paper: Spanking the Children

[Jim Zondag]

Thread created automatically to discuss a document in CD-Media.

Spanking the Children by Jim Zondag

[Jay Meldrum]

Very interesting read. Cool to read some of the history behind the penalties in past FIRST games.

I would agree with you on the 50 point penalties this year. They pretty much end the match for any team that gets one.

In Waterford, we received 50 points due to the other alliance ingesting our ball. This happened when someone on our alliance was trying to pick up the ball and pushed into the opposing teams intake. The crowd was very upset when they were given this penalty because it was clear they did not try to intake the ball, it was pushed up into them, even though the team that pushed it into them was trying to pick it up themselves.

Though the rules are very clear on this; G12..."A BALL that becomes unintentionally lodged on a ROBOT will be considered POSSESSED by the ROBOT. It is important to design your ROBOT so that it is impossible to inadvertently or intentionally POSSESS an opponent’s BALL."

This does make a very difficult design constraint for teams.

[Thad House]

That was a very interesting read, and alot of good information for us that have only been around a few years.

One big thing I've noticed is alot of the scores seem really inflated. Last time penalties had a large influence on games, it took away from your score, instead of giving it to the other team. So it does look like the scores are inflated, but if they worked like the old way, I bet we'd see alot more 0 score matches.

Im my opinion, low number of game piece games will always be harder to correctly penalize. Each piece has a much larger outcome on the final score, and its really hard to calculate penalties for this. Last year most penalties handed out, at least where I saw, were worth 1 game piece, where i'd guess about 15 disks on average were scored. This year, the most common penalty is worth 1 full cycle, but alliances have much more trouble scoring that, so the penalty to scoring ratio is much higher this year then last, and in fact any game I can remember since 2010.

[magnets]

This is very cool.

Robots with 6 cim two speed drives, which now can be bought, are very, very fast this year, and high speed collisions between robots going 16 fps are destructive, especially if a ball pickup device is involved. Back before districts, some teams would play between 16 - 19 matches at one event to qualify for the championship. Now, teams who will end up at CMP will have played over 20 rounds (not included replays) at each of at least two district events, plus matches at the district chamionship. Reliability is a great feature for this year.

EDIT: There's one thing I disagree with. As a spectator game, 2003 was better. Autonomous was exciting, with lots of robots going quickly, smashing through bins, getting air off the ramp, and smashing into other robots, and teleop just turned into battlebots.

[rees2001]

Great paper Jim. I still remember your reaction at Championship in 2008 when the opposing alliance crashed into your drivers station wall in autonomous and knocked your controls off the shelf. When your driver team jumped in to catch it, your. Team was penalized. You went OFF!

[cadandcookies]

As someone who didn't compete in FRC until 2011, this was a very interesting read, especially for the history. Thank you for sharing this.

I suppose I hadn't completely realized how good the last three games have been.

[ErvinI]

Violence was still pretty much unchecked and strategies of pounding the offense into the rack while they attempted to score abounded (ask Karthik about this sometime).

Can't wait to hear this story .

As for 2013, the only penalty gripe I really had with it was the massive penalty for inadvertently touching a robot in the act of climbing, which was worth around 50 points. Fortunately, this happened rarely, although it did happen to us at Buckeye (our back end swung into 1551 while they were hanging in QF 3-3 and we were trying to get to our end of the field. They shook on the bar for what felt like an eternity, but ended up staying on the bar. They got assessed an automatic climb (+30 pts), we were assessed a technical foul (+20 pts), and their original 10 pt. climb was still valid).

[jimwick]

Part of the problem this year is that the referees just had too much to watch, and too many of the penalties were close judgement calls. It was impossible for the referees to see all they needed to see and quickly make fair judgements. We saw numerous referee errors, and who can blame them?

I agree that it made a miserable game. We lost a number of matches that we should have won, and vice versa, because of penalties.

Games really need to be designed so that penalties aren't a big factor. It is possible, as Jim recalls in his paper.

And maybe 2 CIMS are enough motor for a drive train, (4 CIMS if mecanum drive). This constant bashing doesn't really improve the game.

[orangemoore]

Can anyone elaborate what the "Wildstang Pinata" is from 2004 that would be great.

