I 100% agree. Playing good quality defense against strong teams in quals is probably the best way for a low-scoring team to differentiate themselves. This past year on 3061 we literally watched back quals matches of teams we saw play defense and the quality of that defense played a big factor in our 3rd pick rankings.
confused team that reused their KoP chassis:
I am going to have to object to this based on my team’s personal experience this year. At NE district champs things didn’t go as we would like in so far as alliance selection. We placed 15th in rankings which typically means you will be picked from the events I saw. Compared to other robots on the field, ours was lacking in features: we had a well-scoring shooter, but without any climber (we tried). However, we probably had the best defense bot because of our excellent driver (not just my biased perspective we got many remarks about it). Based on talk with our drive coach we had a strong suspicion we wouldn’t be picked since we have a low opr when we defend most matches. So we wanted to show how we can score. We tried to express this fact to drive coaches of stronger teams on our alliance, but due to their assertiveness we still ended up playing defense most of the time. In matches we didn’t defend we still scored well, but in other matches, with defense it just lowers the opr. I think if we didn’t play defense we would have been more favored, and still retained a good enough ranking. We had one team (5687) say they were going to pick us at districts, but teams we are less connected with didn’t seem to notice/value our performance in defense at such a big event with teams we played before. In smaller events we attended where most teams knew each other our excellent defense was noticed. We were able to shut down the dominating teams of almost every alliance at the Pease district event playoffs and got very many compliments from teams (let me say we don’t have a red-card-prone defense bot). This is different for other events and teams. As mentioned we didn’t have a climber which likely also lowered us on the scouting sheet. Nonetheless, next year we won’t be defending.
I was referring to the District Event and Regional Event level - I agree that at the district championship level it probably won’t be enough by itself. In our MSC division the (IMO) most skilled defensive bot in the division (5538 if anyone cares) went unpicked, and we ourselves had them as the third bot on our defensive bot pick list because they could not manipulate cargo, and the 2 bots above them on our list could.
a lot of this is decently specific to 2022. Also, personal opinion but it might be more helpful to get good at driving backwards then having a button to flip your entire control scheme lol.
Makes sense about being able to drive backwards well. Driver should learn to control the robot competently in all orientations.
I will say that in 2019 we had a flip button, but that was because we had one end of the robot for playing disks and one end for balls. in that context I think that the flip button can add value.
Also agree that since we all know nothing about this upcoming game, general non-game specific driving tips are more helpful.
Defense is usually a thing in FRC games (often a valuable thing), but who knows what 2023 will bring…
It helps to form a good relationship between the driver and programmer, and they can sort out what style works best for them. We’ve had flipped buttons (also, 2019) and this year they just backed up to the fender. I think since it wasn’t a long or extra precise movement and target that was the right call, and it seemed the new drivers picked it up faster than the hassle of an extra button. In 2019, you had to flip the entire mindset for a more precise and longer driving time, so a button was reasonable.
This brings up another topic in some ways, but I remember once being involved in a team where there was a push (valid in some ways) to give many students the competition experience by constantly switching drivers. While well intentioned, it quickly became apparent that it was not the right move competition wise and in a way unfair to the entire team.
After some problems with our driver last season, we are making changes to how we manage our drive team. One thing we want to do is have more than one driver with competition experience, so we don’t again find ourselves in a situation where our driver feels so indispensable that he can do whatever he likes, with no consequences.
Has anyone done something like this successfully? We don’t want ‘many’ students to have driving experience, just two. Current thinking is to have the backup driver do offseason competitions for sure, but we’d also like them to have at least some current season experience as well.
I hear stories that one year we fielded two drive teams, used in alternation during qualification matches, with one chosen for eliminations. Apparently we didn’t repeat the experiment.
Last week we took two robots to an off-season competition. That gave us a good opportunity to run two separate drive teams and (mostly) two separate pit crews. Lots of learning opportunities all round.
To share an example of truly excellent defense - 5846 was far and away the best defense robot at NE DCMP and managed to slip to the Titanium 1 seed. Even though they have a swerve chassis, there are plenty of lessons to be had about general positioning, field awareness, penalty avoidance, and taking advantage of your opponents’ weaknesses. The Titanium #1 alliance of 176 and 238 made an excellent choice in that pick, 5846 was a critical component in winning the event.
