PDA

View Full Version : [Official 2007 Game Design] Radical Tournament Ideas


dlavery
05-15-2006, 03:41 PM
This thread is a spin-off of this discussion (http://www.chiefdelphi.com/forums/showthread.php?p=501399#post501399), and has been started to focus on radical tournament structure changes. This thread is intended to collect innovative ways to structure tournament play. Using previous years as an example, this might include ideas to add human players to a robot-only format, or to change the three robots playing at once to a three-team alliance format. Sizes of alliances, lengths of matches, number of matches at a tournament, etc are all open for discussion. Like the above thread, this thread is meant to collect creative ideas that can be applied to any game concept.

-dave

Freddy Schurr
05-15-2006, 03:55 PM
Well, I have an idea for the Championship Competition. I was thinking a field only for the Regional Winners/Finalists, that way the best of the best compete on one field and move to compete against the other divisions.

Also for the finals for regionals and championship, add an additional robot to a three team alliance. This would help bring more robots into play and create interesting matches.

Example

Quarterfinal Match 1

Red Alliance : 25, 195, 1114 ( Robot on the side: 204)
vs
Blue Alliance: 486, 365, 341 ( Robot on the side: 395)

Quarterfinal Match 2

Red Alliance: 25,204,195 (Robot on the side: 1114)
vs
Blue Alliance: 365, 341, 395 (Robot on the side: 486)

Tomasz Bania
05-15-2006, 10:56 PM
Well, I have an idea for the Championship Competition. I was thinking a field only for the Regional Winners/Finalists, that way the best of the best compete on one field and move to compete against the other divisions.

Also for the finals for regionals and championship, add an additional robot to a three team alliance. This would help bring more robots into play and create interesting matches.

Example

Quarterfinal Match 1

Red Alliance : 25, 195, 1114 ( Robot on the side: 204)
vs
Blue Alliance: 486, 365, 341 ( Robot on the side: 395)

Quarterfinal Match 2

Red Alliance: 25,204,195 (Robot on the side: 1114)
vs
Blue Alliance: 365, 341, 395 (Robot on the side: 486)

That sounds like a great way to do that 4 team alliance without making the field bigger, but could introduce some unfairness.

I Believe there should be three alliances of either two or three teams.
It would be an interesting twist as determining who's the winner could be harder depending on the game.

Tomasz Bania

Cory
05-16-2006, 12:48 AM
Get rid of the serpentine draft.

Bill Moore
05-16-2006, 07:23 AM
Well, I have an idea for the Championship Competition. I was thinking a field only for the Regional Winners/Finalists, that way the best of the best compete on one field and move to compete against the other divisions.

Also for the finals for regionals and championship, add an additional robot to a three team alliance. This would help bring more robots into play and create interesting matches.

Example

Quarterfinal Match 1

Red Alliance : 25, 195, 1114 ( Robot on the side: 204)
vs
Blue Alliance: 486, 365, 341 ( Robot on the side: 395)

Quarterfinal Match 2

Red Alliance: 25,204,195 (Robot on the side: 1114)
vs
Blue Alliance: 365, 341, 395 (Robot on the side: 486)
I would also encourage the selection of the "spare" bot by having an additional playoff partner, rather than the current method. Going back to the older method of having one bot sit out each playoff match allows 8 more teams to compete in the playoffs. More teams are rewarded, and the bot on the bubble doesn't sit there and wait -- not wishing bad luck on anyone, but hoping to get an opportunity to play.

Cory
05-16-2006, 01:26 PM
I would also encourage the selection of the "spare" bot by having an additional playoff partner, rather than the current method. Going back to the older method of having one bot sit out each playoff match allows 8 more teams to compete in the playoffs. More teams are rewarded, and the bot on the bubble doesn't sit there and wait -- not wishing bad luck on anyone, but hoping to get an opportunity to play.

It also means that with a 3 team alliance, the minimum number of teams at a regional has to be 32, and not 24.

I can't see that happening, because there are numerous regionals that can barely get 24 teams, letalone 32.

Rick TYler
05-16-2006, 01:41 PM
I would like to suggest a rule for this thread: no criticizing or critically evaluating ideas. Let the weird concepts flow!

Jon Jack
05-16-2006, 01:56 PM
I really don't see the benefit of going 1-8,8-1 in alliance selections. Is there data showing that this drafting method helps lower seeded teams? How many regional winners were 1-4 seeds in 2005 vs 2006?

What would be an interesting twist is not allowing an alliance captain to pick another alliance captain. This is probably a better way of 'leveling' the field for the finals. I would imagine that many alliances that win regionals consist of 2 top 8 teams.

Rick TYler
05-16-2006, 02:03 PM
You want some out of the box ideas?

1. Robot size restrictions changed to a sum. Rather than 28x42 (or whatever) make it limited to a footprint of a certain number of square inches. FIRST inspectors are clever people -- they'll figure out how to measure them.

2. Create a total weight limit for the robot or robots fielded by a team. Only one robot can be radio controlled, the others have to be autonomous. For example, you might build an 80-pound "main" robot with two 20-pound autonomous slaves as helpers.

3. Have a "house" robot on the field, running under autonomous control that:

Plays defense
Plays offense
Just drives around aimlessly
Is attracted to other robots and gently nudges them
Something else?


4. Place random obstacles on the field, which move from match to match. These range in size from Bots dots to shoe box in size. Some of these light up and can be used as scoring objects during autonomous.

5. Have powered, mobile scoring objects. You don't just pick them up -- you have to hunt them down first.

6. Have a variable number of teams on each alliance. The tournament has three rounds instead of two, with four robots per alliance in round 1, three in round 2 and only two in the finals.

7. Replace the finals best-two-of-three single-elimination format with a double-elimination tournament. All alliances which lose in the first round finals games are put back in a pool, and new alliances are picked from the first-round losers. These newly-formed alliances are then reinserted into the "loser's bracket" of the double-elimination tournament. The winning alliances in the first round are left intact.

