Tuesday, 22 November 2022

USA Sports Betting, MLB 2022 and Marital Discord

The 2022 MLB season is over with the Houston Astros winning the World Series by four games to two over the Philadelphia Phillies. It was an interesting season, with the Phillies reaching the final despite finishing third in their Division behind two teams winning over 100 games. The Astros were the best team in the American League though, so their success wasn't such a surprise, but with four teams winning 100 games for only the second time in history, and the New York Yankees only one short on 99, it's not surprising that there were a record number of 'hot' favorites this season.


Unfortunately the strategy of backing such teams in the regular season wasn't a winning one this year overall, except in inter-league games. The universal adoption of the Designated Hitter (DH) rule may have been disruptive, but this doesn't explain why American League (AL) clubs playing intra-league games would have been the worst performing types of game.

Here are the numbers going back to 2013, and note that in 2020 the DH rule was also not in effect:
Excluding the 2020 COVID season, the number of 'hot favourites' (shorter than 1.5) has doubled in the past five years as parity between the teams has dropped but of the five teams winning 99 or more games this season, only backing the Yankees and Atlanta Braves when hot favourites was profitable.

The majority of selections are Home teams, but ROIs have been slightly higher historically on Away teams with that for the 2013 to 2021 seasons (excluding the COVID 2020 season) at 8.9%, but this season the ROI on Aways was -10%.  
The above image shows the results by League, and while it might not be a surprise to see the National League disrupted in the two seasons with no DH rule, the season just ended was the first where the American League selection has lost (other than that very small loss in 2016). 

One other possible explanation is the huge increase in sports betting in the US in recent years. Just five years ago, Nevada was the only state with legal U.S. sports betting but a Supreme Court ruling in May 2018 gave states the power to legalise and regulate sports betting. It's a constantly increasing number but more than 30 states have legalized some form of U.S. sports betting, with likely more to come.

It could be a blip of course, but as markets open up, and billions of new dollars pours into markets, of course there is going to be some disruption. I'm in California at the moment, (it's Thanksgiving this week, although my in-laws are both currently sick and turkey plans may well be disrupted), and today's New York Times had a feature on the topic: 
Four years ago, betting on live sports was illegal in most of the United States. Now, fans watching games or attending them at stadiums are barraged with advertisements encouraging them to bet on matchups, not just watch as spectators.

This transformation in sports betting started nearly a decade ago, at first with the explosion of wagering on fantasy sports. Then in 2018, the Supreme Court cleared the way for states to legalize wagers on live games. Today, 31 states and Washington, D.C., permit sports gambling either online or in person, and five more states have passed laws that will allow such betting in the future. Professional sports in the U.S. now are part of a multibillion-dollar corporate gambling enterprise.

This shift represents the largest expansion of gambling in United States history.
At least $161 billion in wagers have been placed since sports betting was broadly legalized in the United States. This explosion of gambling is just the start. Betting companies have made clear that the ultimate goal is to bring so-called iGaming to states across the nation, where customers can use their mobile phones to play blackjack, poker and other casino-style games.

California is one of the minority of states with no legal sports betting, but TV shows now openly discuss odds, lines and totals, and 'bad beats' are often featured and while sports betting has always been a part of life here, it's never been so openly discussed. 

When the last World Cup was played, just three states (Nevada, New Jersey, Delaware) had legal sports betting markets, but for the current one:
20.5 million American adults plan to bet $1.8 billion on the 2022 World Cup, per a new report from the American Gaming Association. Three in 10 (29%) of American adults who plan to watch the World Cup intend to wager on the tournament.
Of course, not all the money is too sharp:
If given $50 to bet, most Americans would put their money on the U.S. (24%) to win...

Systems that have worked well for several years in the NBA and NHL are also off to poor starts this season, but touch wood, American Football seems to be its usual self. It's unlikely I'll get to see a game live while I'm here, but with Friday being a holiday for most Americans, and the game starting at 11am local time here, bars should be lively for the US v England game. Whether or not my wife will still be talking to me afterwards remains to be seen. 

Thursday, 3 November 2022

Stupid Small Eyed Pig

Backing the 'Dog in the MLB Playoffs had another winner last night, with the Houston Astros 'officially' winning at evens but backed at 2.06 on Betfair, following on the heels of the Philadelphia Phillies win on a delayed Tuesday night game three at an official 2.05 but again, backed higher on Betfair at 2.16. Betfair take their 2% commission so we're not comparing apples to apples here, but 2.16 is equivalent to 2.137. 

With a maximum of three games left, this post-season is guaranteed to end up in profit. I'll look at the MLB season in its entirety later this month. It's been an unusual one with the significant rule changes, but it's nice to end out the season positively.