To illustrate just how incredible they were, see Titanium v. Calcium match 1.
- reduced 230 to 2 high goals and 2 low goals total in teleop
- drew an intake penalty off of 230
- drew a free traversal climb by pushing 230 into the hangar with <30 seconds remaining
- prevented 230 from making it to the hangar while 5846 climbed to the mid rung.
Additional highlights include:
- avoiding any penalties called against them due to their driver’s excellent field awareness
- pushing the offense robot away from their launchpad
- using open-field pinning, pushing in a perpendicular direction to an opponent’s tank drive without triggering the pinning timer
- isolating them to the far end of the field where visibility is weakest for the offense robot, and best for the defender.
230 was no slouch either - being the Calcium #1 seed, with an OPR of 51 points. In Finals 1, 230 contributed 18 points (4 high + taxi) in autonomous with a further 6 (2 high + 2 low) in teleop. This means that even ignoring drawn penalties, 5846 reduced 230’s offensive output by 27 points, more than 50%, and nearly eliminated their teleop contribution. Including all penalties/climbs, 5846 had a contribution themselves of 46 points.
While this is an extreme example, this is the power of truly top-tier defense. It’s very rare to see teams with such level of defensive skill at the championship, much less earlier in the season.
While the swerve vs WCD engagement made the defense significantly more effective, 230 in this match is also a good example of what NOT to do when you are being defended. 230 didn’t make much of an effort to create distance between themselves and 5846 resulting in 5846 being able to easily catch up. 230 kept on driving slowly and carefully when under defense which is really bad as it just makes it easy for the defender to catch up. 230 also overcommitted to trying to reach the launchpad from the back of the the field when 5846 was there when it would’ve been a better idea to try crossing the front of the field to prevent driving straight into 5846. Defenders always win pushing battles they waste significant time of the offensive robot even if the offensive robot breaks through.
Also the match shows why swerve defense is so effective. A well-driven swerve bot can easily reorient themselves around offensive bots to either push them in their least-preferred direction or push on a corner to control rotation (and driving direction for 6wd bots). Once a swerve bot is engaged with an opposing bot’s bumpers, they aren’t leaving until a pin count starts or the offensive bot reaches a protected zone.
I believe they were the best defensive team I have ever played with (no offense intended to the many other wonderful teams I have played with over the years including 2648 who was a great defender for us at Pease). Here are some things I think they excelled at (some of these echo your list):
- They knew the rules, and all of the implications (e.g. how far they needed to stay from a LAUNCHPAD to not be in danger)
- They understood the gameplay patterns of a typical match and how to predict where opponents would want to go (e.g. looking at game piece locations, hangar, etc.)
- They understood the physics of defense (e.g. how T-Bones work and pushing on corners to spin the opponent)
- They were both willing and effective at incorporating information and suggestions from 176/238 scouting into their defensive strategy (without being pushovers and just trying to mindlessly do what we said)
- They were communicative in the match about what they were doing when it varied from the initial plan. This didn’t end up mattering much because neither us nor 176 ever suffered a failure requiring a switch to defense but would have been very helpful if we did, we would have instantly known what they were doing to know where we would need to go as a 2nd defender.
It’s very difficult to know if you would have been drafted if you had focused on offense. You would have had to been among the top 16-20 bots offensively to gain sufficient attention. In the NE Champs that means you would had to score reliably in auto and add an average of a dozen goals per match if you didn’t have a climb, which would be a very impressive performance–perhaps even the first pick of a lower seed. If you were aiming to be a second pick, then standing out on defense is your best strategy. The alliance captains at District Champs are much more likely to have good scouting and they will see that. You played good D in the match I watched, but not having a climb is likely the main reason why you weren’t picked. I’ll bet that almost every robot on the playoff alliances could make a high climb, maybe with a mid climb for exceptional cases.
Going off of this a bit:
You always want a minimum of 2 people who can cover driver, operator, human player. This is in case someone takes sick (or other stuff happens), so you can swap in a new person and keep going.
Where you guys messed up was swapping them every match in quals. If there’s a difference in performance, it shows up, and scouts notice. It also doesn’t help overall performance… practice matches, offseasons, go for it, but stick to one primary drive team once quals start, at least for that event.
