I've started an October tradition. I'll take a couple of days off, hop on a flight to vegas, and spend the weekend prepping the BoxScore Geeks database for the upcoming season. And while I am there, I'll put some money where our data is. This year, I placed a total of $7500 in bets on the 2014/15 NBA season, and I'll be posting regularly about them so that you can follow our progress.
Before we get to the actual bets, I should note that I followed a couple of rules:
- No single bet could be more than 15% of the bankroll, even if it is the best bet on the board. This is to mitigate the risk of ruin, a core bankroll management that says that diversification and hedging are often more important than pure expected value, because if you have an edge in a betting game, it is of paramount importance that you preserve your ability to keep playing the game.
- Put at least 5% of the portfolio on some "long shots"
- Avoid the Lakers, no matter what the model says!
First, let's take a look at the bets:
| Bankroll | $7,500 | |||||||
|---|---|---|---|---|---|---|---|---|
| Team | Type | % | $ | Odds | Payout (W) | Sim | $EV | ROI |
| SUNS | OVER 43 | 14% | $1,050 | 1.91 | $2,004.55 | 87% | $1,743.95 | 166% |
| CHI | UNDER 55 | 10% | $750 | 1.91 | $1,431.82 | 88% | $1,260.00 | 168% |
| PHI | OVER 15.5 | 5% | $375 | 1.91 | $716.25 | 86% | $615.98 | 164% |
| WAS | UNDER 48 | 5% | $375 | 1.91 | $716.25 | 85% | $608.81 | 162% |
| MIN | OVER 27.5 | 10% | $750 | 1.91 | $1,432.50 | 83% | $1,188.98 | 159% |
| SAC | OVER 30 | 10% | $750 | 1.91 | $1,432.50 | 80% | $1,146.00 | 153% |
| BRK | UNDER 42 | 7% | $525 | 1.91 | $1,002.75 | 77% | $772.12 | 147% |
| CHA | UNDER 44 | 5% | $375 | 1.91 | $716.25 | 76% | $544.35 | 145% |
| DAL | UNDER 49.5 | 5% | $375 | 1.91 | $716.25 | 76% | $544.35 | 145% |
| MIA | UNDER 44 | 5% | $375 | 1.91 | $716.25 | 70% | $501.38 | 134% |
| SAS | OVER 57 | 5% | $375 | 1.91 | $716.25 | 67% | $479.89 | 128% |
| TOR | WIN DIV | 3% | $225 | 1.70 | $382.50 | 81% | $309.83 | 138% |
| SAS | WIN DIV | 5% | $375 | 1.60 | $600.00 | 73% | $438.60 | 117% |
| CLE | WIN DIV | 4% | $300 | 1.50 | $450.00 | 76.50% | $344.25 | 115% |
| SUNS | WIN DIV | 3% | $225 | 13.00 | $2,925.00 | 32.40% | $947.70 | 421% |
| MIN | WIN DIV | 1% | $75 | 101.00 | $7,575.00 | 1.60% | $121.20 | 162% |
| POR | WIN DIV | 1% | $75 | 3.80 | $285.00 | 31.40% | $89.49 | 119% |
| ATL | WIN DIV | 2% | $150 | 6.00 | $900.00 | 33% | $297.00 | 198% |
| Totals | 100% | $7,500 | $24,719.11 | $11,953.86 | ||||
First, I have some general notes about probabilities and EV calculations:
- The probabilities are based on excel simulation that "played" hundreds of thousands of games. They are a work in progress. I like to think of these as similar to the revenue projections that an entrepeneur might bring to a meeting with some venture capitalists -- everybody knows they are probably a little inflated, and there are big margins for error. If projections were perfect, the entrepeneur wouldn't need a VC, but just need a banker (i.e. there'd be no risk factor and therefore rewards would be small and predictable).
