as you assume everyone is looking at RA2 as a competition or display of skill like you, which just isn't the case for the newer generation of members.

edit: a more concise way of saying what I mean: We think of RA2 as a game in itself to play, improve at and compete within. IRL builders think of RA2 as a way to make bots and fight them as a replacement for an official RW/BB game. The majority of the active building/hosting community have the second mindset, which is why DSL-S and Stock are dead.

[TheRoboteer]

You're absolutely right that IRL bots suck at fighting compared to DSL-S bots (or at least they should if people are actually sticking to what IRL is all about: making sh** that looks cool), but to say they can't provide entertaining fights is wrong IMO.

When they're fighting other IRL bots they can provide some great fights. I think a good analogy would be something like IRL is your regular run of the mill car while DSL-S bots are like a Formula 1 car. If you put them against each other the racing/fighting will be sh** because the DSL-S bot/F1 car completely outclasses the other, but as long as you only fight/race against cars/bots of the same type they both provide great spectacles.

I don't really see how subjectivity is an issue as long as folks are aware that it is inherently subjective. Every time I enter an IRL tournament I accept the possibility that my bot is gonna be rejected, or that the host's requirements for what is IRL are more lax than mine and that my bot will be wrecked. As long as all bots are held to the same standards then I really don't see the issue there.

This is also why I think the current effort to try to objectively balance components for an inherently subjective meta is, to be frank, dumb. People talk about nerfing the mech disc or typhoon teeth, or buffing bursts, but that doesn't work because as you say, everyone has a different definition of IRL. A far better solution is for hosts to just set rules based on what they think is IRL and for them to ACTUALLY ENFORCE those rules and reject sh** that doesn't follow them (which far too many hosts seem to be sh** scared of doing ATM) and for people entering said tournaments to simply build to the rules of said tournament.

I absolutely agree with what (I think) is the crux of this post though which is (I think) that if you're concerned about winning and you want to build efficient sh** you should be playing DSL-S rather than trying to crowbar 16 typhoon teeth onto an IRL bot and bank on hosts' refusal to reject sh** to win tournaments.

[pokebro14]

Your right. IRL is just not suitable for actual competitive use.
I make bots that are not efficient at winning I do this because the only way to win is to have a bunch of teeth or an overpowered weapon like mechavore. There is almost no way of solving this apart from delete IRL from existance.
I've never really got the appeal to make an OP bot as it's more fun to see a weird and unconventional designs do well in tournaments. It's why I enjoy the lower weight classes as aside from overhead spinners there isn't an actual meta for it and so people create some pretty neat stuff

[Dreamcast]

It's why I enjoy the lower weight classes as aside from overhead spinners there isn't an actual meta for it and so people create some pretty neat stuff

That's a meta. In fact, LW an lower IRL will be a worse than normal meta if those robots go on to win because it'll be composed of them and only them.

[freeziez]

we should all go back to 2.1 IRL am i right guys

[Hoppin]

Bajur brings up a good amount of points, in that IRL is an advocate for people who don't necessarily have the means to create their own combat robot, the IRL meta allows them to make something plausable and accessable, whilst being interactive and I think that's why Ra2 is more appealing to something like ARC, it has an intractability that gives defined results in matches.

Spectator sport is the best way of describing IRL, there is no real want for competitiveness, outside a few outliers. However, there will always be effective ways of doing stuff due to it being a game that is balances by few factors and a lack of changing meta. As touched on by TheRoboteer, the comparison between the competitive meta's is ridiculous and balancing should be done for each meta, not under the version of the game, this is why IRL tournaments have all these limitations, mainly due to having a wider range of designs and not allowing cheap ways to win that undermine the core or IRL. Finally, on this mess of points, it's taboo to by tryhard as by your own logic, tryharding doesn't create cool looking bots which, again, like you've said, isn't what IRL is.

