Sports Betting in America Part III
Betting Season Win Totals
Now that I've covered the basics of how to win at sports betting, it's time to get into some specifics of identifying good bets. Right now, many sportsbooks have put up lines for odds to win the NBA championship. I haven't really thought much about it, so I'm not going to talk about it. However, you can check out Pinnacle's lines if you're interested. Right now, the Wolves are listed at about 32:1. For a while, they were listed at around 20:1, which was just ridiculously low. The new line of 32:1 seems a lot more reasonable.
The other topic which I'll cover in some depth is season win totals. Of course, there will be a plethora of bets similar to these. You'll see bets on the rookie of the year, MVP, and possibly other prop bets such as who will lead the league in scoring. The thing that all of these bets have in common is that they're long term bets.
If you wish to bet on any of these, you'll have to give your money to the sportsbook now and you won't be able to use it again for the rest of the season. Say you have given yourself a bankroll of $1,000 for the NBA season. Suppose you find 5 bets that look appealing, and decide to bet $20 on each of them. You've now bet $100 on bets that won't be decided until the end of the season. In essence, you've shrunk your bankroll by 10%.
The problem with this is obvious. If your bankroll is non-replenishable, your risk of ruin has increased significantly. Even if you end up winning all 5 of your bets, you won't get this money until the end of the season. Even if you pick a team with a really low win total and bet the over and they start off extremely hot and pass the over in December, you still won't see this money until the season is officially over. In case something happens and the season is cancelled, your bet is cancelled and you don't win anything.
If you have a replenishable bankroll, you can be more liberal with these bets. Of course, I still would like to have a larger than normal edge on these bets in order to make them. This should of course be a consideration when making these bets.
Ok, now onto the meat.
I've been reading Basketball On Paper by Dean Oliver. For anyone who enjoys basketball and statistics, this makes for quite an interesting read. I especially recommend it to any of you baseball fans who like sabermetrics and the work of Bill James.
There is a chapter of the book about parity in the league, and the general trend of a team to approach the .500 mark. This means that a team with a winning record will generally experience a drop off in production the following season. Likewise, teams that are under .500 will generally increase their winning percentage the following season. Furthermore, the closer a team is to the extremes, the greater the expected change should be.
Dean doesn't speculate too much about the reasons for this, but my intuition gives me a few ideas as to why this might be true. Teams with higher winning percentages are often lead by a superstar who might be having a career year. They'll also tend to be more veteran than other teams. The following season, we might not expect the superstar to once again perform as well the next season. The veteran players might have diminishing skills (see 2004-2005 Timberwolves) and their age might lead to them being more injury prone.
On the flip side of the coin, a team who doesn't perform during one season might have been hampered by the injury of a superstar the previous year. In addition, a team who isn't performing will often infuse the team with young talent. A year of experience for young players can equate to a large increase in the number of wins for that team.
In a nutshell, the reasons can be expressed as this. For a team to have a good season, a lot of things have to go right. For a team to have a bad season, a lot of things have to go wrong. In general, all these things won't happen two years in a row (though of course, the can).
Let's take a look at the numbers from the last two years:
Last season was kind of weird because of the HUGE decreases by 3 of the top 4 teams in the league (also, the addition of Charlotte into the league may have inflated other teams win totals somewhat). Indiana had half of it's teams suspended for half the year. The Lakers lost Shaq. The Wolves veterans got old and crappy in a hurry. Still, the general trends held. Of the 13 teams that had records over .500 in 2004, 7 teams decreased. Of the 13 teams with records lower than .500 in 2004, 9 of them increased their winning percentages.
One season of course doesn't prove anything, but if you go back, you'll see that the general trend holds. Another thing to note is that teams near the very top of the league tend to resist this trend more than others. This is generally because the teams will have good talent or a good system, and they will do anything possible to keep these things in place for as long as possible.
Although season totals haven't been listed yet, I imagine they will be amongst the first NBA bets I make this season. The Wolves have gone through a complete overhaul and my guess is that there season win total will be listed at something between 39-41. This number may get inflated due to KGs presence as well as the public's perception of the Timberwolves.
There are several categories to consider. I'll go through each of them using the Suns as an example.
Last Year's Record
The Suns won 75% of their games last year(62-20). In general, a team that wins that many games will experience about a 5% dropoff, which is about 4-5 wins. All other things being equal, we'd expect them to win around 57-58 wins this season.
How Lucky Were They Last Year?
It's nice to look at points for and points against rather than just W-L records. A team might win a higher percentage of their games than you'd expect just by having their points scored distributed in a good way. Teams that rely heavily on the three for scoring may be liable for a big dropoff if they won a lot of games last year. Also, teams that score more points on average generally need a greater gap between points scored and points allowed for the same expected winning percentage. In the case of the Suns, they outscored their opponents by about 7 points on average. Most of the other top teams were at around 6. However, the Suns tend to score a lot more points than the other top teams, so they probably won more games than they should have last season. It's close, but this counts for probably about 1-1.5 wins fewer this season.