Which leads me an important point for all teams–be able to accomplish at least the minimal end game task. So often a bonus ranking point is dependent on 2 or more bots scoring in the end game, and during playoffs the added points are more valuable than potential scoring from all but the best offensive bots, with end-game defense usually being risky.
Since I don’t see any fundamental technical driving advice, I want to add that many robot drivers don’t understand how to stop accurately(assuming manual driving without vision/slightly swerve-focused).
For example, at the beginning of the season, the 4499 driver would miss an X on the ground at full speed coming at it straight (>15 inches) and now he can within <3 inches. When our vision was broken, he would miss the target by 30 ish degrees. We fixed our vision, but he still practices tracking the hub without it.
This leads to the following issues
- They missed the target by x distance
- They missed the target by Θ angle
- They wasted 2 seconds moving from a position they could score to a different position they could score leading to dropping 2 cycles over the 10 cycles they achieved in the match.
I think that these show that a driver isn’t comfortable with the robot they are driving.
Benefits of fixing this
- Get that extra cycle or 2.
- Practice is more beneficial now that the robot is controllable.
- Getting it right the first time so that they don’t have to try and rescore over and over again.
There are a few ways to deal with it.
Personally, I like 85-100% translational speed and 50-65% rotation speed for our robots. Usually, most people have rotation speed way too high. (Think victory spins but twice as fast)
We implemented 2 drills. The first was the stopping on the X mentioned above and the other was the tracking exercise.
a)Point stopping (what I called it)
You go left-right or toward-away on a line at full speed. There is an X in the middle of the line. You attempt to stop on the X where the middle of the robot is above the X.
The trick is to go in the opposite direction at full speed when you think that you’re going to be over the x for a split second. This way you don’t drift into the position at a slower speed.
b) Orbit (no not 1690)
This exercise is where you continuously point the front of the robot at a point on the field(the hub) while moving around. This usually shows that the rotation speed is too high.
This was a big part of our driver training in an attempt to have our driver drive better after evaluating driving over our competitions. And there was a lot more than this.
“Don’t be afraid of the robot” = If you’re not going to drive it like you stole it, you’re probably the wrong person to be driving the robot. Practice hard, get good, be confident.
We’ve also been using the 180-the-robot button trick for many years. It helps a lot even if you don’t have manipulators on the back of your robot…why take the time to turn around if you don’t have to? It’s an easy bit of code and works great on the kitbot or any WCD drive train.
As for playing defense or not, we try every year to have a robot that can play vigorous defense as well as the best offense we’re able to design for, and in strategy meetings we’re more than happy to defend rather than score if that’s what’s going to maximize our chances of winning the match. (A good example of this was in Destination: Deep Space. Late in qualifiers, as the only undefeated robot competing neck-in-neck for the top seed (which we lost by one cargo over the course of all qualifiers in the tie-breaker), we were paired with two great scorers against two great scorers, and the alliance made the decision that we would play defense rather than putting cargo in containers. We played very hard defense, had a lot of fun, got the W. It was totally the right move.)
So overall, I think you’ve got good advice, except for the defense bit. I would change that to, “Leave your ego at the door and play whatever role best fits that particular game to the best of your ability.”
Lots of good tips here, including multiple mentions of practice. We had a great facility to practice at for a little while with plenty of space. That went away. We can get away with some limited practice in the gym and hallways now.
Does anyone have any tips/tricks for driver practice with a limited practice space? How did you get good without having a full (or even half) field? Or is it going to be near impossible to achieve your full potential without a full size field?
We have, i would say about a half field or so to practice. Especially this year, when we really could only practice on the right side of the hub. The left side had computers and printers already there, so they stayed in front/behind/on the right of the hub.
I would definitely say that if you can, get another robot to play defense. Having even a little experience helps train the driver to make quick decisions, such as shoot now, or go to launchpad, etc. It also helps the driver and operator become more connected as they communicate to each other.
But also have some solo practice time just to get used to handling the robot. If you have the time for it, a few drills just moving around the field to certain spots are also good to do.
If there are other teams with fields in your area, you could try and scrimmage with them in season. You can also do drills and other practice without a field, but its hard to beat a (even partial) field.