haesy

sports bettor | rugby league quant | fulltime data scientist 🇦🇺
not enough people bet the draw in soccer

[haesy @tmhaes]

@Adambjorn @RoxyLasVegas What brought you to the nations capital? And was there a big racing scene there at the time ?

[haesy @tmhaes]

@1WeakGuttedDog yes getting these, super random sports - post 10pm hahah what is this logic

[ReadingThePlay @ReadingThePlay]

One job. 😳

[haesy @tmhaes]

@ReadingThePlay last week might've been a round for the favourites and the public..... 3/3 on spreads so far, let's keep rolling

[Underdog Chance @underdogchance]

I have never met a successful sports bettor who could not use spreadsheets.

Some of them eventually build advanced models.
Some create apps, databases, automations, or their own odds screens.

But almost all of them learned the foundation in a spreadsheet.

Not because spreadsheets are magical.

Because spreadsheets force you to stop speaking only in betting stories.

Before:

"They're playing well."

"The quarterback is motivated."

"This line looks wrong."

After:

"I estimate their probability at 58%."

"The sportsbook price implies 52%."

"The gap is large enough to consider betting."

That change is bigger than it looks.

You are no longer trying to predict one result. You are estimating how often something should happen and comparing your estimate with the price being offered.

Your opinion becomes visible.

You can calculate it, test it, track it, and later improve it.

You will also find out that many bets you strongly believe in have no measurable edge.

That lesson alone can save you from a lot of bad wagers.

Later, better technology can make the work faster. But if you do not understand the formulas underneath, the technology can simply help you make mistakes more efficiently.

A spreadsheet is often the first place a bettor learns to think in probabilities.

And once you start pricing bets instead of guessing winners, it becomes difficult to go back.

[Todd Fuhrman @ToddFuhrman]

Here are your key numbers when betting NFL Totals

[haesy @tmhaes]

@BennyScarf what if you hold 5% over them and they break the contract

[haesy @tmhaes]

this happens everywhere in corporate not just gambling companies. unfortunately from a business perspective, they don't factor in your degree of addiction and there is significant upside to keeping your biggest whales around. There isn't much you wouldn't offer to someone turning over 20 million a month on multis and favourites

[haesy @tmhaes]

black and white, ying and yang, we’ve seen two versions of the game we thought we knew this year

[The Rugby League Eye Test @LeagueEyeTest]

Almost like the competition has been played under two different rule sets...

[haesy @tmhaes]

can you get kicked out of @bet_bash if you prefer decimal odds? asking for a friend

[haesy @tmhaes]

@JHumphreys28 Cows +6.5 1.9 / 2.74 ML amazing bets on the cricket to be honest. Let’s try get a tone next time.

[Quant Chad @Autonomous_Chad]

SmartOdds sports syndicate confirmed to covertly operate on Polymarket & Kalshi

their Quant job offerings was updated to mention the need for deployment on "prediction markets".

For a reminders SmartOdds is the main rival to StarLizard, the #1 sports syndicate, estimated to make $600M/year.

SmartOdds earnings are not public, but to give you an idea :
- When Starlizard's founder Tony Bloom (on the left) bought Brighton FC, SmartOdds founder Matthew Benham (on the right) countered by buying Brentford FC
- Matthew Benham is actually the quant that built all of Starlizard's models, he was fired after a dispute and went on to build SmartOdds

they are a huge operation, in is the same league as Starlizard, maybe even bigger already.

So If you are market making Sports on @polymarket and @kalshi, expect the level of competition to rise.

Your are not betting against sports degens anymore, but also established syndicates with dedicated quant desks and hundreds of PHDs on their payroll.

alternatively you could just apply for a job there, they are looking for talents with Prediction Markets experience.

If you can't beat them, join them.

[BettingIsCool @BettingIsCool]

Can a model still find value against Pinnacle's tennis prices? A University of Bristol paper (Clegg & Cartlidge, Oct 2025) is an unusually honest data point.

Their graph neural network targets intransitive dominance patterns — A beats B, B beats C, C beats A — the structure that Elo-style ratings compress away.

As a forecaster it does NOT beat the market: 65.7% accuracy / 0.215 Brier vs 69.0% / 0.196 for Pinnacle's implied probabilities.

But restricted to high-intransitivity matchups, betting the model returned +3.26% ROI over 1,903 bets with Kelly staking (p = 0.005), and +1.14% over 2,063 bets flat (p = 0.022).

The takeaway is not "GNNs beat the market". It is that inefficiency survives in structured pockets that the aggregate line does not fully price.

Source: https://t.co/qWXwWZLUcC

[haesy @tmhaes]

@ReadingThePlay Roosters by 2
Dolphins by 11
Bulldogs by 9
Warriors by 4
Panthers by 5
Broncos by 4
Sharks by 16
Eels by 9

[haesy @tmhaes]

Not all closing lines are worth collecting. You want closers from the sharpest book for YOUR sport.

How do you know who's sharpest? Three tells:

they move first, everyone else follows
they take real limits
their closers historically split ~50/50 against the spread and total

Get that dataset. Even just closers + scores. It's the raw material for everything including reverse engineering the market, building targeted samples, actually modelling instead of guessing.

Most punters won't bother. That's the point.

[haesy @tmhaes]

@JHumphreys28 average odds / IP ?

[The Rugby League Eye Test @LeagueEyeTest]

A lot can change in 12 months

After Round 20, 2025 Cronulla were +21 in cumulative net penalties (awarded-conceded), 1st in the NRL

After Round 20, 2026 Cronulla are -34 for net penalties, last in the NRL

A swing of 55 more penalties conceded than awarded in a year...

[haesy @tmhaes]

@JHumphreys28 that's pretty expensive at scale, i was expecting 1c. If you take their juice into account and possible some high credit card fees, it can bite you in the long run - just know the poduct inside and out and what you are getting into

[haesy @tmhaes]

@JHumphreys28 how much juice do they put on top of the raw pinn prices?

[haesy @tmhaes]

@Migmigliore @JHumphreys28 Oh damn. Ok mate that’s passionate from you, there’s got to be some case of wrong doing, keen to hear if more people have experienced the same.