Injuries
How much did injuries affect the team last season? Despite the fact that they ran and ran and ran and...the Suns were generally quite healthy last season. Assuming the Suns keep up the same style as they did last year, I wouldn't expect them to stay as healthy this season. A team that runs than much is just begging for an injury. If Nash gets injured for any significant amount of time, this could really hurt the Suns. I'd say this is about 2-3 wins fewer for the Suns.
Player Gains/Losses
This is by far the most difficult of the categories to analyze. Losing a key role-player could prove to be a much bigger loss than you might expect if the team has nobody to fill that role adequately. On the other hand, replacing a superstar with a somewhat lesser superstar may actually improve a team if the new player fits into the system better (like Nash did last year). The Suns lost a few people during the offseason, most notably Q and Joe Johnson. They've replaced them with a handful of veterans who I'd describe as mediocre at best as well as a few highly athletic kids. I think this will hurt the Suns considerably--to the tune of 3-4 wins.
Age
Obviously an old team that gets older will generally get worse. A young team that gets older will generally improve. The Suns are still quite a young team and this may be their greatest asset. Nash is 31 now, but should still be in the prime of his career (though he may be more injury prone now). Stoudemire is still quite good, but he's so good already, it's hard to imagine him improving all that much. This is probably about +1 win for the Suns.
If you add up the numbers, you get that the Suns should end up with about 5-7.5 fewer wins than last year, which is on a par with the 5% I mentioned at the beginning. However, I mostly made up these numbers on the spot and they probably deserve more thorough analysis. For now, I'll put the Suns total wins at 55.75. With the public's perception of the Suns, I'd expect most sportsbooks to offer a number slightly higher than that, so there's a decent change I'll get a good bet at this one.
Once you come up with a number like the one above, I'll show you how you can determine your EV on a bet once the bookies give you a number and if there's enough of a difference for you to actually make a bet.
Now that I've covered the basics of how to win at sports betting, it's time to get into some specifics of identifying good bets. Right now, many sportsbooks have put up lines for odds to win the NBA championship. I haven't really thought much about it, so I'm not going to talk about it. However, you can check out Pinnacle's lines if you're interested. Right now, the Wolves are listed at about 32:1. For a while, they were listed at around 20:1, which was just ridiculously low. The new line of 32:1 seems a lot more reasonable.
The other topic which I'll cover in some depth is season win totals. Of course, there will be a plethora of bets similar to these. You'll see bets on the rookie of the year, MVP, and possibly other prop bets such as who will lead the league in scoring. The thing that all of these bets have in common is that they're long term bets.
If you wish to bet on any of these, you'll have to give your money to the sportsbook now and you won't be able to use it again for the rest of the season. Say you have given yourself a bankroll of $1,000 for the NBA season. Suppose you find 5 bets that look appealing, and decide to bet $20 on each of them. You've now bet $100 on bets that won't be decided until the end of the season. In essence, you've shrunk your bankroll by 10%.
The problem with this is obvious. If your bankroll is non-replenishable, your risk of ruin has increased significantly. Even if you end up winning all 5 of your bets, you won't get this money until the end of the season. Even if you pick a team with a really low win total and bet the over and they start off extremely hot and pass the over in December, you still won't see this money until the season is officially over. In case something happens and the season is cancelled, your bet is cancelled and you don't win anything.
If you have a replenishable bankroll, you can be more liberal with these bets. Of course, I still would like to have a larger than normal edge on these bets in order to make them. This should of course be a consideration when making these bets.
Ok, now onto the meat.
I've been reading Basketball On Paper by Dean Oliver. For anyone who enjoys basketball and statistics, this makes for quite an interesting read. I especially recommend it to any of you baseball fans who like sabermetrics and the work of Bill James.
There is a chapter of the book about parity in the league, and the general trend of a team to approach the .500 mark. This means that a team with a winning record will generally experience a drop off in production the following season. Likewise, teams that are under .500 will generally increase their winning percentage the following season. Furthermore, the closer a team is to the extremes, the greater the expected change should be.
Dean doesn't speculate too much about the reasons for this, but my intuition gives me a few ideas as to why this might be true. Teams with higher winning percentages are often lead by a superstar who might be having a career year. They'll also tend to be more veteran than other teams. The following season, we might not expect the superstar to once again perform as well the next season. The veteran players might have diminishing skills (see 2004-2005 Timberwolves) and their age might lead to them being more injury prone.
On the flip side of the coin, a team who doesn't perform during one season might have been hampered by the injury of a superstar the previous year. In addition, a team who isn't performing will often infuse the team with young talent. A year of experience for young players can equate to a large increase in the number of wins for that team.
In a nutshell, the reasons can be expressed as this. For a team to have a good season, a lot of things have to go right. For a team to have a bad season, a lot of things have to go wrong. In general, all these things won't happen two years in a row (though of course, the can).
