Ideal Alliance Structure

[Lil' Lavery]

What do you think the ideal alliance structure is?
imho, it would be a rapid fire shooter, a perimeter shooter (aka, a shooter that can fire consistantly with enough range that they don't have to be right in front of the ramp. Preferably with a turret of some sort and the ability to be off to one side), and a bot with excellent storage (20+ easily), the ability to score corner goals quickly, and can pick up off the ground easily (note that this doesnt mean they can't also shoot the center goal). Additionally all 3 robots can get on the ramp.
During autonomous, the herder would run to the corner goal and dump 10 balls in the corner. The perimeter shooter would sit in position 2 and fire the 10 shots it has into the center goal. The rapid fire shooter would move from the back position and fire at the center goal (preferably with camera aim). By having the rapid fire shooter take this path, it would also serve to intercept any bots tryin to hit the perimeter shooter and possibly an opposing robot going for their center goal.

If you won auto, if either shooter (preferably rapid fire) can load from the floor it would be backbot (in order to be in position to fire immediately). Otherwise the herder would start as backbot while the rapid fire shooter got loaded, then they would switch. The defense would stop opposing shooters primarily, and corner shots only as necessary (need to have ammo to replenish shooters later). If the opposing shooters depleted their ammo early in the period, the shooters would assume offensive positions during the defensive period in order to begin to score ASAP.
During the offensive period, the rapid fire shooter would fire as soon as it was in position, attempting to hit as many shots as possible before being dislocated from its shooting spot. After it depletes its ammo, it moved to reload, and intereferes with any defense being played on the other bots on the way out. The perimeter shooter would fire when the rapid fire isn't shooting (don't want balls colliding in the center goal do we? ). The herder would run interferance for the perimeter shooter and gather balls. If the defense on the rapid fire shooter was too intense, it would go for the corner goal in an attempt to pull defense away (or score a whole bunch of points really fast, either way ). If the defense is really light for whatever reason, and the herder also had a shooter, it would fire some shots at the center goal (this is unless, the opponent is getting in position to begin to fire, then it would get in defensive position, but preferably dumping the balls in corner first. If the rapid fire shooter needed to re-load at the HP station, it could instead also play defense on the opponent's shooter if needed).

If you lose autonomous; bots will quickly get what ammo they can. Taking up scoring positions, then they play the offensive period like they would in the other situation. The exception being the herder could try to "lock" any hp loading bots into the hp loading area.
The defensive period they would play very much like the other period, but they would have to be more careful becaue the opponents will likely have more ammo. If the herder "locked" a bot into the loading area, it will play backbot while holding it back there, but would switch with the ground loading shooter once the defended bot crosses mid-court.

During the final period, the shooters will resume shooting with new ammo, and continue until it is depleted once. The herder will play defense (if it still has balls, it should hopefully find an oppurtunity to dump them sometime during the period). After ammo is depleted, the bots return to play defense then get on the ramp.

[Beth Sweet]

Well, at GLR I saw 2 shooters and a defense machine work well.

I think that 2 offensive robots in any form (shooters or good 1-pointers) would be pretty ideal!

[nuggetsyl]

i want someone that can shoot and not be pushed while shooting.


shaun

[sw293]

My easy answer to this question is this: 1 human loader, 2 auto loaders. No place for a defensive robot because scorers can often play defense very well. I am not convinced of the necessity of a corner dumper, but if necessary, one or two teams like 56 that can score both corner and center goals might be considered. There should be two teams that can prficiently score on the center goal.

The more interesting question is how strong alliances might be formed in alliance selection.

[atomikitten]

I think one corner scorer is useful. However, I don't think that human-loading is rather time consuming compared to a bot that very efficiently picks balls up off the floor. I like two shooters, one of which might be defensive, then a seriously defensive dumper; stocky, stable, and powerful. The two shooters might have the ability to dump, but it's not at all a priority. During the kickoff broadcast, they said that "good defense is key this year," but don't they every year? One defensive robot that can also score points is sufficient to two other big scorers.

[sw293]

Earlier I wrote:

Originally Posted by sw293

The more interesting question is how strong alliances might be formed in alliance selection.

And here is what I mean by that:

Suppose at Championships in Newton Division we get the following rankings for alliance selection (assume all of these are center goal shooters):

1. FRC2001 (#3 Human Loaded Robot)
2. FRC2002 (#1 Ground Loaded Robot)
3. FRC2003 (Ground Loaded Robot)
4. FRC2004 (#2 Human Loaded Robot)
5. FRC2005 (Ground Loaded Robot)
6. FRC2006 (Ground Loded Robot)
7. FRC2007 (#1 Human Loaded Robot)
8. FRC2008 (Ground Loaded Robot)

I submit that each alliance has room for only one human-loaded robot.

