Nba nerd

The NBAGeek Goes Back to Vegas

Before the season started, I went to Vegas and placed a few bets. I promised to revisit them as the season went on, and as it happens, I was recently in Las Vegas again, so this seems a pretty good time to do a quick checkup.

So, how do things look?

Team Bet Current Pace Bet Pays Winner?
Knicks Under 50 15 wins $1,000 10:11 Y
Hawks Over 40 41 wins $100 10:11 MEH
Pelicans Under 40.5 41 wins $300 10:11 MEH
Celtics Over 28 30 wins $500 10:11 MEH
Pacers Under 55 77 wins $600 10:11 N
76ers Over 17 27 wins $500 10:11 Y
Jazz Over 25 14 wins $800 10:11 MEH
Warriors Under 51.5 46 wins $600 10:11 Y
Trailblazers Under 38.5 68 wins $400 10:11 N
Celtics Win Division WOOT $40 60:1 Y
Jazz Win Division No Way $40 100:1 N
Grizzlies Win Division Not Likely $60 7:1 N
Pistons Win Division Hell No $20 17:2 N
Cavaliers Win Division Hell No $20 16:1 N
76ers Win Division WOOT $20 500:1 Y
Totals $5,000    

So far, the results look mixed. Rather than grab-bagging it, I'll look at each pick.

The Knicks Under 50

I'd always considered this the "safe" pick, and many told me that I should have mortgaged my house to bet more here. This roster is simply a disaster. Even with Chandler this team had no chance at 50, and without him they are a lottery team. I don't think there's a ton to say here, this team has played exactly as we thought they would. What would really worry me as a Knicks fan is that Bargnani is actually playing at a career high levels (meaning he's "only" bad, not "worst player in the league") and Shumpert is overperforming the projection, and they are still terrible. The second thing that would worry me is the Knicks have virtually no future, since they have traded away a hell of a lot of picks.

The Hawks Over 40

This wasn't one of the safest bets, but conventional wisdom was that Josh Smith was a better player than Paul Millsap, and that's the kind of mistake we try to capitalize on. I think the Hawks are going to get even better as Ayon gets more minutes. I've only caught a couple of games but Korver sure has been amazing for them so far.

The Pelicans Under 40.5

The Pelicans are on pace to win 41 games, but have had a very easy schedule so far. This is way too early to say much about but I still like this bet. Of course, Anthony Davis breaking his hand is a tough break for the Pelicans, but likely a factor here; he has single-handedly kept this team in the hunt, as everything else we were concerned about has come true regarding this roster.

The Celtics Over 28

This bet is now pretty decent. Initially, I was a little worried that the Celtics were actively trying to lose because of their messed up rotations (Humphries was amazing in 21 minutes in game one, then racked up 3 straight DNPCDs). But it looks as though they are trying to win, and they have had a tough schedule. If Rondo comes back and plays, I really like this bet. If they go the Derrick Rose route and keep him out for the year, it will be very close. Other than Jordan Crawford, this team is who we thought they were. 

The Pacers Under 55

I've watched the last several Pacers games and I have to say that in addition to their easy schedule, the refereeing has really been helpful for this team. The Pacers play 90s basketball. They are as physical as it gets, and I've actually been really surprised at what the officials have let them get away with. I think Vogel deserves a lot of credit for recognizing what his team can get away with and pushing that boundary. When teams are getting pounded on constantly, and not getting a lot of calls, they start settling for long jumpers, and the Pacers profit from that tendency.

In any case, I do not think the Pacers are a 70+ win team (as their current pace suggests), but clearly they are a better team than we thought they were. One reason is that when Paul George made the leap this year, he skipped right on over "all-star" and went straight to "MVP". Another is that Roy Hibbert decided to become a top 3 center -- his true shooting is at 55%, a huge improvement over his career average of 49%. And finally, Danny Granger got injured, which was a huge lucky break for ths Pacers, as that meant that Lance Stephenson got a lot more minutes than he might have otherwise. And now, he has played so well that if and when Granger gets back, he won't take any minutes from Stephenson.