[jblay]

This was a truly awesome read and I agree 100%. This game is on the tip of being so good, but it is so bad because of these penalties.

And the worst part is that it hurts the young teams. The experienced teams make sure they are aware of the penalties and risks and that their partners are aware, but a young team with a great robot doesn't have the experience to make sure these things don't happen nor should they. This should not be what wins and losses a team events.

[billylo]

Thanks Jim. This puts thing in perspective (in relation to past games.)

I am summarizing data in support of your pledge for a change. Will share it shortly.

Billy.

[EricH]

Originally Posted by ErvinI

Can't wait to hear this story .

I'll let Karthik fill in the gory details, but suffice it to say that the reffing was extremely inconsistent between fields at Championship. 1114 and 330 played on a field with very lenient reffing--you could tear someone's robot apart and probably get away with it (Again, ask Karthik--I think he's got a better memory on that one that I do). OTOH, on Einstein, a simple robot-robot high collision that was a result of two robots going to the same scoring space and just happened to tip one over resulted in a disable-DQ to the other alliance.

but the game was always compromised by a collection of penalty rules which seemed to have been written by people who had never built or driven a robot.

This particular item rather bugs me. I would REALLY like to know exactly how many current GDC members have EVER been ON an FRC team, in any capacity whatsoever. Some of their rulings in the past have been pretty lousy, to say the least. (My favorite question was dished out by FRC116 in 2011, I suspect by Mr. Lavery himself after he left the GDC, asking exactly what in the rules prohibited a certain activity, given that it had previously been allowed by Q&A and appeared to fit all the rules for the minibots. In the process, the author made the entire GDC look like a bunch of... well, I won't say exactly what. Wish I could find it; it's in an old Q&A system.) With all the "possession" questions they had early in the season this year, it still took two or three updates to clarify an item that could REALLY affect the game--and the fact that clarification was needed was shown by the sheer number of Q&As on the topic.

[Caleb Sykes]

The GDC really put themselves in quite a jiffy this year by only giving alliances one way to score (with the ball) and giving that score a variable value (based on assists). If assists were scored as they occurred, instead of a massive score at the end of the cycle, it would be perfectly reasonable for fouls to be worth 10 points and technicals to be worth 30 points. As it stands though, illegal actions have a drastically variable impact on the game.

If a robot that is not holding a ball gets pinned at the start of the match for 8 seconds, the alliance of the pinned robot is only marginally hurt (certainly no more than 20 points). However, if, with 10 seconds left in a match, a robot holding a ball with 3 assists on it is pinned for 8 seconds, the alliance with the pinned robot could likely be hurt by 40 points.

This is not to say that I would not still be in favor of reducing the value of penalties, but we do have to think of the consequences of doing so.

[Duncan Macdonald]

Thanks for this Jim, really well done.

Originally Posted by ErvinI

Can't wait to hear this story .

http://www.chiefdelphi.com/forums/sh...ght=curie+2007

The short version is that the game had points for elevating your partners so a lot of teams built dedicated ramp bots that had little ability to manipulate the game piece and no way to positively contribute outside the final 30 seconds. Additionally the tube scoring was exponential, so stopping your opponents from placing tubes on specific locations or at all was really valuable for a ramp bot because you could be around 3 ideally placed tubes down on the scoring structure and win with a double elevation.

1114 was a really good offensive robot and 48 was a really good ramp robot specializing in tough defence. (The structure in the middle of the field was over 300 pounds and the robot interaction was moving it around the field)

Here is a video of the match but it doesn't show anything

[P.J.]

Originally Posted by EricH

I'll let Karthik fill in the gory details, but suffice it to say that the reffing was extremely inconsistent between fields at Championship. 1114 and 330 played on a field with very lenient reffing--you could tear someone's robot apart and probably get away with it (Again, ask Karthik--I think he's got a better memory on that one that I do). OTOH, on Einstein, a simple robot-robot high collision that was a result of two robots going to the same scoring space and just happened to tip one over resulted in a disable-DQ to the other alliance.

My team (910) still talks about this one...it's hard to forget about your alliance getting red carded on Einstein.