- The payout column is a gross (i.e. the odds column includes the original bet) and EV is calculated based solely on the odds of winning (again, according to the sim, take that with grains of salt). This isn't exact because over/unders can push. So the payouts are really p(lose) * 0 + p(tie) * (bet) + p(win)*((odds*bet)). But at the beginning of the season p(tie) is a very small number. For some bets, when one or two games are left in the season p(tie) might get very big and p(win) might approach zero (and obviously the reverse will happen for some teams), and the EV calculations will change, but for now, this approximation suffices.
Let's talk about the bets.
The Suns OVER 43
The biggest bet on the board, this is our equivalent to the Knicks under last year. By adding Isaiah Thomas, the Suns added depth at a position where they had injury concerns. The loss of Channing Frye...isn't a loss. We can expect marginal improvement from younger players like the Morris twins and Miles Plumlee. It's hard to see a team that won 48 games last year regress by 5 games even though they added a solid reserve. Yes, there is some concern that they're three-point shooting will cool off after a hot year, but that variance is just part of gambling. If the over/under is at 43 when it should be at 48, you take the bet.
Chicago UNDER 55
Let's be clear, we think Chicago is a good team. But it feels like Vegas is expecting way too much from Derrick Rose here. The list of guys who came back after two years off and played anywhere near their old production levels is....non-existent. It's just never happened. Now, Rose is young and he might someday recover to greatness. But it just isn't smart money to bet on it happening in his first year back.
The addition of Pau certainly seems like an improvement over Boozer, but there is some overlap with Noah and again, everybody is assuming that the 34-year old will have a bounce back year now that he's left L.A. Except that 34-year olds don't generally start playing like 29 year-olds again, even if they were misused by their previous coaches.
If I lose this bet, I think it might be because of Mirotic. It's not a sure thing by any means, but it feels like a solid bet. So far, through the preseason and 3 games, Mirotic looks very good.
Philly OVER 15.5
Again, this is just probability. No one is saying the Sixers will win 35 (I am not repeating that mistake). But Noel looked very good through the preseason. So far in three games he can't score but he's a hell of a rebounder and shotblocker so I have confidence he will be productive. MCW will likely improve from a year ago. It will be hard to prevent winning a few games just on the strength of those two alone.
The Sixers will be awful, and we will sweat this one all year long with key players missing games with "back spasms" or "flu-like symptoms," and the team will likely not play Embiid even once he's healthy, but it is just damn hard to lose 67 games. Rembember that they will have to play plenty of games in January through March against other awful teams.
Washington UNDER 48
This is a case where everyone is assuming that Pierce has more left in the tank than he likely does, and everyone is selling Ariza short. Ariza was a beast last year and his production won't be easily replaced. I love Andre Miller, but he's old as dirt and spent all of last year not playing because he pissed off his coach, and he's not likely to get huge minutes without injuries. I also like Humphries a lot. But it seems like this is a team where the roster changes are a wash (lost a great player, got some good players but not all will get heavy minutes), and that pretty much leaves them in the 40-45 win range just like last year. It won't shock me if they win 49+, but the under is a good bet.
Minnesota OVER 27.5
This one is all about anchoring. The Timberwolves won 40 last year, so it feels like losing Kevin Love subtracts about 15 wins, and this looks like a bad bet..... BUT. The Wolves had the point differential of a team that should have won 48. If you take Love off of a 48-win team, it's a 33-win team. Then you factor in the departure of JJ Barea (addition by subtraction), some increased minutes for Gorgui Dieng, and a possibly healthy Budinger, and you have a team that could "cruise" to 35.
There's always the danger that in January, Flip takes off the coaching hat and puts on the GM hat and decides to blow it up and trades Martin, Budinger, and Pek for Bargnani and Amare and picks.
So far, Flip is underplaying Dieng (their best per-minute big man, and it isn't close), and giving Andrew Wiggins tons of burn despite the fact that he's been awful. But I also don't think Pekovic will continue to be this bad, and Wiggins is likely to either lower his insane turnover rate or get fewer minutes if he can't, either of which will benefit them.
Sacramento OVER 30
Darren Collison is a good addition, Boogie looks like he finally cares, guys like Sessions and Casspri are upgrades because "bad" is better than "terrible". And, they've employed Dean Oliver, which means they might make some smart personnel decisions mid-season.
