This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Messages - UberPyro
Pages: 1 2 [3] 4 5 6 7 8 9 10 11
41
« on: November 27, 2018, 11:10:16 PM »
What Reier showed is more of a top spinner / HS than an FS. He's basically saying it's not worth keeping the FS structure, though if you want to keep it an FS then the crushing teeth should be attached like this: So that the crushing teeth (edit) are facing out (and you probably want 4 of them also). I'm pretty sure that crushing teeth do damage with point. Or you could avoid this issue by using beater bars and/or razors. Having a wedge is also probably a really good idea so that you could gutrip somewhat and not get hamstrung by bots that do have a wedge. (Cutting edges would be better for gutripping though and razors would be the best). In general side plows are more important on designs with a narrow weapon. The FS isn't in dire need of them but they would definitely help (but at the same time don't put too much weight in that stuff). Your chassis armor could definitely be downgraded a little if you add this sort of protection.
42
« on: October 24, 2018, 02:38:08 PM »
Motors get weaker as the bot runs out of power, so in most cases you shouldn't power 2 hpzteks with 2 pinks. Bigger batteries are going to be more efficient in that they provide power for longer.
43
« on: October 22, 2018, 04:26:00 PM »
Can you actually listen to people (including a mod) when they say to put all the bots in 1 post? You seem to have the basics down fairly well. Some general advice for now is that the shiny hub wheels are generally the most efficient wheels, avoid t connectors and y connectors since they are inefficient and ugly and try using the side points on the red and black extenders more for angles, replacing the concave angles in designs (like in some of your crawlers and pokers and popups) with straight ones will help save on weight (by reducing surface area), use this to help you figure out the right number of batteries for your bot, and you could probably improve by learning some heavier glitches. Overall your robots are still better than most new builders.
44
« on: October 21, 2018, 09:53:42 PM »
RA2 is bugged such that all motors draw the same 100 power and there's no way to change it. So HP Z-teks are twice as powerful as normal Z-teks except with an itty bit of more weight. Which means that in almost all cases where they can be used over normal z-teks, they should be used.
You definitely want HPZs on your 36HS because you need the power to move the heavy 9 spike weapon bars. Also, 36HSes typically have HPZ drive. The true 36HSes that people like Sage and Inf built first had HPZ drive in along with the HPZ powered tribars and 2 black batts, but it takes extremely good chassis management to build something like that.
45
« on: October 18, 2018, 09:05:11 PM »
signed
46
« on: October 09, 2018, 02:24:03 PM »
I'll give my top 10:
10) Super Smash Bros Melee (2001) - I played it a lot, did the adventure mode and battled against AI endlessly. Didn't battle so much against other people and obviously I never got remotely good.
9) Kingdoms of Amalur: Reckoning (2012) - In my opinion, this is one of the best single-player RPG games out there. The graphics are wonderful (for a game from 2012..), the skill system encourages hybridization and the combat is super smooth. The main selling points for me is that its super immersive, full of lore, and has a really good story if you're into that sort of thing.
8) Spiral Knights (2011) - This crappy Java mmo shows how cartoony graphics and very simple fighting mechanics can still make for interesting and engaging combat. Each weapon and monster type has a different attack pattern, varying attack speeds, varying areas of effect, etc., and that results in good positioning games and good improvement by remembering how the monsters act.
7) Dungeons of Dredmor (2011) - Cheap rogue-like game with good graphics, a good sense of humor, and a very good (and very wacky) skill system. Where the game really shines is in reading the effects of all the skills and seeing how they synergize together so you can start each game with a strategy. Actual gameplay is turn based and requires you to constantly think about where the game will be 3 or 4 turns ahead, so there's an emphasis on cool downs, regen rates, and positioning.
