Metagame Analysis
Posted: Mon Feb 07, 2011 11:37 pm
I am a big waste of time and space and a drain on society and I hate myself and no one cares so I wrote this
Since this will be the last season of ARC, I think it would be enlightening to take a look at the evolution of the metagame over the years. Through around ten years, thirteen seasons, there has been a wealth of controversy, changes, and development. We'll be taking a look at the nature of the tournament itself, the ruleset, and finally the evolution of designs.
The best place to start would be with the tournament format. Originally CFRC was single elimination, before making the shift to double elimination in ARC: Defiance, and finally the round robin nature of ARC: Revolution. A change that can honestly be said is objectively better. Single elimination has its upsides, most notably that you can host a lot of tournaments in a short amount of time. Unfortunately, whenever something gets over done or overused, it becomes cheapened. One of the reasons why we have such a long break between ARC seasons is to keep the whole competition fresh and excited. Why would anyone really care about losing in a short, single elimination tournament? There will be a new one hosted soon after that. But if you don't care about losing, then you can't really care about winning. This is to say nothing of bracket format, which is an inherently biased system. An untimely forfeit, a bad match up, an unlucky self-knockout, all things that could send you packing in the first round. How can any design truly be called proven if they only end up fighting once? Celestial Slayer is a great example of this, as the first CFRC it won a quick forfeit, before losing to Andrewbot Jr. in a split 23-22 judges' decision. Now, we all recognize Celestial Slayer as a proven design today, but only after being given the chance to win in CFRC and Defiance. Even those tournaments, however, it required some lucky forfeits and knockouts to go its way before winning. In Vengeance, a proven design like Celestial Slayer went three straight 23-22 judges' decisions on its way out of the tournament. What if Celestial Slayer was on the other side of the bracket and got an easy draw to the finals? He would have won because of an easy draw and his title would be marred.
It is for this reason that one could easily say that A-Bomb is the worst robot to win a championship. He won an easy knockout against Meowch, before sneaking passed Adrenaline with a forfeit (a match A-Bomb could have easily lost). He won two more legitimate matches before actually losing with a forfeit against The Box!. He somehow managed to beat Slasher, perhaps his only actual test against a fairly decent robot, before finally winning the entire tournament with two straight forfeits against The Box!. He had faced The Box! three times, and yet never actually fought him which is incredible.
And yet, it is for some of these reasons that knockout brackets are still used. In the round robin tournament, even if you lose you will still have a match next week. In a knockout, there's no guarantee that there is a next week and so each match is vitally important. It is why the NFL use that system for their playoffs, and why the FIFA World Cup has it for theirs. And why it makes perfect sense to put the system at the end of each contemporary ARC season. After ten weeks, we have an idea of who the best designs are, and to win enough matches to get into the post-season, you have to have been role-playing. We have had eighty-two post-season matches over the past five seasons, and only four of those have been forfeits with absolutely none in the finals. This format ensures that something like A-Bomb vs. The Box! will never happen again. You get the combination of the matches having weight in the post season, and the ability for each design to have a chance to prove themselves.
From here it's a good time to look at role-playing, and results. As a writer, I have always looked at three things: Design, Stats, and RP. A beautiful RP will not save a one-armored robot that only has two very large exposed wheels and nothing else. Meanwhile, an RP saying that you are going to drive into a wall will not save Underall. A match between two full-bodied spinners will most likely not come down to RPs due to the nature of design, while a match between Barrier and Shockwave is going to be very RP dependent. One thing that I've always noted in my years of watching robot combat like some Big Dumb Nerd With Nothing Better To Do is that strategy can account for notably little in a match. You don't know when someone is going to overturn, or when you just manage to slide off their wedge, or when a spinner breaks down, or when the shit hits the fan. There are no step-by-step tactics that will allow you to prepare for anything. I believe I was watching a raw edit of Warhead vs. Darkstar. Darkstar's driver was thinking that he'd probably have to avoid the disc and attack from the rear. Of course, Darkstar got the shit kicked out of him. But it wasn't a bad strategy, and in fact it was really the only strategy he had available for him.
