I’ve heard rumors that the GDC won’t make another open field game like 2014 is due to this issue for rookie/low resource teams. It’s not very inspiring for those teams to have their only event cut short due to severe damage that can’t be repaired quickly. Especially now in the age of brushless motors, high speed collisions are going to be more frequent and have a higher chance of causing severe damage. Any remake of the 2014 would have to have speed bumps in some form to lessen the chance of this from happening.
2 Likes
2017 and 2019 had some pretty open areas where rough defense could happen. I’m not sure that it’s too much of a concern for the GDC given the improvements in the kitbot chassis and community projects to create effective designs for rookie and low-resource teams.
1 Like
There’s other ways to limit the damage than speed bumps. It may be time, for example, to revisit the standard bumper design and see if there’s any way to improve it (either in the bumper itself or in the backing). We’ve had the same bumper design since 2006… You could also conceivably limit motors (e.g., you get up to 6 of Falcons and NEOs, combined)–this would be rather unpopular among the ceiling teams and possibly very popular among the floor teams. You could–and those that know me know why I will personally find and deal with anybody actually proposing this–use the 2019 G20, Week 5 calling method, as the baseline robot contact rule. (If you’re not familiar with it, a literal reading of the rule essentially said that any intentional contact between bumpers was a Tech Foul + Yellow Card. When enforced as written, card counts skyrocketed, partly because teams weren’t informed. 2020 saw a much better version.) Theoretically, you could put the Step from 2015 across under the Truss, with a 1-robot-wide gap or two maybe.
If there’s a specific goal, don’t rule out anything just yet, or focus on only one method.
Just going with a straight 2014, there’s a few changes I’d make:
- The FMS lag issues are already taken care of. They were gone in 2015 and haven’t come back. Apparently there was a pile of spaghetti code, so HQ added a staffer who started from the bottom and there hasn’t been a lag issue since. So pedestal not lighting is now fully blameable on the refs not pushing the button (because they’re waiting to make SURE the ball doesn’t bounce back!).
- Zone determination is made by bumpers not wheels. Easier for the refs to determine.
- For that matter, zone lines are cable protectors. (Or defense ramps from 2016, scoring platforms from 2015…Bumps from 2010 even, though that might be a stretch.) Won’t slow you down much, but will make it painfully clear that you’re in a new zone.
- You get up to 3 auto balls. Once you’re down to 1, assists begin counting. (Seriously annoying to watch teams just fumble balls all over the field… and only get 10 points if they managed to get one in.) If you’re down to 0 before the beginning of Teleop, your HP can go out beside the field but NOT enter a ball.
- “Balls” are renamed to “Scoring Spherical Objects” due to this being a high school competition.
- Bumpers are now required on ALL non-intake sides, full coverage. Intake (and, if different, outshot) sides need 8" from each corner. See above comment about revisiting the design.
- We’re using the 2008 balls. They’re bigger. (/s)
- Truss Hang Endgame. 10 points per robot, 5 points per ball in robot, 50 points per ball on top of Truss without being in contact with a robot. Extra balls may be deployed at T-20 seconds but are NOT eligible for assists. If a ball is on top of the Truss before the T-20 second mark and looks like it’s staying there, the HR will relight the pedestal–but if it comes off before the end of the match it needs to be scored before another ball is entered.
That said, I say we just do this game bowling style. Off-field, there’s a number of cutouts of “random people”. (GDC members, HQ staffers, event volunteers, infamous mentors, Dean’s List students…) You can score in the goal for points+assist points, or you can target the cutouts at some random value each (or maybe different values for different “types” of person represented). Your choice. Knock 'em all down, get a bonus.
5 Likes
Scoring Spherical Objects, you mean?
2 Likes
This, but make it so there’s 2 trusses on the field. 1 for hurdles, and the other without any purpose at all. Then don’t announce it until week 6 of the build season. Then let CD run rampant with speculation with the secret purpose of the 2nd truss.
2 Likes
The one-ball comparison to sports is not apt, because at no point in sportsball will you have a crippled/inept/unlucky player get the single playing piece stuck in them, and thus doom their team from scoring any more points for the rest of the game.
11 Likes
Aerial Assist had some major flaws:
- Early season had way, way too many game-changing fouls
- Much of the season saw big, high speed hits that took robots out
But the game was a fantastic spectator sport. In-match tactics were the most important and interesting they’ve ever been. And while by the end of the season strategy converged on two basic ways to play when you had three competent robots on the field (near-side or far-side rebound), there were various smaller ways to play to your strengths or your opponents weaknesses
- Flexibility in roles to swap trusser/inbounder/finisher when one robot is shut down, etc.
- Having a good catch radius and using it to inbound the ball from less expected positions so you can get the truss or high goal off quicker
- Inbounding the ball into a robot that’s already up against the wall in scoring position to remove time from cycles and make the finish less defensible
- Knowing which side your opponent wanted to truss from so they could hit their preferred human player and forcing them to truss from a less comfortable position
Especially when compared to most games (2011, 2012, 2013, 2016, 2017, 2018, 2019, 2020), where the offensive robots largely just played the game separately on the same field, and in some of the games, the third robot barely gets to touch the actual game pieces, 2014 is a clear outlier in all the good ways.
Also if we were doing a 2014 replay, speed bump berms is a very good idea imo.
