hey all, just wanted to drop by to say I’m working on a big update for my scouting software, Roblu (https://www.roblu.net/). I’d like to add additional features that you guys would find useful. I’m trying to build a scouting utility for teams to use that don’t have the time or resources to build their own, so even if you don’t plan on using Roblu, I’d love any ideas you have or things you’d find useful, or cool things you’re doing with your scouting software right now. You can join the Roblu discord over at: https://discord.gg/NBmNSYk. A couple new features coming in the update: Laptop hub mode (a laptop will act as a hub to sync devices and keep an excel document updated), public events & chat (follow, fork, and chat with other Roblu users at events), a fully fledged account & events API for storing scouting data and syncing it seamlessly across your devices, and Android & IOS support. Thanks!
“eliminates the hassle of paper”
Some students would agree, some would not. Nothing wrong with paper.
That being said I hope your app does well.
I have not met a lead scout yet that after I talked to them wanted to do an app (some have tested an app side by side or a scouting group) . They may not like paper and its volume, yet its highly effective. I never tell my lead scouts how to scout just the advantage of paper excel and notes. So far all have chosen paper as the primary scouting method for thier scouts
Its sort of like Vinyl and CD , there are those that prefer Vinyl.
After going trough anywhere from 100-300 sheets/event the last 2 years, I disagree. Way too many trees cut down for comfort.
You are tracking too much if you have 100-300 sheets IMO , theres effective paper scouting and ridiculous paper scouting . Most of what you need digitally is on Blue Alliance anyhow. No need to duplicate efforts.
Information overload is real not only in FRC but in real life, only so many things are truly important and not all teams need to be watched and tracked so there are opportunities for less tracking of every single item.
We simply track those we play with and those that can help us should we have that opportunity along with those we may face in that extended opportunity. That cuts down the workload significantly,
One of the main issues I have with all digital scouting is sample size and data validity …thats why I take notes by hand and use Blue Alliance to verify the notes. A double check system.
You can find our scouting sheets last year in my profile (somewhere…) if you really want to see if we took in “too much” data. We found it to be a good balance, just not enough dead space to make it fit on half a sheet. We found that just about all the datapoints we tracked to be useful in some way, and none of us knew (or knows) enough to currently take advantage of the TBA API for it to be of much use to “check” (see below) without taking up extra time with paper scouting anyways.
District events are around 2/3rd-1/2th of the size of the regionals you went to last year (one of our evens had 28 teams total) which makes it more important than every to scout every single team instead of just those we deem a “threat” or an “ally” since basically everyone will be on the carpet come elims. We found it important to know everyone’s strengths and weaknesses in elims to better combat strategies they may pull, even if it doesnt work out in the end.
Maybe im blind, but can you find on TBA how many cubes 188 scored in the switch in QM9 of 2018oncmp1?
28 teams is our max/min pick list size so districts would make no difference in our scouting preference. If all 28 come to elims so be it. And yes we would now how many cubes 188 scored for sure in that situation assuming they made our radar via scouting notes (and more).
In regionals the rest simply are not rated as high to help us is all. I don’t see much difference.
Can you explain what you mean by this? I feel like sample size and data validation are both huge advantages of digital scouting. Digital scouting allows you to easily leverage larger samples sizes by calculating averages and is far easier to validate your data with those actually recorded by FMS/TBA.
Ok let me try to explain, if you track everything mistakes will definitely occur as you team scouts will be overwhelmed and input stuff that did or did not occur just to move on (no buy in). This calls into how valid is the data ?
Secondly or primarily most competitions I have been at have 9 to 12 matches per event, teams get better or worse during that event and it HIGHLY depends on their partners (and competitors). A good sample size is 100 a better one is 1000 and a great one is 10000 yet we have 12 and that sample is not consistent perhaps they were matched with the top team as a partner and that HIGHLY SKEWS their results…so we have HIGHLY SKEWED results due to partners and questionable accuracy of data input and I say no thanks. I prefer good old notes on what we are tracking as that is per robot and not skewed . If it helps my view on this is based on my degree in this . IMO You need a huge sample size and valid data entry to deduce anything in the same game/event.
As you can tell I value Qualitative scouting over Quantitative scouting and some teams appreciate that. That itself has got us picked as a result (teams see us scout and talk to us in the stands), when not a captain ourselves. They want our list.
Lets back up here.
Nobody said you should try to track everything. I dont expect the scouts to, and if they feel a datapoint is impeding them from getting good data overall im going to do whatever I can to fix it.
If your scouters are overwhelmed, something is wrong with how you’re scouting. You should take a look at your process for collecting data to see whats wrong.
Sure, the data is going to be rough without a large sample size, ill agree there. Saying its useless though? Not so much.
This is where I like the combo of both Quantitative and Qualitative data. We had a spot for comments post-match last year for the scouts to write things down (was a start played that was different than normal, did they break and do defense instead, etc) helps show the full picture farther than what the data alone can.
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.