I read a lot of RP's like this thought. In a perfect vacuum, Yes! You would be able to get around to the sides and flip him. However, in a match both robots are moving around and it is inherently chaotic in nature. Now, I don't want to discourage anyone from getting specific where it's necessary, but I feel like a lot of complaints for results come from the fact that someone's RP wasn't read as supported by the fact that strategy X was not executed. Let me assuage your fears by saying that every writer reads every RP for every result they write. I know this for a fact. Secondly, the results are a means to an end. They are usually summaries of three-minute fights. Chris, Kody, Josh, Philip, and I are not going to write every little boring thing that happens within a three minute period. The highlights are usually what gets written, key points in the fight that swung a decision one way or another, or the lead up to a knockout. In the bigger, closer, or more interesting fights, more detail will be used as necessary. But the outcome for that weeks fight will always be the same as decided by a writer, and whether or not you read strategy X in the result specifically has no bearing. More likely than not, you didn't see strategy X in the fight because the writer felt like it wouldn't have a big enough impact on the fight to swing the result in your favor. THIS IS NOT TO SAY RPs DO NOT MATTER. THEY MATTER VERY MUCH. Countless matches I write are based on RPs as the deciding factor. But it's just as important to recognize that the viability of your strategy is the single most important part of your RP.
Now onto the rules. I have to say our ruleset is the single most complicated thing on the site, through no one's fault however. A lot of numbers, specifics, and bonuses are used in places where it's really supposed to be more of a light guide than anything else. To his credit, John wrote a very nice ruleset and we implemented it before anything was really edited and set in stone. We should have simplified it, instead of me having to constantly remind people that their five point hammer would not actually hit with a force of seven points. So it's easy to see how something like this can confuse new comers, and is the single thing that needs to be changed to a new season of whatever else. The only specifics we really needed for this sport were multi-bots, multiple weapons, self-knockout and opposing knockout. Things like power hammers receiving bonuses doesn't really make a lot of sense because the whole concept of putting points into weapon for a hammer makes it, by nature a 'power hammer'. It will hit hard a slow, as opposed to a one point hammer which will hit weak and fast. Gravity bonuses don't make much sense either, considering the dropping height is so low, and the counter-acting friction. But all of this is really superfluous and not the meat of my concerns with the ruleset.
Dylan devised a great stat system that managed to survive for an impressive amount of time. When I first joined the league I could easily see the simplicity of the system. Four categories, but twenty-two points. This meant that the average, middle of the road point allocation would be between five and six. This has the implication of always having at least one stat that was below average, and always having at least one stat above average. It's a great game styled system, but not entirely realistic. Megabyte isn't that slow, doesn't have that bad traction, and it has fairly decent armor and an extremely hard hitting weapon. So what do we do? Do we have a system that relies on balance, tradeoffs, and picking a specialty? Or do we have a realistic system where everyone is a hard hitting, fast moving, controlled, well armored beast?
I suppose there is the ability to compromise. I like the idea of picking from a list of specialities, that was prevalent in a couple of leagues I played in (I believe Nick hosted them). Pick two or three of Speed, Control, Weapon, Internal Stability, External Armor, Pushing Power, etc. Exclusively do not punish multiple weapon users, scri-mechs, and jack of all trade designs. Use the tournament director's reason when accepting multiple weapon designs so that one does not get a robot with a flipper, a disc, an axe, a clamp, and a saw (ie: no designs that would flourish in Robot Arena 1). Something like Flank Attack would not be able to win a current ARC tournament because to split the stats I would not be able to have the required pushing power, armor, and speed to win fights. Something like Tank or AMP would do well though, because of the simple design that leans on doing a single thing really well.
Hell, let's look at the top ten robots with the most wins (note that this list is a little old). Underall, AMP, Eurypterus, Truth or Consequences, Shockwave, Barrier, Tank, Double Dose, Copperhead, and Zombie Killer. What do these robots have in common? Other than Shockwave, they are all specialized with a disproportionate amount of rammers. Underall (I consider him a rammer more than a thwack, and I personally believe his more sound design was as a rammer), Eurypterus, Tank, Copperhead, and Zombie Killer. I should even consider Barrier because, hell, when considering design in terms of stats, lifters may as well be rammers. They show all the symptoms of rammers, high armor, high speed, high torque/traction, and a non-existent weapon. Now, they require a more controlled strategy than rammers, but they have the ability to be statted like rammers while maintaining the utility of the weapon. I am of the belief that lifters are currently the most overpowered design in the league. If there is a design that could have taken down Underall in its prime, it would have been a lifter. But the lifter's main ability comes from an evolution that probably changed ARC more than anything: the wedge.
