Last active 21 hours ago

  1. 5 days ago
    Wed Jun 13 10:33:51 2018
    A Alex241N posted in V5 Vision Sensor Object Tracking.

    I think this will be the season of insane autonomous routines. Both during auto, as well as in the driver controlled portion. Miss a ball? let the robot recalculate the position. Or let the robot base position off of the flag. Dang.

  2. 6 days ago
    Tue Jun 12 14:09:59 2018
    A Alex241N posted in New qualification Document.

    @TheColdedge Then come up with a better way to find the best team. Taking the top 12 is the best way from the data we are given to find the best 10 or so robots.

    It's interesting that this is the argument you guys choose to make without stating the obvious which is not everyone does skills.

    I personally just found it interesting that the mean qualification rank was so low. It is not unfair to say that being in the top 35 doesn't necessarily mean you are competitive.

    This could all be a symptom of the game though, as a simple mobile goal bot could be highly competitive in skills last year while being fairly meh in the tournament.

    1.) You're making an invalid conclusion, because the data isn't correlated. You can't make that conclusion!
    2.) This isn't the argument we've been making. It's a rebuttal to an argument you've made. We've made plenty of arguments, but you haven't taken the time to find a proper rebuttal.
    3.) Yes, that is correct. But, that conclusion does not imply that high skills scores don't correlate with tournament success- and they obviously have, as 23/35 teams who had a high skills score made it to the elimination brackets. You've stated that qualification rankings are a better indication of robot capabilities than elimination brackets: at least half of the teams in elimination brackets received that spot as result of their qualification rankings. This indicates that elimination brackets are a valid indicator of success.
    4.) Here, you make the assumption that 1.) Building a good skills robot (especially if it only does mogos) does not correlate to building a good competition robot, and 2.) Team's don't rebuild robots to compete better in divisions.

    That would ignore the high level of accuracy required especially in programming skills; in my opinion a good indicator of robotics capability.

  3. Tue Jun 12 13:48:43 2018
    A Alex241N posted in New qualification Document.

    @TheColdedge What we should be able to agree on is that the top 12 teams are competitive and that only 4 were in the top 35.

    Being a division champion does not equal being one of the best 12 teams. It merely means you were the best team in your division! By grouping uncorrelated data to reach a conclusion, you're ignoring the variance in division competitiveness.

  4. Tue Jun 12 13:44:17 2018
    A Alex241N posted in New qualification Document.

    Someone mentioned 28 out of the TOP 35 skills ranking teams (I confessed - I didn’t verify this data) made it to elimination.
    Now... so that doesn’t mean anything to you? But instead you rather choose a data point to show that only 4 out of 12 division champions were from the skills ranking teams.

    What bothers me the most about this argument from @TheColdedge is that it ignores the rest of the elimination bracket. It stands to reason (without checking) that eventual tournament champions that were in the top 35 in skills beat out alliances that also had top 35 in skills.
    Basing an argument on who was tournament champion alone is irrelevant, because it ignores the structure of the elimination bracket, where not every team plays every other team.
    I think @meng touches on this quite well with his follow up argument. You can't use a statistic as a valid argument if it obscures other relevant data!

  5. last week
    Fri Jun 8 13:44:45 2018
    A Alex241N posted in New qualification Document.

    I'm going to be honest: the new changes aren't all bad.

    I enjoy the immediate weighting skills runs have on individual qualifying tournaments; it makes skills runs important at every event of the year, and evens the playing field at regionals by providing a larger barrier to teams that want to use one competition as a skills 'prep' to set themselves up for the highest season skills run in their region, while sacrificing competitive edges in the regular tournament.

    What I don't think is particularly well executed is the effect on worlds-qualification spots previously allocated to high skills runs. Skills was a way to allow exceptionally strong regions to show up at world's proportional to that regions actual capabilities, regardless of size.

    That said, in the last few years, the only region that has been historically small that continues to succeed and punch above it's weight at skills is Singapore. I think that performance should be rewarded. Regions that consistently punch above their weight should be rewarded with more world's spots. Small regions with the potential to grow should also be allocated more world's spots. With these two changes, even small or new regions would have the incentive to expand.

    That's my take on the skills situation.

    As for a smaller aside: signature events. I'm absolutely not a fan of these. The only barrier to competition here is funding, and I don't think that is a good precedence to set. As a result, these competitions will be far from "international", and will most likely be regional or semi-regional at best, with a few odd foreign teams.

    I'd like to see the recf focus on growing and nurturing smaller regions, and incentivising new regions to join.

  6. 2 weeks ago
    Mon Jun 4 17:09:13 2018
    A Alex241N posted in Power Expanders.

    Also of note: the power expander allows each port to only power (supply current to) one motor.
    By ycabling from 4 of the cortex 3 wire ports, and connecting one end of each y-cable to the power expander, and one end to a motor controller/motor, it's possible to separate the load very evenly, without having to use ports 1 and 10.

  7. Wed May 30 12:07:56 2018
    A Alex241N posted in LED Strips - Revisted.

    As far as I'm aware, powering them off the cortex is not competition legal.

  8. 3 weeks ago
    Thu May 24 21:24:20 2018
    A Alex241N posted in Turning Point Lift .

    @Wiredcat Robotics With all this talk about lifts being linear or not, here's my input (as someone who spent the entire season building a basic chainbar).
    You miss.
    You miss a lot and you'll often find yourself having trouble positioning yourself in the right spot so you can hit your target. The few times this season my team attempted positioning ourselves to grab cones of the loader, we would always have to spend a good 5-6 seconds making sure we were in the right spot, and even after that adjusting to the right position would be difficult and usually knocked several cones off the loader. Even though we spent a good time practicing we still had trouble putting ourselves in the right spot.
    Another thing that we also have to account for is space on your robot - cascade lifts can leave more space on your bot for another mechanism to load/shoot balls (if you're going for a robot that can do everything, that is)

    I've done non-linear lifts every year out of my four years. You only miss if your lift is far too large or if you don't practice.

  9. Thu May 24 14:29:19 2018
    A Alex241N posted in Turning Point Lift .

    @MayorMonty I assume the reason most people are wanting to do linear lifts, or reverse double four bars is that they want the lift to be linear, which usually makes driving much easier. But it may turn out that it won't really affect it, only time will tell

    I've found the opposite to be true, with many drivers struggling to line up linear lifts quickly at the edge of fields.

    Also, with the expansion zone this year, the longer it takes to drop your lift the more time you may waste if no scoring objects can be reached from the expansion zone.

  10. Thu May 24 13:46:39 2018
    A Alex241N posted in Turning Point Lift .

    The tallest posts are 32". A robot has a maximum of 18" in height. Double the robot's height is 36".

    I'm extremely confused as to why so many team's aren't considering simple chain bars- attach claw to cap, back up to post, swing arm over, cap settles onto post. Added bonus if your robot can 'slot' onto the post with guides for easier scoring.

View more