Car Nack Predicts 2011-1

[nuggetsyl]

Originally Posted by Kevin Sevcik

'07 was a rather lot easier than '11, though. Here's the rundown of what makes hanging lots harder this year:

1. You have to hang 12 tubes for this challenge, not 10.
2. Most of your tubes are always 50 feet away, not 15.
3. Much less room for multiple robots to work the grid. It's pretty much 1 per half.
4. Lose just one of your tubes to the opponents and this is impossible till you steal the same tube back.
5. The more you succeed at this the harder it becomes as your vision gets blocked.
6. You have a shorter window to accomplish this. You HAVE to try the minibot race because a full rack is only 20-30 more than 1st and 2nd place. An 8th ringer completely dominated ramping.

I'd actually really like to expand on the last point because it's what makes this such an incredibly easy prediction for Car-Nack. It's a pretty simple game theory analysis. There's extremely little incentive for actually completing all 6 logos. Specifically, there's almost no incentive for finishing a bottom row. Allow me to explain. Lets assume that your alliance would take the whole match to finish the grid, averaging 20 seconds per row. There's 2 options, spend the last 20 seconds finishing the last bottom row, or spend it lining up for the minibot race. The choice here is obvious. A completed bottom row is only worth 6 points. The two slowest minibots on the field still add up to 25 points. It's not even really a contest, you HAVE to play the minibot race. So teams are only going to spend about 100 seconds trying to fill the grid, which makes it even more impossible.

This is markedly different from '07. Two ramped bots were worth 50 points, which is a dominant endgame strategy up until you're hanging your 7th ringer in a row. If you had to choose between hanging a 7th or 8th ringer, or ramping, you're spending those last 20 seconds hanging. That's just not the case this year.

I am not saying this will be common if i had to put a number on it i would say less then 5 times will you see the 6 first logos. But never i do no think so. Of coarse i am not CarNack so take it with a grain of salt

[Sean Raia]

Who is Car Nack?
From what i've gathered, he makes one of these predictions each year?

[Kims Robot]

Originally Posted by Sean Raia

Who is Car Nack?
From what i've gathered, he makes one of these predictions each year?

A little browsing through the Car Nack Forum and you will figure it out!

[Chris is me]

Car Knack didn't really get this one head on...

Congrats 148, 1503, 1538.

[NickE]

Originally Posted by Chris is me

Car Knack didn't really get this one head on...

Congrats 148, 1503, 1538.

Also done by 195, 254, 469 in Galileo Q-56.

[Bjenks548]

Also done yesterday at Marc with 67, 245, and 469 with 25 seconds to spare. Q-1

[Koko Ed]

Originally Posted by Bjenks548

Also done yesterday at Marc with 67, 245, and 469 with 25 seconds to spare. Q-1

and again this morning with 67, 2337 and a third team I am unable to recall.

[Adam Freeman]

Originally Posted by Koko Ed

and again this morning with 67, 2337 and a third team I am unable to recall.

51.

Very fun to be a part of both alliances.

[EricH]

Car Nack is one of the few strategy gurus with more than one Championship. To be exact, his team has the most Championship wins of any team in FRC.

If Car Nack says something related to strategy, everyone will be well-advised to listen to him.

[Sean Raia]

Woah, thats really... interesting.
So im guessing he made this position based on the time it takes to bring a logo piece across the field multiplied by 18(number of pieces needed to make 6 logos) and divided by 3(number of teams in an alliance) which gives us 6.

(135 in a match) - (15 autonomous) -(10 for end-game).

so 110/6 which lends about 18 seconds to place a piece (to break his prediction). I think Car Nack is pretty darn good at predictions, but fails to account for the possibility of a robot that launchs pieces across the field to their teammates. Still, i have a feeling hes right.

Note: i didnt account for the placing of ubertubes in autonomous which slightly lowers the time frame needed to place 6 logos.

[Kevin Sevcik]

Originally Posted by Sean Raia

Woah, thats really... interesting.
So im guessing he made this position based on the time it takes to bring a logo piece across the field multiplied by 18(number of pieces needed to make 6 logos) and divided by 3(number of teams in an alliance) which gives us 6.