I think regression to the mean will bring this team back down a bit (raise your hand if you really think this is a 77-win team, which would be GOAT territory), but obviously they aren't likely to crash hard enough for me to like the under on 55 wins.

The 76ers Over 17

Well, I think we all know where this is headed. As a segway from the Pacers, I watched a 76ers team without Thad Young or Spencer Hawes on the second night of a back-to-back in Indiana take the Pacers to the wire. The scary part here is that so far, Young has had a very sub-par season. If he can return to form, and if this team decides to play Noel in January/February, this team will make the playoffs. I also think MCW is for real. He can't really shoot, but you do not just "run hot" over 500 minutes to average 4 steals, 10 assists, and 7 boards per 48 as a point guard. Whenever a player is really good at all the non-shooting categories, and is at least average at scoring efficiently, this is a very good sign.

If I were running the sixers, I'd be shopping Turner very aggressively right now. He gets a lot of (undeserved) credit for thier play so far. The meme is that Turner has finally made "the leap" to stardom, but I think he's still overrated (those turnovers!), and there's a great possiblity that the Sixers could sell high on him right now. Hawes is one they should also shop, as it's highly unlikely that they will want to pay him $12 million + next year, but the reality is that lots of teams aren't going to offer much for a 4-month rental, and it may be more useful to just keep him, let the contract expire, and perhaps sign-and-trade him in the summer.

In any case, as I said before the season, this team doesn't have enough terrible players to lose 65 games.

The Jazz Over 25

Meh. I don't think anyone could have predicted the Jazz would start THIS bad. They're on pace to set all kinds of offensive futility records, and to top it off, Enes Kanter has been awful. At this point, the Jazz will have to run hot for me to win this bet, and the big concern is that it seems unlikely that they will want to. 

I am holding out some hope that Burke becomes a good rookie, and Hayward comes out of his funk, and the team can be good enough to give the bet some legs.

The Warriors Under 51.5

Here's another bet that looks pretty good. One can say that I got "lucky" with Iguodala's injury, but I thought 52 wins was pretty crazy for a team that just ran really hot during the playoffs last year. Of course, then Klay Thompson tried to make me look bad with his outrageously great shooting the first few weeks. Just when I thought he was calming down a bit, he went 8-of-11 from three last night. The Warriors continuing to shoot a nearly-unsustainable 44% from three is about the only way I will lose this bet, and I'm ok with taking that bet.

The Trailblazers under 38

This one looks even more hopeless than the Pacers bet. As I said on the podcast, I do not think the Blazers are "truly" a 50-win team, but they will probably win 50 anyway. The Blazers have had a very easy schedule, but almost all of the credit for the team being 14-3 instead of something like 10-7 (where I might have a chance that they'll cool down enough to get under 38) should go to Wesley Matthews. What the hell happened to this guy in the offseason? Not only is he shooting a ridiclous 51% from three, he's also nearly doubled his career rebounding numbers while lowering his turnovers and fouls. In short, Matthews has been as good as Dwyane Wade. Raise your hand if you predicted that, so I can shoot it off.

Celtics to Win Division (60:1)

Love this bet. There's a reason this division is called the Titanic by Arturo. At 60:1, the chance that Rondo comes back and leads this team to .500 ball and wins the division is delicious. I wouldn't take this bet at 5 or maybe even 10 to one, but at 60 to 1 it is a very high +EV high-variance bet.

Jazz to Win Division (100:1)

I would not take this bet at 1000:1 now. Not much more to say here.

Grizzlies to Win Division (7:1)

The Grizzlies are going to bounce back and get good, I'll say that. But barring a Kawhi Leonard injury, they aren't taking the division crown from the Spurs. Of course, injuries happen, so I won't tear this ticket up.

