View Full Version : pic: High Score in FRC Ultimate Ascent
MikeReilly
20-03-2013, 22:26
[cdm-description=photo]38853[/cdm-description]
Sorry, Finger Lake's got you beat https://fbcdn-sphotos-f-a.akamaihd.net/hphotos-ak-snc6/281806_846866705326_1713411548_n.jpg
:rolleyes:
WeJohnFriedIt
20-03-2013, 22:32
Would you happen to know what caused these scoring glitches? They appeared relatively frequent to me. Was it the new weight-based scoring system acting up?
Jay O'Donnell
20-03-2013, 22:36
I don't know where, but the highest I've heard is 1008 points, or 336 frisbees in the top goal. I'd like to congratulate the team who figured out how to quickly 3-D print frisbees into the goal from their robot :rolleyes:
blueghoul
20-03-2013, 23:15
[cdm-description=photo]38853[/cdm-description]
And that's why the new way of calculating scores isn't the best.
orangemoore
20-03-2013, 23:31
These scores, from what I understand have been contained to mostly the 1st and 2nd week regional. From week 1 to week 2 FIRST implemented a different scoring system. In the blog post they just said it was going to be changed.
Zebra_Fact_Man
21-03-2013, 00:31
These scores, from what I understand have been contained to mostly the 1st and 2nd week regional. From week 1 to week 2 FIRST implemented a different scoring system. In the blog post they just said it was going to be changed.
What is the new system? I haven't heard anything about this change and am rather curious.
Just volunteered at an event this past week, and was this close to doing field reset (so I would have gotten to see first hand what's going on), but I ended up in Pit Admin which was the opposite side of the building. Needless to say, I didn't see much 'til the playoffs.
Haven't really heard/seen any crazy scores since week 1 coincidentally, so this would make sense.
What is the new system? I haven't heard anything about this change and am rather curious.
One of my co-mentors did field reset at KC, and they basically had a small army of field reset and scoring volunteers counting discs as they went in, and hand signaling them to the scoring table. Then they'd count them again post-match to confirm score.
The weight sensors are used for real time scoring, and for most matches they were within 1 disk of the correct score at GTR East. The manual count is the score used in the rankings.
If the real time score is close, they won't publish the score overhead until the manual count is in, but if the real time score has 1 alliance far ahead they will announce the score right away, and make manual changes afterwords.
MikeReilly
21-03-2013, 05:34
Sorry, Finger Lake's got you beat https://fbcdn-sphotos-f-a.akamaihd.net/hphotos-ak-snc6/281806_846866705326_1713411548_n.jpg
:rolleyes:
Bravo, indeed it does!
eddie12390
21-03-2013, 07:37
Would you happen to know what caused these scoring glitches? They appeared relatively frequent to me. Was it the new weight-based scoring system acting up?
It wasn't for this picture but during the practice day at the Pittsburgh regional every blue score showed about ~330 points regardless of how many frisbees they had put into the goal. One of the field reset people told us that a sensor on the blue high goal was broken and would be replaced before the qualification matches actually begin. After that, it was correct 99% of the time with some occasional oddities. (Such as frisbees seeming to disappear from the goal as far as score)
From what I saw volunteering at GTREast:
The weight sensors are small load cells that are relatively fragile. They read out in Volts, and the difference between an empty goal, and a fully loaded goal is ~1V.
The scoring system is then calibrated to the sensors. I presume if the sensor breaks, its output slams to the input voltage (5V? 12V? not sure), which explains the readings being orders of magnitude off.
Dancin103
21-03-2013, 09:41
These scores, from what I understand have been contained to mostly the 1st and 2nd week regional. From week 1 to week 2 FIRST implemented a different scoring system. In the blog post they just said it was going to be changed.
The scores have not been contained to weeks 1 and 2. I wouldn't be surprised to see them off this week as well. While the spread isn't as bad as it was in week 1 it still is not correct. In some instances the regionals aren't re-calibrating the scoring system and they are not hand counting the frisbees, solely relying on the scoring system. We re-ran match footage, included penalty scores, and the final scores were still off by ~20 points...
