Scouting: What to look for in a robot?

I’m thinking of putting together a simple spreadsheet for my team to scout. I need a list of what to look for in a robot, so I can rate teams on individual aspects (reliabilty, speed, etc.) What do you think is the most important thing in a robot? Give a list (if you want) from the most important and to the least important things in a robot.

Ex.

  1. If it has a cool team name?
  2. Do they talk to their robot on a daily basis?
  3. Can their robot do back flips?
  1. Is the robot powered by cold fusion?
  2. How quickly can the robot fly?
  3. What chance does this team’s robot have of world domination?
    :stuck_out_tongue:

Drive:
-Gear ratio
-# motors
-Wheel configuration
-Holonomic capability
-Pushing power
-Breaker issues
-Drive over sacks

Arm:
-Gear ratio
-# motors
-Max height (high goal capability)
-How many sacks can they raise
-Stability (lots of arms rock back and forth as you raise/lower a full load)

Intake:
-General category: (roller, shovel, claw, etc.)
-Scores over the back or same direction as intake?
-Single sacks
-5 stacks
-Rows of sacks
-Descoring ability (keeps the sacks or just throws them out)

Programming:
-Types of sensors used
-Different autonomouses for different situations?

General:
-Tippable?
-Driver ability (make sure to differentiate between driver and robot failures)
-Coach/driver/operator synergy (is there a clear leader? are there miscommunications?)

I wrote this pretty fast, there’s probably a lot more I missed. A lot of this isn’t objective and would need a pit run/interview to get the information.

2 phases to scouting (at least for my teams)…

1st phase (pre-tournament)…

  • Do a run through the pit and collect as much technical details as possible… motors configuration… gear ratios… in take system, etc…
    Most of the info required has been covered by SweetMochi in the earlier post.

2nd phase (during tournament)…

  • Strategy, autonomous programs…
  • Any blind spots to exploit…
  • Drivers and coach communications and ability, etc…

What do you mean by trippable. I know we consider a bot pushable if they have just omnis.

How do you all set it up? Do you use Microsoft Excel or Microsoft Word?

We have always used Microsoft Excel, but you can’t fit very many questions on it.

I agree with all of what @SweetMochi said.

-Drive Under Trough

-How many do they score?
-How blockable are they during their autonomous?

-General patterns that they follow during the game.
-How well does their robot complement yours?

Just added a couple of things, again, i’m sure there are more that we haven’t considered yet.

Not “trippable,” he said “tippable,” as in a robot’s tendency to tip over.

We use paper, its really hard (at least for us) to have enough team members/time to have a separate analysis from our input, however that being said I do know that some teams do some sort of excel database.

General: were you a tough team to go against last year?
Are you a widly known team?
did you win anything previous years at worlds? (if applicable)
are you in the top 8 seed?
:stuck_out_tongue:
pretty much that was our ‘mental’ scout list last year :stuck_out_tongue:
And we noted down ‘surprise’ robots that were unexpected as well

So glad you started this post. we struggle with scouting every event. good to see other teams are looking at the same questions we do.

This is a very lazy and flawed way to scout and I wish teams wouldn’t pick others by reputation, especially from previous seasons. New teams and special robots suffer a lot from teams using that mentality, and the alliance captains might miss a really good partner if they only look at the top seeds or teams that were good last year.

Here are 2 examples from the 2012 WC:
-2W was almost ignored in alliance selection at Worlds, since they were seeded low and weren’t particularly well-known. Their robot wasn’t made to win qualification matches, and they were the 3rd member of the 4th alliance in their division. And then… they won Worlds. Not because they were famous or high-seeded, but because they had a good robot.
-44 was the first pick of the 1st seed in the Math division, and although their robot was impressive, it wasn’t really game-breaking like their Round-Up robot, and it could’ve been executed better. As a wallbot, it was fatally flawed in that it could be pushed while it was expanded. In the division finals its drive motors stripped out in a pushing match, and their alliance was eliminated 2-0.

When scouting, I think you should judge all robots equally and impartially. A team might be really good one year, and not so much the next, whereas a different team can improve by orders of magnitude over the course of a season. The best way to scout is to watch as many matches as you can, and look for robots and teams that have strong designs and strategies, and decide whether they would be successful if you were their partner (regardless of whether they won matches paired with a random ally). That way you can find the best possible partners and have the best chance of doing well.

