Re: In-depth Strategy Analysis for your perusal - Team 2374
 

Good post, but naturally I am going to disagree with one thing. You stated that you can't believe the number of people who are in favor of defensive strategy, and that winning alliances usually have 3 strong offensive robots. I feel that history actually states that it is not so one sided...

Lets look back at the world champions of the years since I've been in the program and rate them either offense or defense depending on what won it for them.

1995 - Defense - In Ramp n Roll, if you didn't play defense, you wouldn't take the mountain to make the best score. Raychem and Woodside won by knocking Gael Force down the hill and playing catch up.

1996 - Offense - Tigerbolt scored a perfect score in each of its final matches.

1997 - Offense - Beatty's cap ability and tube scoring were unmatched...though Delphi's goal block did win them a match...which is more than anyone else could say about facing Beatty that year

1998 - Offense - Technokat's roller claw was offensive.

1999 - Defense - Juggernauts and Aces high used pushing power and strength to bowl their way to the top of the puck for the multipliers...and defended the top of it to the best of their ability.

2000 - Defense - A resounding defense on this one as the finalists team 25 Raider Robotix and 131 CHAOS used purely defensive ball stealing strategy to roll through the best offensive robots.

2001 - n/a - 4 v 0

2002 - Offense - Beatty just grabbed 3 goals and walked to a championship

2003 - Defense - Wildstang would bowl its way to the top of the mountain and defend it with its double wedge design.

2004 - Defense - One word...Martians

2005 - Offense - Went completely offensive...lots of scoring.

2006 - Offense - Autonomous and hoarding of balls from the human player won it all.

2007 - Defense - A heavily defensive alliance overwhelmed a more high powered offensive alliance and relied on the bonus at the end and a well placed tube to block a ramp.

2008 - Offense - Defense was heavily restricted

2009 - Offense - Turrets and fast delivery systems would take the day.

2010 - Defense - With all due respect to 177 and 67, that alliance does not beat the high powered scoring machine of 1114, 469, and 2041 without the amazing defensive, ball clearing effort of 294.

I see 8 - 7 in favor of offense...not really definitive if you are trying to determine which way to go each year.

With that said...I will go on the record as agreeing with you about this year. For the first time ever, I am thinking complete offensive domination is the way to go!

Good luck teams!

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

I disagree with alot of what was OP stated, but this is the biggest flaw I see. If you push me in I should be exempt from the penalty under the standard "can't cause the opponent to get a penalty" rule(unless they removed that this year). Also, by what means must I get out? If I am pushing full force directly for the line and you happen to be in between me and the line, who's fault is it that I am struggling to get out? I mean, sure I could go around...or you could let me out. Either way, I don't see anywhere in the manual that states that I must let you go immediately upon entering and go around you in half a second or that I get a penalty when pushed in. I'm not positive that defense will win it this year(I certainly think it will), but I'm 99.9% sure that going through isn't the best way to deal with a defense bot. As stated above, going around will be the faster of two ways.

I'd just remind you that some years they go penalty crazy, but other years they are relatively lax. I really think this one is going to be divided by regional as to how often you get it called. Some regionals will be fine with getting pushed in, while others will call it if your wheel so much as touches the line.

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

We mostly agree. Points of contention, since agreeing with 90% of what you said would take forever, are:

1) As everyone above me said, floor pickup is more important than it seems. Let us not forget the lesson of Lunacy. If you miss, or if your opponent drops something, and you can't pick it up, you have an inherent disadvantage. Alliances need at least one member who can pick up tubes from the ground or the opposing alliances will just throw all the tubes halfway across the field and take them home.

2) Our team feels that around, not through, is the superior option. For one, it isn't specified that a team on defense that gets pushed into your zone has to leave, ever. For two, going around can be just as quick as going through with a good driver, because you don't have to worry about the other team pushing back. For three, we already had a solid mecanum drive built from last year and our decision saved resources. :P

3) I'd argue that defense has a role to play. It's better to play defense on your opponent who has a tube that could complete a high goal than to, say, try to score on the low goals. Plus playing defense gets you closer to your towers for deployment.

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

Picking up from the floor is essential. The one difference I see from 2007 is the human player sending it more often through the slot than throwing it over due to diffferent field design, zones, and scoring being on the opposite end. Trivial nonetheless.

