What did we learn from week 2 of the 2013 season?

[Alpha Beta]

Originally Posted by Becca334

All cameras from teams can be turned off if FTA thinks cameras cause field issues. Even teams with vision tracking had to turn their cameras off at NYC.

Can someone provide more information on this situation (or direct me to it if it's already been posted)?

Was there something specific about the New York venue that contributed to this issue?

Is there a chance teams will show up at another regional this year or Champs and not be allowed to run a camera?

Seems unprecedented that such a well supported sensor for FIRST would suddenly be disallowed carte blanche.

[Becca334]

Originally Posted by Alpha Beta

Can someone provide more information on this situation (or direct me to it if it's already been posted)?

Was there something specific about the New York venue that contributed to this issue?

Is there a chance teams will show up at another regional this year or Champs and not be allowed to run a camera?

Seems unprecedented that such a well supported sensor for FIRST would suddenly be disallowed carte blanche.

What happened was NYC was having extreme communication issues all throughout Thursday, and FTA pegged it as many teams not having the updated of the driver station that was recommended, not mandatory. So late Thursday they had an inspector go around and just make sure everyone then had that update installed before they were allowed to go on for qualifications. They did that.

Then on Friday, the field seemed alright. Here and there were some communication issues which they quickly solved. Great job FTA.

Then once Eliminations came on Saturday afternoon, for the first qualification match everything was buggy. Throughout the match both the red and blue alliance kept gaining and losing comm unexpectedly. That match had to get replayed. After the replay FTA went to the 24 teams participating in eliminations and turned off everyone's cameras and their dashboards. Leaving many teams who've had been practicing for almost 3 days straight now with their cameras and dashboards, now without them in eliminations. I know from the alliance my team was in - 1635 was unable to shoot without their camera, and 375 could shoot from the pyramid, but was unable to accomplish their full court shots without the camera. I also have a friend from 3419, which won an award on Friday night for their vision tracking program, now unable to use it at all in their elimination matches. I'm sure these weren't the only teams effected in eliminations, as I've seen some other NYC competitors on here upset with that change.

All FTA told us was that some team with a camera was using too much bandwith, but they could not peg exactly which team it was and they decided to just shut down the cameras and dashboards for all participating teams. I understand FTA for coming to this, since during one match in eliminations we had a 30 minute delay due to field issues, but I feel if the cameras were still allowed - it would've changed the outcome of NYC because many many of those robots out on the field had cameras that helped them greatly.

I hope this doesn't effect any other regionals or Champs at all.

[Nuttyman54]

Originally Posted by Becca334

All FTA told us was that some team with a camera was using too much bandwith, but they could not peg exactly which team it was and they decided to just shut down the cameras and dashboards for all participating teams.

This statement concerns me. One of the outcomes of the Einstein investigation was to impose bandwidth limits on every team to prevent the system from overloading and cutting off comms to other teams. My understanding is that if a team was using too much bandwidth (from camera or otherwise), they are the only team affected. It sounds like entire alliances were having issues from one team using a camera and possibly using more than their allotted bandwidth. Was this the case? If so, it indicates to me that the bandwidth limits were not working as designed.

[Becca334]

Originally Posted by Nuttyman54

This statement concerns me. One of the outcomes of the Einstein investigation was to impose bandwidth limits on every team to prevent the system from overloading and cutting off comms to other teams. My understanding is that if a team was using too much bandwidth (from camera or otherwise), they are the only team affected. It sounds like entire alliances were having issues from one team using a camera and possibly using more than their allotted bandwidth. Was this the case? If so, it indicates to me that the bandwidth limits were not working as designed.

Yes, one team was affecting everyone on the field with more than their allotted bandwidth.

[Siri]

Originally Posted by Nuttyman54

This statement concerns me. One of the outcomes of the Einstein investigation was to impose bandwidth limits on every team to prevent the system from overloading and cutting off comms to other teams. My understanding is that if a team was using too much bandwidth (from camera or otherwise), they are the only team affected. It sounds like entire alliances were having issues from one team using a camera and possibly using more than their allotted bandwidth. Was this the case? If so, it indicates to me that the bandwidth limits were not working as designed.

I don't know what happened with the cameras at NYC, but we had similar issues at Horsham last week. The FTA advised several teams (including us) to play without our dashboard, explaining that it was our choice but we were likely to experience continued lag (lost packets) if we did not. We turned it off for the match and later down-res'd, lowered FPS and fixed our code--afterwards we and our alliance partners experienced no problems, even with us running vision processing and an second camera and others running 1-2 of their own.

However, this does, to my untrained brain, contradict another FTA statement from Horsham: that when we and another team turned off our dashboard for the match in question, everyone's trip times got better. At least, I that's what I thought he said. It seems counter though, so unless someone else has knowledge of a similar situation, I'll assume I misinterpreted him.* Nonetheless, they did not start the match until after this discussion, and it took quite some time to determine it. (They also found another router that needed to be turned off before this dashboard investigation began. Both elements took some time, not that I fault anyone at all for it.)

