Narrowest What Real-Time Gambling Data Reveals About Sports: Introducing Gambletron 2000 by ATodd 50

I.e., the effect on profitability is even worse in the other sports

This video is processing – it'll appear automatically when it's done.

Lowers transaction costs What Real-Time Gambling Data Reveals About Sports: Introducing Gambletron 2000 by ATodd 50

This is very similar to the concept of “Slippage” in financial markets

This video is processing – it'll appear automatically when it's done.

It would not be surprising if this deviation occurred completely randomly What Real-Time Gambling Data Reveals About Sports: Introducing Gambletron 2000 by ATodd 50

Some quick math:

  • There are 221 observations where a team had between 10-20% pre-game win probability
  • Mean win probability for those teams was 15.5%
  • 40 teams actually won, or 18.1%
  • If each game were an independent coin flip with p = .155, we’d expect to see 40 or more wins 16.4% of the time using the binomial cdf
  • This isn’t exact because each individual game isn’t p = .155, but it’s a close enough approximation to show that this could have happened completely by chance

This video is processing – it'll appear automatically when it's done.

Detroit/Dallas comeback game What Real-Time Gambling Data Reveals About Sports: Introducing Gambletron 2000 by ATodd 50

This video is processing – it'll appear automatically when it's done.

Wire-to-wire domination What Real-Time Gambling Data Reveals About Sports: Introducing Gambletron 2000 by ATodd 50

Another question we can answer, “if the favored team begins the game with an X% win probability, what are the chances that team will be probabilistically winning for the entire game?”

For example, the data shows that a team that starts the game with a 75% win probability has about a 50% chance to lead for the entire game

This video is processing – it'll appear automatically when it's done.

The NFL and NBA have similar rates What Real-Time Gambling Data Reveals About Sports: Introducing Gambletron 2000 by ATodd 50

Here’s a graph showing that the NBA generally has the highest % of games that meet a high hotness threshold, though the NFL is not far behind:

NBA gets a bit of a short shrift here because the NFL data includes playoff games, while NBA data does not, and playoff games are probably hotter on average because the teams are selected to be better and more evenly matched.

This video is processing – it'll appear automatically when it's done.

Florida State’s comeback victory over Auburn What Real-Time Gambling Data Reveals About Sports: Introducing Gambletron 2000 by ATodd 50

http://www.gambletron2000.com/events/871/florida-state-v-auburn

Auburn was involved in 3 of the top college football games of the year, although they lost to Florida State in the championship, they knocked off both Alabama and Georgia in two very memorable games:

http://www.gambletron2000.com/events/433/alabama-at-auburn

http://www.gambletron2000.com/events/248/georgia-at-auburn

This video is processing – it'll appear automatically when it's done.

Syracuse’s buzzer-beating win over Pittsburgh What Real-Time Gambling Data Reveals About Sports: Introducing Gambletron 2000 by ATodd 50

This video is processing – it'll appear automatically when it's done.

Both of these factors lead to a large number of lopsided college games with little excitement. What Real-Time Gambling Data Reveals About Sports: Introducing Gambletron 2000 by ATodd 50

This video is processing – it'll appear automatically when it's done.

The different structure of hockey betting What Real-Time Gambling Data Reveals About Sports: Introducing Gambletron 2000 by ATodd 50

A note on hockey and soccer

Hockey and soccer are somewhat more complicated than the other sports, because they allow ties, and their betting markets are structured so that you can bet on either team winning in regulation, or a regulation draw. If you bet on the draw, you win your bet if the teams are tied at the end of regulation, even if one of them wins in overtime or shootout. Consequently, if you bet on a team to win, you lose your bet if that team only manages to win in overtime.

This gives us three probabilities to track, which results in a more complicated graph. Here’s what the USA vs. Russia Olympic hockey game looked like:

http://www.gambletron2000.com/events/1741/usa-v-russia

Although the USA eventually won the game in a shootout, for gambling purposes, both the USA and Russia lost. If we add in the probability of a regulation draw, we see the winning outcome:

Because hockey and soccer betting markets work differently from the other sports, it can be difficult to perform apples-to-apples comparisons. For this reason, some of the following analyses omit hockey and soccer, but Gambletron 2000 continues to track them alongside the other sports.

This video is processing – it'll appear automatically when it's done.