With the series tied at two after a rare no-hitter last night, tonight's Game Five is hugely important, with the Phillies a home underdog at 2.46. 

With my wife away, I've had some extra time on my hands, and with excellent timing, I've been entertained by a crypto scammer in recent days. It started with a text from an unknown number addressed to "Kevin" referencing interest in a property deal, and asking for the contract. 

Being a decent chap, and assuming positive intent, I responded clarifying that I was not Kevin. I'd hate for someone to miss out on an opportunity because of a genuine mistake, and the reply stated that "my assistant saved the wrong number digits" adding "I hope I didn't cause you any trouble".

"None whatsoever" I replied, to which I was told "You are very gentle... Let's make a friend, How old are you?"

And so it began. By nature, I'm a sceptic, and 99.9% sure this now wasn't a genuine wrong number, and sure enough within a few messages my new friend was saying how they were "analysing the crypoto currency market" in their spare time.
"My aunt is a finance expert from a Japanese university, majoring in financial market analysis, quantitative investment and sound asset management, He (sic) is trading short-term Bitcoin call options. She will notify me in advance every time there is a good market quote, so I have been able to make rich profits in the cryptocurrency market for the past two years.
The target for next year was $800k. 

The two years later grew to four, with 3 to 4 trades per month on average and when pressed on the strike rate and position size was told 100% and 30% respectively.
Not a single trade of a total of approximately 168 made had ever lost. Very impressive? 

Even more remarkably, the profit on each was 40%. When I pointed out that this seemed rather unlikely and that with compounding they would now literally be a billionaire even starting with a very modest bank, I was told that I'm old and don't understand trading.
Well, I am certainly old, although I prefer the term 'experienced' and after being told I was a 'stupid pig' with an 80s / 90s mindset my new friend is no longer reaching out to me every hour to see if I've opened an account yet. Hopefully I wasted plenty of their time, and saved someone else from potentially losing money. 

Tuesday, 1 November 2022

October in Perspective

It seems that when writing a weekend, weekly or monthly review that I often use the phrase "good in parts", with my first reference to the curate's egg all the way back in January 2010, and it's hardly surprising.


For those of us who have a number of systems or strategies, it's to be expected that over any given short-term period, some will be up and some will be down. If each individual system has a positive EV, the overall trend will be positive, but the more systems you have, the rarer it becomes for a month where everything was up, or down for that matter. 

While the big picture financially last month was positive, a welcome change to how much of the year has been going, the sports investing division of Cassini Holdings wasn't the cash cow I was hoping for, with an 'official' loss of 10.77 units, a negative ROI of -3.8%.
The EPL Draws extended their losing run to seven, with the one selection this weekend (Brighton and Hove Albion v Chelsea) failing. Longer losing runs when you're backing the draw are to be expected though, and hopefully the record of 15 will not be broken this season. 


The two seasons that started last month (NBA and NHL) haven't started off well, and while it's always a concern that something has changed in the close season, I've not seen anything along the lines of last year when NBA rule changes directly impacted the number of points being scored.

Hopefully with a few more matches in the ledger this month, things will have settled down. Handling losing runs never ceases to be difficult, but looking back at past results is always helpful, and implementing a stop-loss after a drawdown is one strategy where you reduce or even stop betting and just monitor results until your confidence returns. 

Overall though, October was a good month, the second best of the year after March, but still down almost 13% from my high, and 9.2% down year-to-date. The major indexes all had good months with the FTSE100 up 2.9% but the S&P 500 was up 8% and the narrower DJIA almost 14%. It also helped that my company's share price hit a new all-time high on Halloween, with those stock options nicely in the money.

It was also a good month for exercise and weight. I covered 226.5 miles on foot, slightly fewer than September, but well on track to average 5 miles a day for the year, and despite my wife going off on holiday, I was more self-disciplined than I thought and only consumed alcohol on three days resulting in a losing month on the scales. 

The worst part of the month by a long distance was losing one of my dogs, which put into perspective a few losing bets. She was 12, had congestive heart failure and was starting to suffer, so we had to make the tough decision to put her to sleep, and it was a lot tougher process to go through than I thought.

An animal rescue charity in the US runs a 'poorly drawn pets' fundraiser where you send in a photo of your pet with a donation, and in return get a drawing. I was rather pleased with mine:
If you're not a dog or pet owner, you won't understand how upsetting the loss of one can be, but I have two more with me who are probably wondering what the heck is going on after my wife has also disappeared from their lives, although she should be back this weekend.

Life goes on, and here's to a winning November, with a major football tournament to turn our attention to.

Wednesday, 26 October 2022

World Series Miscalculation

In baseball, the World Series is set with the Houston Astros to play the Philadelphia Phillies in the best-of-7 series starting on Friday. 

The Astros, last season's runners-up, are the top seed in the American League and have so far won every one of their seven playoff games, while the Phillies are the number six seed (i.e. lowest) in the National League but have won 9 of their 11 playoff games. 

Not surprisingly the Astros are favourites to win the series at around 1.57, but worth noting that backing underdogs, particularly at home, has again been profitable so far this post-season. 

In the last ten years of playoffs, Home 'Dogs have an ROI of 9.5%, although this includes games played on neutral grounds during the pandemic where the team was designated as Home based on seeding and batted in the bottom of the innings. 

David Purdum wrote an interesting article about a supposed odds calculation error by Sportsbook BetMGM who for two weeks early in the season offered 2,500-1 on the Astros defeating the Phillies in the World Series:
A fortuitous bettor in Colorado is in position to win big on the World Series, a potential score magnified by a bookmaker's miscalculation.

On April 13, one week into the Major League Baseball season, a bettor in Colorado placed a $50 futures wager with BetMGM on the Houston Astros to beat the Philadelphia Phillies in the World Series at 2,500-1 odds. Nearly seven months later, the Astros and Phillies are in the World Series, and the bettor is position to win $125,000, if Houston prevails. The bettor also could lock in a profit by placing a hedge bet on the underdog Phillies. A $10,000 wager on Philadelphia at current series price of +160, for example, would secure a $15,950 profit if the Phillies pull the upset or a $115,000 payday if the Astros win off their original $50 wager.

The bettor declined interview requests but provided some quotes to BetMGM.

"It's been surreal living and dying with the Phillies from April all the way through 'Philtober,'" the bettor said. "Shoutout to the Astros for taking care of business, and to the Dodgers for choking as usual."

The bettor added, "I'm not hedging."

Regardless, the bettor is in for a World Series sweat that could be more lucrative than it should be -- because the odds on the wager were longer than they should've been. Instead of 2,500-1, the odds on the Astros' win over the Phillies in the World Series should've been closer to 250-1, if not shorter.

The most basic way to create odds on an exact outcome of a World Series, months in advance, is to multiply the projected champion's odds to win the World Series with the other team's odds of winning the pennant. At the time of the bet, the Astros were 10-1 to win the World Series, and the Phillies, despite a slow start, were around 10-1 to win the National League. Using the traditional method, the odds would've been around 100-1, not 2,500-1. Different approaches might have produced longer odds but unlikely anywhere close to 2,500-1. In comparison, the odds offered on the Texas Rangers beating the Miami Marlins in the World Series were also 2,500-1 at BetMGM.

"We were probably a little bit of aggressive on those," a good-spirited Jason Scott, vice president of trading at Bet MGM, told ESPN, acknowledging a misstep in the oddsmaking process.

Egregious odds errors, often referred to as palpable errors, can be a controversial topic in the betting community. Data entry errors or typos can cause sportsbooks to post the wrong lines, sometimes making a big favorite an underdog, for example. Sportsbooks often include stipulations regarding egregious odds errors and, in the past, have fought from having to pay out on bets made on the bad lines. Scott, however, made no indication that BetMGM would go that route.

BetMGM offered the 2,500-1 price for two weeks but only took six bets on an exact Astros-over Phillies World Series outcome. Scott said the $50 wager accounted for "about 90%" of the total amount bet on the inflated odds.

"I'm more worried about Mattress Mack beating us than him, to be honest," Scott said.

Jim "Mattress Mack" McIngvale, a Houston furniture store owner known to place big bets to mitigate risk from promotions, placed a $2 million bet with BetMGM on the Astros to win the World Series. In total, McIngvale has approximately $10 million riding on the Astros in the World Series with a chance at a $75 million payout. He'll need it: After placing his seven-figure bets in June, McIngvale offered to refund customers who spent $3,000 at his store Gallery Furniture double their money back if the Astros win the World Series.

"It's not pleasant, but we can deal with it," Scott said of the liability caused by McIngvale's bet on the Astros.

BetMGM said it will be sending the Colorado bettor to Friday's Game 1 of the World Series in Houston.

A glitch in the Matrix on Monday night caused some confusion for a couple of subscribers, with the Killer Sports site throwing up qualifiers for one of the NBA systems who were most definitely not qualifiers. The problem was soon rectified, but it was a reminder that it's always a good idea to double-check the data before investing any money. 

Results over the weekend were mixed. With the end of month approaching, I'll write up a summary at that time, but the EPL Draws had a blank weekend with the two losses extending the losing run which started in midweek from 4 to 6, while the Segunda (+0.77) and Bundeslayga Systems (+2.13) were both profitable.

My findings on the 'seeded and rested' teams that I've mentioned here recently has sent me down a bit of a rabbit hole. After looking at seeded and rested clubs in the National Leagues (the English football ones, not the MLB one), and rested, but not seeded, NFL teams coming off a regular season bye week, my interest in time zones was rekindled. Some of you may remember my NBA findings on Eastern teams playing in the West published in 2019, but most of you likely won't. 


Saturday, 22 October 2022

Equinox Knocks

There was another Sports Equinox in the US on Thursday, but it wasn't a profitable one. The Thursday Night NFL game went Over by a mile, and the 2.5 points the New Orleans Saints were getting was nowhere near enough.

On top of that, the two NHL selections both lost, but at least the Los Angeles Clippers picked up a win on the road. Well, technically on the road, since the win came in their home stadium which they share with the Los Angeles Lakers.

A profit yesterday from the Bundeslayga though, and the NBA's two systems were both in profit.  

Regarding my observation that top clubs resting on a bye historically underperform in their next game or series, Can't Win With Kids took the idea down a different path asking the question:
Were there not an odd number of teams in the National League when Macclesfield went bust a couple of seasons ago, therefore one team per week who effectively had a bye week, may be worth looking at especially with Saturday-Tuesday-Saturday game weeks.
There have been quite a few leagues played with an odd number of teams in them, going back to 1893-94 when Division Two had 15 clubs.

Why does this season sound familiar, I hear no one ask. Well, I can tell you. It's because Small Heath went the entire season without drawing a game.  This wasn't a first however, with two clubs - Sunderland and Aston Villa - managing this feat in the Football League of 1891-92, and something not repeated in the top division since Stoke managed it in 1895-96. 

While I'm on the subject, I should mention that the last time a top league had a club with no draws in a season was in 2009-10 when RKC Waalwijk finished bottom of the Eredivisie without a single draw.

Back to the odd number question, and other leagues with an odd number of clubs are:

1931-32 Division Three North (Wigan Borough resigned after 12 games and had their record expunged)
1961-62 Division Four (Accrington Stanley resigned after 33 games and had their record expunged)
1987-88 Divisions One and Two had 21 and 23 clubs respectively
2019-20 League One (Bury expelled without playing any matches)

The season Can't Win With Kids is referring to is the 2020-21 National League season when Macclesfield folded before they had played any matches, and later saw Dover Athletic refuse to play matches from January before they had their results expunged in March. 

The 2021-22 National League season was played with 23 clubs, and Dover Athletic were relegated after having 12 points deducted for sulking the previous season, while the National League South also had 23 clubs that season, which means that there have been several years where clubs have had a bye week.

While finding the Match Odds data from 1893-94 Division Two might be rather challenging, the data is there for those more recent seasons, but it's a relatively unusual scenario so I'm not sure it's worth trawling through the data for something that might not happen again for some time.

These 'ordinary' games remove the element that the rested team is always the, on paper at least, stronger team, but if we're looking at this scenario, we have some good data from the NFL where teams are given bye weeks, so for example this weekend is Week 7 of the season and:
Unfortunately, historically there is no edge here, with an almost 50/50 record against the spread:
 

Wednesday, 19 October 2022

Football Byes, Lost Games and NBA Tanking

After writing my last post on the topic of whether a bye is the advantage it's often made out to be, it occurred to me that in the English National League since 2018, they have had a six team play-off format with the top two seeds having a nice long rest before hosting the winners of the Quarter-Finals in the Semi-Final round. 


An interesting case study and while we should have five seasons of data, COVID hit the Northern and Southern sections in 2021 and there was a problem with Concord Rangers in 2019 who, after finishing sixth, were barred from the play-offs due to failing ground size regulations.

Probably more disappointing for that club than for me, but 25 games is better than nothing.

Of those 25 matches the rested, higher seeded club won just eight, with eight more games Drawn and the other none going to the tired and supposedly weaker side, and laying the home club in every match has an ROI of 37%, +9.36 units, and in matches where the Home side is odds-on the ROI is 70% (+9.85 units from 14 matches). 

I think I should add this to the 'Sacred Document' and hopefully get more comments like this one after the last update:
Many thanks for this most excellent quality after sales service!
We aim to please. 

How did the weekend go, I hear you ask? 