Pistons or Cavaliers to Win Division

I may as well tear these up.

76ers to Win Division (500:1)

WOOT. I liked this bet before the season and now it's probably the best bet on the board. No, the 76ers are not a good team. However, neither is any other team in this division. I don't think the 76ers have great chances at this, but let me remind you that Toronto is currently in first place with a 6-10 record! It is entirely possible that 37 wins will take this division. And the key here is that the payout is huge, making this a big +EV bet.

To put it another way, who wants to give me 500 to 1 that the 76ers won't win 37 games? Anybody? Anyone at all?

Bueller? Bueller?

One small quibble: if the Warriors pace has them at 5 1/2 games under and you're counting that as a Yes, the Jazz's pace at 11 games under should be a No, not a Meh.

Yes, there's a lot more room for the Jazz to pick up those 11 wins than there is for the Warriors to grab another 6, but the Jazz are also on pace to win only a little over half the required games.

Put another way, the Warriors need to outplay their current performance by a fair amount (which may be unlikely, since Iguodala's been playing at a top-10 level when Curry's in), the Jazz basically need their roster to double its production over the rest of the season.
Looking just at the over/unders and ignoring the division winner bets, and accepting DooDoo_Jump's quibble (which I think is correct), it looks to me that you have $2100 in bets in the money, $1800 out of the money, and $900 Meh. Which I guess is not bad.
Part of the comparison should be the original probabilities as they seemed way off during the pre-season and thus far they still seem way off.
Can I just say I called it on the Pacers? Because I called it...
Still hating on the Pacers I see. Seriously, would you please let us know what you guys have against the Pacers? I've watched four times as many games as you and have seen missed calls going both ways. This happens every night in any sport being played across the country! Give me a break! It will be interesting to see what excuse you guys have when the Pacers are 25-5 going into January or 36-10 going into February. Those records are assuming they lose a few games I don't actually think they will, but a road game like what happened at Chicago is always possible.
It's interesting to me that you think me calling Paul George an MVP and saying that Frank Vogel deserves a lot of credit for recognizing a trend and exploiting it as "hating".

Basically, I do not think the Pacers are "really" a 70-win team. If you call that hating...well, troll on, brother.
The Warriors might not be able to keep shooting 44% from three as a team, but it may not plummet either.

Klay Thompson and Steph Curry are 1st and 4th in the whole league in three point attempts as of today, they have combined to hoist up 246 of them so far, more than half of the team's total attempts.

Curry is at 44.3 3FG% right now, right at the team average level, that's actually lower than his ridiculous 45.3% from 12-13. Thompson is at 47.3% in 13-14, it seems unlikely he can keep that up, but he did shoot over 40% for the season in 12-13 so he may not fall that far.
Brian,

Agreed. Hoisting up a shitton of threes is frankly never a terrible strategy if you have 2 or more guys who can shoot at that clip.

But Iguodala and Barnes shooting 48% from three? And Daymond Green at nearly 40? I'll bet against that happily.

Also, for reference, a grand total of 14 teams have ever shot > 40% on 800+ attempts, and only none have ever shot > 42%. So I like my chances :)
DooDoo,

In principal I would agree, except my reasons for optimism are:

1) Burke looked like a good rookie, which projects him in the .075-.100 range. This is a monumental improvement over Lucas III.

2) I expect Hayward's shooting to regress to the mean

3) Favors is already regressing to the mean. I expect Kanter to as well.

In short, if you look at what those players have done historically, the data suggests that they will improve A LOT from their current level of play (i.e., it's not just a blind hope). The big question mark is Burke.

Lastly, let's not discount the inevitable "team fires the coach because the team sucks and suddenly the players rip off 5-6 wins" factor.