Let's take a look at the numbers from the last two years:
| Team | 2004 Win% | 2005 Win % | Change |
| Indiana | .744 | .537 | -20.7% |
| Minnesota | .707 | .537 | -17.0% |
| San Antonio | .695 | .720 | 2.5% |
| LA Lakers | .683 | .415 | -26.8% |
| Sacramento | .671 | .610 | -6.1% |
| Detroit | .659 | .659 | 0.0% |
| Dallas | .634 | .707 | 7.3% |
| Memphis | .610 | .549 | -6.1% |
| New Jersey | .573 | .512 | -6.1% |
| Houston | .549 | .622 | 7.3% |
| Denver | .524 | .598 | 7.4% |
| Miami | .512 | .720 | 20.8% |
| Utah | .512 | .317 | -19.5% |
| New Orleans | .500 | .220 | -28.0% |
| Milwaukee | .500 | .366 | -13.4% |
| Portland | .500 | .329 | -17.1% |
| New York | .476 | .402 | -7.4% |
| Golden State | .451 | .415 | -3.6% |
| Seattle | .451 | .634 | 18.3% |
| Boston | .439 | .549 | 11.0% |
| Cleveland | .427 | .512 | 8.5% |
| Toronto | .402 | .402 | 0.0% |
| Philadelphia | .402 | .524 | 12.2% |
| Phoenix | .354 | .756 | 40.2% |
| Atlanta | .341 | .159 | -18.2% |
| LA Clippers | .341 | .451 | 11.0% |
| Washington | .305 | .549 | 24.4% |
| Chicago | .280 | .573 | 29.3% |
| Orlando | .256 | .439 | 18.3% |
Last season was kind of weird because of the HUGE decreases by 3 of the top 4 teams in the league (also, the addition of Charlotte into the league may have inflated other teams win totals somewhat). Indiana had half of it's teams suspended for half the year. The Lakers lost Shaq. The Wolves veterans got old and crappy in a hurry. Still, the general trends held. Of the 13 teams that had records over .500 in 2004, 7 teams decreased. Of the 13 teams with records lower than .500 in 2004, 9 of them increased their winning percentages.
One season of course doesn't prove anything, but if you go back, you'll see that the general trend holds. Another thing to note is that teams near the very top of the league tend to resist this trend more than others. This is generally because the teams will have good talent or a good system, and they will do anything possible to keep these things in place for as long as possible.
Although season totals haven't been listed yet, I imagine they will be amongst the first NBA bets I make this season. The Wolves have gone through a complete overhaul and my guess is that there season win total will be listed at something between 39-41. This number may get inflated due to KGs presence as well as the public's perception of the Timberwolves.
There are several categories to consider. I'll go through each of them using the Suns as an example.
Last Year's Record
The Suns won 75% of their games last year(62-20). In general, a team that wins that many games will experience about a 5% dropoff, which is about 4-5 wins. All other things being equal, we'd expect them to win around 57-58 wins this season.
How Lucky Were They Last Year?
It's nice to look at points for and points against rather than just W-L records. A team might win a higher percentage of their games than you'd expect just by having their points scored distributed in a good way. Teams that rely heavily on the three for scoring may be liable for a big dropoff if they won a lot of games last year. Also, teams that score more points on average generally need a greater gap between points scored and points allowed for the same expected winning percentage. In the case of the Suns, they outscored their opponents by about 7 points on average. Most of the other top teams were at around 6. However, the Suns tend to score a lot more points than the other top teams, so they probably won more games than they should have last season. It's close, but this counts for probably about 1-1.5 wins fewer this season.
Injuries
How much did injuries affect the team last season? Despite the fact that they ran and ran and ran and...the Suns were generally quite healthy last season. Assuming the Suns keep up the same style as they did last year, I wouldn't expect them to stay as healthy this season. A team that runs than much is just begging for an injury. If Nash gets injured for any significant amount of time, this could really hurt the Suns. I'd say this is about 2-3 wins fewer for the Suns.
Player Gains/Losses
This is by far the most difficult of the categories to analyze. Losing a key role-player could prove to be a much bigger loss than you might expect if the team has nobody to fill that role adequately. On the other hand, replacing a superstar with a somewhat lesser superstar may actually improve a team if the new player fits into the system better (like Nash did last year). The Suns lost a few people during the offseason, most notably Q and Joe Johnson. They've replaced them with a handful of veterans who I'd describe as mediocre at best as well as a few highly athletic kids. I think this will hurt the Suns considerably--to the tune of 3-4 wins.
Age
Obviously an old team that gets older will generally get worse. A young team that gets older will generally improve. The Suns are still quite a young team and this may be their greatest asset. Nash is 31 now, but should still be in the prime of his career (though he may be more injury prone now). Stoudemire is still quite good, but he's so good already, it's hard to imagine him improving all that much. This is probably about +1 win for the Suns.
If you add up the numbers, you get that the Suns should end up with about 5-7.5 fewer wins than last year, which is on a par with the 5% I mentioned at the beginning. However, I mostly made up these numbers on the spot and they probably deserve more thorough analysis. For now, I'll put the Suns total wins at 55.75. With the public's perception of the Suns, I'd expect most sportsbooks to offer a number slightly higher than that, so there's a decent change I'll get a good bet at this one.
Once you come up with a number like the one above, I'll show you how you can determine your EV on a bet once the bookies give you a number and if there's enough of a difference for you to actually make a bet.
