Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Rules/Strategy (http://www.chiefdelphi.com/forums/forumdisplay.php?f=6)
-   -   How to win Steamworks? (Weeks 1/2) (http://www.chiefdelphi.com/forums/showthread.php?t=155154)

Brian Maher 12-02-2017 03:40

How to win Steamworks? (Weeks 1/2)
 
Now that everyone is starting to get a more realistic feel for what their robots can do, here are a few questions about what you think the typical winning Week 1/2 alliance will look like:
  • What roles will its members play/what will the overall strategy be?
  • Will there be a dedicated defender or will all three robots be primarily offense-focused?
  • Will it be able to score all four rotors?
  • Will higher-ranked alliances be more or less favored than usual?

GeeTwo 12-02-2017 06:02

Re: How to win Steamworks? (Weeks 1/2)
 
I see the most likely winning alliances for the early weeks pretty much like I did at the end of the first week of build:
  • One fuelbot (full robot width floor intake plus ability to get from hoppers)
  • One gearbot with floor pickup, consistent and likely fast
  • One passive gearbot that is fast with a good/great driver

At least two, and likely all three, will climb.

This alliance will be able to score both the fuel40+kPa and four rotors against light defense.

As weeks go by, the mix will move towards two fuelbots and one gearbot, or a fuelbot, a gearbot, and a defender.

messer5740 12-02-2017 10:16

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by GeeTwo (Post 1643634)
I see the most likely winning alliances for the early weeks pretty much like I did at the end of the first week of build:
  • One fuelbot (full robot width floor intake plus ability to get from hoppers)
  • One gearbot with floor pickup, consistent and likely fast
  • One passive gearbot that is fast with a good/great driver

At least two, and likely all three, will climb.

This alliance will be able to score both the fuel40+kPa and four rotors against light defense.

As weeks go by, the mix will move towards two fuelbots and one gearbot, or a fuelbot, a gearbot, and a defender.

I agree with this, except for the 40 kPa. I think I
it will be close, but will not result in the extra playoff points. I think getting all four rotors up and turning is more plausible
, because many passive gear bots can make about 6 cycles in 2 minutes
Edit: also many alliances will likely have multiple autos so they can get 3 gears on the pegs before the 15 second auto period ends, giving the alliance an even greater chance of winning( because of the massive amount of points) and it will bring them even closer to starting up all four rotors.

6101 Robert 12-02-2017 10:22

Re: How to win Steamworks? (Weeks 1/2)
 
I don't think we'll see a lot of dedicated defenders outside of those deemed not suitable for other tasks by their alliance partners. Instead, alliances will likely use more hit-and-run tactics to try to disrupt their opponents without going too far out of their way. Alliances that finish their gears or fuel before their opponents will often opt to play defense.

This year, I think many mid-level alliances will do very strongly. This is a year that requires 3 fully-competitive robots on each alliance. First and second seed alliances that are forced into picking less-than-competitive partners will struggle greatly to compensate, especially because powerhouse teams will likely be shooter-focused. No reasonable amount of fuel can account for not getting 4 rotors turning.

Personally I think any alliance worth their salt will get their rotors turning, likely with some time to spare. If I'm misjudging how difficult the rotors will be, alliance will be wise to invest more robot time into the task. In playoffs, is red alliance gets 4 rotors turning and blue does not, red will win the match almost all of the time.

JamesCH95 12-02-2017 11:14

Re: How to win Steamworks? (Weeks 1/2)
 
-Strategies are not worth much speculation as it will depend heavily on the teams involved in a particular situation
-See above
-Getting 4x rotors turning in weeks 1 and 2 will be a rare event. In all of 2016 the eliminations tower capture average was about 30%... I expect the average of 4-rotor success will be significantly lower than 30% for weeks 1-2 of Steam Works.
-Top ranked alliances will be a mixed bag. Teams that can rank effectively may not always have the potency to survive in eliminations. Their effectiveness will depend heavily on getting a 3rd pick that can reliably climb. Other top-ranked teams will transition nicely to eliminations and dominate, even with a weak 2nd pick.

Marinara: https://www.thebluealliance.com/insights/2016

Looking at what is often predicted on CD vs the actual results in years past I am highly skeptical of the optimistic predictions in this thread.

