Who is going to be on Einstein?

[Isaac]

I got into a conversation earlier with the captain of my team about Einstein. We had a long discussion about the merits of various robot types and their advantages in this game. I am curious as to the Chief Delphi community's expectations for this year. Without suggesting specific team numbers or names (as the field will likely contain at least SOME of the usual suspects), what will the alliances that compete on the championship stage (and ultimately the finals) be composed of?

In my opinion there are four to five basic robot types this year that are anyway capable in the game, and some combinations thereof. Based on the three regional events I have attended (Autodesk and CWA with my team and Seattle as a spectator), and more match videos then I care to admit (generally watched during English class), it seems that Human Loaders with fast drive trains have been dominating the field and I would be surprised not to see one in the winning alliance. Paired with a strong full court shooter and a good defender, they seem unstoppable. That said, the endgame is essential. If the same fast human loader is able to climb and dump for 50 in the last 30 or so seconds, I can't imagine an alliance that can shut them down easily.

• Human Loaders In order for a human loader to be effective, it seems like they must have a drive train to match. Teams that are able to drive quickly and put up the big points have dominated every regional I have seen. In Seattle, it was 1983, in Portland 955, and in CWA 1318, although they were knocked out in the Semis. 1114, while of course always incredible, by consistently making trips from the feeder to the pyramid and back (along with their 30 pt climb & 20 pt dump) is absolutely unstoppable.
• Full Court Shooters While both easy to defend and hard to dial in, a GOOD full court shooter can put up serious points. In Seattle, 948 absolutely destroyed the competition by hanging out the in the protected feeder station and putting Frisbee after Frisbee into the 3 point goal. Teams that attempted to defend them with hastily constructed shield were easily repelled by defense, and in elims, 1983 made it very hard for the blue alliance to shut down both robots. In ellensburg and portland, 1425 played commandingly, making themselves very hard to defend and outscore. A team that can both full court shoot and shoot from the pyramid has a double advantage here. 118 is perhaps best equipped to fill this spot on Einstein.
• Climbers This game lends itself well to specialization. Most 30 point climbers have a very specific focus and it is challenging for them to both shoot and climb even if they have the capability for both. It seems as if a 30 point climb, while essential to win, is only valuable if accomplished quickly. All scoring options have a theoretical cap. Assuming that the time for a trip of disks is low, human load shooters are only bounded by the number of disks. 30 point climbers, even if they are the fastest in the world, can only gain 30 points, 50 if they dump, per match. Only 4 teleop trips by a fast shooter can outscore this. In portland, 955 was making more than that and 1114 certainly has. I do not believe that dedicated climbers have a space on Einstein, but a team like 254 or 1114, who can both shoot and climb effectively are in a very very good position.
• Floor Pickup At the regionals I have attended, I have seen very few floor pickup machines. I don't know if this a function of strategy or difficulty of execution but it seems very rare. Although many of the powerhouse teams have the capability, they use it infrequently. It seems like after the discs in the center of the field are depleted, it is not worth the extra time and ambiguity to attempt to floor pickup. This is something I am very unsure about however, and it could have a much larger role in STL than I anticipated with teams such as 1986 and 2056 concentrating on it.

My guess for Einstein is that the strongest alliance will be composed of a robot that can human load and then rapid-fire shoot and then go for the 30 point climb and 20 point dump, a robot that can both full court and pyramid shoot effectively, and then a robot that is able to play very strong defense and possibly floor load.

What have you guys seen at regionals you've attended and what do you hope to see on Einstein?

[EricH]

If I was picking an alliance likely to go far (and I do mean far), I would pick the following:

1) Defense/climber/dumper, preferably with an autonomous shooting mode. I rank them first because they are the rarest of the rare--a fast 30+20 climb that can drive around and defend much of the match. I've seen just about one or two. (If none are available, take a #2 robot instead.)

2) Shooter with floor pickup. Human load, no problem. Floor pickup, even better--starve your opponents and get your first cycle in faster. However, an excellent automode is a MUST for this one. Fast speed for quick cycles, and a 10-point hang round out this one. Think 1114 and 1717 for this one--something on that order. (Also known as a "cycle runner".)

3) Either a second (third?) one of #2, or a dialed-in full-court shooter. I'll go with the FCS here; objective is to get the discs into the goals, draw defense, and if you can't get them in the goals to get them to where robot 2 can pick them up. Have robot 1 play some blocking on occasion. Ideally, able to run cycles instead of going long if they want to, 10-point hang. Oh, and good automode--no sense wasting discs.


I would say that there will be at least 1 alliance like this on Einstein; that is, one climb specialist, one cycle runner, one full-court shooter that can run cycles. I also see three good cycle runners on an alliance; that can go very deep. Two cycle runners and one climb or full-court specialist will also do well.


What I expect not to see is alliances with more than one climb specialist, or more than one full-court shooter. (Exception: FCS robots that can run cycles not counted in FCS counting.)

[bkahl]

1. A floor feeder is a necessity in a good alliance. There will always be frisbees on the ground for them to pick up. Good climb perferable

2. Full Court Shooter- not many out there, and consistent shooters are even slimmer. A good full court shooter paired with a floor feeder will put every frisbee in the goal. 10 point climb?

3. Defense/Climb Bot- a defense bot is almost a necessity to keep other robots from blocking the FCS. The full court shooters are ineffective with a piece of lexan sitting in front of it. This bit can also be used to combat another FCS. A climb is preferable

[Cory]

No alliance will make it to Einstein without a floor loading robot. If they do, they've already lost before teleop begins. You can't start in a 24 point hole.

