Inside sports wagering: A Q&A with the trader setting your odds (Part Two)

Welcome to volume 4 of Winning Formula.

Today, we continue our Q&A interview with a sports trader who sets the odds for some of your favourite sports. You can check out part one of this interview here.

Winning Formula also hit a nice milestone since the last edition with 200 subscribers. Thank you all for joining! Please feel free to share this newsletter with a friend or colleague.

Table of Contents

1. The Rise of Player-Focused Bets

2. Changing betting trends

3. How player odds are calculated

4. Common biases among punters

5. The rise of marketing and promotional bets

6. How betting companies balance their books

The Rise of Player-Focused Bets

Q: Tell me about the rise of player markets.

A: The trend in sports betting has evolved significantly toward player-related markets. 

People are starting to care less about who wins; they want to bet on whether Jeremy Cameron kicks three goals or gets more than 20 disposals.

When they’re betting, they’re not betting on their team to win or lose.

Q: Is this related to the rise of multi-bets?

A: I think it's partly that, but as a bettor myself, I find it's more fun to cheer for something specific. The excitement you get when Cameron takes a mark and is walking back to take his kick is much more engaging than needing Geelong to win when they're already up by 40 or down by 20, and the game is effectively over.

The more specific we get with our betting markets, the more fun customers have. 

Players are more relatable - you can follow them on Instagram, and they have personalities. People are getting more attached to individual players. They buy jerseys with a player's name on the back and want to bet on that player.

There will be people who bet on Cameron to kick three goals in all 24 rounds because they're loyal fans. They think he's the best and can kick three goals every game. They want to cheer for him specifically.

Q: The social media connection is interesting. What about the impact of shows like Netflix's sports documentaries?

A: Absolutely.

When you watch a documentary series, you develop connections with certain players. If you like a few players from the show, you might be more inclined to bet on them to finish in the top ten in their sport. The players you've never heard of or who aren't featured, you have no attachment to.

Q: Do you think there are opportunities for betting companies to get involved in creating media content featuring players to increase interest?

A: Potentially, but there are a lot of regulatory concerns.

In basketball and tennis especially, there have been investigations into players' behavior when someone was betting on their behalf or influencing their performance.

If a tennis match between a top-ranked player and a much lower-ranked opponent, and we see big money coming through from unexpected sources, we would flag that for investigation.

Q: Where does that information go?

A: Most companies have fraud teams that handle those concerns. They would investigate who the customer is, whether they have connections to the athlete or team, whether they have access to inside information, and whether they could potentially influence a player's behavior or a team's performance. 

Q: How significant has the shift toward player-focused betting been?

A: It's been a rapid evolution. About 50% of bets now contain a player-related element, whereas ten years ago, it might have been just 10%. 

One reason is that bookmakers didn't offer many player markets back then - it was just who would win and whether the total score would be over/under a certain number or whether a team would win by a specific margin.

Now, bookmakers offer many more player-related betting opportunities, and customers have responded enthusiastically.

Q: How do you determine the pricing for player-specific markets?

A: We're always looking for data. For player markets, we first look at form. Depending on the sport, that could be their last game, their last three games, or their last ten appearances.

Then, we analyse the matchup. Is it a positive matchup for the player? Is it neutral? Is it a bad matchup? 

For example, in AFL, if a forward is up against a strong defender or a weak team's defence known for conceding plenty of points, it significantly affects their goal-scoring potential.

We also examine past performance against that specific team and investigate other relevant factors. If someone has gone from averaging 10 points in basketball to 20 in the last ten games, we need to understand why. 

Maybe LeBron's been injured, so someone else has to step up. It doesn't mean that the player is suddenly a superstar. it just means they're filling a temporary void.

Q: Are there any advanced statistics you use for certain sports?

A: Absolutely, especially in basketball. Beyond the standard points, rebounds, assists, and three-pointers, we look at metrics like potential assists. When a player passes to a teammate who shoots but misses, it's still recorded as a potential assist.

This gives us a more accurate picture of performance. For example, one player could have ten assists because every pass resulted in a made shot.

In contrast, another player might have just one assist despite creating 20 shooting opportunities that teammates missed. The surface stats would suggest the first player performed better, but the advanced data tells us the second player created more opportunities.

