Starve Them!
 

Why is it that team defense amounts to hammering on the three point shooter, then leaving them alone to go get more ammo once they've shot their load?

I could see how pride would work to force the teams to score a point or three, but leaving those Howitzers to reload and score ten times that just does not make sense.

Ram (err - bump - depends on who, where, and when) them in autonomous, then stick to them like glue. At least a chance to win beats no chance at all!

Re: Starve Them!
 

You are 100% right Jack. That was supposed to be our strategy in the Detroit finals, but if we stopped the Chickens, the Guerillas would get by us, or vise versa. We couldn't seem to stop them both.

Matt B.

Re: Starve Them!
 

Jack,
After losing autonomous to the highly skilled and accurate 3 point shooters - You are forced to try and score during the next period (yours to score in and theirs to defend).
Side note-Which a flaw in the game design as the team that won autonomous is already rewarded with bonus points, and the team that lost is immediately required to reload taking time away from their offensive time length, all while the team winning autonomous is able to have their robots go and reload while playing defense. It would have been more interesting to see the team winning the auton period immediately go on offense, someone has to deal with the fact that you shot the balls and must collect more - why place that added burden on the team that is already 10+ points behind?
Back to your topic -
The robots cannot be in 2 places at once - they MUST score while still trying to stop the opposition from reloading. Take your choice, but in most cases the pill is bittersweet eitherway.
That strategy might work if that is the only way to even have a chance at winning - but with so many robots tipping over, the difficulty of the ramp climbing, the fact that teams can push your robot up the ramp and keep you there, the corner trap zones, and penalties, there is no assurance that the "don't let em' reload strategy would work either.
So, trying and dog them the entire match ends up being "pointless" to that team.

Re: Starve Them!
 

That is not a "flaw" in the game design. It is exactly what was intended. Play out the inverse in your head - really think it through. You will see that having the team that wins autonomous immediately go on offense would be the wrong thing to do.

-dave

Re: Starve Them!
 

Dave -
I have thought about it - alot. I have watched many matches and both Great Lakes and Detroit. It seems that the "lost time to reload" is penalizing teams that try to win auton but fall short. My observation is obviously just my opinion about the dynamics of the game period sequence.
I am not part of the game committee, so I cannot argue "intent". But, I do not see why it would be "the wrong thing to do". Please explain to me what you obviously think I am missing here. PM me if you want to - I have no problem discussing this that way either.

Mike

Re: Starve Them!
 

A couple of good points are made. Team 135 has a good robot, being able to move in just about any direction, picking up balls looked very easy for them, and they just rocketed the balls in to the 3pt goal. They also went undefeated into the elimination rounds.

But in the semifinals at Boilermaker they were defended pretty affectively. I can't remember which team was defending them, but they did a good job. Also, It's hard picking up balls when another robot is messing with you, which happened to them.

As for the alternative game play, I think what we have is a good reward for winning autonomous. Because everyone that shoots or dumps balls in autonomous will have to reload; consequently, some teams don't do anything in autonomous so they have ten balls to shoot right away.

Anyhoo, I think a strong alliance will usually have at least one good or great 3pt shooter, one good 3pt or 1pt shooter, and someone just for defense. But I might be wrong.

Re: Starve Them!
 

Mike,

I think as it stands, we can all see the large, large advantages to winning autonomous. 10 points, plus defense, offense, offense. It's a powerful combination. So, think about the inverse. The team that wins autonomous, call it Blue, goes on offense first. They've just shot all their balls to win auto, even if the Red team doesn't do anything. Worse, if they scored all the balls, then Red has complete control over those balls, and Blue is left with the 10 remaining balls in the bins. So, Blue gets maybe 10 more balls to work with, plus the reloading penalty. Red gets to starve Blue and be on offense 2 periods in a row. Worse, Blue can't reload for their final offense period until Red has scored points.

Overall, I think all that adds up to a disadvantage much larger than the 10-point bonus. So you'd have the odd situation of the really highly capable teams NOT wanting to win autonomous. It might be interesting to see robots aiming for opposing goals in autonomous trying to score just enough points to put the opposing alliance on offense first, but it wouldn't be right.

Re: Starve Them!
 

OK, good. That's what I thought.

I mean, if you win autonomous and go directly into offense, why bother shooting in auto mode?