The wedge is such a beautiful and perfect weapon. If you are underneath your opponent, they cannot do anything to you. Much like tackling in rugby or American football, if you tackle their legs you remove their ability to run. If they can't move, they can't do shit. And so wedges have evolved much in the same way as an arms race. Who has the lower wedge is almost certain to be a deciding factor in a fight where RPs and stats are equal. Gila, Reflector Shield, and a couple other old-school robots might not pack the same impact in today
Since this will be the last season of ARC, I think it would be enlightening to take a look at the evolution of the metagame over the years. Through around ten years, thirteen seasons, there has been a wealth of controversy, changes, and development. We'll be taking a look at the nature of the tournament itself, the ruleset, and finally the evolution of designs.
The best place to start would be with the tournament format. Originally CFRC was single elimination, before making the shift to double elimination in ARC: Defiance, and finally the round robin nature of ARC: Revolution. A change that can honestly be said is objectively better. Single elimination has its upsides, most notably that you can host a lot of tournaments in a short amount of time. Unfortunately, whenever something gets over done or overused, it becomes cheapened. One of the reasons why we have such a long break between ARC seasons is to keep the whole competition fresh and excited. Why would anyone really care about losing in a short, single elimination tournament? There will be a new one hosted soon after that. But if you don't care about losing, then you can't really care about winning. This is to say nothing of bracket format, which is an inherently biased system. An untimely forfeit, a bad match up, an unlucky self-knockout, all things that could send you packing in the first round. How can any design truly be called proven if they only end up fighting once? Celestial Slayer is a great example of this, as the first CFRC it won a quick forfeit, before losing to Andrewbot Jr. in a split 23-22 judges' decision. Now, we all recognize Celestial Slayer as a proven design today, but only after being given the chance to win in CFRC and Defiance. Even those tournaments, however, it required some lucky forfeits and knockouts to go its way before winning. In Vengeance, a proven design like Celestial Slayer went three straight 23-22 judges' decisions on its way out of the tournament. What if Celestial Slayer was on the other side of the bracket and got an easy draw to the finals? He would have won because of an easy draw and his title would be marred.
It is for this reason that one could easily say that A-Bomb is the worst robot to win a championship. He won an easy knockout against Meowch, before sneaking passed Adrenaline with a forfeit (a match A-Bomb could have easily lost). He won two more legitimate matches before actually losing with a forfeit against The Box!. He somehow managed to beat Slasher, perhaps his only actual test against a fairly decent robot, before finally winning the entire tournament with two straight forfeits against The Box!. He had faced The Box! three times, and yet never actually fought him which is incredible.
And yet, it is for some of these reasons that knockout brackets are still used. In the round robin tournament, even if you lose you will still have a match next week. In a knockout, there's no guarantee that there is a next week and so each match is vitally important. It is why the NFL use that system for their playoffs, and why the FIFA World Cup has it for theirs. And why it makes perfect sense to put the system at the end of each contemporary ARC season. After ten weeks, we have an idea of who the best designs are, and to win enough matches to get into the post-season, you have to have been role-playing. We have had eighty-two post-season matches over the past five seasons, and only four of those have been forfeits with absolutely none in the finals. This format ensures that something like A-Bomb vs. The Box! will never happen again. You get the combination of the matches having weight in the post season, and the ability for each design to have a chance to prove themselves.
From here it's a good time to look at role-playing, and results. As a writer, I have always looked at three things: Design, Stats, and RP. A beautiful RP will not save a one-armored robot that only has two very large exposed wheels and nothing else. Meanwhile, an RP saying that you are going to drive into a wall will not save Underall. A match between two full-bodied spinners will most likely not come down to RPs due to the nature of design, while a match between Barrier and Shockwave is going to be very RP dependent. One thing that I've always noted in my years of watching robot combat like some Big Dumb Nerd With Nothing Better To Do is that strategy can account for notably little in a match. You don't know when someone is going to overturn, or when you just manage to slide off their wedge, or when a spinner breaks down, or when the shit hits the fan. There are no step-by-step tactics that will allow you to prepare for anything. I believe I was watching a raw edit of Warhead vs. Darkstar. Darkstar's driver was thinking that he'd probably have to avoid the disc and attack from the rear. Of course, Darkstar got the shit kicked out of him. But it wasn't a bad strategy, and in fact it was really the only strategy he had available for him.
