i write

random thoughts and personal experiments

This is Part 4 of my AI PGA Tour golf betting series. In this experiment I have started with a $100 bankroll and I plan to bet $5 each week using the Kelly Criteria to distribute the $5 most effectively across my picks. I have a prompt built to have AI do research on the tournament and pick 3 players to win. If my bankroll grows I will add $2.50 to my weekly bets for every $50 additional in my bankroll. However, if I run out of funds I will end this experiment so at worst I will have this for 20 events.

Results So Far:

  • Starting Bankroll: $100
  • Current Bankroll: $90.23 (not including the bets for this week)
  • Outright Winners Picked: 1 (1/8 if you count all my picks)
  • Expert Winners Picked: 1 (1/27 if you count all unique expert picks)
  • Last Week’s Picks: Sony Open 2025

Last Week’s Results

Last week our AI picks chose Hideki Matsuyama, Keegan Bradley and Corey Conners. It’s tough to win back to back and Hideki just barely made the cut before finishing strong on Sunday though he was never in contention. Keegan Bradley started in contention on Sunday but after an early round 4 bogey he never was able to get the momentum going that he needed to close out strong. Conners…didn’t even make the cut.

Keegan and Conners were both picked once by the experts, but out of the picks they made 12 unique players picked by the experts, none of them won. It was ultimately Nick Taylor who prevailed as the winner in a playoff against.

A Glimpse into the 2025 American Express Tournament

It’s another notoriously easy course with the last 17 winners scoring -20 under par or better. Here the players cycle between 3 par 72 courses in La Quinta, CA before finish on PGA West La Quinta for the final round

The course details:

  • Course: Three Courses
  • Location: La Quinta, CA
  • Par 72
  • Grass Type: TifEagle Bermudagrass
  • Past Winners: Nick Dunlap – 29 (2024), John Rahm -27 (2023), Hudson Swafford -23 (2022), Si Woo Kim -23 (2021), Andrew Landry -26 (2020)

Who The Experts Picked To Win?

TBD when all the picks are published.

Who I Picked To Win?

Based on the analysis, the AI determined Xander Schauffele, Patrick Cantlay and Sungae Im were the most likely to win. This makes sense considering they are the top ¾ players with the best odds to win. However, after placing my original bets Schauffle withdrew and I had to rerun the analysis. Here are the implied percentages based on additional data AI used to determine its picks.

  1. Patrick Cantlay: 10%

  2. Justin Thomas: 8%

  3. Sungjae Im: 7%

Using the Kelly Criterion for the $5 weekly bet it broke down the bet sizes to:

  1. Patrick Cantlay: $2.33

  2. Justin Thomas: $1.47

  3. Sungjae Im: $1.20

The American Express 2025 Picks

Personally, not accounting for the AI analysis I would be looking at Si Woo Kim, Will Zalatoris and Adam Hadwin.

This is Part 3 of my AI PGA Tour golf betting series. In this experiment I have started with a $100 bankroll and I plan to bet $5 each week using the Kelly Criteria to distribute the $5 most effectively across my picks. I have a prompt built to have AI do research on the tournament and pick 3 players to win. If my bankroll grows I will add $2.50 to my weekly bets for every $50 additional in my bankroll. However, if I run out of funds I will end this experiment so at worst I will have this for 20 events.

Results So Far:

  • Starting Bankroll: $100
  • Current Bankroll: $95.23 (not including the bets for this week)
  • Outright Winners Picked: 1 (1/5 if you count all my picks)
  • Expert Winners Picked: 1 (1/26 if you count all expert picks)
  • Last Week’s Picks: The Sentry 2025
Read more...

In this series of indie hacking recaps I’ll give an overview of what I was able to contribute to my side hustles throughout the previous month and the impact that it has had on the websites. As of right now I have two projects live, a skin care aging tool and a token calculator for OpenAI API.

Read more...

This is Part 2 of my AI PGA Tour golf betting series. In this experiment I have started with a $100 bankroll and I plan to bet $5 each week using the Kelly Criteria to distribute the $5 most effectively across my picks. I have a prompt built to have AI do research on the tournament and pick 3 players to win. If my bankroll grows I will add $2.50 to my weekly bets for every $50 additional in my bankroll. However, if I run out of funds I will end this experiment so at worst I will have this for 20 events.

Results So Far:

  • Starting Bankroll: $100
  • Current Bankroll: $100.23 (not including the bets for this week)
  • Outright Winners Picked: 1 (1/3 if you count all my picks)
  • Expert Winners Picked: 1 (1/10 if you count all expert picks)
  • Last Week’s Picks: Hero World Challenge 2024
Read more...

This is Part 1 of my AI PGA Tour golf betting series. In this experiment I have started with a $100 bankroll and I plan to bet $5 each week using the Kelly Criteria to distribute the $5 most effectively across my picks. I have a prompt built to have AI do research on the tournament and pick 3 players to win. If my bankroll grows I will add $2.50 to my weekly bets for every $50 additional in my bankroll. However, if I run out of funds I will end this experiment so at worst I will have this for 20 events.

Results So Far:

  • Starting Bankroll: $100
  • Current Bankroll: $95 (includes the bets for this week)
  • Outright Winners Picked: N/A
  • Expert Winners Picked: N/A
Read more...

Enter your email to subscribe to updates.