8. Instead of doing alliance selection as they now, allow the top seeded teams to bid on partners using virtual bucks. Give each team $100 virtual dollars and have a silent auction for alliance partners. Team numbers and a picture of their robot are posted on a wall (or other conveniently flat surface) and the top-seeded teams then enter their bids on a bidding computer. A top-seeded team (alliance captain) on which other teams have bid can choose to allow themselves to be "sold" or not. If they choose not to accept the high bidder, they are "off the market" and will be an alliance captain. The high bidder gets their bucks back and can rebid them. Maybe this is done in real-time from eight bidding computers and the current bid amounts are displayed on the big screen. I know this seems complicated, but I want to noodle on it a bit. I like the games theory aspect, and it would introduce teams to a lot of modern mathematical theory they don't normally see in FIRST -- but it needs work. I promise to post a more fully-developed version later.

9. Instead of just scoring won/lost records, allow the accumulation of "team points" similar to some auto racing series. This might work best if there were four alliances of two robots each on the field at the same time. The first place alliance would get 12 points, second 6 points, third 3 points, and fourth 0 points.

GeorgeTheEng
05-16-2006, 02:52 PM
It also means that with a 3 team alliance, the minimum number of teams at a regional has to be 32, and not 24.

I can't see that happening, because there are numerous regionals that can barely get 24 teams, letalone 32.

There a number of interesting ideas flowing here, but I would like to provide an additional caution similar to the above. More robot per match generally requires more floor space... Some venues are close to maxed out with the current field. It may not seem like a lot, but 2 or 3 feet in some cases makes a lot of difference.

That being said, how about a format similar to this year's periods with the difference that in each period something on thie field (goals, end zones, etc) or even robots themselves change alliance randomly.

Ryan Foley
05-16-2006, 03:16 PM
You want some out of the box ideas?

1. Robot size restrictions changed to a sum. Rather than 28x42 (or whatever) make it limited to a footprint of a certain number of square inches. FIRST inspectors are clever people -- they'll figure out how to measure them.

or for even more fun, simply express the starting size as a volume, for example, a 28x38x60in robot has a volume of 63840 cubic inches. A little harder to measure, but could spawn some very creative ideas.

Some other twists:
1) Call me crazy, but why not have house rules? At every regional, or just each weekend of regionals, the game would be slightly different than the other regionals/ weekends. Why? It would be harder to figure out basic strategies just from watching another regional. Then, at championships, all 4 divisions have house rules, maybe the same ones as some of the regionals, with Einstein being some brand new house rule, to give the division champions something new and challenging. It could be length of rounds, how to win automonous, point values, anything (just not robot rules). Oh, and don't reveal the house rules for that weekend until the Monday or Tuesday of each week, to really keep teams on their toes. Then Dave Verbugge could change the target colors on Einstein.

2) Someone suggested this one of the prior years, but what about tag-team alliances. Then you could get 4 robots on an alliance at a time, but only with 3 active. Teams could switch which robots are active manually, or the field would if the 4th robot hasnt gone active within a certain time limit. For even more fun, let the field randomly choose a robot on each alliance to deactivate at the 1minute remaining mark and activate the 4th robot on each alliance.

3) Don't give each team a list of who is in every match and on which alliance. Simply give teams a list of which mathes they are in, without their partners or opponents listed. This way, teams will need to be able to strategize with alliance partners in the "on deck" stage, with only 2-5 minutes before they go on the field. It's more fun that way.

EDIT in response to post 12 from Andy,
Ok, yes #3 isnt a "new" idea, but it would certainly be a radical change to the complete match lists we have had since 2002. The lists FIRST gave out in 1999-2001 had a list of 12 teams on it, so you had a slight idea of who you might be with. #3 suggests giving teams absolutely no idea who they are with. It's another way for FIRST to raise the bar.

Andy Baker
05-16-2006, 03:16 PM
That being said, how about a format similar to this year's periods with the difference that in each period something on thie field (goals, end zones, etc) or even robots themselves change alliance randomly.

Ooooh... now there is an interesting idea, adding onto last year's "periods" introduction. I'll add on to it:

Make one (or more) of the periods "contact free".

For example:

Period 1: autonomous period (could be contact free, or not)
Period 2: Full contact, free-for-all scoring
Period 3: non-contact, free-for-all scoring
Period 4: Full contact, free-for-all scoring

Period 3 would be the "weenie period" where the offensive robots would dominate, and the box-on-wheels would wimper. Once period 4 started, the offensive robots would have to definitely keep their eyes open.

Andy B.

Andy Baker
05-16-2006, 03:24 PM
3) Don't give each team a list of who is in every match and on which alliance. Simply give teams a list of which mathes they are in, without their partners or opponents listed. This way, teams will need to be able to strategize with alliance partners in the "on deck" stage, with only 2-5 minutes before they go on the field. It's more fun that way.

(must... resist... critiquing... Ryan's idea... aaaah!)

While this may appear to be "fun", it is not new. FIRST did this to us already, from 1999-2001.


Andy B.

Lil' Lavery
05-16-2006, 04:17 PM
I really don't see the benefit of going 1-8,8-1 in alliance selections. Is there data showing that this drafting method helps lower seeded teams? How many regional winners were 1-4 seeds in 2005 vs 2006?

What would be an interesting twist is not allowing an alliance captain to pick another alliance captain. This is probably a better way of 'leveling' the field for the finals. I would imagine that many alliances that win regionals consist of 2 top 8 teams.
I'll explain why that wouldn't work. Even though it isnt GP, it would happen. Teams would purposely throw matches to get out of the top 8 so that a high seeded team could pick them.

Lil' Lavery
05-16-2006, 04:27 PM
Variable number of alliance partners per match.
Match 1 is 2v2
Match 2 is 4v4
Match 3 is 1v1
etc.

