Value & Credibility of 4th robots on an Alliance

[thatprogrammer]

Originally Posted by waialua359

Here's the real challenge. How does an alliance politely nd respectfully ask an alliance captain to sit out, when its obvious or best the other 3 robots play in a specific match? I would guess that there are many instances every season where this should happen, but alliances limit themselves because its not an option anyone would consider on the table.

This depends on the alliance captain and if they are willing to switch out. I'd bet a large number of teams are willing to switch out, but the possibility never crosses their mind. Perhaps bringing up the possibility while noting the reasons for not wanting to play that team in a particular match would ve sufficient for most alliances.

[GKrotkov]

Originally Posted by waialua359

It worked out for you folks since you got to Einstein in 2015.
Same for us. In 2015, we got to Einstein where are 3rd alliance partner played only the 1st elimination match. The next 8 were played by our 4th alliance partner. No one was broken. We just decided to play the 3rd pick instead.

Here's the real challenge. How does an alliance politely and respectfully ask an alliance captain to sit out, when its obvious or best the other 3 robots play in a specific match? I would guess that there are many instances every season where this should happen, but alliances limit themselves because its not an option anyone would consider on the table.

I imagine this very much depends on the personalities of the alliance captain's key players, and the team's style of captaining an alliance. Seriously, this is probably pretty personal* and it'd be hard to give blanket advice for this.

For example, I distinctly remember a conversation we had with our Champs alliance partners. The subject of switchouts came up, and Sean's exact words were: "If someone's coming out, it'll probably be us." There are also alliance captains who built their alliance around a particular strategy, which may specifically involve their robot, and think that this gives the entire alliance the best chance of performing in eliminations. I've never had experience with a captain who runs an alliance like a direct democracy, but I'd be really curious to see that. Different styles of running an alliance aren't universally better or worse, just different. And probably tailored to how the captaining team works.

Given that, chances are the best way of bringing up subbing out the captain (assuming the captain didn't) involves etiquette, social situational awareness, and solid reasoning (perhaps backed up with data).

Tl;dr: How does an alliance ask their captain to sit out? Tactfully and with understanding of how their team works.

* regardless of whether it should be

[JesseK]

Originally Posted by waialua359

Here's the real challenge. How does an alliance politely and respectfully ask an alliance captain to sit out, when its obvious or best the other 3 robots play in a specific match? I would guess that there are many instances every season where this should happen, but alliances limit themselves because its not an option anyone would consider on the table.

This is a tough one. Hypothetically, you're not just asking the captain to sit - you're also asking the people in the stands to live with the decision as well. This year saw plenty of cases cases where the captain was one of the strongest qualifiers, but also one of the weakest in head-to-head elims.

[ASD20]

Originally Posted by Cothron Theiss

As EricH mentioned, I would word it as the alliance that plays all 4 teams would get the bonus, not just the team that plays the 4th team. Otherwise, you might see the second and third rounds of Alliance Selections turn into an absolute mess.
Either way, I'm not a huge fan of just an objective point bonus for playing all 4 robots, but depending on the game, it might work. I think an ideal solution would be for games to be strategically diverse enough and all teams capable enough that it becomes strategically worthwhile to play different teams. Or (I'm scaring myself as I type this) even a scoring system that changes from match to match. For example, there could be two *equally* difficult end games for teams to choose between at the end of the match. One end game would be worth, say, 5 points and the other worth 10. These point values would swap from match to match somehow, making it more advantageous for alliances to select 3rd and 4th robots that can perform different functions and swap between them.

But then again, this comes with the drawbacks of being hard to implement, hard to design for the GDC, lots of rules, lots of work. But hey, maybe things will move in that direction.

Right now, the best robots can perform pretty much every task the GDC comes up with. I think the best way to encourage making strategic picks and changing out robots is to create a game where the best robots actually can't do everything.

For this to happen, the GDC needs to create a game with conflicting tasks, like something that requires a large robot and something that requires a small robot or something like lifting your partners, so you have lifters and liftees. I know that they've done this second one before and I'm curious as to how many robots could do both). Even these examples are not great and I can already see a team like 148 taking on and off half of their robot between matches. Even if some teams are able to out-engineer the GDC, if the frequency of robots that can do everything is more like the number of robots at Champs who had a 2-ball auto this year, compared to the number who could go over 4/5 defenses, shoot a decent number of high goals, and climb, I think we will see a lot more strategy in eliminations.

[dirtbikerxz]

I just remembered a question I had from all of this talk about 4th robots.
People give me advice on this :

Say for example my team was picked by another team for their alliance. We get to a match where I STRONGLY believe that a field setup of the 2nd, 3rd, and 4th bot would have a better chance of winning that match (the captain bot sits out), how would I approach the alliance captains and propose my idea to them without sounding rude?

[Ernst]

Originally Posted by dirtbikerxz

I just remembered a question I had from all of this talk about 4th robots.
People give me advice on this :

Say for example by team was picked by another team for their alliance. We get to a match where I STRONGLY believe that a field setup of the 2nd, 3rd, and 4th bot would have a better chance of winning that match (the captain bot sits out), how would I approach the alliance captains and propose my idea to them without sounding rude?

I think that's where good scouting data can come into play. It's one thing to go to an alliance strategy meeting with just a strong belief; it's another to go with a strong belief backed up by spreadsheets and binders full of robot and match data. It's a lot harder to argue with someone who can actually support what they're saying.