The Jazz have had a really tough schedule, and although they've stunk, they could easily have 6 wins right now instead of 3. The bet obviously does not look good but I still refuse to believe the Jazz are actually this bad. Again, that will make this bet closer. I expect that even if the Jazz improve it will come down to the wire.
With an over/under as low as 25 (or 17), taking the over is just playing the odds. The high variability of +/- 5 wins when teams will be dropping games they shouldn't in January (midseason grind) and March (tanking or resting for the playoffs) makes it a safer bet unless the team is obviously tanking (which the Jazz do not seem to be doing).
Those are reasonable points, Patrick, but I think you could give the Dubs the same kind of generous analysis: they had to sit one of their two best players over the last 5 or 6 games. I'm not going to look at the tiny margins in those losses (Mavs, Spurs, Thunder) and claim, "Of course if Curry had played in the Spurs game, or if Iggy had played in the others, they would have scored 3 more points", because that's silly, but I do think that they very well would have won some of those games. And if they'd won two more games, they'd be a Meh, but they'd be a Meh on the other side.

Although I totally agree about Iggy/Green/Barnes's 3pt shooting being unsustainable, I think there's a decent chance those three settle in somewhere around 40%, as they're typically taking only wide open threes. Barnes was already above average in his rookie year. And, in fact, despite being pretty terrible from 3 his rookie year, Green improved a lot over his college career (ending up at 38% on 100+ looks, and 36% on 100+ the year before). It's possible that he's adjusted to the NBA 3pt line, though definitely not to the tune of 50% (or even 45%) shooting.

Tiny sample size, but when he's playing with Curry, Iggy's shooting amazingly well, but absent Curry, he's shooting terribly. To dig up some ancient WoW history, it might (might might might) be the case that Curry has the mythical 'Carmelo Effect', at least when it comes to 3's. Heck, if you look at this year and last year (and even his rookie year) almost all of the top 5 3pt attempters on the Warriors have shot well above average (and something like 80%+ of all their 3pt attempts). Over his career, the worst year they've had from 3 they still shot 38.8% (and that was his injury season). I'm totally happy to admit that this is all fan storytelling, but it's such a pretty story I can't stop: when you have a player who's as dangerous against single coverage as Curry is, who's as good a passer as he is, and your team is built to space the floor, it's very possible you're going to end up getting better 3pt shots than other teams.

With Curry/Thompson (both proven 40% guys) taking half the threes, it's definitely possible they end up hitting 40% again. They are, after all, one of the 14 teams to shoot above 40% over a season (last season), if only just.


The hating comment comes from the perception you give that the Pacers have the record they do because of the referees. That is a joke. Congratulations for noticing what I saw George and Hibbert doing in the playoffs last year. You don't get credit for noticing this now when it should have been obvious based on last year, their age and naturally expected progression, and the fact that those two in particular work on their games in the offseason. Thanks for watching a game or two, seeing some questionable calls (shocking!), and turning that small sample size into your solution for being so wrong about your expectations for them. I was just hoping to see more objective analysis on why your expectations have been so different from the actual results, but please, go ahead and ignore the underlying bias and throw out some more name calling...brother.
cpism2

I think everybody reasonable, Patrick included, agrees that the Pacers success is due in large part to young players improving, and great coaching by Vogel.

I also think that everybody reasonable, Patrick included, agrees that the Pacers get away with murder on defense, which plays a significant role in their success. We saw the beginnings of this in last year's playoffs, with Hibbert's "verticality", which basically allowed him to foul anybody as long as he made a show of sticking his arms straight into the air while doing it.

Referees having a significant impact on winning isn't a phenomenon unique to Indiana. The most esteemed NBA gamblers incorporate referees bias into their predictions. Simply pointing out that this "bias" affects Indiana's ability to win, isn't irrational "hating".
So what is the expected payout so far?
In terms of model validity, you should be pretty concerned that a lot of your 90+% sure things are looking like busts.
Yeah... Jazz/Pacers/Warriors/Blazers were all 90% (and the model gave the Jazz over 25 the highest probability of all the bets at 99.1%).