Going through the traffic rather than avoiding traffic sounds like the smarter play. However, I would add that trying to avoid spinning your robot around from grabbing game pieces to scoring would make things a whole lot quicker and efficient, if your robot can ? :)

Its a given that top row scoring will be first priority vs. lower scoring rows.
However, I'd bet that the ability to score on the top row will be vastly different between rookie and veteran teams (2007 and prior). And that will be the single biggest factor as to why teams win vs. losing. So what is your minibot gets to the top right before mines? All I need is a 5+ or 10+ higher score than you prior to the endgame.
In other words, I'd make sure you can score on all grids effectively and efficiently before trying to figure out how to be the fastest deploying minibot and speed. But its still important.

Those that played and were successful in 2007 have an added advantage. Comparing it to any other game in previous years that I've seen, coached, and studied doesnt help, IMO. Well, except for 2006. :)

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

I disagree with this statement. There are many superior drivetrains that we know of and will see this year that can plow through any defense with power AND speed. They wont waste time playing defense, because they'll be too busy scoring a ton on offense.

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

Emphasis mine.
I totally agree with this as the main reason they won the World Championships. When we won IRI this year with 469 and 330, we could not figure out why the opposing alliances didnt spend more time getting the balls out of their far zone back to their middle or near? The other alliances in the semis and finals clearly had better offensive machines, but I saw successful blocks a lot rather than kick a lot out of the zones.

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

Probably because if we had broken off 469 to clear, 359's stellar "protective defense" would have slowed us down so much that we'd be ineffective at clearing, AND be in a position where 469 was free to pick us apart. :) We almost had you with the blocking strategy... no regrets.

-John

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

I definitely agree. The biggest difficulty with beating 469 was that their accuracy was so high that you had to block the shot to kick it out and remove it from the cycle. I would like to watch the videos of the IRI finals and semi-finals right now, but TBA is down and I'm not even sure if those matches are up! I don't recall TBA having our match videos from IRI, so I don't think any of IRI is online.

Getting back to the initial thread...

1. I also agree with the large role of the defense in this year's game. Yes, efficient scorers are required to going far, but there will be effective defenders that make an impact, too.

2. Winning the mini-bot consistently (or having an alliance that does) is critical to going far. Placing an "Uber-Logo" on the top row is 30 points... that'll probably take at least half the match for a good scorer, a team with the fastest endgame can do that in under 10 seconds. What can they do for the rest of the match? Score against or defend against your top scorer! My point being that this year, like 2004 and 2007, is a year in which the end game of one robot can have a huge impact!

3. I'm quite certain that if one robot pushes an opposing robot into their protected area, they will not incur a penalty... unless the opposing robot chooses to remain there! If the opposing robot has good opportunity to escape, and doesn't... there will be no leniency! I'm skeptical that this rule will be as forced with as much variation as some say... The rules state clearly that one alliance can not force the other alliance to receive a penalty. There is a warning line... the refs will not be forgiving towards a robot that "accidentally" drives into the zone.

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

Regarding "defense vs. offense," there really isn't a clear separation in many games. They often aren't mutually exclusive activities. Sure, in games like 2006 and 2008 the line was generally pretty clearly defined between defense and offense, but in many other games the two were interlinked and often aided eachother.

Was 294's ball clearing purely defensive? Did they not also score some of those shots and supply 67 and 177 with balls to score? Was 469's design not both offensive and defensive, as it both starved the opposition of balls and kept piling on points?

Was battling over a spider peg in 2007 offensive or defensive? Galileo division winner 1902 called their strategy "doffense," as they played a strong, physical game in order to keep spider pegs clear of opponent's tubes not only to reduce their opponents score, but allow their alliance to use them to build their own rows. Similarly, placing tubes in strategic areas was often done as a defensive strategy, in order to eliminate the opponent from building large rows. That's why many of the more competitive matches had lower scores than you would traditionally expect when higher-end teams got together.

Was breaking up a row in 2005 defense? Or was it offense because you were scoring and trying to build a row for yourself?

Was pinning a team against a wall defensive or offensive in Lunacy? It both prevented them from scoring, and allowed your partners (or yourself in some cases) to score on them more easily.

What about "ball stealing" in 2004? Capturing the hill in 2003? Controlling the goals in 2002?

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

http://www.youtube.com/user/FRCHOTTeam67

The Finals are on Youtube somewhere else - SF1.1 is my favorite though. Double buzzer beater hang to win the match.