Brooklyn UNDER 42
Several factors are at play here. First, losing Pierce and Livingston will hurt a lot. Second, many of their core players are fragile either because they are old (Garnett, AK47, Johnson) or are standing on wobbly ankles (Lopez, Williams). And finally, there is a very real probability that they blow it up soon. Last year's insane spending spree didn't bring results, and the team may shift gears.
Charlotte UNDER 44
We love us some Lance Stephenson. He is a great addition. Unfortunately, we also loved McBob, and losing him hurts the team enough that it is probably a wash. They also lost key role players Tolliver and Douglas-Roberts. Unless MKG's new jumpshot pans out soon, this team is going to have huge troubles on offense, because Lance and Al can't do it all alone.
Dallas UNDER 49.5
Yes, we love Tyson Chandler and we like Chandler Parsons. But we also loved Jose Calderon and Shawn Marion, and while Dalembert was no Chandler, he didn't suck, so the upgrade was incremental. It's hard to imagine Tyson Chandler gets 2500+ minutes, but If he does, I will kiss this money goodbye.
Miami UNDER 44
Everybody thinks that just because Bosh had to change his game to play with LeBron, that means that Bosh will now revert to the Toronto Raptor's version. But, again, 30-year olds generally just don't magically morph into their 26-year old selves. It's ridiculous to assume he will be that good. Similarly, it seems a safe bet to assume that Wade will have trouble logging heavy minutes. He's proved people wrong before, but the under seems like the smart money here.
At the time of this writing, of course, Bosh is challenging Anthony Davis for Eastern Conference player of the week. So, yeah, very smart money.
San Antonio OVER 57
More simple math. The team won 62 last year, didn't lose any players, and somehow managed to grab the best NCAA player in the draft with the 30th pick. The only reason this isn't 15% of the portfolio is that we all know there will be plenty of back-to-back games where Parker, Ginobli, Leonard and Duncan don't play. Having said that, the Spurs' bench would be a playoff team anyway, so it's still a solid bet.
Toronto to win the Atlantic
The payoff odds are not great, but really, the Atlantic is so bad that Toronto is a huge favorite here. This bet seems self-explanatory.
San Antonio to win the Southwest
The Spurs play in a much tougher division but they still seem like the clear favorite over the field. Again, not too much to say here.
Cleveland to win the Central
The only competition here is Chicago. It is a small payout but once again, I'd consider them a favorite over the field.
The Suns to win the Pacific
I think both the Clippers and the Warriors are more likely to win the Pacific, but the 12:1 odds were way too good to pass up. I'd call this a "fair" bet if the odds were more like 4:1. One of the best bets on the board, but obviously high variance.
The Wolves to win the Northwest
Now we're talking. Every year, there is a long shot +EV bet on the board. These are great bets for a small portion of your portfolio because when they hit, they can transform bankrolls. This was essentially a bet that Gorgui is a franchise big, and Flip gives him the minutes, and Wiggins exceeds expectations. That's obviously a long shot, but it didn't feel like a 100:1 long shot.
After seeing Flip's rotations and coaching through 3 games, I'd actually love to take this bet back, but I still think it was a great bet given the knowledge we had at the time.
Portland to win the Northwest
These are pretty good odds that the second-best team in the division will break through to win it. It seems like the payout here should have been a bit lower. It seems that Vegas did not adjust far enough after the Durant injury.
Atlanta to win the Southeast
This is one of my favorite bets. It seems like the Southeast is Orlando and then four teams duking it out. If you think that Atlanta is in the same talent level as Miami-Washington-Charlotte (and we do), then the "correct" odds should be 3:1, not 5:1. This bet was 8:1 at one point, a fantastic value, but sadly, I didn't make it to Vegas in time to take advantage of that.
I'll be charting the progress of this "portfolio" about once a month as the season goes on. This will give you plenty of opportunities to make fun of what we get wrong (my initial "I've-got-a-bad-feelin-about-this" vibes are for the Miami and Chicago bets), or to jump on the bandwagon for the teams we get right (we're all Suns fans this year, right?).