KelleyCook 12-02-2017 11:34

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by Brian Maher (Post 1643627)
Now that everyone is starting to get a more realistic feel for what their robots can do, here are a few questions about what you think the typical winning Week 1/2 alliance will look like:
  • What roles will its members play/what will the overall strategy be?
  • Will there be a dedicated defender or will all three robots be primarily offense-focused?
  • Will it be able to score all four rotors?
  • Will higher-ranked alliances be more or less favored than usual?

I'm fairly certain that this is how the playoff winning alliances will be:

* Three (possibly only two) teams all capable of doing Gears. Point values -- especially the 100 in the playoffs -- dictate this.
* Four rotors will almost always be spinning in the playoff rounds (I fully expect in well under a minute in the Michigan state championship). Its very difficult to win if you don't get this.
* After the fourth rotor, those gearbots will also be able to shoot high to get (at least) the remainder of the 40kPa.
* The 3rd alliance choice will probably be a defensive bot to stop the other team from shooting. Depending on how good the other two are at gears, this third team may play defense the whole time.
* Finally that third bot might not even bother attempting to climb. Since preventing a climbing bot from getting to the rope to climb is the same value as climbing yourself.

Kelley Cook

EricLeifermann 12-02-2017 11:45

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by KelleyCook (Post 1643685)
* Four rotors will almost always be spinning in the playoff rounds (I fully expect in well under a minute in the Michigan state championship). Its very difficult to win if you don't get this.

Kelley Cook

Not a chance this happens in under a minute EVER. The coordination of 3 robots scoring gears, because it's going to take 3 robots scoring gears to get 4 rotors the fastest, isn't going to allow it to take less than a minute. Not to mention how much driving is needed.

Roboshant 12-02-2017 11:45

Re: How to win Steamworks? (Weeks 1/2)
 
I think there will probably be about 2-3 alliances per event that are capable of getting 4x rotors, while pretty much all the other alliances will be capable of doing 3. I think we'll see some killer defense from the 3 rotor alliances to prevent the 4x alliances from getting 6 more gears on.

The 4x alliances might not even be the high ranked ones. Considering that the second pick of each alliance will be the only team with a good view of the loading zone, teams that get an earlier second pick due to serpentine might be at an advantage. There will most likely alliance captains and 1st picks that can utilize cameras well, but they will probably be focusing on fuel for the first ~1 min of the match to get 40kpa.

Another week 1-2 prediction:
Really bad gear control from some teams. A lot of teams are going to show up with sub-par, ill tested passive gear mechanism. Combined with the fact that 2/3 of the teams on each alliance will not have a good view of the loading zone, and most of these teams will probably not have a camera, we'll probably see a lot of dropped gears in quals.

Edxu 12-02-2017 11:54

Re: How to win Steamworks? (Weeks 1/2)
 
Having an alliance of 3 gear-focused robots and at least 2 climbers will definitely win you any week 1/2 event.


Since you need 12 gears, and presumably 3 less because of auto programs, you need only 9 gears to finish all the rotors.

At a cycle time of 12 seconds, it should take around ~40 seconds for three organized and efficient gear-focused robots to finish all four rotors. This means that 40 seconds into the match, they have 60*2(auto) + 40*2(teleop) + 100 (Playoff Rotor Bonus) = 300 points.

This means that you have 95 seconds to play extremely hard defense on the opposing alliance to stop them from finishing their rotors, as it's basically impossible to make up the score difference by scoring fuel.

With 2 hangers, you can increase your scoring potential to 400.

TL;DR: If you finish your rotors and your opponents don't, you win the match. Fuel doesn't matter.

Poseidon5817 12-02-2017 11:56

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by EricLeifermann (Post 1643691)
Not a chance this happens in under a minute EVER. The coordination of 3 robots scoring gears, because it's going to take 3 robots scoring gears to get 4 rotors the fastest, isn't going to allow it to take less than a minute. Not to mention how much driving is needed.

If there is an alliance that scores three auto gears, they only need to score 3 gears each for the 4 rotors, or 20 second cycles for each robot to make it in 60 seconds. While definitely extremely unlikely, it could happen once or twice a season if teams try to make it so.