Personally I think if you want to bring unrealistic metas back to life, I'd suggest, hosting more and get ways to make playing the game as a game interesting. With players new to the meta, glitches are such a overburden to learn, you have to know and somewhat master all of them to make something viable which I don't think is particularly fun.


(Only kix uses judge bursts)

[Scrap Daddy]

I agree with mostly everything that has been said here. If I only want to play for 10 minutes, I fire up some replicas or IRL bots and play pretend. And that is probably what a lot of new people want to do, too. My dream when I first found RA2 was to have every BB replica and redo the seasons in RA2, and see who would win. That's why a lot of my first bots were replicas when Sonny was basically the only one doing it, and at the time he was only building RW bots. I think it is true most people just want a functional RW or BB game, and RA2 really is the only option.

As a stock OG, I do understand why it's hard to get into the meta. The glitches are hard to explain, even with videos. The DSL meta essentially became Stock 2.0 once 123savethewhales broke the game. It is insanely frustrating to spend hours building a bot, legitimately using all the glitches, minimizing chassis space, only to find out it havoks or you have AGOD and have to scrap it. Now with the sergepatcher, why would't I at least see if the setup gonna work to save a few hours? Why even bother in the first place? I can tell you for a fact stock is at its precipice. We basically broke stock from the inside with Hax Mode, and there are only so many designs and setups still out there when there are only 10 usable components. I still like the challenge of stock of finding those one or two setups still out there, but I just don't personally see it for DSL/stock with graphics updates.

As for IRL, I found it to be a fun way to get out robot ideas that I would want to build IRL. I remember entering Battlebots 1 and half of the bots were DSL meta. Even back then I think we could all look at them and say they weren't in the spirit of IRL (Mr AS I'm looking at you), and that was enough back then. An unoriginal shell spinner with spike spam is certainly not in the spirit of displaying cool ideas. I don't think a weapon spammy bot ever won a tourny back then, and I think hosts do a pretty good job at keeping them out nowadays without any specific rules. The extender bots have taken this to a new level that I was not even at back then, so I can't really comment on the tiny chassis situation.

One idea I think could work, once all weapons are balanced properly, would be a percentage weigh limit that could be used. I'm not doing any math but lets say only 20% of the weight can be used towards weapons. Maybe it can delve a little deeper and say 33% including weapons and weapon motors. Obviously someone will have to figure out the specific numbers, but I think it would work assuming the balance is done properly.

[UberPyro]

I think I have an important perspective since I'm a relatively new player (joining officially in 2016) that only really cares about building competitive robots and watching them battle. I guess whether or not I'm experienced is subjective, since I have been lurking for ages and that might show in some of my better builds. I'm also am probably (read: definitely) the only user here that never watched any of the tv-airing robotic combat, stumbled upon this sorry game after googling for robot games, and kept on playing the game purely from a gameplay standpoint.

From a purely conceptual and tactical standpoint, I think the game played competitively is actually very good. There is a massive amount of freedom that comes with choosing from a plentiful number of components, being able to place baseplate components pixel perfect, and being able to attach 6 (or more...) components off of those with rotations as fine as 10 degrees. Glitching takes constructing robots to a whole 'nother level of complexity. Even though many components and many design concepts may be bad, there is still room for an endless amount of innovation, engineering, and re-engineering for competitive robots. The variety is truly excellent.

Unfortunately, RA2 in its current state is pretty terrible for bringing in new players and isn't really that appealing in general. This is pretty obvious but let me state some of the reasons:

1) The age shows. It's pretty graphically ugly and the physics engine is frustrating and can be pretty terrible to learn to get around.

2) Most people don't have the patience to sit down and build a competitive robot. Despite me calling the game "actually very good" due to its freedom and expression, building can be painful and boring. Making a skin tight chassis, planning how the baseplate components will be organized, using math (screams!) to know how many weapons to add and of what type, trying to get all those perfectly placed positions. A serious robot takes upwards of 2 hours, by an experienced user, on the first iteration. I sometimes wince in pain when looking through some of the IF Reier AI robots because of how tedious some of the builds must have been.