/randt
My $0.02.
The scores have not been contained to weeks 1 and 2. I wouldn't be surprised to see them off this week as well. While the spread isn't as bad as it was in week 1 it still is not correct. In some instances the regionals aren't re-calibrating the scoring system and they are not hand counting the frisbees, solely relying on the scoring system. We re-ran match footage, included penalty scores, and the final scores were still off by ~20 points...
/randt
My $0.02.
What are you calling "final score"? The score they display at the regional on the screen after the match is not necessarily the "Final score".
As far as I understood it, the FTA's had been instructed to have manual counts done for all matches, not just close ones. In blowouts, they decided to not wait until the count was finished to announce the score. This is why the 1114/2056/1325 alliance's top score of 211 was shown on the big screen as 204.
Chris is me
21-03-2013, 10:31
The scores have not been contained to weeks 1 and 2. I wouldn't be surprised to see them off this week as well. While the spread isn't as bad as it was in week 1 it still is not correct. In some instances the regionals aren't re-calibrating the scoring system and they are not hand counting the frisbees, solely relying on the scoring system. We re-ran match footage, included penalty scores, and the final scores were still off by ~20 points...
Are you absolutely sure they are not hand-counting frisbees? Sometimes they will display a score before they correct it with a hand count, but only if the match outcome is obvious. The updated score in this case is put into the ranking system but not announced.
Dancin103
21-03-2013, 11:18
Are you absolutely sure they are not hand-counting frisbees? Sometimes they will display a score before they correct it with a hand count, but only if the match outcome is obvious. The updated score in this case is put into the ranking system but not announced.
Yes, at one event so far my answer is yes. There is no way the frisbees were hand counted because the final posted score does not match the score when you replay the video of the match...
But it happens and getting these things to be consistent at each event is a challenge.
From what I experienced as a volunteer doing Field Reset at Pittsburgh was this:
FIRST found out the sensors were a bit wonky in week 1 and week 2 so they started having some field reset people and volunteers count the frisbees. The weight sensors would only be used, like people said before, for real time scoring and could very well be off.
What Pittsburgh ended up doing was having the Field Reset people both count during the match and actually take out the frisbees from each goal at the end of the match, count and log them with the scorekeeper. The scorekeeper will then log everything, including penalties and pyramid goals and climb points and such, and then show the "Final Score" (the score that will be the "official score" used for everything needing it like the Qualification system and such). The score that people see during the match and immediately after is the estimated score based solely on the weight sensors in the goals (which can and have been off before)
Going into the future, I think they're probably going to require that regionals hand count frisbees to avoid errors in scoring, especially when a match is a close one.
Chris is me
21-03-2013, 12:40
One minor clarification: FIRST has been hand counting frisbees after the match and having a volunteer give a count during auto since week 1 at BAE. I'm really, really surprised that some events didn't get the memo.
Yes, at one event so far my answer is yes. There is no way the frisbees were hand counted because the final posted score does not match the score when you replay the video of the match...
But it happens and getting these things to be consistent at each event is a challenge.
Were the 20 missing points maybe invalid climbs? May we see the video?
JohnFogarty
21-03-2013, 13:51
I was a field reset volunteer for Peachtree and I noticed that when this happened a Frisbee somehow got wedged in a very precarious position in the goal pushing the lexan "weight plate" down.
Robotmmm
21-03-2013, 13:57
Dancin103
What event? I was at the TCNJ District where you last competed and was curious about if they were counting. I carefully watched the volunteers count and record all the discs after each match. I even counted the shots sometimes and checked it against their count and scoring and it seemed correct in the matches where I did this.
Kristian Calhoun
21-03-2013, 14:05
IMO, there should be two people manually verifying the scores for each alliance.
I would also gladly wait the extra 2-3 minutes for the scorekeepers to input the official manual disc counts into the system before displaying the scores to the audience. Even if the outcome of a match isn't changed, it would show everyone at the event that the scoring is at least being done properly.
KrazyCarl92
21-03-2013, 14:13
IMO, there should be two people manually verifying the scores for each alliance.
