I feel I explained a few posts above, not much more to say
This is our first draft, any thoughts?
Cumbersome to get good data seems like busy work. Not all teams will only do Outer and inner or lower so that is asking for the impossible.
The rest …seems like oh 40 checkoffs each game per team…busy work. Are you asking for a recall of the game from notes? Or do you expect a scout to recall on the fly ? Or change entries on the fly?
Scouting is about overall performance, are they a good partner ? If foe how to defeat?
My 2 cents
It may be a good idea to take a look at what angles you have to be sitting at to see if teams are shooting inner, and if its worth the extra effort on the scouts to try and track it.
The other big thing is tracking shot attempts. If you dont know how many attempts are made, how can you see the average score of balls per cycle? (There are other ways, but it’s a lot easier with this and less assumptions are made.)
Edit: on CP manipulation, you can only select Stage 1 or Stage 2. What happens if they happen to do both?
I count 15. I have some reservations about how ball scoring is done (lots of back spacing and retyping, but I dont have a good solution for that offhand), but it’s a pretty good sheet that gets to the important points otherwise. I dont see how this is considered busywork, unless they just flat out never use the data.
Try doing that sheet over and over , not faster than paper, pen highlighters
Your time savings is in not needing a transcriber. Assuming theres a data connection (needed for gforms to begin with), you have near instantaneous data updates per match.
Assuming scouting was 100% accurate and not just to get through the arduous questions game after game after game
Any “form” without how long did it take? is worthless IMO and 2m30 seconds is not a good answer
Something else I forgot from above that is a good idea to think about tracking, how long their climb is. (Thanks for reminding me boltman.)
For shot cycle time you can divide attempts by length of teleoperated minus climb time to get the cycle time, but imo that’s not as important. You may disagree though.
Cycle time is a good metric, actual individual timed cycle times even better.
IF Cycle time is “not a key factor” what is HANG? Wheel?
One game piece, one goal area and 53% of the scoring seems important IMO
From my scouts during teleop, i want to know how mny PCs a robot scored, in which port and where on the field they were scored from. I dont care about cycles back and forth. Score matters, count a little less for stages. Where they score from tells us what defensive strategies can be used against them or if as a partner we covet the same scoring position on the field. Were using a google form for our scouts. But everyone sjo UK ld do what works for them. Last year we used the blue alliance cargo numbers heavily since that was one aspect of the game we couldnt do.
Although i agree with Boltman about watching matches and gather a qualatative evaluation of each bot. Drive team, defense, other non-quant game play metrics. And everyone exagerates their ability from the pit.
We basically are starting with this form because we will have very few scouts and some will be looking at aspects of all three bots in a match. We like qualitative and quant data, much of the quant data is readily available so we look for scouts to give us their thoughts, a map of where the robot is operating in auto, if they do a specialized task, where they climb so we can match up, a letter grade (subjective but surprising how accurate it becomes in a basic “pile sort” for pick lists, and a target where tally marks and x’s for misses will hopefully give us what we need to form match strategies.
Also with zebra data at all pnw matches, heat maps will gives us better location data than sketchy scout data. We scout only 3 matches per bot.
May I ask when you scout those 3 matches? (beginning, middle, or end of quals)
With a small team (usually 10 students sometimes less) we have started to rely on video scouting for the first day of later tournaments and DCHMPS. We strive for 3-4 matches worth of good data to feed our strategy and initial picklist going into day 2. Then we use the “eye test” to really home in on moving people around on day 2, selecting specific lists of bots that are intriguing potential partners for alliance selection. If TBA gets to real time match video posting, we would stay up all night in between day 1 and day 2 haha
Yep. 1 early, 1 middle’ish, 1 late
Identifying complementary skils isnt hard. TBA and zebra give me better data than somewhat interested scouts. And i watch to add qualitative perspective along with drive team jnput
vast amount of our match strategy comes from heat maps now. Again, later tournaments are much easier. We always play week 1 and it makes that one very difficult to scout in a way that really can drive day 1 week 1 strategy. It honestly just starts as what drive teams (not pits) say they can do in the queue for our first day strategic decisions. “we are planning on doing x, we think we should be able to do y” is our day 1 motto
I fixed the thing about the CP and added other stuff.