View Single Post
  #3   Spotlight this post!  
Unread 11-02-2006, 12:44
Ethulin Ethulin is offline
Too many hats to count
AKA: Erik Thulin
FRC #0492 (Titan Robotics)
Team Role: Leadership
 
Join Date: Mar 2005
Rookie Year: 2003
Location: Seattle
Posts: 245
Ethulin has much to be proud ofEthulin has much to be proud ofEthulin has much to be proud ofEthulin has much to be proud ofEthulin has much to be proud ofEthulin has much to be proud ofEthulin has much to be proud ofEthulin has much to be proud of
Send a message via AIM to Ethulin
Re: scoting program

Quote:
Originally Posted by Eugenia Gabrielov
Let me go from where EricH is talking and build on that. I'll focus on match scouting.

Step 1: What do you think is important to your team?
Before you rush to collect lots of data, try to figure out what your team thinks is most important. If you guys are an all-defensive bot, you're probably looking for a shooter. Maybe you can only shoot into the center goal, so you need a team with a high record in the corners. There are many things to consider. Maybe you can't store balls, so you need a team that can store as many balls as possible.

Step 2: How do you quantify those importances?
In each case, you have found a need for your team. You, as a scouting leader, need to figure out how to draw that data from the match quickly. I suggest this format: a sheet with things you take a tally of at the top (such as balls scored for the robot, or balls stored in the robot), followed by a large chunk of the paper for guided commentary. You may want to include at the top of this blank sheet..."Scouts, please comment on these things." Or you can write some questions and match scouts can answer each one.

You don't need to be programming savvy. Say you just want the average number of balls scored per match by a team. Microsoft Excel will do it for you happily, and is available on just about every computer as well as most PDA's. You can insert numbers for each match and ask it to calculate an average and standard deviation, and you can share that data with your pit. Keep in mind that not every match is balanced: sometimes a good robot may be in a weak alliance, so it may have to act out of character from its usual abilities. You should note that sort of thing on your comment sheets.

Step 3: Analyzing this information
The first people that need to look through this information are the scouts. Quickly! One leader should be assigned that reviews all the scouting data for a match between matches and compiles it. While they're throwing all those balls back in, and setting stuff up, you can be analying match data. Our team's scouting leader will have a large notebook that they will use after each match to record data. Data from this notebook will then be shared with the Pit via a runner. You can have Excel or another computing program running and just enter numbers, and the program does the calculating for you.

Let me know if you need any help setting up a program. I'd be glad to show you ours and walk you through creating your own.

Good luck!
Eugenia Gabrielov
Scout, 461

I like a lot of that, but step 1 I think is a bit lacking. Look at what the goal of your scouting is. If the goal is just to pick alliance partners then the above is fine. If it is to do match prediction then picking what you want to record really does not cut it. We will be recording every thing that happens in the match using STAMP (stampscouting.org) and video taping every match for review later.
__________________
2003 Pacific Northwest Regional Website Award || 2003 Pacific Northwest Regional Finalist || 2003 Pacific Northwest Regional Engineering Inspiration Award || 2004 Pacific Northwest Website Award || 2004 Pacific Northwest Regional Champions (#1 seed) || 2004 Galileo Semi-Finalist || 2005 Pacific Northwest Regional General Motors Industiral Design Award || 2005 Pacific Northwest Regional Champions (#1 seed) || 2005 Galileo Finalist || 2006 Pacific Northwest Regional Chairman's
Stamp Scouting || Titan Robotics