Have mini-alliance selections before each match
Each match has a random team assigned Red Captain and Blue Captain, then each of those picks the rest of their partners from the other teams in the match. They flip a coin or something to see who gets first pick.

Andrew Blair
05-16-2006, 04:28 PM
How about this:

Consider a three robot alliance, all starting the game under driver control-No autonomous...yet. Throughout the match, one robot on each alliance at a time is put into autonomous mode, while the other two/four play normally. You know what period your robot is going to go auto, so you can line it up appropriately, or code a really cool system that allows you to start anywhere.

Problem is giving auto an incentive.

Andy Baker
05-16-2006, 04:30 PM
Andy,
I don't think you quite finished your thought. ;)
JV

True... but according to this advice:

I would like to suggest a rule for this thread: no criticizing or critically evaluating ideas. Let the weird concepts flow!

I was trying to encourage the weird concepts. Dave wants weird. Rick wants to encourage weird. Viva la weird!

Andy B.

santosh
05-16-2006, 04:46 PM
I wouldn't mind seeing an allliance captain in the 1st round to ask to join the alliance of a c ranking above theirs.
For instance:
1st ranked team: team A
2nd ranked team : team B
Random other team: team C

The 1st pick of the draft by team A is Team C. This means that team B is next up. I woudn't mind seeing it possible to allow team B to ask to join team A instead of having to form their own alliance.

That is kinda far fetched I guess, howver I, like Cory would like to see FIRST get rid of the "serpintine" alliance selection.

Andy Baker
05-16-2006, 05:18 PM
OK... another idea on alliance selection, as opposed to the boring 1-8,1-8 and the serpentine draft:

1st round: 1-8, just like teams do now
2nd round: Teams draw draft order numbers out of a hat: it could be 5,8,2,3,7,1,4,6

This 2nd round order selection could take place immediately after the 8th alliance captain picked their first partner. No one would know who was picking next. The #1 alliance still has the advantage and benefit of being #1, but now their chances of having a more dominate alliance is less than it would be for a 1-8, 1-8 selection process. Hmmm... this could be interesting.

Andy B.

Bill Moore
05-16-2006, 07:32 PM
OK... another idea on alliance selection, as opposed to the boring 1-8,1-8 and the serpentine draft:

1st round: 1-8, just like teams do now
2nd round: Teams draw draft order numbers out of a hat: it could be 5,8,2,3,7,1,4,6

yada yada

Andy B.
You could draw the alliance draft from the hat like the NBA.

Top qualification seed has 8 chances in the hat, 2nd seed has 7 chances, 3rd = 6 chances, etc.

Then as the numbers are pulled from the hat for the draft order, the alliance seeds are set. i.e., the first number pulled becomes the first alliance, 2nd number becomes alliance 2, etc. (The 8th quals seed has a 1 in 36 chance to become the top playoff alliance.)

To further increase randomness, if the 2nd alliance captain picks the 4th alliance captain, the remaining alliances (5 through 8) do not "move up", but the 9th quals seed replaces the missing captain at the 4th playoff spot. (Yes, the 9th seed will be picking before the other remaining seeds, in this instance.)

This could really throw a monkey wrench into many teams scouting. It will be hilarious to watch them scramble.

Is this radical enough, Dave?

JackN
05-16-2006, 07:39 PM
You could draw the alliance draft from the hat like the NBA.
Is this radical enough, Dave?

That is cool. It would make it interesting though. What if the #8 seed got first pick, could they pick the 1 seed who has the last pick. it would make things very interesting.

Tetraman
05-16-2006, 08:23 PM
Evolution Field

1/3 of matches have a simple field
next 1/3 will add something to the field and another way to score
last 1/3 will include yet another something and another way to score
Elimination matches will also include something new an another way to score
The Finals matches will include one last thing.

And a game that is 4v2 or 5v1

The possibility to do this:

six robots on the field.
Robot 1 vs Robot 2
Robot 3 vs Robot 4
Robot 5 vs Robot 6
All on the same field, all at the same time.

Andy Baker
05-17-2006, 10:46 AM
OK... another idea on alliance selection, as opposed to the boring 1-8,1-8 and the serpentine draft:

1st round: 1-8, just like teams do now
2nd round: Teams draw draft order numbers out of a hat: it could be 5,8,2,3,7,1,4,6


Better than that...

1st round: 1-8, just like now
2nd round: order is determined by the 8 student pairs (the alliance captain and the person who accepted the 1st round pick) participating in a 2 vs 2 dodgeball tournament. The winner gets the 1st pick of the second round.

Oh boy, I wish we had time for this at IRI.


AB

Travis Hoffman
05-17-2006, 10:54 AM
Better than that...

1st round: 1-8, just like now
2nd round: order is determined by the 8 student pairs (the alliance captain and the person who accepted the 1st round pick) participating in a 2 vs 2 dodgeball tournament. The winner gets the 1st pick of the second round.

Oh boy, I wish we had time for this at IRI.


AB

Students collapse due to exhaustion from dodgeball tourney. Mentors get to drive robots in eliminations. Chaos ensues.

Travis Hoffman
05-17-2006, 10:55 AM
Students collapse due to exhaustion from dodgeball tourney. Mentors get to drive robots in eliminations. Chaos ensues.

And a mentors-only round at official events would be quite a sight.....

mjbrauck
05-18-2006, 12:00 AM
How about in the elimination rounds, instead of the same to alliances playing each other 3 times in a row you rotated teams. so it would go like this…


QF1 1st alliance vs. 4th alliance

QF2 2nd alliance vs. 3rd alliance

QF3 5th alliance vs. 8th alliance

QF4 6th alliance vs. 7th alliance

QF5 Winner QF1 vs. Winner QF2

QF6 Winner QF3 vs. Winner QF4

QF7 Loser QF1 vs. Loser QF2 (loser has lost twice and is eliminated)

QF8 Loser QF3 vs. Loser QF4 (loser has lost twice and is eliminated)

QF9 Loser QF5 vs. Winner QF7 (loser has lost twice and is eliminated)

QF10 Loser QF6 vs. Winner QF8 (loser has lost twice and is eliminated)

this leaves four teams the winners of QF5, QF6, QF9, and QF10 who move on to the semi-finals.