I would also gladly wait the extra 2-3 minutes for the scorekeepers to input the official manual disc counts into the system before displaying the scores to the audience. Even if the outcome of a match isn't changed, it would show everyone at the event that the scoring is at least being done properly.
I volunteered as a scorer at an off season event this year. At least for last year, the FMS system could not begin connecting to the next matches robots until after the score from the previous match was posted. This means that in matches where you wait for the discs to be meticulously counted, waiting for that count becomes the rate determining step in match cycle times. I believe this may be why FIRST elected to post scores ASAP for matches that aren't close; it will speed up the match cycles because they can connect to the next match's robots sooner.
JohnFogarty
21-03-2013, 14:56
I volunteered as a scorer at an off season event this year. At least for last year, the FMS system could not begin connecting to the next matches robots until after the score from the previous match was posted. This means that in matches where you wait for the discs to be meticulously counted, waiting for that count becomes the rate determining step in match cycle times. I believe this may be why FIRST elected to post scores ASAP for matches that aren't close; it will speed up the match cycles because they can connect to the next match's robots sooner.
You are correct, the system still works that way this year.
IMO, there should be two people manually verifying the scores for each alliance.
There are multiple field reset volunteers who manually count the frisbees. The same people who get them out of the goals also count them. It takes a small army to do field reset this year. Every match is manually counted.
Kristian Calhoun
21-03-2013, 16:06
There are multiple field reset volunteers who manually count the frisbees. The same people who get them out of the goals also count them. It takes a small army to do field reset this year. Every match is manually counted.
I'm not debating that point. I've been to events this year that only had one person counting the frisbees for each alliance for the entire weekend. In such cases, it would have been nice to have some redundancy/error checking in the final counts.
mhos1997
21-03-2013, 16:08
We just saw a score of 514 from all disc points in LA. Then we read about the glitches. =(
I'm not debating that point. I've been to events this year that only had one person counting the frisbees for each alliance for the entire weekend. In such cases, it would have been nice to have some redundancy/error checking in the final counts.
That's not good. Do you know if it was a shortage of volunteers, or just disorganization?
Jared Russell
21-03-2013, 18:23
Yes, at one event so far my answer is yes. There is no way the frisbees were hand counted because the final posted score does not match the score when you replay the video of the match...
But it happens and getting these things to be consistent at each event is a challenge.
I saw volunteers manually counting the discs at Hatboro-Horsham and Chestnut Hill. I would have felt more confident with more volunteers to count/verify the score. Are some mistakes/inaccuracies here and there still going to happen? Sure, but this shouldn't be 20+ points. The score that gets displayed on the audience screen isn't necessarily the one that includes the hand count, particularly in a blowout. However, the score that goes into FMS (and shows up in the official FRC match reports) should always be accurate.
akoscielski3
21-03-2013, 22:53
WOOT WOOT 762 points FTW!
Too bad we didn't have 10 more points ;)
I'm not debating that point. I've been to events this year that only had one person counting the frisbees for each alliance for the entire weekend. In such cases, it would have been nice to have some redundancy/error checking in the final counts.
We had two volunteers on each end counting them at the Detroit District last weekend and it worked really well for us. We would tally them as they flew into the goals and take the sheets to the scoring table right after the game ended. If there were any discrepancies between the scorers, we would take the discs out of the goals and count them. It was only really an issue when discs were flying into the goals very quickly (thanks, 469...).
As far as I know, we didn't get any complaints about inaccuracies. I don't know how accurate it was when compared to the results from the sensors, but it seemed like you could tell the outcome of a match as it ended unless the point totals were within 5 or 10 points.
For the record, we had 8 or 9 field reset volunteers, and it worked out well for us. Field reset was definitely more work this time than it has been for the past few years.
GaryVoshol
22-03-2013, 21:49
A person at the score table at Grand Blanc today reports that while discs counted by humans* disagree with the sensor score, in most matches it is only one disc per goal that must be adjusted up or down.
* I believe, but do not know for sure, that they are counting the discs as the goals are emptied. They do a visual count at the end of autonomous.