3) The game's learning curve is sadistic. Especially because glitch knowledge is pretty necessary to have full control of the standard and unrealistic playstyles. Think about how weird this is from an outside player's perspective: some super-tedious ritual-like glitches are preformed in the game, usually to get a desired position on a few components, even sometimes for a dsl-s outcome. Knowledge of how to abuse the old havok engine is needed to build robots in a game that's really trying to simulate the old robot wars or whatever. It's actually really bizarre.

Let me attempt to bring this back on topic somehow: IRL building is less affected by these issues. With RA2CF, #3 is taken care of because extenderbots can easily be built after removing the placement collision meshes. Additionally, where in DSL-S you are allowed to intersect as long as you can cut holes in things without the use of CF (which is kind of undetectable...), the IRL style usually allows this. In short, IRL building does not require glitch knowledge like the other building styles. Additionally, players aren't building unrealistic abominations of what they see on TV.

[removed a paragraph for being wrong]

For #1, this might sound ironic but I think the graphics hurt the IRL style less. This is because, in IRL, the components because art tools and the actual look of the robot is deliberate and controlled. You don't need glitter crans to make good art.

Really I spent all that time to say IRL is much easier for newcomers to get into. I guess it's also relatively fresh and fun to see what others built, I don't really know.

I think a lot of it comes down to this:

If people are more interested in building robots that actually watching them fight (which is true even for myself), then I guess it's more natural to have a building-focused meta rather than a battle-focused one. In other words, it's easy to care more about showcasing your robots and the look of an IRL tournament than really caring about winning. If you care about winning an IRL tour, well, I can't really help you.

Particularly that bit on showcasing. Well-built IRL bots generally get a really good response from the community, which creates sort of a motivation loop as people both build and comment. On the other hand, Reier is the only person to comment on any of my IF (standard) builds in recent times. It feels super dead to me. Well, I guess I'm stuck with what I got .

With all that said, I'm not really sure how to "bring back DSL-S." There has been a small resurgence recently, but that's really just a rebound of some of the older players asking "where the DSL-S at tho" and I don't think it will be sustained. I don't think most newcomers would really be interested, we would really be doing this for ourselves. I guess just encouraging more standard-style tours is a good way to start, and we could judge popularity by signups. Or try to do something with showcases. I don't really know.

Of course, I have dreams of a game that fixes all of RA2's problems and goes beyond, but we're stuck with what we got for now. Hopefully the Robot Rumble game that is being developed goes somewhere. Sometime in the future I might try to make my own version or help out with another.

Well, I realize this was pretty unfocused and incomprehensible, but maybe there were some ideas you found interesting.

[superbot13]

I think tournament hosts should allow what they want and what they think is IRL and then people can enter if they agree with it. And if they don't they don't have to enter

[09090901]

I think tournament hosts should allow what they want and what they think is IRL and then people can enter if they agree with it. And if they don't they don't have to enter

This, but in actual english.

Also anyone who says IRL is easier or takes less time than stock or standard has never built a decent looking IRL bot in today's standards.

[Billy5545]

As an IRL builder, I say that the reason why I got into IRL is because first, IRL bots look good, realistic, and rational, unlike the ugly, unrealistic, and crazy bots of DSL-S and Stock. Second, I want to build bots like I see in real life. And third, as shown in Master of Disaster 2, I think I only build DSL-S or Stock, or Ironforge if there is a tournament about them, and even that, I don't have much interest in stock and Ironforge. But for the MoD 2 tournament, what I built was actually IRL, just that it looked DSL-S because at that time, I didn't know about things like Sergepatcher or the specific different build styles.

[yugitom]

Also anyone who says IRL is easier or takes less time than stock or standard has never built a decent looking IRL bot in today's standards.