SF1 winner QF5 vs. winner QF6

SF2 winner QF9 vs. winner QF10

SF3 winner SF1 vs. winner SF2

SF4 loser SF1 vs. loser SF2 (loser has lost twice and is eliminated)

SF5 loser SF3 vs. winner SF4 (loser has lost twice and is eliminated)

This leaves two teams the winner of SF3 and SF5 who play in the finals.

Note: you could leave out the semi-finals and finish the tournament with

QF11 winner QF5 vs. winner QF9 (loser has lost twice and is eliminated)

QF12 winner QF6 vs. winner QF10 (loser has lost twice and is eliminated)

QF13 winner QF11 vs. winner QF12 (winner has won the entire tournament)

Jherbie53
05-18-2006, 05:32 PM
If there are still three teams on an alliance, then the first round could go 1-8. Then the second round would be a random drawing, either by a computer or the teams drawing from a hat, for the order of the picks. This could also be done twice if there are going to be four robots on an alliance. You could also have a random match list for the finals. After the alliances are selected there would be a drawing, like before, to determine who plays who. It could not only be just for the first round either, the semi's would be random too. Just imagine not knowing who you are playing until after their's or yours match.

If there are three teams on a alliance like this year, at the regionals that are bigger, like the one with two fields in Canada [don't recall the name], you could have four teams on the alliances. Then at the smaller regionals they would have just three on a alliance, instead of four. This would allow more teams to play at the larger regionals, 32 teams, and still allow smaller ones that have trouble getting enough teams, 24 teams, to play.

Last thing right now, On the match lists you could still have what teams are playing in each match, but not who's with who. You would find out the exact teams you are with and playing against at queuing. This would be hard to do strategy, because you just might talk to a team you are facing.

Tetraman
06-13-2006, 06:56 PM
Drivers and Human players are allowed to switch places. But the coach still can't touch a thing.

Billfred
06-13-2006, 07:31 PM
I suppose that a few ideas can be had from "Aim, well, Kinda High", plus a few other ones:

1) Get rid of the serpentine draft. We used 1-4, 1-4 at Mission Mayhem, and things seemed to work themselves out properly.

2) The use of Tournament Rule #4 accomplished its intended goal quite well. I'm sure that with more than four hours of warning, things could be even better.

Teams are expressly prohibited from saying “Team XXXX graciously accepts.” Any team representative who says this may be pelted mercilessly with Poof balls by the field crew until he or she accepts using some more creative (while still appropriate) wording.

3) The issue didn't come up at Mayhem, as all of the teams toughed it out, but the timeline process for calling time-outs and replacements in the tournament confused me as a referee. I shudder to think what an alliance captain with a lot on the line has to do in the way of mental gymnastics. Would there be a way to have some sort of countdown timer to allow alliance captains and referees to see exactly how much time they have to make a decision on something?

4) Suppose that Friday morning, teams were given two or three cracks at an autonomous field, like that of FVC. Teams' average scores in these matches would correlate to bonus time at the end of each operator-controlled match, all the way through the end of the event. (This time would be theirs for keeps, regardless of how much or how little bonus time their alliance partners have.) In a game like Aim High, the conversion might be, say, one second for every five points scored, up to ten seconds.

Jherbie53
06-13-2006, 10:02 PM
Lets say the alliances have an extra robot, one thats not required for a "full field of robots" for a match, then heres a substitution rule. There are four teams on an alliance. After each match the captain team must sub in the other robot. In 2003 and 2004 you had to do this for you second match, but could play any two robots you wanted in the third match. I'm saying that in every match there would to be a substitution. The alliance captain chooses who's in and who's out. It doesn't matter if its match three, or match one against a new alliance, you sub every match.

I was also thinking that with every substitution the robot that was just put in stays in, meaning that the three robots that were selected would rotate, with the exception of the captain team. But that might make it a little confusing :confused: and I don't think that would go over well with teams.

Dan Petrovic
06-13-2006, 10:06 PM
Instead of eliminations being played as 1v8, 2v7, 3v6, 4v5

It would be 1v5, 2v6, 3v7, 4v8

So that 8 doesn't get completely dominated by 1.

Gabe
06-13-2006, 10:44 PM
The possibility to do this:

six robots on the field.
Robot 1 vs Robot 2
Robot 3 vs Robot 4
Robot 5 vs Robot 6
All on the same field, all at the same time.It would be really cool to have more than two alliances one the field. Instead of simply having 1-on-1 play, my idea is:

Six robots on the field.
Robot 1 & Robot 2 = Alliance A
Robot 3 & Robot 4 = Alliance B
Robot 5 & Robot 6 = Alliance C

All three alliances would be on the same field at the same time. :]

Billfred
06-13-2006, 11:41 PM
Well, since I've heard plenty of stories about ganging up on the top team in a 1v1v1 game, suppose that we go to 2v2v2 with a twist.

Matches become three minutes in length, plus whatever autonomous tomfoolery you'd like to add. Once driver control begins, one alliance is disabled for the first minute of the match, and the other two slug it out. Then another alliance is disabled for one minute, and the first alliance is enabled. Then, at the two-minute mark, the third alliance is disabled and the second alliance comes back in.

Assuming that there is no descoring in the game, it becomes very hard for the two weaker alliances to gang up on the stronger one; you're either facing them head-on or facing the other weaker alliance, depending on the phase of the game. It also increases the strategy element, especially if you aren't sure beforehand when you will be disabled. If you were to find out out after autonomous, you then have to adjust your plans on the fly; get the first disable, and your human player might have more time to load you up. Get the last disable, and you're not only going to have to go and hang from that bar a little earlier, you'll also have to figure the best position to stay on the bar.