It would be a bigger bang for the buck if you DIDN'T shoot in auto mode. Why let the tiny computer aim and possibly miss a bunch when you could wait and let the human (with a superior organic computer) aim and hit with higher accuracy later on. And it's not so much the aiming, but the decision to fire. In auto mode, the robot doesn't know it's about to get pushed... but the human can see that and hold off on firing.

Also, if you unload "Rambo style" into the goal and win... you are empty and just made your life harder.

In other words, had it been the other way, people would be "penalized" for having a shooting auto mode.

And I'm sure the GDC likes the increase in Auto modes. (Which, by the way, is really cool this year - best auto modes EVER.)

Re: Starve Them!
 

I'm just glad autonomous means something again after two years of meaningless "victories" (ooo I released all the balls or cool I capped the center goal and got a couple more tetras).
The first year autonomous meant something because you have to clean up all those boxes that got thrown into your side of the field. It changed how you played the game. You see it this year and this is the way autonomous should be.

Re: Starve Them!
 

I think the way the game set up requires more thought into strategy. That way, if you choose to try and autonomous mode, you better win it or you are left with no balls to shoot and have to reload for most of your first offensive period.

There have been a few rounds where we knew we couldn't win autonomous and elected to keep the balls until the final round to score. We would reload our helix and then go and human load all the balls they scored with and then tried to score them in the final round. We would starve their human player so they couldn't reload.

It is risky, and it might burn us against really good opponents (or if you drop a chain like we did in Detroit), but it was the strategy we chose to take for that round.

Re: Starve Them!
 

[quote=Not2B]OK, good. That's what I thought.

I mean, if you win autonomous and go directly into offense, why bother shooting in auto mode?

To get the bonus points silly!

Don,
I agree strategy would be impacted and that was my point in the first place. Teams developing their strategy would have to decide not only how many shooters will attempt to win auton, but also if they need one of their partners to hold onto the balls in case they lose auton.
It just a different angle - and something to think about "what if". I doubt that it would cause the best auton teams any harm, other than reduce the initial auton score. Teams that can score in auton can obviously also score in Offense periods - so it really only is a matter of time and it also slightly reduces the # of blowouts.
The concept is dependent on seeding vs, elimination matches. I was thinking only about seeding matches and NOT about elimination matches.

Re: Starve Them!
 

I guess you missed the part about ramming to force a draw in auto.
And don't pretend is doesn't happen because it's not allowed or isn't GP.
I've seen teams get hit so hard it broke their alliance partner :eek:

Re: Starve Them!
 

I thought that might have been one of your strategies. Especially if you were in an alliance with another shooter that might win auto by themselves. It was an awesome sight when you loaded and then in 2nd or 3rd period threw 25 or more balls up toward the goal.

I must hijack this thread to congratulate you on your performance this weekend in Detroit. (Despite the fact that your 12-0 record meant you defeated 1188 3 times on Saturday. ;) And how did that alliance in last round of qualifying come about, anyway?) Usually there are about 3 tiers of robots in a competition - the good, the average, and , not being mean but realistic, the "thanks for trying". 469 was a cut above, creating it's own tier.

Back to whether winning autonomous mode gives the alliance too much of an advantage, wasn't that what was requested by many after the low benefit of scoring a vision tetra in Triple Play?

Re: Starve Them!
 

Can't argue with that Jack!
Tough one to call though - I was told that high speed ramming is still high speed ramming - even in Auton (by a head ref whom I respect alot)
But, what else can an alliance do other than guess where each others robots are going to end up in auton? We didn't include collision avoidance software programming this year.
In auton you can try and send all 3 out to score and hope they don't knock each other off coarse. You can sit still during auton in preparation for offense or defense - whichever comes next for you, or you can at least try and have one of the alliance partners "block" the way of the opposing robots. Notice I didn't say ram them off their path to their shooting spot, still all in all - I agree that starving the scoring alliance makes sense, if you also have someone on your alliance that can score.

Re: Starve Them!
 

[quote=meaubry]Of course the bonus points are good. But there is more to it than that. It comes down the the hidden value of going on defense first.

If you unload in auto mode and get 10 extra points, you spend the next 40 seconds wondering around. You are almost out of ammo. The op-force has no reason to feed you balls. They are reloading. When it switches, they are now fully loaded. Do you spend the time loading AFTER they score (slow loading) or do you prevent them from scoring? To recover 10 pts, you only need to shoot 3.33 balls - something easier to do if you are flush with ammo.