Roboshant 12-02-2017 12:10

Re: How to win Steamworks? (Weeks 1/2)
 
How to win:
- Start 3 rotors
- Even if other alliance looks like they can get 4
- 4 rotors pretty much requires 3 bots to focus on gears (at least for most week 1 and 2 events)
- If the other alliance is focusing on gears, they probably won't score more fuel than you.
-After scoring 3, have your second pick play defense (best view of the other teams loading zone)
- Try to prevent the 4th rotor from getting on. Even if this team racks up 50 pts of fouls, as long as they prevent 140 pts for the other alliance it's fine.
- If both alliances get 3 rotors it going to come down fuel and climbing.
- Have two bots focus on fuel the entire match (one can assist the gear scoring bot, so that they can start playing defense earlier).
-ensure that you outscore the other alliance in terms of fuel.
- Alternate strategy would be to have all 3 robots go for gears in the beginning of the match to get 6 rotors earlier.
-The earlier you get 6 rotors up, the earlier one team can start playing defense.
- Have 2-3 robots climb at the end
- Climbing this year seems much easier than last year, I think we'll see a lot of 2-3 robot climbs in weeks 1-2.

EricLeifermann 12-02-2017 12:18

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by Poseidon5817 (Post 1643700)
If there is an alliance that scores three auto gears, they only need to score 3 gears each for the 4 rotors, or 20 second cycles for each robot to make it in 60 seconds. While definitely extremely unlikely, it could happen once or twice a season if teams try to make it so.

20 second cycle is what the elites are going to do. It's not going to be something that an average 3 team alliance is going to do. Not to mention any defense will slow it down.

Rangel 12-02-2017 12:19

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by Poseidon5817 (Post 1643700)
If there is an alliance that scores three auto gears, they only need to score 3 gears each for the 4 rotors, or 20 second cycles for each robot to make it in 60 seconds. While definitely extremely unlikely, it could happen once or twice a season if teams try to make it so.

That's true but an alliance of that level shouldn't be wasting time trying to get all 4 rotors in a minute. If they can do it that fast, they should be mixing in defense or scoring fuel throughout the match instead of just the last minute.

Poseidon5817 12-02-2017 12:24

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by EricLeifermann (Post 1643711)
20 second cycle is what the elites are going to do. It's not going to be something that an average 3 team alliance is going to do. Not to mention any defense will slow it down.

I said it will only happen once or twice in the season. I was merely pointing out that teams would only have to run 20 second cycles to make it so, in any one of the thousands of matches this year.

StAxis 12-02-2017 12:27

Re: How to win Steamworks? (Weeks 1/2)
 
I would predict that at least one week 1 event will be won by teams without a shooting robot. By week two your primary gearing robot will need to pick up some slack to free up a potential shooter, but most of the points will still come from gearing. Shooting won't be a major difference maker except in odd situations until later in the season.

Lil' Lavery 12-02-2017 12:44

Re: How to win Steamworks? (Weeks 1/2)
 
3 rotors (1 auto) and 2 hangs.

JABot67 12-02-2017 13:06

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by Lil' Lavery (Post 1643724)
3 rotors (1 auto) and 2 hangs.

This is accurate.

jayred1127 12-02-2017 16:52

Re: How to win Steamworks? (Weeks 1/2)
 
I think everyone is underestimating the challenge of placing the gear on the peg. Especially in week 1-2 there are going to be many dropped gears.

I think shooting fuel will win events for early regionals and district events.

Classified* 12-02-2017 17:20

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by Edxu (Post 1643699)
Having an alliance of 3 gear-focused robots and at least 2 climbers will definitely win you any week 1/2 event.


Since you need 12 gears, and presumably 3 less because of auto programs, you need only 9 gears to finish all the rotors.

At a cycle time of 12 seconds, it should take around ~40 seconds for three organized and efficient gear-focused robots to finish all four rotors. This means that 40 seconds into the match, they have 60*2(auto) + 40*2(teleop) + 100 (Playoff Rotor Bonus) = 300 points.