6) War of the Roses/Vikings (2012/2014) - I never really got into Mount & Blade, but these are just fighting simulators of two different eras. I'm pretty sure the servers are closed down now. War of the Roses was especially fun for me since you could customize your weapon to exactly the way you wanted it, adjusting things like what the sword blade grind was good against, how durable vs how light/fast shafts were, what kind of metal things were made from, etc. In case you couldn't tell, I really like customization and optimization.
5) Robot Arena 2 (2003) - Game sucks too much to be placed any higher. It's only this high because a better robotic combat simulator was never made (though hopefully that will change). I know most people are into realism, but I mostly appreciate the arcade aspects of this game for what they are. Making skintight, clean, effective robots can be a pain in the butt but It's rewarding in the end and it's fun to see your ideas become functional.
4) Team Fortress 2 (2007) - my name much
3) Minecraft (2009) - I really enjoy both the redstone aspects of minecraft as well as modded minecraft. I've used redstone to build things like 8 bit encoders and decoders before I even knew what they were, and had created ways to store and display messages when I was really young. It was the first system through which I could apply my logic talents. Additionally, technical modded minecraft is fantastic because it allows me to create complicated systems of automation by pulling together a bunch of different mods. I can automate the production of a bunch of different things, and because I'm familiar with all the mods, I know exactly what from each mod I need to make the system work. In other words, seeing how all the mods synergize together is really nice.
2) Mabinogi (2004) - This is a crappy and old MMO by the (not so) legendary publisher Nexon. I've played a lot of MMOs in my time but this one remains very unique and different from the others, and there never has been a game like it since. Combat in the game is kind of like rock paper scissors, except its more like a net of very many skills, and there are cooldowns and attack patterns to deal with. There's also an emphasis on socializing with other players and doing things that aren't combat, which is different from most mmos today (and probably where the game would shine the most if the player base wasn't dead). And the 1st main story line is great.
1) Pokemon Showdown (2011) - Pokemon in general has kind of dominated my childhood. For example, I have over 600 hours in Pokemon Diamond. I like the anime. I like the openings to the anime (I think the ones from adv are the best). Pokemon Showdown allows for competitive Pokemon, which I'm sure most of you know about. I've been playing it off and on for like 6 years now so I like to think I've gotten decently good. I enjoy making Pokemon teams the most, because, you know, there's a lot of optimization and synergy involved in getting all of the Pokemon to play nice. I've also had some programming projects in the past try to analyze the metagame and do some optimization work.
Honorable mentions:
Loadout - cartoony, zany FPS with an emphasis on designing the weapons you use.
The Awakening RPG - one of the best flash game RPGs I've ever found, with a wonderful but simple real-time rock-paper-scissors combat system. Like it's a surprisingly good combat system, I basically copied it in one of my own games.
Battlerite - Like a moba game but no lanes, towers, levels, items, etc. Just arena combat with a cooldown-based system. I've been playing it a lot recently just because its free and sort of new and I don't have anything better to play.
Bloons Tower Defense series - this one doesn't need a description
I do think Undertale and DDLC are fantastic games but there's not as personally important to me as the games on my list.
I think Soma, Prey, and VA-11 Hall-A were some of my favorite games to watch.
Oneshot was good but I feel like it could have gone so much farther with the character's reactions to breaking the fourth wall. As far as RPG maker games go, I liked Blank Dream and Ib, with Ib being (imo) the best RPG maker game there is.
47
« on: September 28, 2018, 01:18:14 PM »
Wow, that looks VERY close to the rammer in the bot building guide. It's still a good build though, I wonder if there's anything you could do with the leftover weight.
48
« on: September 28, 2018, 01:13:36 PM »
Orc may have lost that battle but his team was way hotter with that no-speed-boost yanmega. These battles look fun.
49
« on: September 25, 2018, 02:03:15 PM »
Wow that looks like it was a lot of work. Nice job, I'll definitely check that out when its released.
50
« on: September 25, 2018, 12:48:28 PM »
I think Reier has a solid point in here. Considering that all these 10 years I've been part of GTM, Online playing was entirely ignored. I could imagine DSL-IRL being even more fun to play online!
My arms are open for RA2 online and I personally want to give it a try.
I know that the community is mostly oriented around IRL right now and that's perfectly fine. Playing online matches with IRL bots might be kind of awkward though (or not as much fun) simply because the balance between the robots aren't good. Like if you have a robot with a great aesthetic but poor performance, having to control it against opponents with stronger robots and lose could be frustrating. The point of IRL is to have lots of battles with good looks, which comes from bot realism more than driving realism. Driving brings a lot more to a competitive meta, like IF-standard, because now there's huge pressure to drive well in order to win, your driving skills matter, and there can be actual in-battle strategy (like in round 2 of the video). That's why Reier chose IF for these tournaments. If winning is your goal, you might feel that controlling IRL robots is like having a boxing match underwater. You might find after some IRL games you wished your bot was stronger/faster/less awkward/more destructive, and then try out building in IF or standard. It's also worth stating that IRL tournaments have to be made much smaller to fit in an online format. But no one has tried an online IRL tournament and IRL seems to be what people want to do these days. It's kind of dumb to avoid something that we haven't tried yet so take this post more as a warning than as truth. Obviously I'm stuck up the IF camp (and for reasons) but if someone wants to make an IRL tournament it would be interesting to see how it looks in the end.
51
« on: September 19, 2018, 12:05:02 PM »
The weight and chassis management is really good on this one. My biggest concern is with the batteries and motors: each orange outputs 10 amps and each drive motor uses 10 amps, and it looks like you only have 2 oranges. Also, it looks like there is some dark extender work inside the chassis, which could be made light to save weight (though its a little hard to see what's going on there) and that could be spent elsewhere (i.e. adding more batteries or making the arm longer). Personally I would go with a 2 wheeled robot since those tend to be more efficient and try to develop the weapon more but ultimately where to go from here is up to you.
52
« on: September 18, 2018, 01:16:18 PM »
In general, make sure the weapons you put on your bot have enough of their surface showing to have a nice contact area. The iron spikes in front of 1 and in back of 2 aren't doing very much, they should either be longer or removed. Additionally, 3 has a very small range with which it can hit opponents with the disk. I'm afraid that if another robot collides at an angle, it's going to hit the heater skirts and not the weapon.
For bot 1 in particular: The huge treads are probably a waste of weight on a middleweight edit: dang IF weightclasses, and they're causing a problem for your robot: if an opponent's wedge contacts the wheel before your wedge (and they probably will because they tend to be wider) then your robot lost the wedge war. Additionally, if an opponent is around the front end of your tread and stays there, it's pretty much out of reach of your hammer while it can hit your tread. It just seems like the treads are blocking the effectiveness of your weapon. So you could make the treads smaller or make the wedge wider. Also, you don't need so many overlapping wedges, just use the skirt hinges that have 1 connection for a wedge and have 2 of them. Space them out wide because those will be more effective. Also, it might be a good idea to make the hammer longer.
For bot 3 in particular: the skirts on top are not doing very much since they're not on hinges and they're blocking the weapon (they aren't really protecting anything either..). If you want invertible wedges then they should be on hinges also but they probably aren't worth it. More weapons on the disk (and using a disk that allows for more weapons) is probably a good idea, and rotating the beater bars vertically will help it hit low robots.
In general, your robots seem to have heavy armor in relation to their weight. I guess that's a preference thing and it's not necessarily bad, but in cases (especially with the HS) you don't want to find yourself outweaponed and then outpointed by the end of the match. Also follow Thrack's advice. It's really important to build to the weight limit for a competitive robot.
I might have said too much as to how to go about improving your robots. I don't want to control the way you build, so it's OK if you don't follow everything here, but hopefully you'll take some of this advice to make your robots better in combat.
53
« on: September 14, 2018, 07:45:01 PM »
In a terrible, horrible twist of fate my robot seems to have denied a loss. Looks like all the time I spend singing Nordic songs while rubbing the wheels is finally paying off (?!)
