FRC Blogged - Doing the Right Thing

[Hallry]

Taken from the FRC Blog, 3/26/13: http://www.usfirst.org/roboticsprogr...he-right-thing

Doing the Right Thing

Blog Date: Tuesday, March 26, 2013 - 13:06

Over the weekend I attended the Central Washington Regional at Central Washington University in Ellensburg. My experience up until the final round was wonderful. Folks were having fun, the volunteers were great, matches for the most part were running smoothly, I saw some terrific robots, and, I think, I made some new friends. Things did not go as expected in the final round, though, and it caused some significant pain for many present. I’d like to talk about this.

The final rounds were between two very closely matched alliances. The blue alliance won the first match, the red alliance won the second match in a nail-biter – only 6 points separating the two alliances, and the final and deciding match had apparently ended in a tie. The Heads Referee carefully reviewed the rule regarding tie-breakers in elimination rounds in the rule book, and worked out the final score with the Scorekeeper. The Game Announcer announced that we had a tie score, and the venue erupted. He then proceeded to carefully read the rule regarding ties aloud – even with an audience of thousands, you could have heard a pin drop. The first tie-breaker that becomes active is the number of foul points incurred, with the alliance that had fewer foul points being awarded one additional point, and so winning the match. If the number of foul points is equal between the two alliances, there are other levels of tie-breakers employed, but in our case, these were not necessary. The red alliance had a single 3-point foul assessed against them during the match, while the blue alliance had none. When this, and the final score, was announced, the venue erupted again. The blue alliance had won and earned their slots at Championship, after an exhilarating set of final rounds. The Game Announcer later told me his announcement of the tie and the winner was one of the most exciting things he had ever done in that position. I could see why!

Unfortunately, after the winner was announced and the score was displayed, we learned something was wrong. A single disc that would have given the red alliance two points and the victory had been overlooked in one of the goals. This was a red disc, and so was the same color as the front of the goal, making it harder to see than a white disc would have been in the same position. We checked, we double checked, we compared notes. There was no question in anyone’s mind that the disc had been scored by the red alliance according to the rules, but had been not counted in the final score.

I’ve seen some rumors that we reviewed video evidence in determining that the additional disc had been scored; this is untrue. This would have been a direct violation of section 5.5.3 of the manual. I was behind the scoring table the entire time when this was being discussed, and never saw anyone looking at video. At one point it was offered to us, but we declined. We already had overwhelming evidence the additional disc had been scored per the rules.

I want to emphasize at this point that FRC has the most dedicated, most caring, most conscientious volunteers anywhere. But many of their jobs are hard – really hard. And while some of our volunteer jobs have significant requirements showing in their position descriptions, ‘Perfection’ is on the list for none of them. A simple mistake was made, that anyone could have made. The folks involved knew what this meant for the match, and were devastated, but did the right thing by stepping up to the plate to let everyone know what was going on.

So, we made a very hard decision. We adjusted the score in the final match to reflect what we know happened – that the red alliance had actually won the match, and earned those slots at Championship. This was announced on the field. I can only imagine what it feels like to be told you have earned your way to Championship, only to have that opportunity taken away from you a few minutes later. I am very sorry that this happened. No one wants to see an event end like this, but I firmly believe we did the right thing, as difficult as it was.

I want to thank Team 360, The Revolution, Team 2557, SOTABots, and Team 3789, On Track Academy, for displaying the utmost in Gracious Professionalism when it was revealed that they were not, as we believed, the winners of the event. Members of the other alliance came up to me after the event was over and pointed out how gracious they were being in receiving this extraordinarily difficult news. And they were right, of course. These three teams are examples for us all.



I’ll blog again soon.

Frank

[Peyton Yeung]

Hats off to both alliances on this one. That must have been a very tough call and I'm glad everything went ok.

[dtengineering]

What? Wait a minute? Since when did we start deciding tied elimination matches based on obscure rules, rather than playing a tie breaking match?

Okay... I know the rule must have been changed at some point and I must have missed that in the rules book (and likely a lengthy discussion on CD)... but that is the part of this blog post that really bugs me. Bring back "Finals Match 4"!

Sorry to hear how the rest of it went down, but at least everyone agrees that an alliance won by outscoring the other, even if it took a couple tries to get it right. That's better than deciding such a close final based upon a penalty.

Tie-breakers! Tie-breakers! Settle the match the old fashioned way, where someone actually wins on the field!

Jason

[Pault]

Originally Posted by dtengineering

What? Wait a minute? Since when did we start deciding tied elimination matches based on obscure rules, rather than playing a tie breaking match?

Okay... I know the rule must have been changed at some point and I must have missed that in the rules book (and likely a lengthy discussion on CD)... but that is the part of this blog post that really bugs me. Bring back "Finals Match 4"!

Sorry to hear how the rest of it went down, but at least everyone agrees that an alliance won by outscoring the other, even if it took a couple tries to get it right. That's better than deciding such a close final based upon a penalty.