Changing betting trends

Q: What betting trends have you noticed during your three years in the industry?

A: Same-game multis have become extremely popular.

People prefer betting small amounts to win large payouts rather than more significant amounts to win less. It's human nature. We'd all rather throw the “Hail Mary” pass than make multiple short passes.

As a fan, betting on five things rather than one means you're more engaged throughout the game. You have more to cheer for and can stay invested longer. 

Social media plays a role, too. People like showing off when they hit big bets. Either the company posts it, or customers share it themselves.

Q: How do the odds work for same-game multis?

A: We use what we call "correlation”. Let’s look at an example that includes Geelong playing Melbourne.

If I bet on Jeremy Cameron to kick five goals and his team, Geelong, to win, there's a positive correlation. In this example, if Cameron performs well, it increases Geelong’s chances of winning, so the combined odds will be shorter than simply multiplying the individual odds.

Conversely, if I bet on Christian Petracca, who plays for the other team, having a great game with 40 disposals but Geelong winning, that's a negative correlation. If Petracca (a Melbourne player) dominates possession, Geelong is less likely to win.

Some combinations are independent, like players from opposing teams kicking two goals. In those cases, the odds are closer to simply multiplying the individual odds.

How player odds are calculated

Q: How do you calculate these correlations?

A: We use simulations. We simulate a game, say, 5,000 times using a computer. For example, if we simulate Geelong versus Melbourne, we might find that Geelong wins 50% of the time, making their odds $2.00.

Then we ask: in those simulated games where Geelong won, how many times did Jeremy Cameron kick five goals?

Maybe it's 10% of the time. For any combination a customer wants to bet on, we look at how many times that specific scenario occurred in our simulations, which determines the price.

Q: That sounds sophisticated. How exactly do these simulations work?

A: It's actually fairly simple mathematics.

For example, with AFL, you'd map out their 24-game schedule and assign probability values to each game. 70% chance to win this one, 30% chance to win that one, and so on.

Using a formula in Excel that generates random numbers, if the random number is greater than the probability threshold you've set, the team wins; if it's under, they lose.

We run this 5,000 times a second and let the computer decide the outcomes.

The simulations become complex because we're simulating hundreds of variables simultaneously. In one simulation, Geelong might win by 72 points, with Cameron kicking three goals and another player getting 23 kicks. In the next simulation, Essendon might lose by 5 points, Cameron has a shocker, and someone else might get 40 kicks.

We take averages across all these simulations.

Q: Do these simulations account for all the individual player statistics?

A: Yes, we have rows for each team player with their expected statistical outputs based on historical data.

We might say a player averages 18 disposals per game, but there's variance. The simulation accounts for this distribution. 

If someone typically gets 18 kicks, some simulations might show 5 disposals, others might show 50, with most falling somewhere in between following a normal distribution.

We stick to realistic parameters. No one's probably ever had 60 disposals in an AFL game so that the computer won't generate unrealistic outliers like that.

Q: How do you manage risk when unexpected events happen during a game, like a star player getting injured?

A: We don't have inside information. We find out about injuries and lineup changes simultaneously as the public, but we're trained to react quickly when news breaks.

We have several tools at our disposal. We can suspend markets temporarily, taking the game off the board while we recalculate the odds. 

If something requires a significant adjustment, we'll pause betting rather than risk exposure.

For smaller adjustments, we might let it ride because our built-in margins provide some protection.

Some professional bettors specifically target these moments when new information creates temporary market inefficiencies. That's just part of the game. We can be in a vulnerable position when information changes suddenly.

Common biases among punters

Q: Do you see common biases among bettors?

A: Star player bias is huge.

The general public tends to be overly optimistic about star players. If a star player is playing against their former team, everyone expects them to have a "revenge game" and score a few goals. 

We saw a recent example with Luka Dončić, who was traded from the Dallas Mavericks to the Los Angeles Lakers.

If you surveyed 100 punting Australians about how many points Luka will score in his revenge game against his old team, they'd probably all overestimate. No one's going to predict just 10 points. They'll say 30, 40, or 50 points.

We already set higher thresholds for star players because they're better than average, but people still overestimate beyond that.

Recency bias is another big one. People think, "It happened last game, so it'll happen again." 

