DISC diplomacy

[jvriezen]

This is one thing I haven't seem much discussion about. There will be lots of different situations, but here is a good example:

Your alliance in a qual round has the following bots:

Bot 1) A proven highly accurate full court shooter
Bot 2) Has a floor pickup and good short range shooter while touching pyramid
Bot 3) Has no floor pickup, a long range shooter, moderately accurate, but needs some tuning and/or driver practice. Bot 3 wants real field experience to continue to tune and show off its skills to aid in getting picked.

How do you allocate the white discs to human players and to robots (I'm assuming here that human players aren't necessarily going to only feed their own robots)

The best strategy is to give all the discs to Bot 1, and have Bot 2 pick up those that miss. Bot 3 ends up playing D and gets zero discs (other than pre-load) to show off/improve. Basically its a sort of 'starvation' situation, but within a single alliance.

And then, what happens when Bot 1 is heavily defended? How do you dynamically decide among human players (and coaches?) how many discs should go to Bot 2 or 3 instead, maybe on the opposite feeder station?

I suspect discs be split more evenly in practice rounds and this may be the only opportunity for some weaker teams to get a lot of discs on the real field, if they play 'smart' for the win during qualifications. The ones who suffer are those that are inoperable for a few matches and then get things fixed -- you have to convince your alliance mates that you are worth the risk to load discs.

Update: One way is to give each team 1/3 of the discs and if they are agreeable, they can give up their discs (either to another human before/during the match or by feeding a bot other than their own.

[dodar]

Or you just use bot 3 to throw bot 2 discs from the corner.

[vhcook]

My take on this is that you give bot3 some discs to shoot, but not as many as bot1. Bot2 can clean up both their misses, hopefully.

If bot1 is really that awesome, maybe you don't need much D to win. If it's a really tough match, you ask 3 to play more D, and maybe use 2 to help with disc starvation if that's appropriate given the opponents. Or you could use 3 as a passer to 2. There are lots of choices that lead toward victory and give everyone a reasonable chance to show what they've got, as long as everyone is willing to adapt based on what's happening on the field.

[Squeakypig]

If Bot 1 is extremely accurate, then bot 2 won't have enough discs to be a saturated scorer. Bot 2 should pick up missed opponent discs and pick up their own discs when there are enough there to be worthwhile. Bot 3 should play defense against bots who are playing defense against Bot 1 so it's insane accuracy can have full effect.

But, at the same time, it really depends on your opponents. Don't just scout who your alliance is going to be, scout the alliance you are going against.

[PayneTrain]

Usually a setup like that is a rarity outside of traditional regional eliminations or post-district tournaments. If you actually have those 3 kinds of robots allied in a random match against 3 other randomly selected robots, well, I feel sorry for the other alliance because there's a high probability they got a rare draw anyway.

If these three robots are in a situation like this, they are either getting what will likely be an easy win, or the QP bounty for a win at all costs is more important anyway. Most times in qualifiers an alliance like this will have an easy time getting the win anyway so "spreading the wealth" shouldn't be an issue at all.

Championship qualifications is an entire other animal where Bot 3 should have refined the shooter to work right now, and if not, that's tough cookies because you won't even compete with/against over a third of the teams in your own field and your shot % and average points scored/match better be a head-turner for scouts if your QP score, team number or bot design aren't.

[Kevin Sevcik]

Originally Posted by jvriezen

This is one thing I haven't seem much discussion about. There will be lots of different situations, but here is a good example:

Your alliance in a qual round has the following bots:

Bot 1) A proven highly accurate full court shooter
Bot 2) Has a floor pickup and good short range shooter while touching pyramid
Bot 3) Has no floor pickup, a long range shooter, moderately accurate, but needs some tuning and/or driver practice. Bot 3 wants real field experience to continue to tune and show off its skills to aid in getting picked.

How do you allocate the white discs to human players and to robots (I'm assuming here that human players aren't necessarily going to only feed their own robots)

The best strategy is to give all the discs to Bot 1, and have Bot 2 pick up those that miss. Bot 3 ends up playing D and gets zero discs (other than pre-load) to show off/improve. Basically its a sort of 'starvation' situation, but within a single alliance.

And then, what happens when Bot 1 is heavily defended? How do you dynamically decide among human players (and coaches?) how many discs should go to Bot 2 or 3 instead, maybe on the opposite feeder station?

I suspect discs be split more evenly in practice rounds and this may be the only opportunity for some weaker teams to get a lot of discs on the real field, if they play 'smart' for the win during qualifications. The ones who suffer are those that are inoperable for a few matches and then get things fixed -- you have to convince your alliance mates that you are worth the risk to load discs.

Update: One way is to give each team 1/3 of the discs and if they are agreeable, they can give up their discs (either to another human before/during the match or by feeding a bot other than their own.

Luckily, that's a pretty rare Quals alliance you've put together there, so this sort of conundrum won't come up all that often. Still, if I was Bot 3, I'd take the role of running interference for Bot 1 and do the best job I could at it. Bot 1 is pretty likely to end up picking, so you obviously want to make a good impression. You're not making an impression on anyone as a good shooter, but then again you're not at the moment. You ARE making an impression as a competent and skilled defender/blocker etc., with potential scoring ability.

The only time this strategy doesn't payoff is if Bots 1 and 2 aren't the super duo they claim because they were lying about their capabilities. Not fun because your robot looks even less capable because you didn't do anything despite the apparent inability of your teammates. I've had that happen a few times, notably at 2011 Champs when we had an untested minibot deployment and deferred to other teams that claimed they had a sure thing. Who then completely failed to get off a minibot because they forgot to or their system wasn't nearly as reliable as claimed. Since we were better than average at tubing, I'm pretty sure we were passed up for elims because we didn't have a demonstrated reliable minibot system. Sometimes honesty hurts you, but I'd still rather be upfront with my alliance mates than be one of those teams that people write-off as unreliable and shady.