FRC2001 wants to pick FRC2002 (best ground-loaded robot). FRC2002 would rather be allied with FRC2007 or FRC2004, who have better human-loaded robots. FRC2002 knows that FRC2001 is probably not going to pick FRC2007 because their robots are incompatible (both human-loaded robots), and even if that selection is made, FRC2004 will still be available (still a better human-loaded robot than FRC2001). Therefore, when FRC2001 picks FRC2002, FRC2002 should decline, right?

But what happens if FRC2001 then picks FRC2007? FRC2007 ought to decline, because an FRC2001/FRC2007 alliance would have two human-loaded robots (=inefficient). So then FRC2001 picks FRC2004, who declines for the same reason. In this scenario, FRC2002 is worse off for having declined FRC2001's offer of alliance. So by declining FRC2001's offer of alliance, FRC2002 is taking a gamble.

However, if FRC2001 were smart, they would have picked pick FRC2007 and FRC2004 first, and after they had declined (as expected), FRC2002 would have accepted FRC2001's subsequent offer of alliance because their two top teams would no longer be available.

The flaw in my argument is the assumption that all teams will make rational decisions in accepting or declining alliance offers.

I hope that was intelligible. Please tell me if it's not and I'll try to clarify it.

[pyroslev]

VCU's winning alliances and high scorers typically had two shooters and the third varied. The third was a shooter but also had strong blocking and distracting capabilities.

VCU Winning Alliance
1610- Strong autonomous and shooter
343- Good shooter and turreted capabilities, heavy ball holding capabilities
1598- Shooter capacity and mobile blocking capability

Just a good alliance example. Shooters win, imho.

[MattB703]

Everyone is talking about a defense specific robot, but in the championship divisions there will be enough depth of talent that you should be able to get defense bots that can score as well. Take a look at 818 - The Steel Armadillos. The were among the top 5 shooters at Detroit and were arguably the strongest defense bot as well.

I guess my point is that at Nats we will be able to have our cake and eat it too.

[Tim Delles]

aaa, but comes the argument of which is better offense or defense?

Well when i'm thinking of alliance this year you can look at it 2 ways.

1 - 3 great shooting robots. for example match up FRC2001, FRC2002 and FRC2003 (I'm just picking these teams because they were the first shooters that came to mind) They all have good drive trains, and great shooting ability. Now you have only 2 robots that are playing defense of these guys, so thier should be one that is left open to do as they please with the center goal. And then on defense since thier drive trains are all good they can cleanly play defense. Now a quick break down from what I have seen.

FRC2001 - Human Player Loaded. Good solid shooter.
FRC2002 - Ground Loaded. Good solid shooter.
FRC2003 - Ground Loaded. - Good solid shooter.

2 - 2 great shooting robots and a great defensive bot. Once again take FRC2001, FRC2002 and FRC2004 (Yeah they play offense but they are still a great defensive team in my book no matter what). Now since only 2 robots can play defense you have FRC2004 (the defensive bot in this case) play defense on your opponents best defensive bot (making it so that 1 robot is still free to shoot).

FRC2001 - Human Player Loaded. Good solid shooter.
FRC2002 - Ground Loaded. Good solid shooter.
FRC2004 - Great Defensive Robot.

Now I only listed what these teams are good at. Some teams may be able to play defense and offense really well, but in my opinion one of those 2 setups are what you are going to see win all the divisions in Atlanta, because as much as defense is great, I don't see it winning by itself like it could in previous years.

[Ethulin]

Though I understand how this thread is the "Ideal" alliance structure, I wonder how people are taking into account the new picking system in their planning, going 1-8 then 8-1.

It seems that seed #1 really could get another good shooter on its side, but after teams 7-8 have picked TWICE (so 23 robots gone) do you really think another uber bot will be available?

So, to partcialy re-phrase the question:

What is your ideal but feasable alliance structure?

[MattB703]

Originally Posted by Ethulin

Though I understand how this thread is the "Ideal" alliance structure, I wonder how people are taking into account the new picking system in their planning, going 1-8 then 8-1.

It seems that seed #1 really could get another good shooter on its side, but after teams 7-8 have picked TWICE (so 23 robots gone) do you really think another uber bot will be available?

So, to partcialy re-phrase the question:

What is your ideal but feasable alliance structure?

I agree that if you are selecting as the #1 alliance you will likely end up with two uber-shooters and the best defense bot that is left over, but you might get beat by the 6,7,or 8 alliance who chose three offense bots.

I like this year's serpentine selection system!

[petek]

Originally Posted by Ethulin

It seems that seed #1 really could get another good shooter on its side, but after teams 7-8 have picked TWICE (so 23 robots gone) do you really think another uber bot will be available?

Look no further than the oft-mentioned NJ regional where 15th-seeded Cold Fusion was still in the pool when the #1 alliance got to choose their third. Still don't understand how that happened...