Or they'll fixate on matchup history: "Three years ago, Cameraon kicked 10 goals against the Melbourne, and he's playing them again today."

People will focus only on the data that supports their preconceptions and ignore everything else.

Surprises within sports wagering

Q: What has surprised you most since being in the role?

A: What's fascinating about my job is how probability plays out daily. Bets with 1000-to-1 odds win every day somewhere. 

Events with 2-to-1 odds also lose every day. That's just how math works.

Within a single day, there are so many possible outcomes and combinations. 

Today, 1,000 customers collectively won $50,000 betting on some unlikely multi. But at the same time, 50,000 customers collectively lost $10,000 betting on a team to win when they were up by 15 points with five minutes to go, thinking it was a sure thing. 

Q: What's the biggest win you've seen from a customer?

A: For a casual punter, like someone betting $20, I've seen them win a million dollars. 

You see these kinds of stories about once a month at the office or on social media. Some punter is hitting an incredible multi with astronomical odds.

High rollers and how do you place a million dollar bet

Q: Talk to me about high rollers. how do sports betting companies handle punters who want to better large

A: In terms of big individual bets, some companies have customers who'll bet a million dollars on a single outcome. But you can't just place a bet that large through an app as it would exceed the platform limits. 

For bets that size, you'd need to speak with a manager who would verify your identity and ensure you understand what you're doing.

The traders would also need to confirm that the price is correct. If someone's trying to bet a million dollars on what might be an error in the odds, that's different from betting a million on a price we're confident in. 

Some books are risk-averse and might refuse certain large bets; others are more willing to take them on.

Q: What type of person typically places these million-dollar bets?

A: Just people who have money. That's all that matters from our perspective.

It's interesting to see how much different people must bet to get the same psychological satisfaction. Some people bet a dollar and cheer just as enthusiastically as someone betting a million dollars. It's all relative to the individual.

The rise of marketing and promotional bets

Q: As a trader, how much influence do marketing promotions have on your work?

A: Very little. Traders are trained to focus solely on the price based on the data. We set the odds; marketing handles promotions separately.

If the Olympics were happening and the marketing team wanted to promote an Australian athlete's chances of winning gold, they might ask, "What's the price? Can we bump it up?" We'd say, "No, our price is $2. That's what the math says."

If you see a betting advertisement mentioning specific odds, those are still the trader's odds. Marketing can't arbitrarily change them. 

What marketing can do is offer separate promotions like "money back if your horse comes second" or boost odds as a special promotion, but that's handled by a separate generosity team with their own budget.

How betting companies balance their books

Q: Do major betting companies need to balance their books across outcomes?

A: Most big companies don't balance books the way people think. 

Major bookmakers can ride the wave with significant liquidity, knowing they'll win in the long run, thanks to their margins. They'll have good days and bad days, but it evens out across all the events they offer.

Smaller books don't have that luxury because they lack the financial security. For them, the weight of money moves prices more significantly. 

If 1,000 people bet on Geelong to win, a major book wouldn't care, but a smaller book might quickly adjust its odds and try to attract money on the other side.

In the industry, we have terms like "red book" (meaning you'll lose no matter what outcome occurs), "black book" (you'll win on one outcome and lose on another), and "green book" (you'll win regardless of the outcome due to the margin built into the odds).

Take a coin toss, for example. We might offer $1.91 on both heads and tails, creating a green book with a 5% margin. Over time, that margin ensures profitability.

Predictions on the future of sports betting

Q: How do you see the future of sports betting evolving?

A: Bookmakers will increasingly try to flip the current dynamic where we hold all the keys. We want to create the perception that the customer has more power and choice.

If you wake up thinking, "Jeremy Cameron's going to score exactly three goals and four behinds this weekend," you might not be able to bet on that because the market doesn't exist.

We want to move toward a model where customers can propose their markets and odds, similar to what prediction markets offer. This empowers the customer and creates a more engaging experience that could be the future of sports betting.

Well, that’s part two of my interview with a sports trader. Let me know what you think.

These newsletters grow with personal referrals, so please forward them to someone who loves sports, business and in this case, punting on sport!

If you have any feedback or ideas for a newsletter, don't hesitate to contact me on LinkedIn, reply to this email or email me here or by replying to this email.

‘Til next time,

👋 Will