Scouting Robot Cycle Speed

Through scouting district events and district champs, I realized that there were problems with conventional scouting methods. It seems that quantitative measurements of power cube amount on each plate seemed lacking, with great drive teams changing their strategy based on how the match was turning out. On top of that, it was hard to properly figure out speed in auto, with certain teams setting themselves above the rest with lightning fast autonomous modes.

Did anyone think of a way to simply retrieve information on the speed of cube cycle times, or any other way to keep numeric averages consistent with robot performance, removing outside factors? I’ve thought about trying to take cycle times for cube delivery outside of auto, and how fast they delivered each cube in auto, but this doesn’t seem to remove my concern of removing data bias, especially due to the difficulty recording time accurately while paper scouting over multiple scouts. I’m guessing there isn’t any easy solution, so I’m mostly just wondering on hearing the general consensus on issue.

My team thought about doing his for a bit. We have our stuff all done using custom VBA forms. We were going to add a hidden timer to count the time between the scouter’s taps on the tablet. We shot it down though because most of the time, the tap doesn’t line up with the actual scoring. Additionally, trying to quantify that is very hard. A great team might fill up the scale quickly, or have alliance partners that do, and all of a sudden, they take 5 extra seconds during their cycle so as to not knock off other cubes during placement.
The biggest issue is scouter error. What you might want to do is have a separate scouter or group take times, and rank the cycle time average on 1-5 after the match. This way you get a general idea of who plays better. If you could combine this with their total cubes scored per match, you’d get a pretty accurate idea of how quickly the team cycles.
In short, it’s not a bad idea to do, but trying to work it into a reliable scouting system is really hard. If you can find a way to solve most of the problems we ran into with that, you should do it!

We use shorthand and time on the actual game clock at time of score:

S= Scale
oW= Away switch
C= Climb (P=ParK)

So a single bot game match may look like:

SA13 S121 W113 S95 oW75 E52 E45 E36 E24 C

Three “E” is another C

We put slashes through failed attempts

Easy to deduce both cycles and priorities of scoring “tells the story”
We figure out max game as well using this method over several observations.

Good old pen/paper and highlighters. Report in excel sheets.

I highly recommend listening to what Oliver has to say.

It’s very relevant for this game.