This means that you have 95 seconds to play extremely hard defense on the opposing alliance to stop them from finishing their rotors, as it's basically impossible to make up the score difference by scoring fuel.

With 2 hangers, you can increase your scoring potential to 400.

TL;DR: If you finish your rotors and your opponents don't, you win the match. Fuel doesn't matter.

I think you are way underestimating cycling times. We have an extremely fast robot (18.5 fps) plus an active gear mech that is also very efficient. This weekend during driver practice we were getting at the very least 11 seconds per cycle, but most were more like 15-16 seconds. Our chute was a lot closer than the real field, so I would add 5 seconds at the very least to that. I highly doubt you will get three gearbots that are that fast on an alliance. Most teams gear for around 14 fps. There is also the problem of traffic with all three robots cycling gears which will add at least another few seconds onto that. I would think 25 seconds is a reasonable estimate for a real match.

You can watch our robot cycling in this video.

I'm not saying 4 rotors won't happen. I really think it's feasible, but it will take the better part of the match.

Also, having three robots with a gear auto is going to be rare. The center peg is relatively straightforward (pun intended :p) but the side pegs are not. They either require vision tracking or very precise encoder/gyro code. Not something that every team has.

Chris is me 12-02-2017 17:50

Re: How to win Steamworks? (Weeks 1/2)
 
The alliance with three consistent hanging robots that also score at least 3 rotors will be enough to win the majority of week 1/2 events.

Bob Steele 12-02-2017 19:06

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by Lil' Lavery (Post 1643724)
3 rotors (1 auto) and 2 hangs.

You got that right....

Teams that can consistently get 3 rotors (120 points) + one auto rotor (+20)
and 2 hangs... (100 Points) score 240 points with no fuel. This wins nearly everything....

winning score week 1 and 2 .....

sprinkle in 90 balls for 30 points....
270 points... wins regionals and definitely districts weeks 3/6

District Champs and half champs: This plus another 180 balls... plus another hang.... 400 points...

(Defense will be HUGE for the fourth rotor points....it is simply ... let the other team get to 3 rotors... and then

I think that actually 2 rotors and an auto and 2 hangs wins almost everything the first couple of weeks.

200 points....

possibly even less.

Harshizzle 12-02-2017 20:13

Re: How to win Steamworks? (Weeks 1/2)
 
I dont see a consistent hanger being left for the second round of picks (although I would love to be mistaken).

I think two hangs and 3 rotors will be achiveable by the top alliances, with the differentiating factor being teleop/auton fuel, and whether the first gear is done in auton or teleop.

Defense and no match practice for drivers makes me think the 4 rotor alliance will be very rare, at least in weeks 1 and 2.

This actually makes me think that alliances that are seeded 5-8 might have a better shot over other years. If they can get a third pick that climbs consistently, they are at a massive advantage, that can only be offset with a ton of fuel, or an extra rotor. If they know that they only need to keep the opposition within 50 points, they'll make sure that fourth rotor doesn't happen, and then it comes down to restricting the fuel the opposition scores.

6101 Robert 12-02-2017 23:02

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by Harshizzle
This actually makes me think that alliances that are seeded 5-8 might have a better shot over other years.

I definitely agree with you here. Good gear and climbing robots will likely be in insufficient supply. Some smart 5th-8th alliances may even pick gear and climbing robots over better suiting fuel-bots in order to deprive the upper seeds of a solid third partner. You really need 3 robots good with gears to do well, this is another reason why top seeds will find it difficult to carry their alliance partners.

Parker Brotman 13-02-2017 10:04

Re: How to win Steamworks? (Weeks 1/2)
 
Here's what I expect to see in Chesapeake:

There might be a few times in which an alliance gets all 4 rotors, but it will be a rare sight and they likely won't be able to replicate it.

Far fewer than 16 robots per event will be able to climb not just weeks 1 and 2, but for the entire season up until DCMP. I'd actually expect the number of climbing robots at each event to be about 8. Because of this, I expect the higher-ranked alliances to have 2 climbing robots and the lower-ranked to have maybe 1 but probably 0 climbing robots.