Making the best and most efficient stock bot you can possibly make is much more excruciating than IRL building, due to the amount of practice and learning you have to do. That isn't to say that IRL building doesn't take ages or require skill, stock is just much more difficult and complex to build in, if you're aiming to build the best bot you possibly can that will win tournaments. IRL building is relaxed with rules to allow people to build the best looking thing they can as easily as possible. This makes building IRL bots much more fun, even if it may end up taking just as long as it would to build an amazing stock bot.

On a related note, I actually love stock, despite hosting and being interested in mostly IRL stuff before. I like DLS-S less because it just seems like a watered-down version of stock, as a spectator. Both strive for maximum efficiency and the like, just DSL-S does it without as many glitches. I will admit that DSL-S is way more accessible because of that and I definitely don't think it's a bad metagame. I'm really enjoying seeing stock tournament fights nowadays, watching the best designed and best built bots succeeding. Stock has actually refreshed some interest I've had in the game because I, as was well-put by Badger, used to view RA2 mostly as the best platform for making the best looking robots and actually have them fight each other, something CAD nor any robot combat game could do better. I hope more people get (back) into stock, as it's one of the greatest competitive formats in a game that I've seen. My new view on stock is probably due to playing Garry's Mod for so long since I first got into RA2. It made me appreciate the limits RA2 has as a game. If you're gonna start breaking the rule of 7 and stuff, you may as well shift platform to something like GMod that can achieve much better visuals and much better fights, with the same amount of restrictions on competitive play.

[09090901]

Also anyone who says IRL is easier or takes less time than stock or standard has never built a decent looking IRL bot in today's standards.

Making the best and most efficient stock bot you can possibly make is much more excruciating than IRL building, due to the amount of practice and learning you have to do. That isn't to say that IRL building doesn't take ages or require skill, stock is just much more difficult and complex to build in, if you're aiming to build the best bot you possibly can that will win tournaments. IRL building is relaxed with rules to allow people to build the best looking thing they can as easily as possible. This makes building IRL bots much more fun, even if it may end up taking just as long as it would to build an amazing stock bot.

Except nobody builds stock legit anymore, and even then we already know what bot types are the best. All the hard work for stock was done years ago

[yugitom]

Also anyone who says IRL is easier or takes less time than stock or standard has never built a decent looking IRL bot in today's standards.

Making the best and most efficient stock bot you can possibly make is much more excruciating than IRL building, due to the amount of practice and learning you have to do. That isn't to say that IRL building doesn't take ages or require skill, stock is just much more difficult and complex to build in, if you're aiming to build the best bot you possibly can that will win tournaments. IRL building is relaxed with rules to allow people to build the best looking thing they can as easily as possible. This makes building IRL bots much more fun, even if it may end up taking just as long as it would to build an amazing stock bot.

Except nobody builds stock legit anymore, and even then we already know what bot types are the best. All the hard work for stock was done years ago

I'm just saying that is the case if you do build stock 'legit', which most tournament hosts will enforce/encourage. And I can't argue that all the hard work for stock hasn't already been done but it doesn't change the fact that any given newcomer would probably take much longer getting the hang of stock, without giving up on it.

[Thyrus]

I'm not reading all of this but I agree with badger and dosagree with scrap on the idea of a robot only being allowed to use a certain percentage of weight for a weapon as shellspinners would break that rule quite quick imo

[FOTEPX]

I'd be behind the whole "building IRL solely for spectacle" idea if building IRL wasn't such a ballache. Why would I bother building something superfluous that takes 10x longer to build and is, at best, maybe a quarter as effective, when I could just build something in less than an hour that would keep the crap out of any IRL bot I know of? This is why I support DSL-S over DSL-IRL; it's the fundamental bedrock which all the DSL-IRL bots are built off of.

It's the main problem I've had since starting Complete Edition, really, and the only way to solve it in a way that makes both parties happy is to make two separate mods. Balance all the motors and weapons to DSL-S standards, people'll complain the motors aren't realistic enough. Balance all the motors and weapons to IRL standards and the DSL-S meta pretty much becomes Ironforge. It's a no-win situation, because I simply don't have the time required for the upkeep of two separate branches of DSL. So, if I had to pick one...