JackN
06-14-2006, 11:15 AM
I was thinking about the random alinace choosing and here iare latest Ideas.

First Pick-#1
Second-Seventh-Random
Last Pick-#8

Second Round Picking goes based on combined record both teams 1-8

It makes sure the best team still gets the first pick. But other teams get a chance at having high pick

Other Idea

Captains 1-7 Pick then the second round again 1-7 then a random aliance of however many robots is picked, Highest seeded team not picked is automaticly the captain.

This creates a true underdog aliance, and gets more teams involved.

Last Idea

Pre-set aliances

Every team has same aliance during the whole competition. You pick your aliance before you attend and you and your group battle through together. It creates the ultimate teamwork and Co-Opertition form because you know who and how your partners will play. Scouting is much easier because you can scout a whole aliance instead of individual teams.

santosh
06-14-2006, 11:48 AM
How about this.
Instead of only the top 8 alliances getting to play in the playoffs, everyone gets to play.
You are seeded by your rank at the end of qualification matches.
The top 8 teams select based upon the same way as Mr. Bill Moore said.
So that right there takes out 24 teams out of the pickings.

Now the next top 8 seeds (if there are atleast 24 teams left) that haven't been picked yet, get to choose teams that are left over in the same fashion, and so on and so forth till all the teams get picked.
The number of alliance captains in each of the sets of the pickings depends on the number of teams left to pick. Does that makes sense? Kinda confusing.

I relaly would like to see 4 teams alliances like JHerbie53 said. Kinda like the pre 6 bots on the field era. Like where we had 3 teams per alliance, but only 2 played on the field at once, except we would have 4 team alliance where 3 played on the field.

josh s
06-14-2006, 04:37 PM
Ooooh... now there is an interesting idea, adding onto last year's "periods" introduction. I'll add on to it:

Make one (or more) of the periods "contact free".

For example:

Period 1: autonomous period (could be contact free, or not)
Period 2: Full contact, free-for-all scoring
Period 3: non-contact, free-for-all scoring
Period 4: Full contact, free-for-all scoring

Period 3 would be the "weenie period" where the offensive robots would dominate, and the box-on-wheels would wimper. Once period 4 started, the offensive robots would have to definitely keep their eyes open.

Andy B.

This would definitely allow the best teams shine. I also like the idea of adding a fourth team to the alliance to give teams more flexibility with their strategy to hopefully make for more competitive matches.

Andrew Blair
06-14-2006, 07:21 PM
Ooooh... now there is an interesting idea, adding onto last year's "periods" introduction. I'll add on to it:

Make one (or more) of the periods "contact free".

For example:

Period 1: autonomous period (could be contact free, or not)
Period 2: Full contact, free-for-all scoring
Period 3: non-contact, free-for-all scoring
Period 4: Full contact, free-for-all scoring

Period 3 would be the "weenie period" where the offensive robots would dominate, and the box-on-wheels would wimper. Once period 4 started, the offensive robots would have to definitely keep their eyes open.

Andy B.
Perhaps, if we add in a "weenie period", we could add in a full contact-no dynamic scoring period. For example, take Stack Attack. The weenie period could be in the beginning of the game, but at the last, say, 30 seconds of the game, you would have a perfect opportunity to fight for the ramp. It would certainly make the last minute ramp/bar battles more exciting, but at the cost of an exciting last minute score. Spectators love violence...

UCGL_Guy
06-21-2006, 02:17 PM
5 minute periods - teams would be forced to manage their power and would give more total field time with less traveling between. Maybe include a 1 minute halftime period to make adjustments (no battery change), change partners - during the second half you are partnered with the other alliances bots so you make a new parternership.
Autonomous selection of alliance partners - robots can be programmed to activate another machine to be on their alliance and vice versa - with default selections. this could be done with light colors or shapes or "playing tag"
Weird yes but I like weird

Freddy Schurr
06-21-2006, 02:35 PM
This is just a radical idea coming from my mind.

I think that the Top 8 should not be able to pick in the 8. It would make the competition fairer and enjoyable to watch.

Nuttyman54
06-21-2006, 02:44 PM
5 minute periods - teams would be forced to manage their power and would give more total field time with less traveling between. Maybe include a 1 minute halftime period to make adjustments (no battery change), change partners

I'd be concerned about matches taking way too long. While it would be interesting, you'd have to make the tournament last all week in order to give everyone enough matches


during the second half you are partnered with the other alliances bots so you make a new parternership.

A play on this that one of my teammates mentioned after the changing light thing at champs: Switching alliance partners with 2v2v2. You play the 1st period with team A verus B and C versus D and E. Second period (no breaks in between), the first parnter (You, B and D) all rotate so that it's You and C verssus B and E versus A and D. one more period (and rotation) makes the cycle complete. There are flaws, i know (You, B and D never play together), but it would be interesting to try to play.

lukevanoort
10-05-2006, 09:20 PM
This idea would only work for the larger regionals (like VCU and GTR) or the Championship. The top sixteen teams each pick a partner who is NOT a top sixteen member. (picks going in order 1-16) Then a 2v2 version of the game is played bewteen alliances in a 1vs16, 2vs15, 3vs14, etc. pattern. One match is played and the winner gets to pick a team from the losers alliance to join them as a full three member alliance. Then the winners play a normal elimination ladder, where the teams are ordered by the lead teams seed. For example, #4 seed Team 13 picks Team 1121 and #13 seed Team 1517 picks Team 37. #13 beats #4 and chooses to take team 1121. The #13 alliances then is 1517, 37, 1121. Lets say 1, 2, 3, 13, 11, 10, 6, 8 alliances win. Eliminations are then played as 1 vs 13, 2 vs 11, 3 vs 10, and 6 vs 8. From there on it follows the current format.

Billfred
10-24-2006, 04:19 PM
One thing that I would really like to see officially sanctioned in the manual are some rules for when events are ginormous (like GTR or some of the other larger regionals) or really tiny (read: under 24 teams). I know that GTR had twelve alliances (and thus a few eighth-final rounds), so it's not unheard of.