It's a wonderful problem to think about. So complex, and so many variable. Time to reload, time to pick balls off the floor, time for human player to toss the ball, heck - even time for the balls to run down the ramp and tube. If you are starving at the beginning, you don't practically feed again until 60 seconds into the game.

For my own curiousity, I checked the GLR scores. 18 out of 83 matches were decided on 10pts or less. (Which was more than I thought it would be.) 10 pts are important, but are they important to stop playing for almost half the game?

I like it today. It favours a strong auto mode. Grant it - not something 862 has, but I still like it.

Of course, I could have this totally wrong. I mean, I'm not Bill Beatty!! :)

Re: Starve Them!
 

Mike A,

I agree with you on most things, but not this one. I have to take Lavery's side on this one. Having the auton winner go on defense is an incentive to win auton. If it were the other way around, we probably would not have focused on auton as much.

If you do not think you can win auton, then don't shoot. Get in position and rapid fire as soon as your offensive round starts. You will overcome much of the auton deficit. I agree that winning auton is a big advantage, but not just because of the 10 points. The bigger advantage is the 80 seconds of continuous offense and the transition game that has to happen in the last 40 seconds for the team that loses auton.

I like all the different strategies that are involved in this game, but my head hurts after every match due to the incredibly fast pace. This is the hardest game I have ever had to coach. Thankfully, my student drivers are a lot smarter than me and covered for me on several occasions.

Re: Starve Them!
 

At Midwest, this is exactly what team 71 did in the last match of finals. They were against team 111, who has a vicious automode, which is nearly impossible to stop due to their turret, and were winning auto every time, while 71 was wasting balls, as they were less accurate in auto than in human control. Thus they tried the strategy of saving their balls to have a huge first round, but they ended up in trouble because after emptying their stores, they went back to human load, and were blocked, not pinned, into the corner, and never made it back across half court.

Its an interesting strategy, but I am of the opinion that you must win auto mode if you want to win the match.

Re: Starve Them!
 

Paul, I strongly agree with you on this point. I was mentally exhausted after the finals yesterday.

Re: Starve Them!
 

I also agree. It's very hard to tell exactly where and when the opposing alliance is going to hit you. This gave our team an incredibly hard time trying to find the right point in time during the match to stop accumulating balls and start scoring. Too early, and we wouldnt score as many as we'd like. Too late, and we would have about 30-40 balls in our robot and get pulverized with defense.

Re: Starve Them!
 

I think its safe to say that ramming happens when a moving bot hits a bot that is standing still. If two bots are racing to the same field position and they collide, then who rammed into who?

Nobody!

Shooters do not own the spot in front of their center goal. Both teams have the right to try to be the first robot in that controlling field position. If they collide on the way, then neither team should be penalized.

Now if you have a shooter that stays in its starting position and shoots from there, and a bot flys across the field and slams into it, THAT is ramming.

Re: Starve Them!
 

This post needs no further support. Bravo! Far too many times have robots been allowed to waltz to the human player station to reload without any interference. IF THEY DON'T HAVE ANY TEETH, THEY CANNOT BITE!!!!

However, I prefer the get in front of them method to the bumping (never "ramming"!!!) method during autonomous, because if they have a turreted shooter, they can still compensate. If you're in front of them, in most cases, you still have a chance to block any shot they take. Pay close attention to where a robot stops in autonomous and compensate accordingly.

I'm sure the offender will also be keeping watch over the defenders and adjusting their autonomous routines to react to the blocking techniques - counter-countermeasures. I've seen some great adjustments during auton this season - makes the game more exciting!

Re: Starve Them!
 

That strategy might work grat against some teams (maybe us for instance), but try it against 469 or 217 and you will get burned. Thier camera auto aim and turret mounted shooters make it nearly impossible to mess up thier autonomous.

Like Mike A said, You end up having to play a game of offensive catch-up.

Re: Starve Them!
 