Now, it's fine that those aren't looking like sure bets yet. It's a model, and it's still early in the season. But, if the current trends continue, it's definitely a sign that the model has some big flaws, at least when it comes to its confidence.

Think of it like this: if you made 10 guesses (the over/unders) and you said that 6 of those guesses each had a 90% chance of being correct, if 3 of those 6 are wrong, you're not necessarily a terrible guesser, but your level of confidence in your guessing ability is suspect, right?

All that said, I think it's a really difficult proposition.
Who knows the reason Barnes and Green are shooting so well, is it the system reasons mentioned above, because they are the age where players have the potential to improve a lot, or just luck?

Either way, they (along with Iguodala) have taken a combined 111 threes. Those three guys would REALLY have to go in the tank to erode the 246 threes that Curry and Thompson have shot so far (at a combined 45.9%).

good conversation going, guys. As a someone old to the game, but new to advanced stats, I'm very appreciative for the insight.

I would ask Patrick one question: Of all your bets that are not currently looking great, which do you still have the most optimism to turn around in your favor?
Definitely getting into sports betting next year once I learn a bit more. No brag, but I've personally been right too many times to not be taking advantage (lot of that credit has to go to WoW for helping me view the game better).
I will say that I am a bit surprised with New York. I expected them to be bad, but not 3-13 bad. I was expecting Prigioni to get a lot more minutes and was not counting on Tyson getting hurt. When you consider the opportunity cost of not playing him (more minutes for Barg, Melo at the 5, and KMart), Tyson's 2 month injury could set this team back several wins.
When will the power rankings come out?
Cpism - Chill out. Quit taking this referee comment as a personal attack. I'm sure because you've watched every game that you have capitalized on the $5000 you bet on the pacers over.

Art- There is such thing as daily sports betting. You don't have to do full season prop bets.

Jason- Enjoyed this article. Lets get a T-puppie win over the spurs tonight. Vegas says Spurs 65% (-185) chance at winning tonight. I'll roll with my bias, WOLVES +160!
Cpism,

"Brother" is now "name calling"? Dude, have you read my stuff? I swear like a fucking sailor. Believe me, when I want to insult you, you'll notice, and "brother" is not an insult, but I guess it makes it easier for you to sell your bullshit argument that I am "hating" if you paint a picture where I'm insulting you.

Again, I called George an MVP. I very specifically called out that we thought he'd improve to be a star, but that he got even better than that. Most people on the internet consider that a compliment.

Again, when I pointed out the refereeing, I did not say anything like "that's so unfair", and instead I gave Vogel a slow clap for teaching his team to exploit this. This is good coaching, just like Marc Jackson getting his team to shoot a lot of threes is good coaching -- they have caught on to trends in the modern game that other teams are still figuring out.

And one last time for the fucking slow kids: If I say the Pacers look a little better than they are because of their schedule, I'm saying that they are not really a 77-win team. This is not "hating". Check the record books.
@ Patrick. When you guys put together the initial projections for the Pacers, you guys had a very high confidence interval on the Under. Do you guys just go with the model or ever try to contextualize things (like I attempted to do in the first post on that article?) A lot of the flags in your article seemed to be almost too long sighted given the general youth of the starting lineup. I also think you guys should give some serious credence to "Super Star Calls" and the importance of perception in the league in your models (George). I really love reading the admission of oversight but any plans in the future to prevent them? I think a lot of the stuff you missed wasn't all that hard to predict with confidence the other way. I think a major miss was overvaluing stats of players when they're injured, especially if the sample is about 50/50 and the means between the two are huge (Hibbert). Also, role changes matter, players typically play even better with consistent minutes ie Starters than as bench players (Hill), as far as the research I've done, the weaker competition argument is BS. WP increase is positively correlated with minutes increase. In either case, I enjoy reading the work you guys do :)

Sign in to write a comment.