*EDIT: I guess so then. Is it possible that the bandwidth limits only identify the team that is exceeding them, rather than stopping them all together? Then again, I don't know why we would have had the choice to leave ours on if it was affecting others (and we certainly wouldn't have had we known, not that we didn't anyway). And apparently NYC didn't know who was causing it. Ok, so scratch that idea.

What I learned from Week (1-)2: apparently there's still more education needed to sort out FMS bandwidth issues.


In other news - I learned to make sure I remind drive teams to look for discs fallen and stuck on your robot (count towards your 4), and also that G30 only applies when you contact your loading zone carpet--not when you break the plane. I also learned that pyramids come apart up to 1/4" in normal match play, and that you absolutely, positively have to wait for the green lights before entering the field. Oh, and that if no one warns them, the lighting crews may want to do some really crazy (cool) light tricks on the vision targets, in autonomous, in the finals.

[Yipyapper]

I learned that, no matter how many times they win a regional, the thoughts of "maybe they won't win this year" won't come true for 2056.

15 regionals attended since their creation. 15 wins in that span.

You guys are so amazing to follow year after year.

[ThomasClark]

As far as scores go, this sums up what I've learned. Basically, what I found most suprising is having an amazingly high pyramid score of 60 points, which less than 0.3% of alliances achieved, and no disc points makes your alliance only slightly above average. It will be interesting to see how powerhouse climbers and shooters rank against each other later in the season.

Also, we've collectively hit 10,000 foul points already.

[dcarr]

Originally Posted by ThomasClark

As far as scores go, this sums up what I've learned. Basically, what I found most suprising is having an amazingly high pyramid score of 60 points, which less than 0.3% of alliances achieved, and no disc points makes your alliance only slightly above average. It will be interesting to see how powerhouse climbers and shooters rank against each other later in the season.

Also, we've collectively hit 10,000 foul points already.

Wow, that's a very nice chart.

I'm not seeing how to read this directly - what are, say, 50th, 75th, and 90th percentiles in terms of overall scores?

[ThomasClark]

Originally Posted by dcarr

Wow, that's a very nice chart.

I'm not seeing how to read this directly - what are, say, 50th, 75th, and 90th percentiles in terms of overall scores?

Thanks.

The 75th and 90th percentiles are 70 points and 95 points, which are written on the right. I didn't include the median, but it was 46 points.

[dcarr]

Originally Posted by ThomasClark

Thanks.

The 75th and 90th percentiles are 70 points and 95 points, which are written on the right. I didn't include the median, but it was 46 points.

Oh okay, thanks.

[Gregor]

Very interesting chart.

The highest scored was actually 211, not 204, scored by the alliance of 2056, 1114, and 1325 in SF 1-2.

The score published right away was lower, but after the manual count it was updated to 211.

[mashygpig]

You shouldn't rely on climb and dump teams, we picked one as the #1 seed at Superior and they never even climbed. A good strategy can score you lots of points even with bad robots. defense really comes to play in finals.

[MattC9]

Originally Posted by mashygpig

You shouldn't rely on climb and dump teams, we picked one as the #1 seed at Superior and they never even climbed. A good strategy can score you lots of points even with bad robots. defense really comes to play in finals.

Can you please explain?

[ttldomination]

Originally Posted by mashygpig

You shouldn't rely on climb and dump teams, we picked one as the #1 seed at Superior and they never even climbed. A good strategy can score you lots of points even with bad robots. defense really comes to play in finals.

While I'm inclined to agree, I think this is a gross generalization.

I'd take a consistent climb and dump team over an inconsistent shooter any day (e.g 4451). The key is consistency and making sure that they know that it's up to them to get their 50 points.

- Sunny g.

[Bryan Herbst]

Originally Posted by mashygpig

You shouldn't rely on climb and dump teams, we picked one as the #1 seed at Superior and they never even climbed. A good strategy can score you lots of points even with bad robots. defense really comes to play in finals.

For those who weren't at Lake Superior: 2220 (#1 seed) picked 2177 as their first alliance partner. 2177 was a stellar team that was able to consistently climb to the top of the pyramid and score 3-4 frisbees by the end of qualifications.

Unfortunately, the #8 seed proved a very challenging defense to overcome. 2177 couldn't get to the feeding station to load up on colored frisbees, and in my opinion spent a little too much time trying to get those frisbees (kudos for coming up with a strategy and sticking to it though).

So what I learned in week 2 is that it isn't too difficult to block the feeding station. In general, defense is playing a far larger role in this game than I thought it would. No offensive strategy is without an opposing defensive strategy.