[Tim Delles]

Originally Posted by Ethulin

Though I understand how this thread is the "Ideal" alliance structure, I wonder how people are taking into account the new picking system in their planning, going 1-8 then 8-1.

It seems that seed #1 really could get another good shooter on its side, but after teams 7-8 have picked TWICE (so 23 robots gone) do you really think another uber bot will be available?

So, to partcialy re-phrase the question:

What is your ideal but feasable alliance structure?

Who said the #1 alliance was the best shooter? Who said they were the best anything. The top 8 could unfold like this:

#1 - FRC 2001 (Okay defensive bot. Okay at collecting and dumping balls.) (7-0)
#2 - FRC 2002 (Great shooter. HP load only) (7-0)
#3 - FRC 2003 (Good shooter. Ground load only) (6-1)
#4 - FRC 2004 (Good Collector. Great dumper. Okay defense) (6-1)
#5 - FRC 2005 (Good shooter. Collect and HP load) (6-1)
#6 - FRC 2006 (Great Shooter. Collect and HP load) (6-1)
#7 - FRC 2007 (Good Collector. Great shooter. Okay dumper) (6-1)
#8 - FRC 2008 (Great Defensive bot. Okay shooter. HP load only) (6-1)

Then if you look down on the list you find

#9 - FRC 2009 (Amazing shooter. Collect and HP load)*** In all reality this could really be the best robot but got stuck in 1 match were 1 or more alliance partners didn't come out . Should be the number 1 seed because of higher QP points but lost the match were their alliance partner didn't come out.***
#12 - FRC 2010 (Great Shooter. Collect and HP load)
#13 - FRC 2011 (Great shooter. HP load only)
#20 - FRC 2012 (Good shooter. Collect and HP load)
#21 - FRC 2013 (Great shooter. HP load only)
#24 - FRC 2014 (Great Defensive bot. Okay shooter. HP load only)
#26 - FRC 2015 (Great Shooter. Collect and HP load)
#27 - FRC 2016 (Good shooter. Collect and HP load)
#31 - FRC 2017 (Great shooter. HP load only)
#33 - FRC 2018 (Good shooter. Collect and HP load)
#34 - FRC 2019 (Great Shooter. Collect and HP load)*** In all reality this could really be the best robot but got stuck in 3 matches were 1 or more alliance partners didn't come out***
#37 - FRC 2020 (Great shooter. HP load only)
#41 - FRC 2021 (Great Defensive bot. Okay shooter. HP load only) *** In all reality this could really be the best robot but got stuck in 4 matches were 1 or more alliance partners didn't come out***
#42 - FRC 2022 (Good shooter. Collect and HP load)
#44 - FRC 2023 (Great shooter. HP load only)
#50 - FRC 2024 (Good shooter. Collect and HP load)
#61 - FRC 2025 (Great Defensive bot. Okay shooter. HP load only) *** In all reality this could really be the best robot but got stuck in 5 matches were 1 or more alliance partners didn't come out***
#68 - FRC 2026 (Great shooter. HP load only) *** In all reality this could really be the best robot but got stuck in 6 matches were 1 or more alliance partners didn't come out***

Now this is only 26 teams, that are either good at offense or defense. Now i'm pretty sure a division at the Championship will have atleast 26 robots. So i'm not seeing any alliance having a weak alliance unless they were 'carried' into seeding.

So basically its not about the draft, it is still all about scouting.

[AcesPease]

Originally Posted by Tim Delles

Who said the #1 alliance was the best shooter? Who said they were the best anything...

So basically its not about the draft, it is still all about scouting.

I heartily agree. Scouting is very important. And the best robots do not always end up very high in the standings. Here is an example: At UTC #1 (126) picked #3 (20), #2 picked someone in the top 7 and the new #3 picked someone in the top 7. The #4 team tried to pick the new 6 and new 5 teams, who declined. Team 177 apparently did some good scouting. They were now #5 after finishing #8 and they picked the #22 robot (176, a good 3 and 1 point scorer with large ball collecting capacity) and then the #34! robot (1124, who was just beginning to show 3 point ability Saturday morning). The #1 seed used their last pick on the #15 seed (571, a robot with potential to score a lot of 3s, that had not demonstrated any scoring during the qualification rounds). 177's alliance went on to win their quarter final and then faced the seemingly unbeatable alliance put together by 126 and 177's alliance went on to win the regional, despite going with low seeded teams.

Good scouting and having three robots that can score and get on the ramp (including one that can score in the corner) is the way to go. Any decent robot can play defense in this game.

[Rick TYler]

We should use chess notation when writing about alliance picks:

2101: 2343 2712??
2703: 2113! 2976!!
2404: 2212? 3012??

&c.