DSL-IRL will adapt and change to get the most out of the DSL-S toolbox. Just look at how DSL-IRL changed from 2.1 to 2.2, and then from 2.2 to using DSC's multicoloured components. It's a symbiotic, constantly evolving process. For DSL-IRL to keep evolving, DSL-S needs to be placed first because it's what everyone's using to build DSL-IRL. There is no DSL-IRL pack out there - the last one that attempted that was Avalanche's attempt, and we all know how spectacularly that went down. Hate to break it to you, DSL-IRL fans; you're in our yard, and if I needed a better example of this, TheRoboteer summed it up best-

This is also why I think the current effort to try to objectively balance components for an inherently subjective meta is, to be frank, dumb. People talk about nerfing the mech disc or typhoon teeth, or buffing bursts, but that doesn't work because as you say, everyone has a different definition of IRL.

This is a fine example of the confusion people find themselves in when trying to balance DSL-S to DSL-IRL standards. In DSL-S, there is no debate that Typhoon Teeth, Mechavore Discs and Minion Discs are overpowered, but in DSL-IRL it's allowed to slide, granted that the builder doesn't exploit it's OP-ness. So, let's theorize, and say that DSL 2.3 nerfs those 3 criminal weapons. What'd happen then to DSL-IRL? Answer: Actually very little. You might see an uptick in those weapons being used, but overall, the meta would shift very little. Even if it was a great shift, DSL-IRL would still stay DSL-IRL. You cannot break that - where there's a will, there's a way. People've been doing replicas since the Stock days, for god's sakes. If people can manage in as limited a toolbox as that, I can guarantee you that balancing 3 components isn't going to ruin everything for you.

TL;DR - Play the game however you want. Just keep in mind what allowed you to play the game the way you do in the first place.

[8bean]

Competitive metas imo are daunting to a new player. When I discovered GTM (2010 ish), the first thing I noticed was the competitive stock and DSL bots. Even if you google search Robot Arena 2 today you'll still find crazy stuff built by the beetlebros hanging around. When I compared these complete powerhouse machines to the lava textured box with 6 maces on it that I considered the best thing ever, it pushed me away. This is where IRL comes in.

DSL IRL is a showcase of a builder's creativity and imagination rather than a competition. It has a much more gradual learning curve that is very friendly to new users. It also doesn't require an extensive knowledge of glitches to perform due to CF and allows for easy correction of mistakes through just deleting and replacing a few components rather than having to rebuild the entire thing from the ground up. I certainly feel that this is what has brought a resurgence of new members into the community recently, including myself. This doesn't mean that DSL IRL is a n00b paradise however.

Like geese said, today's IRL standards are very high and building one takes a large amount of skill as well. Trying to fit all of the electronic, drive, and weapon systems into a bot while using only the attachment points on extenders can take some time. Along with armoring the bot and making it look nice requires even more skill and patience. This is what has kept the experienced community on DSL IRL for such an extended period of time now. While a Stock, IF, or DSL-S bot may have been an absolute nightmare to build, the only way it is worth it is if it can win. IRL on the other hand pays off once it's finished with how clean/cool/original it looks. The element of competition is not whether or not you've got the lowest wedge or the most iron spikes but rather how much everyone likes your bot which is usually affected by how much time and skill you put into it. IRL though is very loose and open which in turn makes it less competitive in general.

RA2 was made to be competitive, and ever since the beginning of this game people have constantly been trying to outdo each other. This cycle of competition encouraged more gameplay and more out of the box thinking as a result. Glitches started to arise and players flourished upon them. Bots became so vicious that angling a wedge a whole 10 degrees higher could mean the difference between winning the tournament or losing the first round. In this truly competitive state, being the champ of a tourney actually meant something. It proved that you had the superior strategy and would in turn bring a whole new wave of bots to counter that and evolve the game further. Unfortunately, we are now stuck in a state where so few competitive tourneys are held that the incentive to build better is almost nonexistent. Why spend a huge chunk of time building and polishing a competitive design when all eyes are on the IRL tourneys instead. This leads to my final point and opinion on this matter.