Another thing I'd like to see return is that of the clock-stopper, like in 2001, perhaps with the final score of an alliance divided by the number of robot-seconds used. Lots of the stuff getting crammed into the heads of USC business majors has to do with measurements per machine-hour, labor hours, units produced, et cetera. I can only assume that they're teaching us stuff similar to that found in the Real World. (You could extend this further by penalizing teams who waste their raw materials, like their tetras from 2005, but we'll save that for another time.)

Finally, one that might be interesting to see is a game that involves something like (American) football-style plays, with a bit of starting and stopping over the match. Suppose, for example, that when a robot on an offensive robot makes contact with the side barriers, play stops and both sides assemble again on a line of scrimmage near the site of contact and play resumes.

LordTalps
10-24-2006, 07:32 PM
Three upper goals, using the shifting lights, to change which one you need to hit for max points. Talk about programming having more fun and more impact :D

Andrew Blair
11-04-2006, 11:18 PM
I'd like to see the human player engaged more in physical competition. I know that FIRST isn't a physical competition, but too often, the human player gets stuck in a secondary role, where they do little, let alone affect the outcome of the match. 2004 and 2006 were exceptions.

An engaging game would include something like an obstacle course, or button reaction and subsequent action, where the first human player or alliance to complete the task, or to continue completing the task gets a multiplier, or a more lucrative scoring opportunity. Spectators like to see physical involvement, because it shows human side of the competition a bit better. If we want FIRST to act more like sports- we need to include sport.

Steve W
11-04-2006, 11:34 PM
I know that GTR had twelve alliances (and thus a few eighth-final rounds), so it's not unheard of.




GTR actually had 16 alliances. We played like a regional on each field (8 alliances) and then the winners from each field met for the finals.

Peter Matteson
11-06-2006, 07:49 AM
I'd like to see the human player engaged more in physical competition. I know that FIRST isn't a physical competition, but too often, the human player gets stuck in a secondary role, where they do little, let alone affect the outcome of the match. 2004 and 2006 were exceptions.

An engaging game would include something like an obstacle course, or button reaction and subsequent action, where the first human player or alliance to complete the task, or to continue completing the task gets a multiplier, or a more lucrative scoring opportunity. Spectators like to see physical involvement, because it shows human side of the competition a bit better. If we want FIRST to act more like sports- we need to include sport.

Along this line, and I know its too late for this year, but I always wanted to see a the human player have to run on a treadmill or ride a staionary bike and maintian a minimum power output, say 150 watts. If you fell below this your robot would be handicapped by limiting max voltage available to the motors.

Billfred
11-08-2006, 11:03 AM
I thought about this while running to my business law class (remember kids, don't forget to make sure your alarm will actually go off in the morning): What about a "knockout punch" that, if achieved, would end the match immediately in favor of the alliance that achieves it?

The nearest example I can think of is if the center goal in Aim High was cleared by a human player instead of a field crew member. (Yes, I realize this introduces other issues, like using the stick to remove balls from the goal, but play along here.) If one alliance clogged the center goal to the point that the human player could not clear the clog within ten seconds, the match would end immediately.

In Triple Play, an example might be maintaining ownership of all nine goals at one time.

Lil' Lavery
11-08-2006, 03:44 PM
I thought about this while running to my business law class (remember kids, don't forget to make sure your alarm will actually go off in the morning): What about a "knockout punch" that, if achieved, would end the match immediately in favor of the alliance that achieves it?

The nearest example I can think of is if the center goal in Aim High was cleared by a human player instead of a field crew member. (Yes, I realize this introduces other issues, like using the stick to remove balls from the goal, but play along here.) If one alliance clogged the center goal to the point that the human player could not clear the clog within ten seconds, the match would end immediately.

In Triple Play, an example might be maintaining ownership of all nine goals at one time.

Remembers variable length matches in 2001..... oh god ;)

65_Xero_Huskie
11-14-2006, 10:50 AM
You want some out of the box ideas?

8. Instead of doing alliance selection as they now, allow the top seeded teams to bid on partners using virtual bucks. Give each team $100 virtual dollars and have a silent auction for alliance partners. Team numbers and a picture of their robot are posted on a wall (or other conveniently flat surface) and the top-seeded teams then enter their bids on a bidding computer. A top-seeded team (alliance captain) on which other teams have bid can choose to allow themselves to be "sold" or not. If they choose not to accept the high bidder, they are "off the market" and will be an alliance captain. The high bidder gets their bucks back and can rebid them. Maybe this is done in real-time from eight bidding computers and the current bid amounts are displayed on the big screen. I know this seems complicated, but I want to noodle on it a bit. I like the games theory aspect, and it would introduce teams to a lot of modern mathematical theory they don't normally see in FIRST -- but it needs work. I promise to post a more fully-developed version later.



I would think that this is considered gambling. Im not sure, but it sure sounds like it, and it would make the teams imbalanced even more than they are.

im not knocking it, i would love to see it happen, it would be a nice twist.

WE could also have the teams that arnt selected vote on the matchups for the playoffs so that they can get to see the matches they want.

Steve W
11-14-2006, 05:09 PM
GTR actually had 16 alliances. We played like a regional on each field (8 alliances) and then the winners from each field met for the finals.

I am bad. I was one of the people fighting for 12 alliances and my mind just .....

Getting old I guess. Here is the link to the ladder that I posted (THANKS Richard)

http://www.chiefdelphi.com/forums/showthread.php?t=46331

Billfred
11-19-2006, 03:55 PM
One change that might make for an interesting dynamic: Allow each team one five-second timeout, to be used over the course of the event. When the timeout is invoked, all play stops on the field to allow the alliance to strategize or regroup, particularly after a significant shift in gameplay happens. (The offensive muscle of the match tipping comes to mind.)