Tie-breakers! Tie-breakers! Settle the match the old fashioned way, where someone actually wins on the field!

Jason

Maybe I'm wrong, but I'm pretty sure this rule was in place last year, and I doubt that they changed it at the beginning of this year only for it to be changed back in an update.

Edit: Section 5.4.4 of the 2012 game manual does describe the same tie breaker system in place now. Can't say for the beginning of this year.

[EricH]

Originally Posted by dtengineering

What? Wait a minute? Since when did we start deciding tied elimination matches based on obscure rules, rather than playing a tie breaking match?

Okay... I know the rule must have been changed at some point and I must have missed that in the rules book (and likely a lengthy discussion on CD)... but that is the part of this blog post that really bugs me. Bring back "Finals Match 4"!

After the 2010 IRI eliminations, IIRC. You know, the one that had 3 or 4 ties in the elims at various points, leading to hours-long delays in ending due to the replays (I think one of the semis went to 5 or 6 matches)? I think after that is when FIRST said "You know, I think we can do this better" and implemented the tie-breakers, for the 2011 season. Of course, we were all so riled up about the minibots that year, and the coop bridges (and Einstein) the next year, that there was no lengthy CD discussion about the tiebreakers.

[dtengineering]

Originally Posted by EricH

After the 2010 IRI eliminations, IIRC. ....about the minibots that year, and the coop bridges (and Einstein) the next year, that there was no lengthy CD discussion about the tiebreakers.

Ah... that would explain it, then... 2010 was the last year that I was actively involved on a team, so I haven't been following the tournament rules as closely since then... and there was certainly enough 'new' in 2011 to talk about aside from something so rare as a tied elimination match.

I can't deny the efficiency of settling an elimination series this way, but really... in all the FRC, FTC and VRC events that I've attended that have gone "over time", I don't think the "tie breaker match" has ever been the main cause. No doubt it happened at IRI once, but it seems weird that an anomaly at an off-season event should impact 'real' FRC events.

Anyway, thanks for the update... and I should really be focusing on the GP shown last weekend, as described in the blog post... but I will be cheerfully grumpy for the next while about how "it was better in the old days when we settled ties the honorable way!"

Jason

[Nuttyman54]

Originally Posted by dtengineering

I can't deny the efficiency of settling an elimination series this way, but really... in all the FRC, FTC and VRC events that I've attended that have gone "over time", I don't think the "tie breaker match" has ever been the main cause. No doubt it happened at IRI once, but it seems weird that an anomaly at an off-season event should impact 'real' FRC events.

IRI was the last straw, but not the only instance of this occurrence, and it was not an anomaly. 2010 was a particularly bad year for ties due to the low "resolution" of scores that year. The WPI regional had two different quarterfinals go to 5 and 6 matches respectively that year. Many other events had similar situations. Like others have mentioned, after an extra hour of quarterfinals matches, everyone just wants to be done with it.

There is a fairness aspect to it as well. Once again going to the WPI quarterfinals, the other two alliances both had essentially an hour+ of downtime waiting for their opponents to be decided. Meanwhile, the #2 seed had to go straight into semis after a grueling 3 extra matches, the last of which with only the standard 6 minute break in between. Their first semifinal was a tie as well, leading to another two back-to-back matches before they made it through to finals. 1735's drive motors overheated in the finals after having played 9 nearly consecutive matches. When all was said and done, 1735 and their partners 2370 and 663 played 11 matches in eliminations, only one less than they did through qualifiers.

This was an extreme example, but it and other similar episodes around the country did prompt the tiebreaker rules to come into play for the sake of the teams, as well as the audience and volunteers. I guarantee that alliance would have performed significantly better in the finals had their robot and driveteam not run through the equivalent of a second regional in just 3 hours. While the games since have not had the propensity for ties that 2010 did, it has come up a few times (including Galileo Semifinals in 2011).

I agree with you that it's much more exciting (in general) to play out ties...but after having been through WPI 2010, I understand the necessity of having tiebreakers in place.

[Joe Ross]

Originally Posted by Nuttyman54

2010 was a particularly bad year for ties due to the low "resolution" of scores that year.

In 2010, ties occurred in ~10% of qualification matches
In 2011, ties occurred in ~4% of qualification matches
in 2012, ties occurred in ~3% of qualification matches
So far in 2013, ties have occurred in ~1% of qualification matches

[dtengineering]

Originally Posted by Nuttyman54

IRI was the last straw, but not the only instance of this occurrence, and it was not an anomaly. 2010 was a particularly bad year for ties due to the low "resolution" of scores that year....
I agree with you that it's much more exciting (in general) to play out ties...but after having been through WPI 2010, I understand the necessity of having tiebreakers in place.

I went and took a look at the first 15 regionals of 2010... after all, 2010 was regarded as "a particularly bad year" due to the game design. 12/15 had no tiebreaker matches. 2/15 had one tiebreaker match. WPI had five tiebreaker matches.