DSL IRL is an important meta for the survival of GTM and RA2 as a whole. It creates an even playing field for all parties that is easy to sink your teeth into but hard to master. On the other hand, competitive metas create a divide which is incentive to do better and be better as a builder as well as truly push the game to its limits at the expense of a brutally difficult learning curve. These metas are even more important to the longevity of the site and the game because they constantly evolve it. Unfortunately, these two play styles will not mix which has shown in the overwhelmingly large IRL playerbase compared to the small competitive playerbase. I feel that in order to bring a resurgence of competitiveness and balance out the sides, there has to be an incentive for people to build again. One of the three metas would have to have a frequently running popular tournament. Something like a GTMCS or a BBEANS that one truly strives to achieve.

so yea thats my take that probably could've been written better

jesus christ bean I've already read like 6 goddamn books on this thread alone! just get to the point already.

[Ra2Winner999]

You're absolutely right that IRL bots suck at fighting compared to DSL-S bots (or at least they should if people are actually sticking to what IRL is all about: making sh** that looks cool), but to say they can't provide entertaining fights is wrong IMO.

When they're fighting other IRL bots they can provide some great fights. I think a good analogy would be something like IRL is your regular run of the mill car while DSL-S bots are like a Formula 1 car. If you put them against each other the racing/fighting will be sh** because the DSL-S bot/F1 car completely outclasses the other, but as long as you only fight/race against cars/bots of the same type they both provide great spectacles.

I don't really see how subjectivity is an issue as long as folks are aware that it is inherently subjective. Every time I enter an IRL tournament I accept the possibility that my bot is gonna be rejected, or that the host's requirements for what is IRL are more lax than mine and that my bot will be wrecked. As long as all bots are held to the same standards then I really don't see the issue there.

This is also why I think the current effort to try to objectively balance components for an inherently subjective meta is, to be frank, dumb. People talk about nerfing the mech disc or typhoon teeth, or buffing bursts, but that doesn't work because as you say, everyone has a different definition of IRL. A far better solution is for hosts to just set rules based on what they think is IRL and for them to ACTUALLY ENFORCE those rules and reject sh** that doesn't follow them (which far too many hosts seem to be sh** scared of doing ATM) and for people entering said tournaments to simply build to the rules of said tournament.

I absolutely agree with what (I think) is the crux of this post though which is (I think) that if you're concerned about winning and you want to build efficient sh** you should be playing DSL-S rather than trying to crowbar 16 typhoon teeth onto an IRL bot and bank on hosts' refusal to reject sh** to win tournaments.

I kill dsl s bots all the time with gyroscope blade, Cw-The matrix Hammerblade Revamped, Unintentional Tombklone.

[Dreamcast]

The "Sub-par DSL" mentality of yesteryear is not the driving force behind people playing IRL. RA2 players gravitate to what is comfortable and relatable, and that is rarely the veterans' brand of RA2.

When a lurker is from early 2007 and they see some robot from "Luigica" that has a simple yet interesting setup, the lurker's mind starts thinking. Ideas like "Can I build that?" or, "I've seen other parts, can I mix and match them?" go through the lurker's head, and they are now interested in vanilla RA2. The lurker downloads a video file of a competitive match and goes, "Wow! That robot got pwnd, Mako is cool." and they're now interested in DSL-S.

But today is not 2007. It's harder to go from "Big robot combat fan" to "RA2 fan." The standard has slowly gone up, the complexity for those who don't cheat has slowly gone up. In addition, resources to learn about RA2's mechanics or its history are weak, causing more of a disconnect between old and new. Naturally, after these people declare Stock and DSL-S, "confusing hogwash," these people will start to use IRL as a competitive meta.