FYI, the --real-- high score as of week 4 is Waterloo's 277-81, scored in Q28 by 1114/2056/1846 beating 1241/2200/4039.
FYI, the --real-- high score as of week 4 is Waterloo's 277-81, scored in Q28 by 1114/2056/1846 beating 1241/2200/4039.
That match had 89 foul points in the 277.
The 1114/2056/4069 alliance put up 274 in the semifinals with no fouls.
That match had 89 foul points in the 277.
The 1114/2056/4069 alliance put up 274 in the semifinals with no fouls.
Actually, the highest score according to FIRST data is 243 in second quarter final match...still impressive;)
You seem to be right, but I'm sure I remember 274 going up. Perhaps a miscounted RTS score. Weird.
Ivan Malik
24-03-2013, 13:35
A person at the score table at Grand Blanc today reports that while discs counted by humans* disagree with the sensor score, in most matches it is only one disc per goal that must be adjusted up or down.
* I believe, but do not know for sure, that they are counting the discs as the goals are emptied. They do a visual count at the end of autonomous.
As one of the counters at Grand Blanc I can confirm this. We not only had people counting the discs as they entered the goals, but also exiting in case of bounce outs, that way we were 100% sure. Elims saw all discs counted out at the end of a match by hand twice by two different people, Quals only if a visual count was hard to do at the end (stacked discs and what not.) Auton discs were counted both during and after auton visually. We tried to think of every angle. With 3 people counting per side, sometimes 4, if the real time was off then we knew the real score.
AluminumNarwhal
29-03-2013, 00:50
During Inland Empire's Practice Match 25, somehow the real-time scoring reported the final score as 42-2817. Interestingly, 2803 of those points were listed as foul points.
xSAWxBLADEx
29-03-2013, 01:40
During Inland Empire's Practice Match 25, somehow the real-time scoring reported the final score as 42-2817. Interestingly, 2803 of those points were listed as foul points.
https://sphotos-a.xx.fbcdn.net/hphotos-ash3/521348_10200454135383186_1161579226_n.jpg
...The picture in the post above...
Go home live scoring system,
You're drunk
But really, the foul points? I...what? Unless some ref didn't realize that his phone was set to vibrate right on top of the foul button, how is an automated system supposed to ever add foul points?
markmcgary
29-03-2013, 04:05
Go home live scoring system,
You're drunk
But really, the foul points? I...what? Unless some ref didn't realize that his phone was set to vibrate right on top of the foul button, how is an automated system supposed to ever add foul points?
Our bot went rogue and was getting friendly with the blue pyramid. [j/k]
Our bot went rogue and was getting friendly with the blue pyramid. [j/k]
If only there were 196 less foul points, then I'd be impressed.
my guess? the refs decided to have fun and see how many times they could push the techfoul button in a match.
JoeWithTheSpecs
29-03-2013, 22:34
Pretty certain most people have seen this practice match score already, but here it is anyways;
http://f.cl.ly/items/0J3V0J3V3v1138312p1a/Image%202013-03-07%20at%208.54.08%20PM.png
2803 Foul points here as well? Conspiracy I tell you.
Zebra_Fact_Man
29-03-2013, 23:48
Pretty certain most people have seen this practice match score already, but here it is anyways;
http://f.cl.ly/items/0J3V0J3V3v1138312p1a/Image%202013-03-07%20at%208.54.08%20PM.png
2803 Foul points here as well? Conspiracy I tell you.
28 v 26, that's a pretty close match.
But to look at it another way, we should be impressed that these alliances were able to invoke a 20pt penalty more than once a second for 120 seconds. That's actually a rather impressive feat.
28 v 26, that's a pretty close match.
But to look at it another way, we should be impressed that these alliances were able to invoke a 20pt penalty more than once a second for 120 seconds. That's actually a rather impressive feat.
Well at least this one is only 16 off
GaryVoshol
30-03-2013, 07:13
Those Blue refs are just slow.
It was a bad call, anyway. :yikes:
vBulletin® v3.6.4, Copyright ©2000-2017, Jelsoft Enterprises Ltd.