CTO l Building Responsible AI solutions | Speaker I ML research I Networking technologies I Digital solutions | Sustainability | Board Member | Advisor | Mental Health Advocate
#AI #predictions vs. #Betting odds At the beginning of yesterday’s The Championships, Wimbledon Men’s final, the #AI based predictions were spot on predicting that #Alcaraz had almost a 50% better chance of winning than #Djokovic Compare this with the bookies betting odds (with millions at stake), that only marginally favored Alcaraz (by 15%). What could explain the big difference between the two? My guess is that the AI model is likely placing more emphasis on recent performance data and statistical analysis, while the betting system must be balancing this with historical performance and other intangible and human sentiment factors. Let’s dig a little deeper into the potential reasons why AI might have predicted a higher chance of winning for Alcaraz: 1. Recent performance and form – Alcaraz’s superior record in 2024 (33-6) vs. Djokovic (22-7). 2. 2024 titles – Alcaraz 3 vs Djokovic 0. 3. 2024 Grass court performance – Alcaraz has a slightly better record on grass (8-1 with one title) vs. Djokovic (6-1 with no titles on grass). 4. Return game – Alcaraz has a significantly better return game on grass (35.7% overall, 32.5% on grass); Djokovic (30.8% overall, 23.7% on grass). 5. Age factor – 16 yr gap – The AI may be favoring the younger player in a potentially long and grueling match (Their past few head to head matches have been long and hard fought). 6. Injury and recent surgery - Novak had a knee surgery just 4 weeks ago, which could have been a factor in the AI algo. On the other hand, the bookies’ betting odds only marginally favored Alcaraz. Let’s examine why: 1. Djokovic's historical performance – As an all time great and multiple-time Wimbledon champion, bettors were likely giving more weight to his experience and past success. 2. Service game strength – Djokovic's service game winning percentage on grass (89.0%) is higher than Alcaraz's (84.0%), which could have been be a significant factor not fully accounted for by the AI. 3. Human factors – Betting odds typically incorporate factors like public perception and historical betting patterns, which the AI system likely did not consider. This could explain the discrepancy between the AI prediction and the betting odds. #wimbledon #sport #analytics #datascience #tennis Cache AI
Tricia Friedman
3mo
You have me curious to see if certain sports are easier for AI prediction, I’d guess tennis has less variable that team sports but that could be very wrong
I help reveal the treasure hiding in your data to boost your growth | Data Insights Expert | AI-Startup & Business Advisor
3moI think you’re spot on with your analysis of why AI did a better job of calculating the odds, Malur Narayan My money would’ve been on Alcaraz too (if I was a betting person) based on the historical data of male tennis players in Wimbledon finals for over 20 years.