Each year on a Saturday in early summer I dust off my seersucker suit and attend the Belmont Stakes, the final leg of horse racing’s Triple Crown. This year, while having my Sunday morning Bulletproof Coffee and exporting data off my Basis B1, I realized I had unintentionally captured a data points that could quantify my day of gambling at the track! I have been meaning to conduct some experiments around quantifying “thrill”, so I threw the following chart together:

Quantifying the thrill of gambling

Pretty cool!

I didn’t set out to structure this as an experiment, so unfortunately I had limited data to work with and this is all more observational and to be taken with a grain of (Himalayan pink) salt (but still fun!). I threw out most of the data points collected by my Basis B1 (steps, calories, skin temp, air temp) as they didn’t appear to correlate with anything or change significantly. Most research shows that galvanic skin response (GSR), heart rate, and heart rate variability (HRV) give the best indication of excitement/thrill/arousal, so I plotted heart rate and GSR.

Gambling and Biometric Observations

Gambling and biometric observations

(blue = heart rate, red = galvanic skin response)

  • Heart rate elevates to maximum levels in 3, sometimes 4 instances – at time of placing bets, right at the start of the race, and after the race is over and I won money (interestingly, hear rate tends to decrease while the race is happening). There is also a spike when I go to cash out my winnings
  • Losing does not result in any post-race heart rate spike
  • Galvanic skin response increases as the race is happening(excitement and arousal?), usually starting from a baseline and peaking at the finish (especially if I stood to win big!)

Issues, and what I would have done differently

  • I am only able to obtain 1-minute average readings off of my Basis B1. Ideally I would like to have 1 reading per second, especially during races, which typically last no more than 2 minutes start to finish
  • I would have properly synchronized my time readings with the “official” clock at Belmont (my timestamps appear to be about 2 minutes off the “official” race start times, so I shifted my data accordingly)
  • I would have worn a chest strap to better monitor heart rate (the optical bloodflow sensor on my Basis was really flaky, and had significant dropouts)
  • This would have also allowed me to capture HRV, which would have been ideal
  • I didn’t track the times I was researching/placing my bets, but looking at the charts it seems like it resulted in elevated heart rate
  • I had a few drinks throughout the day (3 to be exact, spread over the day) so not sure if/how that had any impact

This has piqued my interest in thinking of ways to better construct experiments to quantify thrill/arousal, especially as it relates to gambling. To be continued!