(135 in a match) - (15 autonomous) -(10 for end-game).

so 110/6 which lends about 18 seconds to place a piece (to break his prediction). I think Car Nack is pretty darn good at predictions, but fails to account for the possibility of a robot that launchs pieces across the field to their teammates. Still, i have a feeling hes right.

Note: i didnt account for the placing of ubertubes in autonomous which slightly lowers the time frame needed to place 6 logos.

Ubertubes don't count towards making a logo. Also, remember that a robot feeding its partners isn't scoring, and that you have exactly as many tubes as you need to complete both grids. Your feeder better not accidentally hand one off to the other alliance.

Plus, I think your time estimate is off. If you're only leaving 10 seconds for the endgame, you're in trouble. The poles are across the field, and you'd really wanna have time to line up before the last 10 seconds. Cutting 20 seconds off is probably reasonable for top flight teams that'd be trying this. Most teams will probably start setting up at 30 seconds.

[Alpha Beta]

If time is the only reason that 6 logos won't happen, it is worth considering alternate (closer) sources for game pieces. You could steal a game piece from the oponnents lane for a -3 penalty. If that piece then completes a logo on the top row it would gross you an additional 12 points and net you 9 after the penalty (an even bigger point bonus if ubertubes are involved.)

On a side note. It is possible to deploy your mini-bot from your lane where theoretically the defense is forbidden to go. But why wouldn't you take a 3 point penalty to prevent a 30 to 10 point score.

I actually like a game where smart penalties play into strategy. Great job GDC.

[EricH]

Originally Posted by Alpha Beta

On a side note. It is possible to deploy your mini-bot from your lane where theoretically the defense is forbidden to go. But why wouldn't you take a 3 point penalty to prevent a 30 to 10 point score.

Because it's not just a 3-point penalty. Messing with a robot deploying a minibot, especially if the deploying robot is contacting the base, is an automatic red card if it happens during the end game. No matter if your alliance wins, you don't!

During the end game, that would be a 3-point lane violation, a possible 3-point contact with a robot contacting tower violation, red card for the previous violation being intentional, and a red card for interference with deployment. 3-6 penalty points and 2 red cards for one move that prevents a 10-30 point score. That is not a smart penalty to take.

[Alpha Beta]

Originally Posted by EricH

Because it's not just a 3-point penalty. Messing with a robot deploying a minibot, especially if the deploying robot is contacting the base, is an automatic red card if it happens during the end game.

You are absolutely correct that interfering with a robot that is contacting the tower during the end game would be a poor choice. That would be the worst way to interpret the "smart penalty" concept. No one wants any part of a red card. Now defending the robot from getting to the tower even if it means going into the opponent’s lane to do it and eat the -3 is still an option. I expect to see hard defense in the end game. If I can delay a superior mini-bot with a penalty just enough to come in 2nd while my partner comes in first that is a 7 point swing, even more if I can delay him enough to not reach the top. Can you do that without drawing a red-card? Up until <G23> and <G25> comes into play.

[EricH]

Originally Posted by Alpha Beta

You are absolutely correct that interfering with a robot that is contacting the tower during the end game would be a poor choice. That would be the worst way to interpret the "smart penalty" concept. No one wants any part of a red card. Now defending the robot from getting to the tower even if it means going into the opponent’s lane to do it and eat the -3 is still an option. I expect to see hard defense in the end game. If I can delay a superior mini-bot with a penalty just enough to come in 2nd while my partner comes in first that is a 7 point swing, even more if I can delay him enough to not reach the top. Can you do that without drawing a red-card? Up until <G23> and <G25> comes into play.

If they touch the tower at any point during the end game, it's an automatic 3-point penalty in addition to any lane penalty. If the refs deem that you are interfering with deployment (i.e., they're starting to swing a deployment arm, and you hit them or shove them, even if contact was already begun), it's an automatic red card.

Up until the endgame, it's a viable strategy. But if you're doing that, I'll go do the same thing to one of your partners, or make a logo.

I also have another couple of tricks that could get me first in the minibot race despite your defense. And trust me, they'll be used by somebody. Smoke screens aren't always smoke...sometimes they're mirrors.