On the other hand, I think lower-ranked alliances will be able to score more gears due to serpentine alliance selection. This is a year where (with the exception of climbing) it won't matter whether you can complete the tasks as much as it will matter how efficiently you can complete them. Because of this, it will be important to have 3 capable robots on an alliance.

M.O'Reilly 13-02-2017 10:43

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by 6101 Robert (Post 1644003)
I definitely agree with you here. Good gear and climbing robots will likely be in insufficient supply. Some smart 5th-8th alliances may even pick gear and climbing robots over better suiting fuel-bots in order to deprive the upper seeds of a solid third partner. You really need 3 robots good with gears to do well, this is another reason why top seeds will find it difficult to carry their alliance partners.

+1

I'd almost rather be seed 5, 6, or 7 to get a consistent 2nd pick. In week 1/2, fuel bots will be top ranked teams. Gear bots will be the top picks. Climbers will be at the back end of the snake draft.

IMHO, 3 rotors + 3 hangs wins most competitions week 1/2. Seeds 5-8 have best shot at having 3 climber bots. Top alliance with 2 climbers MUST put up 50 points in fuel to match 3 climbers, a tough order for week 1/2 methinks. Certainly some can do it, but not most.

Daniel_LaFleur 13-02-2017 11:48

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by Brian Maher (Post 1643627)
Now that everyone is starting to get a more realistic feel for what their robots can do, here are a few questions about what you think the typical winning Week 1/2 alliance will look like:
  • What roles will its members play/what will the overall strategy be?
  • Will there be a dedicated defender or will all three robots be primarily offense-focused?
  • Will it be able to score all four rotors?
  • Will higher-ranked alliances be more or less favored than usual?

My predictions:
1> Members of the winning alliance will consist of: 1 gearbot, 1 hybrid gear/fuel bot, and 1 defender/fuel bot. All 3 will climb.
2> While many alliances will go for 3 offensive bots, the winner will have a dedicated defender
3> Yes there will be 4 rotors turning ... once. Once an alliance proves they can turn all 4 rotors, heavy defense will be applied and it will not happen again.
4> The winning alliances will come out of the 1,2,and 8 positions.

Rangel 13-02-2017 12:07

Re: How to win Steamworks? (Weeks 1/2)
 
Overall I think it will totally depend on what the top 2 robots can do at the event. If the top teams aren't too much better than the rest, lower seeded alliances will likely have the advantage. That said, if the top teams can mix lots of ball scoring in without significantly reducing gear scoring output, I think they will reign supreme. Two really good shooter/gearbots will be able to overcome a lack of a third climb. The real question at that point is will they be able to get all four rotors. I definitely don't think any alliance at early events will be unbeatable because of how time consuming four rotors are.

Edit: All these predictions all seem to be working under the assumption that alliances will select the best robots available during their turns to pick. Rarely ever do the top alliances get the 24th best robot in the event. They usually get a robot far better in reality.

hydraskull 13-02-2017 14:17

Re: How to win Steamworks? (Weeks 1/2)
 
I'd be surprised if a regional winning alliance doesn't feature 3 consistent climbers. Missing a climb puts an alliance at a disadvantage, and it's a mechanism that may be the easiest to cheesecake this year - besides a passive gearage. Week 1 and 2 competitions will probably be won by an alliance captain with a hybrid robot, and two gear focused picks. Consistently turning 3 rotors and 3x climb will probably win.

Taylor 13-02-2017 14:47

Re: How to win Steamworks? (Weeks 1/2)
 
Quote:

Originally Posted by Daniel_LaFleur (Post 1644224)
My predictions:
4> The winning alliances will come out of the 1,2,and 8 positions.

The 8th alliance captain has a very interesting advantage, particularly within the first three weeks.

pandamonium 13-02-2017 17:11

Re: How to win Steamworks? (Weeks 1/2)
 
I think that winning alliances will be capable of 4 rotors and 40kpa at many events yes even week 1 and 2. Here is a list of some of the teams competing weeks 1 and 2:

125 175 233 359 118 624 33 67 195 341 126 16 1197 3309 1538 3476 180 148 2468 987 2122 973 1678 610 1114

I fully expect some of these teams to be capable of 40kpa autonomous modes as well.


All times are GMT -5. The time now is 17:30.

Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi