Smoke the Books
@SmokeTheBooks
Followers
2,471
Following
705
Media
170
Statuses
821
Hedge Fund Manager building sports betting models and brainstorming random betting ideas. Nothing to sell. Don’t bet the rent money people!!!
Cherry Hills Village, CO
Joined November 2022
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
México
• 1132444 Tweets
Morena
• 679195 Tweets
casillas
• 402441 Tweets
Flamengo
• 330644 Tweets
Xóchitl
• 324435 Tweets
緊急地震速報
• 194779 Tweets
The Loyal Pin is Coming
• 184005 Tweets
#ปิ่นภักดิ์Q51ใกล้เสร็จแล้ว
• 180219 Tweets
Stars
• 153924 Tweets
Junior
• 132231 Tweets
Whindersson
• 131373 Tweets
PREP
• 111484 Tweets
地震大丈夫
• 82653 Tweets
Florencia
• 80627 Tweets
#CH3Apologize
• 67780 Tweets
LEALTAD FURIOSA
• 65855 Tweets
Televisa
• 51845 Tweets
Oilers
• 43316 Tweets
地震警報
• 41779 Tweets
Mario Delgado
• 39015 Tweets
Soto
• 31086 Tweets
アラーム
• 28614 Tweets
Tigre
• 27161 Tweets
Panthers
• 26983 Tweets
Natalie
• 26537 Tweets
アラート
• 26066 Tweets
Borja
• 24044 Tweets
Stanley Cup
• 20370 Tweets
#HomeForHomeless
• 19351 Tweets
Bucaramanga
• 19182 Tweets
Pereira
• 17870 Tweets
#MissUniverseColombia
• 17528 Tweets
渡辺くん
• 15014 Tweets
McDavid
• 12503 Tweets
Barranquilla
• 12249 Tweets
Santa Fe
• 12153 Tweets
しょっぴー
• 10160 Tweets
The efficient market hypothesis in sports betting (or any other market) is not binomial, it’s a continuum. While the stock market in the US today is more efficient than, say, the stock market in Canada in the 1970’s, it can still be beat. While it is highly efficient and almost
Have you heard of the Closing Line Value (CLV) Theory but aren't sure what it is? Here I will explain:
1. What CLV is
2. How to calculate the EV of all your bets
3. If it applies to Tennis (with my own sample of 4511 bets)
4. How you can use it to win money betting 👇
• The
17
15
245
11
14
135
Bankroll simulators are very helpful for understanding and visualizing variance, staking methods and bankroll management, so we put one online.
You can run up to 1,000 simulations at a time, each with up to 10,000 bets. The graph will randomly show 10 of
7
8
79
We built an NFL Sports Betting AI. It's free and the link is below. Further down in this post is more info on what it is and what it isn't.
We've been playing around with custom AI GPTs built on OpenAI's ChatGPT. The impetus here wasn't to make a free NFL Betting AI. Totally
6
4
78
With the exception of Circa & Caesars, the current state of the legal sports betting market is that only alcoholics can go to liquor stores.
3
15
77
The efficiency of betting markets is complex. In many ways, they are extremely efficient. If you look at 10,000 soccer games with no-vig lines of -300/+300 almost exactly 25% of the dogs will win. On the other hand, there simple things that are just continually mispriced.
For
9
9
77
Here is an analysis of variance within NBA games. The chart shows the results of 1,000 simulated games of evenly matched teams against the actual results of the last 1,000 games where the spread was 1.5 points or less.
This simple simulation is the by-product of something we’re
9
8
75
What are the odds that you've lost money after 1,000 bets due to variance? To answer that you also need to know the average odds of your bets and your EV. If you somehow actually know your true EV, this chart is the answer.
@SmokeTheBooks
Curious if you can generate some sort of graph that shows the likelihood of all ROI ranges. Maybe using buckets?
2
0
4
7
14
67
@DK__DFS
The simple fact that
@FDSportsbook
let’s him bet $2k or $5k on a player prop tells you everything you need to know
3
0
68
Another issue I have with the efficient market hypothesis religion is how easily the lines at sharp books can be manipulated. The limits on all but the biggest sports are still very small in absolute dollars. Mavs tip in 8 hours and Pinnacle's limits are $1,500 on moneyline &
4
5
66
Following up on the last post, here's what variance looks like for different odds with same +EV. Red is 10 simulations of 1,000 bets at +500 odds with 5% EV and blue is 1,000 bets at +100 odds with 5% EV.
6
5
62
One last post on variance. Both graphs show the same 10 simulations winning 55% of your bets. One graph is the first 1,000 bets and the next is same simulations for 10,000 bets. Look close and you can see the first 1,000 are the same in the 10,000 graph. Stupid variance
5
14
57
Remember when stock brokers used to be rich in the 80's? That's where we're at with sports gambling. Story time!
Before the interwebs, doing a stock trade cost about $100 and retail investors had to go through a broker. A stockbroker churning 100 trades a day was making real
13
6
57
In response to egregious cashout offers, we made a calculator that shows cashout EV and whether you’re better off betting the other side instead if you really want out of the bet. It’s outrageous that cashouts are often worse than betting the other side.
5
1
46
Here are distribution curves comparing the expected outcome of 1,000 bets using Kelly Criterion versus always betting 2% of your bankroll. Full Kelly gives the possibility of enormous profits at the expense of wild $$ variance.
6
7
45
Here are 1,000 bets simulated 10 times with a 55% win rate. The outcomes here are pretty close to what you'd expect. You have a 90% chance of making $1,300 or more, a 50% chance of making $5,500 or more and a 10% chance of making $9,700 or more. There's a 4.6% chance of losing
5
4
42
Variance doesn't only apply to your bet tracking. For sports betting model builders, it also applies to player/team stats. Here are a couple examples:
1. A batter that truly hits .250 has a 25% chance of getting a hit per at bat (obviously). 100 at-bats into the season, there
4
2
42
If you're a recreational bettor, the juice (-EV) of parlays is insane. But, the opposite is true as well. If you really have an edge, the ROI of parlays balloons with EV and number of legs... as does variance.
This is intuitively true as well. Imagine you magically knew of
4
7
40
This quote from
@Gingfacekillah
when asked if markets are perfectly efficient is actually not overstated:
“This is such an annoying argument that I’m so tired of hearing… I really think it’s starting to waste people’s time on earth, which I think is kind of precious.”
If
4
4
37
Returning to my series on ways to mathematically make money betting on sports... long post!
Unlike the methods in the last two posts, "Line Comparison" betting is easy and takes no specialized knowledge. It also definitely works. Right now, it’s most associated with
@OddsJam
8
6
36
The chart below shows the results of playing every NFL home favorite vs. playing every road dog back to 1977 including this year to date. The combined record of the road dogs is 3672-3421-194. The ROI on road dogs is -1.1% versus -7.7% for home faves. If the lines were perfectly
4
4
35
Caesar’s live lines are tricky to beat for US sports. FanDuel isn’t as square as everything thinks all around. Those two are almost in their own catagory, in between Bookmaker/Pinnacle/Circa and all the others
7
2
32
Here's another way to visualize Kelly Criterion vs fixed % of bankroll vs fixed $ amount wager. 50 simulations of 1,000 bets below. The randomized win/loss sequence is the same for the 3 methods in each sim, only difference is bet size. 50 seconds long. Peep the vertical axis!
4
6
32
I lied... another variance post. Variance increases with odds even as your ROI stays the same. As odds get longer, the percent of bets you have to win goes down of course. But, the odds that you've lost money after X number of bets increases.
4
1
31
This was a really great episode of
@robpizzola
@CirclesOffHQ
podcast, interviewing Matthew Trenhaile who is Head of Growth Sports at
@Pinnacle
... lame that it was only an hour and a half
3
7
31
Method
#2
to beat sports betting, what I call Historically Mispriced Scenarios. Here’s an example: In NCAAF, betting on an unranked team against a ranked team that lost their last game by 14+ points has a record of 117-65-3 since 2007 with an ROI of 24.1%. The market has
5
4
30
Billy Walter is betting the Chiefs small ($500k to $1mm) and explains his reasoning in a rare interview here. Good luck to all
2
5
30
In the first 3 years of legal sports betting in Colorado, there were $12.1 billion of bets made. Gross profits to the sportsbooks totaled $817mm. Two things jump out here. First, look at the books' profit on parlays. Second, who tf is betting table tennis like this?
12
3
30
There are two main categories of methods to beat sports betting: grinding/scalping lines and originating bets. Originating can be data science, art or a combination… it’s almost always difficult and secretive. Grinding & scalping is simple math x speed x time spent.
@oddsjam
4
4
27
@jasonfurman
She did no such thing and you’re abetting her. The president of the soft and sorry home of safe spaces and so-called anti-fascism suddenly calls for a wide-birth for freedom of expression, won’t say that neo-Nazis aren’t welcomed or that calling for the genocide of Jews is
3
0
27
I’m really amazed by some of the year end posts from the +EV community. My New Year’s resolution is to spend more time calling out the frauds that refuse to post a
@betstamp
or
@pikkitsports
. My 2024 dream is to find more people trying to straight up beat sides and totals.
4
2
26
This is an incredible tome. Massive amount of work to write this. About 300 of 400+ pages are a meticulous analysis of hundreds of random sports betting touts and there are a lot of great insights throughout that portion. No offense to Mr. Buchdahl, but it’s not exactly a
3
0
26
Unpopular opinion, but my mindset has evolved. The recreational books are great, despite their limits. No one could offer their betting menus without limits.
@AlexMonahan100
would be a billionaire by now… compound returns and all that. I wish they were more upfront about it and
6
3
25
My day job centers on structured credit bonds (think The Big Short). It is very seductive to try to math/model them to death because they beg for multi-regression, option adjusted spread, Monte Carlo, etc. In my experience, those traders can’t see the forest for the trees.
My
6
3
26
Back-testing sports betting theories requires huge databases that can be quarried. Baseballs retain humidity for some time and I had a theory that totals weren't accounting for this, so I back-tested it.
You can use our database for free:
5
5
29
@blknoiz06
My friends brought seven black women to my club and didn’t ask me why I talked to them
3
0
25
Here's a chart showing the likelihood of making less than X% ROI after 1,000 bets at +100 odds based on your EV with flat stakes. How you read this graph is shown by the arrow/text. Notice that each EV line crosses 50% probability at their respective ROI.
@SmokeTheBooks
Another interesting chart. Thanks.
One should keep in mind that "Odds of making money" would refer to (I assume) being +$1 or better.
So a bettor with 75% chance of "making money" would be more like 65% chance of making more than a few units.
1
0
1
3
8
25
Hey
@BarstoolPrez
, what kind of a “sportsbook” are you running over there? Max bet wont buy my family Chic-Fil-A, but you’re chirping
@AlexMonahan100
. So lame…
5
4
21
This graph is the math answer to this Kelly Criterion question below. Full Kelly is mathematically sound and leads to the highest probable profit over large sample sizes. Emotionally, it seems way high to a lot of people (myself included) and the variance is hard to stomach. The
@SmokeTheBooks
based on this, would it make more sense to QK into +300 odds and lower then 1/8 kelly anything above +300? (numbers are just examples but hope logic makes sense)
0
0
1
5
2
22
Long Tweet: I’ve been trying to think of easy ways for ordinary, non-sharp bettors to win betting on sports without any special knowledge or cost (or at least not lose). This one isn’t perfect and I don’t know for sure how well it will work. I don’t know why I write these posts
6
0
22
I believe this is out of print, but still available on Kindle. It’s a great gambling story and I highly recommend giving it a read.
6
1
22
@UmpScorecards
Does your website show Called Ball Accuracy and Called Strike Accuracy or is it just on the Twitter infographics?
PS: Thanks for making an incredible and free website and Twitter... this is so cool for baseball fans.
2
0
21
This graph shows variance by EV using 10 simulations of 1,000 bets each at 0%, 2.5% and 5.0% EV. They are working themselves out, but after 1,000 bets they are not close to organizing themselves by color correctly.
0
3
21
That was an annoying 6 days of red ink. MLB is still mistreating me. WNBA has been ok, but not a lot of games. Our MLB data is almost complete, and then we’ll focus on all the soccer leagues. MLB-only beta website almost ready. Retweet for free login.
0
17
16
I doubt many will get this, but here’s the actual key to
@oddsjam
. The math isn’t complicated… it’s like 7th grade level and it works (post coming trying to simplify this). It’s access to the data in real time that is so hard to replicate. We’ve tried to do it on a small scale
6
0
19
@JasonDRobins
@DraftKings
Please stop limiting wagers. You advertise that we can win so much, but if you win you get severely limited.
I’ll help you set lines if you want, but don’t post lines and only let me bet $7.12 on them
0
0
19
What percent of the money that FanDuel saved do they have to kickback to the NBA and who gets it? The league? The Nets? Jacque Vaughn? Dennis Smith Jr?
It’s illegal, they’d go to jail, end their career and they’re already rich so they’d have to be paid a lot!
How many people
Let’s say this was true….
Dennis Smith Jr is on a vet min of 2.5m/ 1 year…. He was playing the best game of his season and NEEDS to show teams he’s good
He was “done for the night” with 14 pts/ 7 rebounds/ and 8 asts
Most people needed:
15 pts
8 rebs
10 asts
20
3
53
2
0
19
I switched this from the odds of losing money to the odds of making money, but here's an animation that shows the likelihood that you're profitable after 1,000 and 2,000 bets based on the average odds of your bets and your EV.
2
5
19
This simply isn’t true. I’m friends with a dozen current and former pro athletes and coaches. They could give a flying fuck about spreads. They care about winning and their next contract. There’s no secret red phone between the Nets and DraftKings.
Serious question
What’s the difference between insider trading and NBA teams having PARTNERSHIPS with Sportsbooks and the books CLEARLY having information not made to the public
Why is one illegal and the other legal
91
271
2K
13
0
19
The point here is that even with a spectacular 55% win rate, the range of outcomes after 1,000 bets is huge. It takes A LOT of bets to know how good you really are. It can seem like you're hot or cold, but you aren't... it's just variance.
0
1
18
I had friends over tonight in Denver and we all decided to put in my buddy’s player prop parlay. Even though I’d estimate this bet at -25% EV, this is how it went. The current state of legal gambling is gross. Exceptions are
@CircaSports
,
@CaesarsSports
and
@sporttrade_app
3
4
18
Respect to
@CircaSports
for keeping the old tradition of dime lines in baseball alive (with $20k limits)
1
0
17
There is some overlap amongst the sports betting community (esp +EV crowd) and stock/option trading crowd. I have some thoughts.
1. The stock market isn’t a zero sum game. They’ve gone up over time forever. Throw darts at the S&P 500 names and you’ll make money over time
2.
3
1
17
I believe there are some basic ways to improve your +EV betting. Here's the simplest and most obvious example in the world. In the NFL, home teams and favorites have been consistently overvalued forever. They've covered 47.7% of the time over the past 20 years with an ROI of
4
1
17
@JasonDRobins
@DraftKings
When your day comes, you gonna look at Jesus and say you were in the entertainment business? You’re a money changer who bans anyone that is good at it or gets lucky. Stand by the lines you deal and stop advertising that you can make so much money because it’s a blatant lie.
0
2
16
Fandual actually lasted quite awhile for me, but now it’s totally gone. I still have a few left in Colorado that I haven’t used, but more and more I’m going offshore or Circa (Caesars still alive too).
3
0
16
It is now fairly well known that home-field advantage in the NFL is not worth 3 points. In the past 20 years, the home team won by 2.2 points on average. Exclude 2020 Covid and its 2.3 points. The chart below shows actual average winning margin versus home spread margins.
4
0
17
Square books rooting Niners and sharp books rooting Chiefs. DraftKings has spent untold millions curating a customer base that sucks at gambling!
Good morning world! 👋 Heading into the office after a helluva busy night at
@CircaSports
. Niners by 1 is the nuts but otherwise we are rooting for Taylor Swift erm I mean the Kansas City football Chiefs…as of now. Long way to go and lots can change when you take 500k a click 😳
38
13
159
1
0
15
This is exactly correct. The books’ interests are perfectly aligned with the states’ because taxes are a function of net gaming revenue.
@d_feustel
Legalized sports books are a scam, and the states are willful partners in limiting sharp bettors.
0
2
13
4
0
15
Our model shows history of betting opening or closing lines. When we build these, we look for consistency of the lines because it’s a tell that you’re on to something. Usually you see it. When you don’t, it’s concerning. This one is betting on MLB pitchers with bad ERAs facing
3
1
14
This is NFL visiting dogs of 6.5 or less since 2018. Whether you played at the open or close was random until November 2021, then it diverges and playing the close is much more profitable.
6
1
14
Wild they’re this far apart. Also seems like DraftKings always pushes favorites in NFL, especially shortly before kickoff.
2
0
14
I don’t normally play parlays, but Caesars had a parlay boost that made throwing darts EV+
🤑🤑🤑
0
0
12
Some simple NFL cover stats past 20 years:
Dogs - 51.0%
Road dogs - 51.8%
Road dogs w/ <30% record - 54.4%
Dogs off 20+ point loss - 56.5%
Dogs off 26+ point loss - 63.6%
Dogs that scored <10 points last week - 55.8%
Road faves off a bye week (13+ days) - 64.5%
0
1
13
Umpire report from Rockies home opener, that somehow still went over 13. I’ll be watching this ump to play unders. This is really unusual… guy might cause electronic balls and strikes by all by himself.
1
1
13
PSA: This should be shared far and wide with recreational bettors, gambling addicts and the degens over on Reddit. They may not care but at least they'd know, as I'm sure the vast majority of my followers already do
1
6
13
I spent 3 minutes making a matrix of estimated odds of the final score. This is read as "1 in _____" odds that a specific team will win by this score. Small value on these 3, but I'm tailing
@grpwins
because karma
KC 34-10 +40,000
KC 27-7 +50,000
KC 27-10 +27,000
0
1
13
Deeper dive into my last post on "trends" here. NFL home teams and favorites have been overvalued in the betting markets for decades. In the last post, I mentioned that playing all NFL home favorites since 1977 gives an ROI of -7.7% while playing all road dogs gives an ROI of
1
1
13
Crossed $1mm in bets since I started in October. OddsJam tracker lost $15k of profits at PointsBet a few weeks ago, so ROI closer to 5.5%. My main strategy is not Positive EV (scalping lines), which is why I don’t beat CLV nearly as much as OddsJam community. Let’s go!
3
1
13
We track all NBA and NHL team flights. We moved those dashboards to a public website and hope this is helpful to some of you. This is the entirety of the website. Happy Holidays.
1
1
11
NFL spreads are pretty accurate. Here's a graph comparing closing spreads vs actual average winning margins. My dad always told me to check my work before laying 2.5... maybe he was right.
Dots would line up better with bigger sample size, especially past 7.5 where there just
4
2
12
Incredible article here. Sharp gambler using government run sportsbook in DC causes “Emergency action necessary to promote the immediate preservation of health, safety and welfare of DC residents.” That’s the language they used to limit him.
3
1
12
We use a bet screening tool for a variety of purposes. We moved it to the internet & it's free. 100 users for now & if it doesn't break, we will add more users.
Please don’t use it to build stupid “trends” or “systems”
3
0
13
I put a new baseball in a container with 85% humidity for 2 days and took it out to see how long it’s take to dry out. Weight decreased by 0.5% in an hour and 3.21% in 24 hours. Obviously, I’m doing this for a theory on betting on totals.
1
0
11
Do +EV bettors think that lines are sharpened by a bunch of PhD physicists from MIT with supercomputers? I glean this is a contentious issue on gambling twitter.
7
0
12
“I have dreamed a dream, but now that dream is gone from me.”
Story time!!!
Sportsbook related eventually…
In Colorado we were first to legalize weed. Some thought it was a great business opportunity because drugs sell themselves and margins were huge, both of which were
2
3
12
"In 2022, New Jersey sportsbooks won $450.4 million off $2.4 billion in parlay bets"
That hold boggles the mind... stop making hugely negative EV parlay bets people!
0
2
11
I get the hyper-focus on CLV. If I could think of a better model to beat CLV than the tools that are out there, I'd thoroughly enjoy building it. I grind lines when I'm bored too. But you're beating the market because others are sharpening it. That sounds like a challenge
1
1
11
NFL unders are getting ridiculous this year at 83-53-1. Even excluding this season, simply playing all unders has a slight positive ROI over the past decade at 0.90%. Playing them at the open gets it to 1.97%.
2
2
10
This is an historically mispriced scenario out of our software based on one single thing going back to 2010. These things can stop working at anytime as others discover them, but look at that beautiful straight line. One. Single. Thing.
1
0
10
Why don’t all of the sportsbooks just instantaneously copy
@Pinnacle
lines? It’s not like they haven’t thought about it. I’m genuinely curious what you all think
13
0
10
$245 billion is the handle, not the hold. Gross profit <10% of that. Net profit is ~$0. Gross is about half of what we spend at the movies (inc. Covid years). Gambling addiction is concentrated and very serious but no one bitches about movie revenues.
2
1
10
@BeauLWagner
Why does fanduel let you bet $2k or $5k on a player prop? They don’t let me do that.
4
0
10
The odds of a fan catching two balls on back-to-back pitches aren’t nearly as low as you think.
There are about 300 pitches per game and about 20 of them are hit into the crowd (1 out of 15). Average attendance is about 30,000, but a fan sitting low and not behind the safety
Me: i've been to hundreds of games and never caught a foul ball 😢
This guy: I just caught 2 on back-to-back pitches 🤷♂️
86
628
6K
1
1
10
Is there any conceivable reason why favorites would win at such an outsized rate in English Premier League in weeks 25-31? Last 22 seasons covering over 1,500 games and an ROI of 9.0%. If this is a coincidence, it's a really weird one.
3
0
10
Not sports gambling, but since media coverage of the banking crisis has been so terrible, I thought I'd share what really happened:
1. Covid stimmy generated the largest inflow of deposits ever... many banks grew deposits 50% or more when interest rates were at about 0%. Deposits
1
3
9
@Hitman428
@JasonDRobins
@DraftKings
This is 100% right. No doubt that 90%++ of the people with severe limits are squares that hit a lucky run. No honor when you’re advertising that you can make so much money.
0
0
10
I got a free Norwegian cruise, which should come in handy never.
4
1
10
This is the Kelly Criterion in bombastic terms
Would you rather have:
* 2% chance to make $1 million (EV is $20,000)
* 80% chance to make $5,000 (EV is $4,000).
7
0
9