I read a lot of RP's like this thought. In a perfect vacuum, Yes! You would be able to get around to the sides and flip him. However, in a match both robots are moving around and it is inherently chaotic in nature. Now, I don't want to discourage anyone from getting specific where it's necessary, but I feel like a lot of complaints for results come from the fact that someone's RP wasn't read as supported by the fact that strategy X was not executed. Let me assuage your fears by saying that every writer reads every RP for every result they write. I know this for a fact. Secondly, the results are a means to an end. They are usually summaries of three-minute fights. Chris, Kody, Josh, Philip, and I are not going to write every little boring thing that happens within a three minute period. The highlights are usually what gets written, key points in the fight that swung a decision one way or another, or the lead up to a knockout. In the bigger, closer, or more interesting fights, more detail will be used as necessary. But the outcome for that weeks fight will always be the same as decided by a writer, and whether or not you read strategy X in the result specifically has no bearing. More likely than not, you didn't see strategy X in the fight because the writer felt like it wouldn't have a big enough impact on the fight to swing the result in your favor. THIS IS NOT TO SAY RPs DO NOT MATTER. THEY MATTER VERY MUCH. Countless matches I write are based on RPs as the deciding factor. But it's just as important to recognize that the viability of your strategy is the single most important part of your RP.
Now onto the rules. I have to say our ruleset is the single most complicated thing on the site, through no one's fault however. A lot of numbers, specifics, and bonuses are used in places where it's really supposed to be more of a light guide than anything else. To his credit, John wrote a very nice ruleset and we implemented it before anything was really edited and set in stone. We should have simplified it, instead of me having to constantly remind people that their five point hammer would not actually hit with a force of seven points. So it's easy to see how something like this can confuse new comers, and is the single thing that needs to be changed to a new season of whatever else. The only specifics we really needed for this sport were multi-bots, multiple weapons, self-knockout and opposing knockout. Things like power hammers receiving bonuses doesn't really make a lot of sense because the whole concept of putting points into weapon for a hammer makes it, by nature a 'power hammer'. It will hit hard a slow, as opposed to a one point hammer which will hit weak and fast. Gravity bonuses don't make much sense either, considering the dropping height is so low, and the counter-acting friction. But all of this is really superfluous and not the meat of my concerns with the ruleset.
Dylan devised a great stat system that managed to survive for an impressive amount of time. When I first joined the league I could easily see the simplicity of the system. Four categories, but twenty-two points. This meant that the average, middle of the road point allocation would be between five and six. This has the implication of always having at least one stat that was below average, and always having at least one stat above average. It's a great game styled system, but not entirely realistic. Megabyte isn't that slow, doesn't have that bad traction, and it has fairly decent armor and an extremely hard hitting weapon. So what do we do? Do we have a system that relies on balance, tradeoffs, and picking a specialty? Or do we have a realistic system where everyone is a hard hitting, fast moving, controlled, well armored beast?
I suppose there is the ability to compromise. I like the idea of picking from a list of specialities, that was prevalent in a couple of leagues I played in (I believe Nick hosted them). Pick two or three of Speed, Control, Weapon, Internal Stability, External Armor, Pushing Power, etc. Exclusively do not punish multiple weapon users, scri-mechs, and jack of all trade designs. Use the tournament director's reason when accepting multiple weapon designs so that one does not get a robot with a flipper, a disc, an axe, a clamp, and a saw (ie: no designs that would flourish in Robot Arena 1). Something like Flank Attack would not be able to win a current ARC tournament because to split the stats I would not be able to have the required pushing power, armor, and speed to win fights. Something like Tank or AMP would do well though, because of the simple design that leans on doing a single thing really well.
Hell, let's look at the top ten robots with the most wins (note that this list is a little old). Underall, AMP, Eurypterus, Truth or Consequences, Shockwave, Barrier, Tank, Double Dose, Copperhead, and Zombie Killer. What do these robots have in common? Other than Shockwave, they are all specialized with a disproportionate amount of rammers. Underall (I consider him a rammer more than a thwack, and I personally believe his more sound design was as a rammer), Eurypterus, Tank, Copperhead, and Zombie Killer. I should even consider Barrier because, hell, when considering design in terms of stats, lifters may as well be rammers. They show all the symptoms of rammers, high armor, high speed, high torque/traction, and a non-existent weapon. Now, they require a more controlled strategy than rammers, but they have the ability to be statted like rammers while maintaining the utility of the weapon. I am of the belief that lifters are currently the most overpowered design in the league. If there is a design that could have taken down Underall in its prime, it would have been a lifter. But the lifter's main ability comes from an evolution that probably changed ARC more than anything: the wedge.
The wedge is such a beautiful and perfect weapon. If you are underneath your opponent, they cannot do anything to you. Much like tackling in rugby or American football, if you tackle their legs you remove their ability to run. If they can't move, they can't do shit. And so wedges have evolved much in the same way as an arms race. Who has the lower wedge is almost certain to be a deciding factor in a fight where RPs and stats are equal. Gila, Reflector Shield, and a couple other old-school robots might not pack the same impact in today