Welcome back to my series on professional sports bettors.
If you are landing here for the first time, I recommend reading through Episode 1 of the series.
At this point, you understand the concept behind what a professional sports bettor does to capitalize on a perceived edge in the market. How they do it is a whole nother story involving the wearing of multiple metaphoric hats.
As they say, it’s best to start at the beginning.
Hat #1
Data Engineer, collecting and cleaning data.
Data is everywhere. Literally everywhere. Data is messy. Very messy. Humans are lazy. Very… Ok, you get it. Most everybody will attempt to follow the path of least resistance to arrive at their destination: Psychology 101. The same can be said for the collection and dissemination of data into a sports market. Low hanging data will be picked up first, chewed, and consumed by the market quickly. This data has no barrier to entry. Open up your favorite sports reporting website. Copy & paste the box score into your spreadsheet. You have data.
Unfortunately for you Mx. Copy/Paste data person, the efficient market theory exists. Since the data you gathered has no barrier to entry, copy/pasting data from ordinary box scores into a simple linear regression model is really easy. Historically, it has been done many times before. Therefore, that information is included in the creation of the sports book’s odds. Inherently, this data has no “added value.”
How does one overcome this? They need to gather an exceptionally large amount of data as quickly and painlessly as possible from multiple sources. Web scrapers and tapping into APIs are typically the fastest way to gather large amounts of data.
A system needs to be built well in advance of tonight’s prime time match up to gather this information in time to be evaluated. A data pipeline must be created, deployed, and maintained. Sports websites love to make large sweeping changes to their website in the middle of a season. Maintaining a data pipeline with little to no control over the source data can be a highly frustrating experience.
Not only is it important to get accurate data, but it is also important to get data quickly. Assuming your down-stream analysis is sound and your predictions are in-fact, more accurate than the market, speed is the second most important factor.
The efficient market theory suggests as more money gets funneled into a specific market, the more efficient the price/odds become.
Well, this would imply the sooner you can make your prediction, the less efficient the betting line will be. This concept only works at a macro level, of course. Any individual event can capture negative variance, but when you average all the positions together over time a savvy bettor will, on average, “beat the closing number.”
What does, “beat the closing number” actually mean? It means the average price you enter the market is lower than the average price available when the sporting event began.
For example:
Let’s say it is Monday night, shortly after the final NFL football game has concluded. The next game is not set to be played until Thursday. Roughly 70 hours after the final whistle on Monday night.
At this point, sports books have begun releasing the odds/prices for all the games for the next week on the schedule. The opening line is influenced by a team’s previous performances and the market’s perception of those teams until this point in the season.
However, Team A has likely not faced Team B. So the direct comparison between these two teams has not been ironed out. The “opening number” set by the sports book has not been hardened by the betting public.
Specifically.. The line is set at San Francisco 49ers -3 over the Las Vegas Raiders. Meaning, San Francisco has to win by more than 3 points for a wager placed on SF to be graded as a winner. Think of it as subtracting 3 points from SF’s final score or adding 3 points to Las Vegas, should the bettor choose to back the underdog.
Four possible outcomes can occur for a bettor backing San Francisco 49ers -3.
1) SF Wins by less than 3
Actual Score: SF 21 — LV 20
Adjusted Score: SF 21 -3 = 18 — LV 20
Wager Grade: Loser
2) SF Wins by exactly 3
Actual Score: SF 23 — LV 20
Adjusted Score: SF 23 -3 = 20 — LV 20
Wager Grade: Push (tie)
3) SF Wins by more than 3
Actual Score: SF 24 — LV 20
Adjusted Score: SF 24 -3 = 21 — LV 20
Wager Grade: Winner
4) LV Wins by any amount
Actual Score: SF 20 — LV 21
Adjusted Score: SF 20 -3 = 17 — LV 20
Wager Grade: Loser
The inverse is true for a bettor backing Las Vegas Raiders +3.
Sports book economics
If Bettor 1 places a $110 wager to win $100 on SF -3 (a typical price), the sports book has a liability of $100. Bettor 2 also likes SF -3, but they like to bet larger amounts. Bettor 2 drops $1,100 to win $1,000 on SF -3 to win. Liability for the Sports book is now $1,100 should San Francisco win.
Ideally, the sports book will have $1,100 in liability on SF and $1,100 on LV and have taken in $1,210 on SF and $1,210 on LV. In this balanced scenario, if SF wins by 4, the $1,100 liability on SF will get covered by the $1,210 placed on LV. SF bettors get paid $1,100 the sports book is left with a profit of $110.
What happens when the money isn’t so evenly balanced? What if the 49ers continue to get backed by the market? More and more money is pouring in on SF -3. The casino needs to entice the betting public to make more wagers on Las Vegas to balance their books. Enter: price discovery. The sports book can move to line (the price) to San Francisco -3.5.
The scenarios listed above get adjusted slightly. SF still needs to win by 4, but now if they win by 3, what previously would have been a tie (a refund) is now a loss. It also means, for Las Vegas bettors, should the Raiders lose by exactly 3, a bettor holding a LV +3.5 has a winner instead of a tie.
Regardless of where the number settles, be it -3, -3.5, or all the down to -7, on average the final number will be more accurate than the opening number. Advantage bettors will influence the market, via their wagers.
We’ve come a long way from talking about data pipelines, but it is time to bring this conversation full circle. Data pipelines allow a bettor to gather and clean their data quickly. Speed means getting to market sooner. Getting to market sooner means getting a better price. Getting a better price means pocketing a larger profit at the end of the season.
Ultimately, getting to market sooner equates to a larger hypothetical edge.
Check back soon for Episode 3: Data Analyst
An important note not relevant for the vast majority of those looking to place a bet. The maximum wager accepted increases as the time before the event decreases. Meaning many serious betting groups hold back on placing their wagers until limits are north of $10,000. This provides a window for many casual bettors with a small edge to place their wagers before the large sharks have their feast.
I’ll write more on “closing line value” at a later date. It is a fascinating nuance to the sports betting landscape.