Melbourne Cup tips from data analysts

By Chris Pash | 5 November 2019

AdNews went to the serious data analysts to try to work out the best bets for the Melbourne Cup.

These are the people who know how to read deep data sets.

The horses getting repeated mentions: Surprise Baby, Vow and Declare, Finche, Constantinople.

Jake Lyme, data scientist at customer experience agency LIDA, says he is usually reluctant to help when people in the office approach him for assistance predict ng the winner of the Melbourne Cup.

His response: “If I could predict sporting outcomes, I probably wouldn’t be here.”

However, this year he's decided to apply his data analytical skills to form on the track.

He looked at key metrics: turf (soft to hard), distance (short to long) and jockey performance.

"By analysing 1000+ races from the last year, I’ve found that these metrics together give the greatest indication of performance," he says.

"Then for a little added spice, I included a few more metrics: horse weight, field size and barrier number.

"But I wanted more. I wanted the in-the-know data, like how the jockey is feeling? Did the horse have a rough plane ride? Do European horses prefer longer distance races?

"I wasn’t sure how to collect this kind of info because it isn’t traditionally quantified, and there isn’t a consistent source to collect and analyse these criteria.

"Then it occured to me that a bunch of people had already done this work – the experts, the form columnists, ex-jockeys, punting groups, and all the fans. So, I collected tips from 50 experts ranking their 1st/2nd/3rd picks, and analysed the trends."

Comparing the two data sets, he came up with these three horses: Vow and Declare, Constantinople and Finche.

However, he ran these past the resident punter in the office, who said; “These are solid picks, but you’re missing the magic of the sport. It’s the 100-1 win of Prince of Penzance, the hope that the 10’er turns into a night out.”

So Lyme went back over his data set, not only looking at the forecasted win rate, but also the expected payout, finding which would be the best bet for risk and reward.

Based on that, the two longshots are: Surprise Baby and Twilight Payment.

"Now, please note that taking this as gospel is in itself a gamble," Lyme says.

"The outcome of sport is determined by so much more than what I’ve analysed. And sometimes you’ve just got to follow your intuition and bet on Surprise Baby because you like the name – that’s my plan." 

At Macquarie Bank, the Quantitative Strategy team, more used to analysing companies to determine whether they are a good investment risk, transformed racing stats into traditional quant signals to build a model that ranks the runners in the 2019 Melbourne Cup.

Horses at the top of the model: Surprise Baby, Vow and Declare, Finche, Cross Counter, Mer De Glace and Constantinople.

"Each year we look to the market to guide us, making slight tweaks to our model to boost our chance of success," the team says in a note to clients. 

"For investors to stay at the top of the game it is critical to incorporate continuous innovation. We take a similar philosophy with our Melbourne Cup model and this year the folks at Ladbrokes have provided access to their big bet data." 

The team has been running the model since 2007. The early days saw some healthy profits however more recently it’s been a lean run.

This is a reminder, the analysts say, that the model is not intended to be taken seriously. The Quant team has limited knowledge of horse racing and the purpose of the analysis is fun.

For the Melbourne Cup, the Macquarie analysts looked at:

  1. Value – cheapness, what you get for the price. Measured by average prize money per race.
  2. Momentum – what is working, what is getting better? Movement in the odds from opening prices until the latest available odds.The horse’s form in its last five starts is also a good indicator. 
  3. Sentiment – what everyone is looking for right now, what people like. The last available odds before publication as this is representative of the mood of the market.
  4. Quality – what is the best, the most reliable? Odds dispersion amongst bookmakers to indicate the variance of the perceived value of the horses. The analysts also look measure the horses winning percentage to get a hit rate of the horse.
  5. Innovative Data – Other data sources to give an information edge, including big bets placed by punters, the barrier draw.

macbank

 

Have something to say on this? Share your views in the comments section below. Or if you have a news story or tip-off, drop us a line at adnews@yaffa.com.au

Sign up to the AdNews newsletter, like us on Facebook or follow us on Twitter for breaking stories and campaigns throughout the day.

comments powered by Disqus