[dubiousSwain]

Originally Posted by dirtbikerxz

I just remembered a question I had from all of this talk about 4th robots.
People give me advice on this :

Say for example my team was picked by another team for their alliance. We get to a match where I STRONGLY believe that a field setup of the 2nd, 3rd, and 4th bot would have a better chance of winning that match (the captain bot sits out), how would I approach the alliance captains and propose my idea to them without sounding rude?

I have personally struggled with this question before. I am of the opinion that the alliance captain has earned their spot and their judgement should not be subordinated, but I it is perfectly appropriate to bring this up in a strategy meeting. As long as you are respectful and have the best interests of your alliance as a whole in mind, you owe your alliance the diligence of making your opinion known.

However, once the decision is made, it should not be publicly questioned. Doing so hurts the unity of the alliance and ultimately decreases your chances of winning.

This is not just a situation for robotics. You should keep this in mind whenever dealing with superiors and subordinates. I remember a great quote from Kingsmen: The Secret Service:

Originally Posted by Merlin

No no no. You don't talk to me like that. If you have a complaint, you come here and whisper it in my ear.
...
You need to get that chip off your shoulder.

This is especially important when dealing with children. There is nothing worse to a child than being embarrassed in front of their peers. My personal motto is "Praise in public, discipline in private"

Ultimately it comes down to tact. Tl;Dr: Don't air your dirty laundry in public.

[Kevin Leonard]

Originally Posted by waialua359

Here's the real challenge. How does an alliance politely and respectfully ask an alliance captain to sit out, when its obvious or best the other 3 robots play in a specific match? I would guess that there are many instances every season where this should happen, but alliances limit themselves because its not an option anyone would consider on the table.

This is an interesting topic.
In 2014, at the Archimedes Division, Team 20 was on an alliance where we wanted to bench the alliance captain.
The Alliance was 4077, 195, 20, and 4265.

Things got very heated during the strategy meeting, and in the end we had 4077 play the first match, and when that didn't succeed, we tried subbing them out. We ended up losing the quarterfinals series on foul points, and to this day I wonder how we could have done that differently, because on paper our Alliance was excellent.

This past weekend, at IRI, we had a discussion about potentially benching 3620 in favor of 3683 after we lost our first semifinal match. In the end, we decided to keep 3620 in, but there were definitely some upset individuals involved.

There's a combination of factors in play here, ranging from pride to the differing motivations teams have for playing. I don't know the best way to approach it, or anything, but I know you have to be careful when people's prides and emotions are at stake.

[TheBoulderite]

The fourth robot isn't always a pickup from the very bottom of the capability pool or rankings. For example, our alliance at IRI consisted of 1241, 133, and 868. 1241 was ranked 30th, 133 was ranked 29th, and 868 was ranked 20th. While we had 1241 and 133 play, 868 could easily have substituted for one of the two. 868 could put up a lot of high goals, on par with the other robots in our alliance.

[Ernst]

Originally Posted by TheBoulderite

The fourth robot isn't always a pickup from the very bottom of the capability pool or rankings. For example, our alliance at IRI consisted of 1241, 133, and 868. 1241 was ranked 30th, 133 was ranked 29th, and 868 was ranked 20th. While we had 1241 and 133 play, 868 could easily have substituted for one of the two. 868 could put up a lot of high goals, on par with the other robots in our alliance.

um

Originally Posted by TheBoulderite

at IRI

[TheBoulderite]

Originally Posted by Ernst

um

It may have been at IRI, but it goes to show that the fourth bot isn't completely useless.

[EricH]

Originally Posted by TheBoulderite

It may have been at IRI, but it goes to show that the fourth bot isn't completely useless.

A 4th bot at IRI is an early second-round pick at worst anywhere else.


I do know of some cases of an AC sitting out, but the one I most remember happened back in the 2v2, 3/alliance days when the AC swapped out with the 3rd robot and the 2nd robot played every match. That one resulted in a regional win (AZ '04).

[BrennanB]

Originally Posted by Ernst

um

Originally Posted by TheBoulderite View Post
at IRI

As opposed to the hundreds of regional/district events that have 4 robot alliances?

Like what the heck? You can't use the regional level baseline and apply that to 4 robots on an alliance, because they aren't remotely comparable. Obviously.

[EricH]

Originally Posted by BrennanB

As opposed to the hundreds of regional/district events that have 4 robot alliances?

Like what the heck? You can't use the regional level baseline and apply that to 4 robots on an alliance, because they aren't remotely comparable. Obviously.

*chuckle*

OK, serious question: How many total 4-robot alliances were there at the regional/district level (and DCMPs) this season? I can account for one (at Ventura, semis IIRC).

Now, how many were there at CMP? (64)
IRI? (8)

I don't think I'd be willing to bet on IRI having more 4-team alliances than the entire regional/district level, but I'll go with IRI having more than the regional portion of that.

[blueyoshi256]

Just wanted to weigh in on the "When should the captain sit out" discussion.

We were the captain of the 4th seed on Tesla. We played exactly 1 match in Elims, because that was the best strategy.
Both 2415 and 3130 could shoot from the outerworks. 3042 was a crazy good defense bot (see SF for proof). Our bot had only rarely been enough of a threat to warrant defense, so we weren't great against it (we got in off of the waitlist, and made a massive upgrade). It made sense to run 3130, 2415, and 3042 in elims, so we did.

As an additional bonus, our robot has technically never played in a match and lost at champs (more from luck than anything else).