Overall not bad at all. American Football made us 3.71 units from 7 College and one NFL selections, while the EPL Draw System found two winners from two selections for a nice 4.85 unit gain. 

Not so good in Spain or Germany where the systems combined for a 4.05 unit loss. 

Backing the 'Dog in MLB playoffs over the weekend through to last night added 4.6 units to the pot, and while it is early days in the NHL, the system there is down 1.49 units after 10 games. 

Back to College Football, and in this post last week, I wrote about the differences that we see when sourcing our bets from different places.

This past Saturday, the 'official' system had 5 winners from 7 selections but Frédéric had twice as many picks and almost twice as many winners:
By the way I went 9-5 this weekend on the NCAAF. Maybe you see your selections just by clicking on "football" and going from there, vs me actually selecting each league (NCAAF, NCAAF FCS, and sometimes NCAAF added)? Still puzzling as to how we can come up with so many variations..

He very helpfully listed his 14 selections, and it's some good news that the official seven were all present, but what of the other seven extra qualifiers? Nowhere to be seen on Killer Sports. 

One such team was the Central Arkansas Bears who were on the road at the Kennesaw State Owls, and despite both teams being FCS teams (Football Championship Subdivision), the game is not in the database. 

Something appears to have changed after the 2019 season since a check on another team, Elon Phoenix, shows them 11 times in 2019, and just twice since. 

Honestly with so many games each week, seven is a lot easier to manage than 14, but if I am leaving money on the table, it appears I need to source my bets elsewhere. 

Unfortunately the Odds Portal site also doesn't appear to show any of those seven missing games, although they do have 48 matches which is two more than Killer Sports. Which begs the question, where are the markets for these games if they are not on Odds Portal? More to come I suspect.

The NBA season started last night, albeit with no official selections, and it's looking like a competitive season with no clear favourite, and five teams priced between 7.0 and 12.5 (Golden State Warriors, Boston Celtics, Los Angeles Clippers, Milwaukee Bucks and Brooklyn Nets) but with an outstanding prospect coming up in the draft (Victor Wembanyama), tanking may well be an area of opportunity this season.

This ESPN article explains it well, and this line was interesting:
...tanking teams have a bigger impact on in-play wagering and "will cough up leads at greater rates than typical double-digit dogs."

Over the past three seasons, the teams that ended with five worst records during the regular season covered the spread in approximately 45.7% of games, excluding pushes.
As for who those five teams might be, the article mentions the contenders for the worst in the NBA, stating that:
Five teams - the Oklahoma City Thunder, Indiana Pacers, San Antonio Spurs, Houston Rockets and Utah Jazz - have season-win totals of less than 25. That's the most in any 82-game season in the last 20 years, according to sports betting archive SportsOddsHistory.com.

I'll try to remember to check the records of these five ATS from time to time. 

Monday, 17 October 2022

Bye - Gift or Curse?

I touched on this topic in my last post, and it's not a new idea, but in sports where teams performing well in the regular season are rewarded with a bye, is this more of a curse than a gift? 


The argument is that teams become rusty whereas their opponents do not lose, and arguably gain, momentum by coming through that extra round or series.

The play-off format for MLB this season is new, but the Wild Card round has been around since 2012, though not in 2020 when COVID rule changes resulted in a 16 team play-off field with no byes.

From 2012 to 2021 (9 seasons excluding 2020) each League had one Wild Card game with the winner advancing to play the team with the best Regular Season record, so the number of extra rest days for the top seed was minimal.

Nevertheless, of the 18 series between the Wild Card game winner and the number one seed, the Wild Card team advanced 50% of the time, while winning 43% of individual games.  

The format this season changed with the addition of an extra Wild Card Team in each League, and instead of a single elimination game, the Wild Card Series games are now a best-of-three series, meaning the resting teams (now the number one and number two seeds) have a longer rest.

So the data for one season is obviously very limited value, but both the #1 Los Angeles Dodgers and #2 Atlanta Braves lost their series in the National League, while in the American League the #1 Houston Astros swept their series, and the #2 New York Yankees are tied at 2:2.

With the additional days of rest, and the ability of the top seeded team to have their best starting pitcher available, it's perhaps not too surprising that in the opening game of a Rested v Wild Card team, the Rested team prevails in 75% of Game 1 fixtures, before the rust settles in. After this, the ROI on backing the Wild Card team is a solid 18.8% since 2016. Unfortunately for 2012 to 2015 the Killer Sports data is incomplete. 

Another major sport that rewards its top regular season performers with a bye week is the NFL, and here the data since 2002 shows that the rested team underperforms with just a 36.4% record ATS. 

Something to think about in the New Year when the NFL Playoffs begin.