The 'ramming' method is a good strategy in working against robots who sit and shoot into the center goal. It cost our alliance, numerous times, the autonomous period. While my teammates and i dislike it, i dont necessarily see it as a bad thing. sometimes, theres no robot for them to ram into, instead they hit the sides at full speed and cause damage to themselves. its a gamble that teams have to be willing to take, these robots were supposed to be built with 'ramming' in mind. thus the opportunity to add bumpers. Also, one team at Annapolis that caused damage, went and worked with that team to fix it in time for their next qualifying. teams should just slow it down so they do minimal damage to themselves and other robots.

Re: Starve Them!
 

I disagree Ken. If a robot is moving along in the "y" direction they have zero relative velocity in the "x" direction - right. So if another robot runs into them at high velocity traveling in the "x" direction, by your definition they are ramming.

Although, it really all boils down to what is considered "high velocity" by the referees. No ref will call ramming on a robot traveling about 5 feet/second or less. How much above that is considered ramming depends on the ref.

The head ref at MWR had a secondary definition for ramming. He said if you repeatedly hit a robot from the same direction 3 times in a short period (pull back and hit 3 times), he would also call that ramming. So he expected robots to hit 2 times and try another location the third time.

Sorry for getting off topic.

Re: Starve Them!
 

In one of our semifinal matches in Trenton against 25 we managed to neutralize them by constantly harassing them. (However, we were smoked by 103 and especially 1279)

Re: Starve Them!
 

Thank you for the compliments. We had a great weekend for sure. Our alliance partners were awesome and helped us finish the weekend with the win. It felt great especially after coming up a little short at GLR where we also went undefeated in qualifying, quarters, and semis, only to come up against two of the triplets in the finals giving us our only two losses of the season.

[Thread Content]

Dan, our lead strategist, came up with the requirements of our robot based on the same discussions we had concerning the game at the beginning of the season. We had to hold 30 balls, be very tough to move, have a turret to track the target, and have a good enough drivetrain to get out of a trapped position. I think our combination of maneuverability and speed in one mode and power in the other makes it difficult to trap us one-on-one. Then, if you put two robots on us, hopefully we have a partner like 217 that can also load up and get to shooting...

Re: Starve Them!
 

I don't mean to call anyone out, but the triplets, 25 and 469 all seem to rely on that human loader a lot. I feel that if you can keep their robot at least 7 feet from the loader station, the HP wont be able to make enough shots. Granted, 469 can pick them off the ground. The triplets have accumulators, but they didnt seem to work amazingly to me (but at waterloo an upgrade maybe :ahh: ) I know i've talked about this with my team, but keep those shooters at least 7 feet away is going to do wonders for your D

Re: Starve Them!
 

Here's my take on autonomous first.
I LOVE IT! The high reward for winning autonomous, combined with the new closer positioning of robots (no longer all the way across the long dimension of the field) has led to some of the most exciting autonomous gameplay I have seen. In 2003, you would see some interaction between the two alliances in autonomous, but primarily that was just two robots running into eachother on top of the ramp. In 2004, some interaction would occur when teams would go for the yellow balls to release the ball dumps early, but for the most part this was rare and wouldn't impact the entire match very much. In 2005 interaction in autonomous was zilch, none, nada. But in 2006, with such a large emphasis on autonomous, it has become an exciting and highly strategic part of the match.
Many examples could be found during the peachtree elimination rounds. At the start of our QF matches, both alliances autonomous strategies essentially mirrored eachother. 116/1533 would run to the corner goals and dump 10 balls. 1139/34 would dead reckon the center goal and shoot. 1369/1242 would try and hit the shooters to keep them from making shots. Here's where it got interesting. 1369 expierienced some errors with their autonomous, so they just spun in a circle. Both 34 and 1139 were never 100% accurate (they would either hit almost all their shots, or none at all, depending on their dead reckoning positioning). The first match, both 116 and 1533 sucessfully dumped all 10 balls. 34 hit 3 shots into the center, giving the red alliance the win in autonomous. Red alliance would advance to win the match. Match #2, 34 would miss all of its shots. 116 dumped all 10 balls, but 1533 only made 9 out of 10, giving blue the win in auto, and the match. Fearing the same result, the red alliance switched 1242's and 34's role in the next match. BOTH robots went directly for their own corner goal to prevent 116 from sucessfully unloading (Unfortunately, 116's autonomous failed anyway because it's stupid coach forgot to plug the drive motors back in after using their timeout to make repairs to the drivetrain :o ). Red alliance won the match.
During the Peachtree finals another outstanding example appeared. The #1 Red alliance consisted of 1261, 1414, and 1057. During most of the eliminations 1414 had run to the corner goal, dumped 10, and 1261 had sat in position 2 and fired balls into the center goal. But, because they didnt move, 1261 was an especially vulnerable target to opposing defenses. In the finals 1057 began "intercepting" opposing robots who were trying to slam into 1261, giving the #1 alliance an even more impressive advantage during autonomous (and allowing them to win the Peachtree regional).