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

My team (one person in particular) has done an incredible amount of number crunching with various strategies, and doing that, floor pickup is an ability that is a MUST to get the most points possible, but quickly becomes obsolete depending on the roles of your other alliance members. With our number crunching, we found that the highest scoring alliance, which also considered other team compositions and won almost every time against them, would include a feeder bot (one who got the tube from feeder station quickly and brought it to the alliance zone), a scorer bot (one who just sits in the alliance zone and hangs tubes brought by the feeder), and a balanced bot (one who picks up tubes from feeder station and hangs it). To make it easier to read without parenthesis: a hanger, a feeder, and a balanced bot. We codenamed this team "White Panda". To have the feeder and the hanger, the hanger needs to be able to pick up tubes from the ground (since bots can't "hand off" tubes to another). We were originaly going to do a feeder bot, but unfortunatly found little evidence of other teams being able to pick up off the ground. As such, we are considering switching to a balanced bot that picks up from the ground. Without other teams that go the extra mile and pick up from the ground, the best alliance (according to our numbers) can't possibly happen. In my opinion however, picking up off the floor is a useful skill in any case.

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

I'm interest in the 'numbers' that you used in determining strategy. Could you explain that in depth?

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

Spider peg, spider peg; does whatever a spider peg does!

...in all seriousness, last year's game lent itself to 'doffensive' strategy -- controlling the middle zone meant both more points for you and ball starvation for your opponent. We used that idea in conjunction with hanging points to guarantee a win in all but one of our qualification matches at FLR (and 469 did it so much better than we even dreamed of...)

This year's game seems to be almost as clear-cut in terms of strategy as compared to last year:

"If you don't stop us, we outscore you like crazy and win a bazillion to nothing. If you do stop us, then we stop you, too, and then win on the endgame anyway."

The difference is in the scoring, of course, and I think this strategy is a bit harder to implement with how the game is set up.

I just hope that our minibot is as consistent as our hanger was!

Me, too.

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

Well, the numbers involved are about 4-5 pages of gibberish and rough estimations on scratch paper. And by rough, I mean very rough estimations. I'm not trying to turn them into the game-breaking calculations it sounded like, and I do realize now that's how I made it sound in my previous post, and I'm sorry I did. :(

Anyway, we basicly (very roughly) estimated that a run time on a good feeder robot picking up a tube and bringing it to the alliance zone should take about 10 sec assuming all goes well. The hanging of a tube being picked up off the ground by a good hanging robot should take about 10 sec as well. We started with placing ubertubes on top rows when possible, and made 10 sec time stamps that showed what each robot was doing.

Using those rough numbers, this hanger/feeder duo would hang 10 tubes (and then feeder goes for mini-bot deploy) in the match. The balanced bot would complete 1 logo on it's own and would still have time for mini-bot deploy, with each logo being the top-scoring possible for the moment.

Also, in our opinions, it would be better than two robots doing 5 tubes each because it means one less robot frequenting the middle section of the field, allowing easier manouevering. This also eliminates three different robots fighting over two feeder stations.

The closest alliance opposition, codenamed "Black Lavender", to beating it lost by two points (64 to 62) by playing a scenario between the two in our heads. Mini-bot scores were not taken into consideration in those final scores. Black Lavender was an alliance composed of two balanced bots, and a defender. We also did this for 3 or 4 other alliance compositions, but these were the best two that were likely to occur. I'm sorry for all of the confusion.

Re: In-depth Strategy Analysis for your perusal - Team 2374
 

So you think defense has no role to play in this game? Wow, harsh. What about in the endgame, when there are two towers for three robots? Assuming the minibots are created equal, doesn't it make the most sense to have the robot with the strongest drive train be the one that's playing defense on other teams to stop them from reaching their towers? All you need to do is slow them down by enough to put their minibot below your alliances' minibots in the scoring. The endgame really starts before the endgame starts, if you catch my drift.

Assuming that the "powerful" bot is as good a scorer as you predict, and assuming that it has a competent partner, then by endgame the top two rows will probably be full. It's entirely unreasonable to expect to score more than three tubes in the last 15 seconds of the match (otherwise, you would have a completely full rack already). Three tubes low + logo bonus = six points gained.

Physical defense by such a powerful drivetrain would probably completely shut down, and at least significantly slow, an opposing minibot. If you can slow them down so much as one place, you will at least reduce their endgame score by five points and increase yours by five. So the least possible return given by playing defense in endgame is almost double the maximum return of offense in endgame. I think that's a pretty solid reason to play some defense.

Even in the teleop game, I think defense will be seen. If your robot fills up the goals faster than your opponents can, then fill up the top two layers and stop them from doing the same. Stopping your opponent from completing a logo by placing one tube on the middle is 8 points, which is worth more than making three trips back and forth across the field to score a logo low.

We have accepted that defense is a role we are not going to be able to play and do not particularly want to play. That doesn't mean defense isn't important.