[JB987]

Originally Posted by Cory

No alliance will make it to Einstein without a floor loading robot. If they do, they've already lost before teleop begins. You can't start in a 24 point hole.

+1

[benherms]

Originally Posted by bkahl

1. A floor feeder is a necessity in a good alliance. There will always be frisbees on the ground for them to pick up. Good climb perferable

2. Full Court Shooter- not many out there, and consistent shooters are even slimmer. A good full court shooter paired with a floor feeder will put every frisbee in the goal. 10 point climb?

3. Defense/Climb Bot- a defense bot is almost a necessity to keep other robots from blocking the FCS. The full court shooters are ineffective with a piece of pecan sitting in front of it. This bit can also be used to combat another FCS. A climb is preferable

An alliance like this is most effective if the floor pickup bot can perform a seven frisbee auto and the defensive bot has a powerful drivetrain that can keep an opposing FCS out of their feeder station. Also, like you said, the FCS MUST. BE. CONSISTENT. Finally, if we're talking about Einstein, only exceptional coaching and performance under pressure combined with a pinpoint strategy will create a sufficient alliance.

[apples000]

From looking at this from a pure offensive scoring position, I think that a climber/dumper with 3 disc auto and some defensive play could be a candidate. This robot would score a guaranteed 68 points every match and if the other partners could do just as well, the total score would be 204 points.

[Abhishek R]

I'm pretty certain that the "pure offense" strategy will not work. After watching several of the regionals from the current season, most of the "defense" was just a robot moving side to side along the centerline and occasionally the tall blocker to attempt and deflect full-court shots. A more risky strategy, requiring a stellar drive team, is the man-to-man defense (or, since CD likes using hockey for this challenge, body-checking) as was displayed by 2789 TEXplosion at the Lone Star and Alamo Regionals. By making the trip back-and-forth from the feeder station a nightmare for the pyramid shooters with a quick and powerful drivetrain, they were able to delay the number of cycles these pyramid shooters were able to make taking it down from 5+ cycles to even just 1 cycle. They also had a removable blocker for those pesky FCS'. Even matched up against the now famous alliance of 1114 and 2056, I think a good defender could completely take down one, if not both robots. I'd imagine the scores in the finals of each division will only be around 150, assuming this strategy of man-to-man defense gets played.

I think the alliance that gets to Einstein will be composed of:

-Extremely mobile pyramid shooter with under-pyramid capability and 7 disk auto

-Full-court shooter with perfect 3 disk auto

-Fast and strong drivetrain defender with good drive team

[Kevin Leonard]

I think that the two most important positions on an alliance are a feeder station robot (whether that be a FCS or just a fast cycler) and a floor loader (preferably with a multi-disc autonomous). The third robot can occupy a multitude of positions- from defensive climber to "offensive lineman"/climber or even another cycle-runner.

I think this year- above all other previous years- is going to see the largest disparity in alliance composition on Einstein.

All of the above being said, sometimes you'll have robots that can fill a sort-of combination niche- like 1986- who has a 7 disc auto, but during teleop plays the role of a fast feeder station loader. So an alliance like 1986-987-3rdRobot would still work because they have complementary multi-disc autos and play different roles on an alliance.

[Woolly]

Originally Posted by benherms

An alliance like this is most effective if the floor pickup bot can perform a seven frisbee auto and the defensive bot has a powerful drivetrain that can keep an opposing FCS out of their feeder station. Also, like you said, the FCS MUST. BE. CONSISTENT. Finally, if we're talking about Einstein, only exceptional coaching and performance under pressure combined with a pinpoint strategy will create a sufficient alliance.

Not to mention, the FCS better be able to run to the pyramid to shoot if need be, and absolutely must hit all three autonomous shots. FCSs that are 1-trick robots will be shut down entirely.

[Rangel]

I see a winning alliance at championships having three particular robots:
1)
7 Disk autonomous
Shoots fast and accurately
50 Point endgame
2)
5 Disk autonomous using centerline frisbees
Shoots fast and accurately
10 point hang(ideally 30 point hang)
3)
3 Disk autonomous
Shoots fast and accurately(yes no defense)
10 point hang

[Gregor]

Originally Posted by Rangel(kf7fdb)

I see a winning alliance at championships having three particular robots:
1)
7 Disk autonomous
Shoots fast and accurately
50 Point endgame
2)
5 Disk autonomous using centerline frisbees
Shoots fast and accurately
10 point hang(ideally 30 point hang)
3)
3 Disk autonomous
Shoots fast and accurately(yes no defense)
10 point hang

Switch the endgame of #1 and 2, and I totally agree, although I can see #2 having a 3 disk auto.

[MrForbes]

I'm just having fun reading all this. Our team lost a regional to two posters, and won with another.

[Billfred]

One robot must do a 50-point play.
One robot must do a 40-point play. (That's like a 50-point play, but with 2 discs since the partner got four. I don't see teams hedging on 45-45.)
One robot must do a floor pickup autonomous.
All three robots must have their autonomous dead-on and timed in sync.
I think you'll see a lot of joint climbs on the practice field, shades of triple balance practice last year.
Someone will have an 84" wall, even if it sits on the cart the whole time.

[dodar]

1. Corner-to-Pyramid
2. FCS
3. Ground Pickup

You will not win this year unless you go full offense.