54
« on: August 26, 2018, 11:09:04 PM »
This is my first time seeing this thread, and wow, these fights are amazing. Positioning seems to be 80% of the battle. I can see how human controlled robots opens up completely new options for building (as it changes what is good).
Also, users of wedged robots really need to pick up the turn-around-and-boop-backwards technique when trying to pit crippled robots to avoid the whole me_irl_at_golf deal seen in fights such as the end of freezies vs cringey (since the back of those robots are relatively flat). Having an inverted set of controls seems important in general here.
56
« on: August 24, 2018, 01:29:26 PM »
Would work even better for DSL since then we wouldn't end up naming half the parts.
57
« on: August 21, 2018, 10:10:45 AM »
From my experience a slower spinner is popup bait, have you done any tests against them?
No because I know that's already true. This isn't really a good build.
58
« on: August 19, 2018, 08:00:10 PM »
Only two, and they can't keep up. I'm starting to realize I have a bad tendency to cram lots of weapons in small spaces, and it never turns out well. Additionally, through lots of painful building I've learned that using ro7 break to do a 30+ stack with back and forth 20cm extenders will ALWAYS turn out badly, even on something like a rammer. Having so many things inside of each other (or something to do with the weight) just makes the game give up and the bot becomes a havok bomb. I'll post some pictures of failed robot gore. (would be kind of cool if the eFFe didn't reset know what I mean) I want to stress that each razor is supposed to be perfectly inside one another but isn't for some reason probably pertaining to the snapper. Even though each razor is connected to the same snapper (there are no degrees of freedom between each razor), the end up spreading out like this whenever the start position on the snapper changes. Also notice the cloned axles (so weird?!). F12 does not show any sign of desyncing. When fired the weapon is just treated as a solid object. The only way to fix this is to move the snapper back. Garage-and-back (and reloading the game) does nothing. The piston couldn't really fire the weapon. Also it was a havok bomb and would trinity when spun in circles. There was a rammer version of this (because for some reason I thought eliminating the piston would fix everything... it didn't). You don't want to see it, it was that bad of a build. I also just want to say that each of these builds were strenuous, exhausting, and painful with multiple iterations that each took 3+ hours to build... I'm not sure why I like wasting my time with this stuff.
59
« on: August 18, 2018, 06:10:29 PM »
anti-agod rebuild Always pause the game during garage-and-back, leave the bot builder, and import/export. Pausing while a battle loads stops it from agoding immediately before battle. Motors are now rotated the right way.
60
« on: August 16, 2018, 09:07:45 PM »
Just got a chance to test this, and I'm already getting a lot of use from this can certainly see myself using this for any tournaments in the future. Though, I know it's too early as it's still a prototype, but might I suggest at least Click's AIessentials since most of those have enough variety and ones where I use the most for AI'ing?
I can give it as many AIs as I want, I just have to program them in. Right now there's sort of a 1:1 ratio between the bot types it can detect and the AI it assigns them (except for VSes which are annoying and can't be differentiated) but there's no reason we can't let the human have a bunch to choose from. One thing I would be worried about is the dropdown getting too cluttered, but adding AI essentials is probably a good idea since the parameters in the template work for most robots, and some of those AIs are situational and don't pertain to any specific bot type (having more than one AI for the same style bot is something I want to try to avoid). I could also have a system where AIs are dropped into a file and they get parsed and put into the program. I'm not really sure yet. I also have to figure out what to do with my system of taking out one odd parameter from the AI, not sure why I implemented it that way. My own code gives me headaches, I might have to take a break from it. this is what i get.
tfw Aier guess SnS as SpinupOmni2
It sees the spin motors and instantly thinks its a spinner. I could fix it though, by giving the wheel stagger detection system priority (wow, that was a mouthful).
Pages: 1 2 [3] 4 5 6 7 8 9 10 11
|
Show Posts