Wow, what a great post. I have read it 3 times and I think I will go back for a fourth after I have some popcorn. At World’s, there were so many teams, that it was easy to just categorise teams on their design type (was it an NZ design or not :smiley: ) and on their wins to losses. We got burned in some qualifying matches using this strategy. I like how you have emphasised doing the hard yards and watching the teams compete, rather than a cursory look at their robot sitting on a table in the pits.

Cheers, Paul

Just to clarify, 44s motors were trashes after holding off another wallbot at a 2:1 torque ratio in semi finals (and not be pushed…), there was essentially no power behind them because of this - also our division were not allowed to cool our motors :confused: We did not pick them because of their reputation - but because we believed (and still do) that their robot complimented ours, a lot of people dont agree with our decision and i can see why, but if we could go back again we would have made the same decision for our first alliance

That being said however, i do agree with your point. A lot of good teams go unnoticed, and a lot of older teams with more of a reputation get picked over them. Scouting is vitally important, if it was a perfect would i would have my scouts no know what team had what robot so they would have to analyse them all equally… I’m not going to go into detail about our scouting list for this season as its all been covered by other people

haha, i didnt say it was our way of doing it for worlds!
when you are at a competition of 36 robots, its not too hard to just “pick out” the good robots by just strolling through the pits
and the ones that are “questionable” can be easily categorized within seconds of seeing it drive (aka, the impressive looking robot that cant drive, or the clawbot with beast drivers)
and in our region, usually the best teams are the ones that have previously been to worlds an won awards :stuck_out_tongue:
and we just seem to “know” their capabilities by personally playing them year after year in tough matches :slight_smile:
dont worry, we do the “usual” scouting at worlds though :wink:

Good point, not so good example. I agree with murdomeek that at local tournaments with only 24-48 this is a pretty good way to scout. However, worlds really does need to be “real scouting” if you want to do well. (please feel free to call me out on this if you want :)) However, using 44 as an example makes me confused since GER had the best wallbot in their division, and LCR thought that that would help them, what’s so bad about picking them? To be honest I think most people would rather be on the winning alliance then with a well-known team. It worked to have both for LCR.

Don’t get me wrong, GER’s robot was great (and to be fair to them, 169A went on to give 2W a bit of a shove in the semis). But the ultimate question with scouting is, could you have been more successful with a different partner?

It’s still a good idea to watch matches at local tournaments, because that is how you get the best results for scouting. For example, at one of our tournaments in the Round Up season, the 1st seed team was undefeated and had a decent SP count. They had a standard clawbot, which we knew to be a sturdy design, and as the 2nd seed team we nearly joined their alliance. What we didn’t know was that they hadn’t faced any good opponents, and in several matches they were paired with a team that loved to farm SPs. The 1st seed robot was actually slow and unreliable. That information was revealed to us by the 3rd seed team, who were gods of scouting had watched every match at the tournament and ranked every robot there by what they had seen. So instead we paired up with the 3rd team, and while we went on to win the tournament, the 1st alliance was beaten out by the 8th in the Quarters.

That goes to show that even in a small tournament, you can easily make a bad choice by looking only at the robot in the pit and the stats on the scoreboard. And even if you know the team from past years, they still may have a great robot with severe electronic issues and a lucky qualification schedule. The ONLY way to know for sure is to watch the matches, and see with your own eyes how they compete. It’s also the only real way to determine their strategy, whether they have a good autonomous mode (or set of autonomous modes), and their driver ability.

It worked out that time for you, but you have to remember that teams sometimes exaggerate to get other teams to pick them. If you were able to verify it for yourselves, great; just remember that they could’ve been making that up so that you would pick them instead.

We considered that, but we figured that if they were lying and the 1st seed team was actually really good, why would they appeal to us over the 1st seed?
Regardless of the good outcome, we decided that the aforementioned situation was too close for comfort. Ever since, we’ve been much more organized with our scouting, and have made a large effort to know how each robot at the tournament might compliment our own.

We use a series of look fors including how well the team works together and how well they coordinate with their alliance partners. We look obviously for the classics how efficiently they score in all three goals.

The biggest challeneg we get is getting good information. It seems everyone sees the importance of good scouting for alliance selection but no one wants to do it. So when we assign scouts it is handled like a homework assignment. With the students doing only as much as it takes to get credit for doing it.

is this a problem other teams see and how do you combat this perception?

I personally think scouting for alliance partners is right up there with design and programming for event success. But I might be off my rocker.