I imagine this being implemented as a second differently-colored button in the alliance station, next to the (individual) E-stop button. The usage could be tracked in scoring software (which also allows for the software to disable the button for teams that have used it). Alternatively, one operator badge could come with a coupon for the time-out attached, which would be given to a referee once it's been used. Stopping the match a second time, naturally, would be cause for a disqualification.

While it might throw off match scheduling a bit, a ten-second time-out used by forty teams (which is bigger than some regionals, and is about half of a divison in Atlanta) would add less than seven minutes onto the day. Regionals take up that much time playing the YMCA. ;)

Andrew Blair
11-19-2006, 08:37 PM
Here's an idea for something that still can be done:

The week prior to Kickoff, give us the game piece! We wouldn't need the hint (though they're so vauge anyway), and the rush to purchase game items, if an issue like the balls, could be spread out over another week. Teams still would have no idea what was being done in the actual game, so no serious work or prototyping could be done, but we could still see how large the piece was, how to interact with it, and how it interacted with other game pieces. It's just an idea, but I really think that it would be cool, get us thinking, and get teams excited and together a bit prior to the actual start of build.

Branden Ghena
11-22-2006, 11:15 PM
Why not make the game rules apply differently for each match, with the number of rules, and which ones, that apply chosen randomly before each match?

Dan Petrovic
11-22-2006, 11:34 PM
Here's an idea for something that still can be done:

The week prior to Kickoff, give us the game piece! We wouldn't need the hint (though they're so vauge anyway), and the rush to purchase game items, if an issue like the balls, could be spread out over another week. Teams still would have no idea what was being done in the actual game, so no serious work or prototyping could be done, but we could still see how large the piece was, how to interact with it, and how it interacted with other game pieces. It's just an idea, but I really think that it would be cool, get us thinking, and get teams excited and together a bit prior to the actual start of build.

How about just a silhouette of an isometric view just to make it even more interesting?

Try guessing the tetra after that curve ball!

Billfred
11-30-2006, 07:46 AM
So here's a notion I had the other day, one that might not be perfect. (dons asbestos underwear)

On Saturday, you tend to draw the largest crowds at regionals. In the afternoon, you also get some of the more exciting matches of a regional. But what about the morning?

Imagine, if you will, that the match list for Saturday wasn't posted until Saturday, like it's done on the other two days. But before the list goes out, the scoring software applies a little black magic to the list to organize the six teams in each match according to their seed. The six teams in a match wouldn't change (that'd be a scheduling nightmare), just possibly their affiliations within the match. So if we've got seeds number 2, 9, 11, 25, 30, and 42 in a match, it would be organized 2/11/30 versus 9/25/42.

This system isn't perfect--247 and 1251 were both pretty low seeds at Palmetto, and they're both wearing gold now--but it might help prevent a total squashing, which (all fairness debates aside) just isn't as interesting to watch.

EDIT: And another idea in a completely different vein, inspired by Rick TYler's post on page 1:
8. Instead of doing alliance selection as they now, allow the top seeded teams to bid on partners using virtual bucks. Give each team $100 virtual dollars and have a silent auction for alliance partners. Team numbers and a picture of their robot are posted on a wall (or other conveniently flat surface) and the top-seeded teams then enter their bids on a bidding computer. A top-seeded team (alliance captain) on which other teams have bid can choose to allow themselves to be "sold" or not. If they choose not to accept the high bidder, they are "off the market" and will be an alliance captain. The high bidder gets their bucks back and can rebid them. Maybe this is done in real-time from eight bidding computers and the current bid amounts are displayed on the big screen. I know this seems complicated, but I want to noodle on it a bit. I like the games theory aspect, and it would introduce teams to a lot of modern mathematical theory they don't normally see in FIRST -- but it needs work. I promise to post a more fully-developed version later.

So imagine if we took this a step further. The Top 8 would be prohibited from picking each other, which I'm normally in favor of allowing, but that would be too painful here. Each alliance captain would have X number of points with which to purchase their alliance partners. Note that I didn't specify how many--alliances in eliminations would not be limited in size, at least in theory. You could just bid on two great partners and hope nobody breaks, or maybe have four or five mid-card robots that you could switch between to suit your strategy.

What this doesn't include is some means for a team to decline (say, their robot's broken beyond hope of repair in time)--I assume you'd just notify the captains that they're out.

Graham Donaldson
11-30-2006, 12:59 PM
EDIT: And another idea in a completely different vein, inspired by Rick TYler's post on page 1:


So imagine if we took this a step further. The Top 8 would be prohibited from picking each other, which I'm normally in favor of allowing, but that would be too painful here. Each alliance captain would have X number of points with which to purchase their alliance partners. Note that I didn't specify how many--alliances in eliminations would not be limited in size, at least in theory. You could just bid on two great partners and hope nobody breaks, or maybe have four or five mid-card robots that you could switch between to suit your strategy.

What this doesn't include is some means for a team to decline (say, their robot's broken beyond hope of repair in time)--I assume you'd just notify the captains that they're out.

I like that idea. That would add another interesting element to strategy and scouting, and alliance selection. Do you want to max out your points on two robots and hope their abilities will work to your advantage, or do you want to have a variable-team alliance where you can switch out teams depending on the alliance you're playing? Hmm... :cool:

Richard Wallace
11-30-2006, 01:43 PM
EDIT: And another idea in a completely different vein, inspired by Rick TYler's post on page 1:


So imagine if we took this a step further. The Top 8 would be prohibited from picking each other, which I'm normally in favor of allowing, but that would be too painful here. Each alliance captain would have X number of points with which to purchase their alliance partners. Note that I didn't specify how many--alliances in eliminations would not be limited in size, at least in theory. You could just bid on two great partners and hope nobody breaks, or maybe have four or five mid-card robots that you could switch between to suit your strategy.This sounds fun. Kinda like a GM building a sports team by signing free agents, while restricted by a salary cap.

But I do think the higher seeded alliance captains should get a bigger budget. Say the number one alliance captain gets 135 bananas to bid with, number two gets 130 bananas, and so on, with number eight getting 100 bananas. Once you've spent your bananas, they are gone -- use 'em or lose 'em.

Joel J
11-30-2006, 01:52 PM
I was thinking about the serpentine system that I hate so much.. and started to wonder about the feasibility of using qualifying points (W-L-T) to determine the draft order in the first round, and then ranking points (32.28938) to determine draft order in the second round. I don't know whether it would be the RP of the alliance captain, average RP, or max RP between the two current alliance members that would be used to rank the alliances for the second round of the draft, but either way. I'm not sure about this idea, but I thought I'd mention it here.. That method of determining draft positions wouldn't really mess with the draft currency idea suggested above (warming up to the idea), as they still have to select in some order.

Peter Matteson
11-30-2006, 03:11 PM
I was thinking about the serpentine system that I hate so much.. and started to wonder about the feasibility of using qualifying points (W-L-T) to determine the draft order in the first round, and then ranking points (32.28938) to determine draft order in the second round. I don't know whether it would be the RP of the alliance captain, average RP, or max RP between the two current alliance members that would be used to rank the alliances for the second round of the draft, but either way. I'm not sure about this idea, but I thought I'd mention it here.. That method of determining draft positions wouldn't really mess with the draft currency idea suggested above (warming up to the idea), as they still have to select in some order.

Interesting idea, because since we went to the current seeding system in 2004 these items have been somewhat at odds with each other. Just winning is the most important item in seeding because you get the QP for that. When you tie and it goes to the tie breaker of opposing score it would say you want a high opponents score but letting them score a lot can bite you with a penalty. So you don't want the score to be too close. Frankly I've never been a fan of using opposing score in rankings because to much involved is beyond the control of your own team/alliance, i.e. no robots for the opposing alliance show up, none of them work, and if you strategy is based on keeping the total score low in your matches it hurts you. Remember in 2003 the best robots didn't qualify well because they kept the score low (I know it was a different qualifying system but the logic still applies), a similar situation can keep a better team out of the top qualifying spot. The current ranking system has favored pure offense to "managing" or "controlling" the game.

In short I like the idea of using something else to determine the second round picking order if you want to mix things up just don't use opponents score. I would prefer it if they want to limit purely defensive play and mix things up to use lowest average scoring differential rather than average opponents score.

P.S. I'm intentionally being contrarian to the game designers intent of keeping the games highly offense based for the last 3 years. I believe low scoring doesn't mean a defensive battle all the time, it can mean efficient control of the field by an alliance.

Joel J
12-02-2006, 11:23 AM
Interesting idea, because since we went to the current seeding system in 2004 these items have been somewhat at odds with each other. Just winning is the most important item in seeding because you get the QP for that. When you tie and it goes to the tie breaker of opposing score it would say you want a high opponents score but letting them score a lot can bite you with a penalty. So you don't want the score to be too close. Frankly I've never been a fan of using opposing score in rankings because to much involved is beyond the control of your own team/alliance, i.e. no robots for the opposing alliance show up, none of them work, and if you strategy is based on keeping the total score low in your matches it hurts you. Remember in 2003 the best robots didn't qualify well because they kept the score low (I know it was a different qualifying system but the logic still applies), a similar situation can keep a better team out of the top qualifying spot. The current ranking system has favored pure offense to "managing" or "controlling" the game.

In short I like the idea of using something else to determine the second round picking order if you want to mix things up just don't use opponents score. I would prefer it if they want to limit purely defensive play and mix things up to use lowest average scoring differential rather than average opponents score.

P.S. I'm intentionally being contrarian to the game designers intent of keeping the games highly offense based for the last 3 years. I believe low scoring doesn't mean a defensive battle all the time, it can mean efficient control of the field by an alliance.
I'm still thinking more about options that are better than the serpentine, but I have to say that your post has made me get off the fence with respect to using the average loser's score as a possible means to determine second round draft order. I am now more fond of the idea.

Putting reality aside for a minute here:

One could find it reasonable to assume that a win of 98 - 4 was probably "easier" to come by than a win of 98 - 75. That is, one could say that the alliance that scored 75 points was probably stronger and more of a challenge than the 4 point scoring alliance. Now, there is a match that ends with a score of 5 - 4. While the winning alliance did less "damage" to get their win, one would have to say that the strength of their opponents was probably not much greater than the strength of the losing alliance in the 98 - 4 match. So, the average loser's score may tend to represent the strength of the alliances a given team has gone up against?

Now, to snap back to reality, one of the problems I see is that a team winning 98 - 4, would probably have been able to win 98 - 75, or 98 - 97, had their opponents risen to the challenge (not necessarily, of course), while a win of 5-4 doesn't get much closer. Anyway, to not ramble, it seems like the one measure better than the strength of the schedule is the strength of a given alliance, which could be measured by their winning score? Sure, but just look again at two examples mentioned above: 98 - 4 and 98 - 75. You could argue that (in general) the two 98's are equal, but who had the harder time getting to 98? Maybe the 98 - 4, because the weaker alliance probably played alot of defense? Or maybe it was the 98 - 75 alliance, because there was a better mix of offense/defense throughout (with the 98 scoring alliance playing defense every once in a while, then scoring, rather than focusing solely on trying to score). Maybe the 98 - 75 alliance would have scored a bit more points, if there opponents were as weak as the 98 - 4 losing alliance. I'm making assumptions (and not even considering scoring for the opponents, etc), but do you see what I'm getting at here? Maybe its time I look at data from the elimination rounds, where the strength of a given alliance tends not to change, while the strength of their opponent does (with the progression from the quarterfinals to the semifinals to the finals). Maybe there are some trends there.

I don't know.. as I said before, I'm getting more fond of using the loser's score.. what are the strong reasons against my line of thinking? What am I missing here?

I really hate the serpentine!

trilogism
12-02-2006, 11:54 AM
Sand.