4 Likes
Two things related to human players:
-
There should be some penalty for inaccurate truss passes to human players instead of the <1 second delay in waiting for the volunteers. A legitimate strategy shouldn’t be “we’ll make enough bad passes that the volunteers will start to position themselves behind our human players to catch our bad passes”. Maybe a 5 second count before the volunteer can hand the ball back to the human players?
-
At the end of the year, robots were essentially getting possession credit for human players bouncing balls off robots. While I get how it was allowed based on how the rules were written (and obviously don’t blame teams for taking advantage), I don’t think this was the intent or spirit of the rules.
4 Likes
I think that this would actually give advantage to the ceiling teams. These are the teams that are capable of building PTO mechanisms and other complex dual function mechanisms. It’s the floor teams that need the extra motors to reduce the design complexity of their robots. The ceiling teams are also better able to calculate required torque/speed for the given application and then can design around the available motors. The floor teams are more likely to need to throw on an extra strong motor (say a Falcon, where a ceiling team could have designed around a 775).
2 Likes
Agreed. Motor limits definitely skew things toward teams that know how to end-around those limits.
[I think the takeaway here is that designing a good game is really quite difficult!]
2 Likes
I think most people would agree that Arial Assist was a very good game from a spectator standpoint. Lost of defense meant interesting matches with unpredictable outcomes, the single ball provided a focal point for the audience’s attention, and there was a good amount of strategic depth to the game.
I’d argue though that it wasn’t such a great game from an engineering challenge standpoint. Robots basically had two tasks: pick up a yoga ball and shoot it into a goal (or over a truss). There was no end game and only one game piece, so most teams didn’t have to make any significant tradeoffs in their robot strategy. There were also no real advanced challenges, which meant the better teams were only separated from the average teams by slim margins. This made for some interesting matches, but meant that more than any other year the best robots were the ones with the most driver practice and not necessarily the best engineered.
The GDC then over-corrected for this in 2015 by making a very interesting design challenge that was super boring for spectators and left very little room for upsets (or interesting matches). I think in general since then the GDC has done a better job of balancing those two aspects of game design. All of the games 2016-2020 have had a variety of challenges of different levels, enough tasks that average teams will have to make some tradeoffs, a decent mix of offense and defense, and gameplay that was interesting for spectators. I’d like to see them continue that trend rather than bringing back Arial Assist.
5 Likes
A little hyperbolic don’t you think? Dead ball calls were certianly detrimental to an alliance, especially when the dead ball got released later on in a match, but far from “never scoring again” territory
I feel like this is not that different than other years… Say 2019? Also what percentage are we talking about of teams being effected? I’d guess somewhere in the one percent range. Was 2014 that much worse on robots than recent years? I’m not sure it’s clear cut.
Everything else you said makes perfect sense. Fouls were a big issue early season that caused a lot of issues. People definitely adapted to the conditions over time. Again I wonder how the fouls of 2014 stack up to other human error fouls/missed scoring eg 2017 with missed dropped ropes/failed rotor spins/forgetting to use populated gear/hand outside the airship. Those might be similar in impact to 2014’s issues, but I think 2017 has a better rep in that regard.
One of 2014s big issues is that running a sketchy auto that failed really hurt your alliance. Games shouldn’t punish you for trying to complete a task.
1 Like
The biggest issue was that the GDC didn’t even think dead balls would BE an issue. It took the teams maybe an hour to realize that it would be.
Then the “dead ball” system relied on being able to get the HR’s attention, the HR to agree and light the pedestal, and the pedestal would take 5 seconds to light on a bad day. If your opponents are any good that’s 50 points you’re down (1 cycle) and have no chance of getting back because your third robot is non-functional and can’t get assists.
And if the dead ball becomes live again, you lose more time getting it off of the field…
A dead ball potentially meant an auto-loss if you weren’t an elite team.
3 Likes
100% agree. Dead balls were devastating in close matchups. You would need some way to call one through fms that immediately removed the assists/lit up pedestal/alerted refs ideally.
The point was the dead ball was not a “never be able to score ever again” situation.
At the same time when was the last time you won a match where you had a completely broken robot on your alliance in a close matchup? The issue of the dead ball being impactful is not an issue unique to 2014. 2014 just tips the scale even more towards the negative outcome.
1 Like
And the teams would continue to bang on the glass for another 10 seconds, without turning around to see that the pedestal light was on.
3 Likes
2018 SVR F2 is a high-profile case of that…
When I was on 330, we tended to assume that we were 1v3 or 2v3 when we were doing strategy discussions Day 1. For some reason, we also tended to do well.
Average teams, though, you’re right, a dead robot is generally quite painful. (Particularly in 2009 if they were in the side slot…)
2 Likes
My twist on aerial assist mashed up with breakaway:
Game object is a large puck that slides across carpet easily. It is illegal to kick the puck into the air but fine to kick along the ground.
Field is divided into 3 sections by alternating slots and bumps. The pucks can fit under the slot but robots cannot. The robots can drive over the bumps but pucks cannot go over them. Bollards (green) in the field make passing across a zone or scoring from the middle difficult to impossible.
3 objects on each alliance in play a time, recirculate from home zone?
I haven’t thought too hard about how this would actually play out. My suspicion is that it would be too congested and easy to lock down. I just like the concept of heavily incentivizing passing without explicitly making passing worth points. For a team to solo score they could have to collect, pass, cross, collect, pass, cross, collect, score.
4 Likes
I wonder if anybody has a 1999-vintage Puck around still… That could be highly entertaining and amusing.
1 Like
10/10 would play
2 Likes