Now onto the "starving" topic:
Starving may well be become a very effective strategy at the Championship event, due to some of the potential devestating shooter combos that may arise. But at the regional level, it has not been necessary thusfar. A vast majority of shooters need to be reletively close to the front of the ramp to shoot accurately. Because of this a "zone" defensive scheme has allowed for singular robots to block multiple shooters (even all 3 if the shooters can't fire fast enough). Running interferance has helped this some, but often it has only contributed to further traffic and less ability to get positioned correctly in front of the ramp. The most effective counter-measure so far has seemed to be having a robot than can score lots of points quickly in the corner goal to force the defense away from the front of the ramp.
By making the shooters miss shots you not only prevent them from scoring points, you allow any herders on your alliance (and their's) to grab the balls. Most highly effective shooters has been primarily human loaded (the exceptions being bots like 1731 and 435), thus when trying to stop an alliance full of good shooters, it is often to your advantage when balls are "loose".
Additionally, because of the human loading shooters, their reloading time is often the best time to play defense on them. Because of the field set up, with the ramp and the edge of the field, teams can be blocked into the area immediately in front of the human player. Many shooters prefer this area to load up, and even if they don't they are often near it when they are loading so they can be pushed into it. You can then seal them into this are (not even needing to pin them).

Re: Starve Them!
 

uhmm.... What ?!?! there are no right of ways painted on the floor of the playfield.

If two cars both run the stop signs at a four way stop and collide, then who rammed who? who was at fault? they both were.

If two cars both try to merge into the same lane and collide, they are both at fault.

Its like playing chicken - if you are both heading for a collision and no-one flinches or backs off, you are both responsible for what happens next.

Re: Starve Them!
 

There's a couple solutions to this. One is to mount the gyro somewhere on your robot.. if you get pushed, it's angle will be off so you can stop shooting. (This is the easiest to do now, as the gyro+wiring weighs very, very little. It also doesn't require any hardware modifications)

Re: Starve Them!
 

This is the "pick" strategy that 293 tried with one other team (we were the shooter) in a qualification match in Annapolis. It was not necessary, however, as our opponent did not try to defend us in autonomous as expected. A tip of the cap to the alliance that pulled it off successfully; I wish I was there to see it.

In elimination rounds, we made at least two shots every time in autonomous (except when we forgot to change the battery before SF2-01) even though we got rammed every time, and on two occasions we made every ball we shot except the first. One of those was the 42-point autonomous period (before bonus) where 293 and 103 combined to make 14 balls in the center goal. Unfortunately, that match was restarted because the light didn't come on after autonomous. :(

Re: Starve Them!
 

I love that FIRST has made a game where winning autonomous matters. What I do think is a problem, however, is that the computer does not always assess correctly which alliance won automode. That the decision is incorrect is often obvious to the teams, the refs and even the spectators.

The effect of an incorrect determination as to which alliance won automode is far greater than the 10 points. It's about the strategy the team plays for the duration of the round. This is true if you think you are ahead by the 10 points, or if you are forced to play offense first because it determined incorrectly that you lost. The 10 point swing at the end when they finally get it right cannot begin to compensate the teams for this.

I am hoping from now, through the end of the FIRST season, the refs will stop the play for however long it takes to manually figure out who won autonomous mode, and even reset the scoreboards, if necessary. Getting it right is far more important than the minute or two we may have to wait.

Re: Starve Them!
 

I am looking forward to seeing team 469 in Georgia, since their design is so similar to ours. It took us a while to get the camera and turret in sync, but when it came together at the Boilermaker, it changed everything. I just want to thank team 1272 for looking past our poor ranking to pick us. Additionally, I would like to thank team 1319 for their great defense. I am thrilled with the way that the autonomous period dictated the strategy of the entire game. It was the primary reason for our accomplishments during the finals at Boilermaker. Looking to forward to West Michigan in a couple of weeks............ :)