I'll admit that five tiebreakers is a bit excessive, and could be seen as being responsible for delaying the completion of the event, however there were a total of 22 elimination matches played, including the tiebreakers. An event should be prepared for 21 elimination matches, assuming each of the seven series goes to three matches. So technically, even with all those extra tiebreakers, the event would have gone one match... six minutes... beyond what would be achieved in a "longest case" scenario using the current rules.

I submit that based on the 2010 elimination rounds that tiebreakers are an anomaly and are rarely (if ever) the root cause for an event running beyond time limits.

I'll admit that there needs to be something in the rules to allow events to finish at a reasonable time, however I'll also submit that saving six minutes is hardly worth giving up the "tiebreaker" match. And while I'm happy to admit that this totally falls into the category of "first world problems", I'll argue that we should settle ties on the field... in the finals at the very least!

Jason

P.S. at the other two events that had tiebreakers, they both ocurred in the finals... the perfect time!

Oh, yeah... full disclosure... I'm a bit of a hockey fan, and playoff games aren't considered "epic" until at least the third overtime period, so there might be a bit of a pre-existing bias.

[Tom Line]

Originally Posted by EricH

After the 2010 IRI eliminations, IIRC. You know, the one that had 3 or 4 ties in the elims at various points, leading to hours-long delays in ending due to the replays (I think one of the semis went to 5 or 6 matches)? I think after that is when FIRST said "You know, I think we can do this better" and implemented the tie-breakers, for the 2011 season. Of course, we were all so riled up about the minibots that year, and the coop bridges (and Einstein) the next year, that there was no lengthy CD discussion about the tiebreakers.

We were one of those teams at IRI. Our matches against 1114 were epic on a level I will never forget. 5 matches of 5 or 6 robots lifting in every match. 5 matches of insanity.

If it were you in those matches, you would want the people in the stands to stay and cheer. Gracious Professionalism means understanding that those teams playing on that field are creating a memory that they'll never lose, putting yourself in their position, and cheering like crazy.

FIRST is about inspiration, and winning through penalties is hardly inspirational.

[pfreivald]

Originally Posted by Tom Line

FIRST is about inspiration, and winning through penalties is hardly inspirational.

Alliances win through penalties all the time...

I hardly think a tie-breaking system that is clearly outlined in the rules destroys the inspirational value of a win.

[Racer26]

So all this talk of tiebreakers has me thinking:

Suppose for a minute that Central WA was being run under the pre-2011 tiebreaking system of infinite replays until a winner is found.

Based on what I've heard of the incident, it took a few minutes to discover the offending disc, likely enough time to have started the tiebreaker match. What happens then?

My guess is that 2pt disc gets counted in auto of F4, and now you have two matches whose outcomes have been affected by a sloppy count.

Do they still go back and award the red alliance based on their legitimate win of F3?

[techhelpbb]

Personally I am pro-replay/tie-breaker but there's a caveat. I am pro-replay because if the issue is possibly the result of unusual man made intentional interference it thwarts awarding the match to the person causing the interference.

On the other hand replaying a match as other's have noted carries with it the same risk of human error and the circumstances for each robot may differ. There is no reason to suspect that all the robots for any given match will be operational at the same levels from match to match. This can result in a match with a different mix of robots or different factors that directly impact the score.

Again, to me, winning the matches has an element of randomness to it and obviously strategy can impact the score as well as the robot designs and build qualities. I think making it to the competition at all with something resembling a finished robot is a major achievement for anyone that does it.

[EricH]

Originally Posted by Racer26

So all this talk of tiebreakers has me thinking:

Suppose for a minute that Central WA was being run under the pre-2011 tiebreaking system of infinite replays until a winner is found.

Based on what I've heard of the incident, it took a few minutes to discover the offending disc, likely enough time to have started the tiebreaker match. What happens then?

My guess is that 2pt disc gets counted in auto of F4, and now you have two matches whose outcomes have been affected by a sloppy count.

Do they still go back and award the red alliance based on their legitimate win of F3?

No. You see, there are a specified number of red discs on the field. When only 5 turn up, none in the robots, there's going to be a hunt for the 6th. When they find it, they try to figure out how it got there. Eventually, the replay is called off due to improper scoring of F3 and the red alliance is crowned the winners.

[Peyton Yeung]

Originally Posted by EricH

After the 2010 IRI eliminations, IIRC. You know, the one that had 3 or 4 ties in the elims at various points, leading to hours-long delays in ending due to the replays (I think one of the semis went to 5 or 6 matches)? I think after that is when FIRST said "You know, I think we can do this better" and implemented the tie-breakers, for the 2011 season. Of course, we were all so riled up about the minibots that year, and the coop bridges (and Einstein) the next year, that there was no lengthy CD discussion about the tiebreakers.

I remember IRI 2010 and it got to a point when I didn't care who won. I just wanted it to be over.