Saturday 28 September 2019

Week 4 Inefficiencies

There are seven NFL teams who have lost their opening three games of the season, and in Week 4 since 2002, such teams have a 33-20-2 winning record (62.3%) ATS when playing teams with at least one loss and one win. 

In Divisional games, their record is even stronger at 15-7-1 (68.2%), and the Washington Redskins (+3.0) at the New York Giants tomorrow are such a qualifier. They are also a qualifier for the Small Road 'Dogs system.

The New York Jets are on a Bye week, the Pittsburgh Steelers and Cincinnati Bengals play each other, and the other three worth considering - if you agree that the public underestimate 0-3 teams - are the Denver Broncos, Arizona Cardinals and Miami Dolphins.

In College Football, the Small Road 'Dogs are already 2-0 up on the weekend after Duke and Arizona State both won straight up on Friday night. Another five likely selections from today's matches hoping to improve the current 15-9-2 record.

Another huge favourite in the MLB today with the Houston Astros yet again (for the 11th time, Justin Verlander for the 6th time) starting at sub -400. The Astros are the only club to start this short on the road since at least 2004, and do so for the third time this season. The Astros lead the Los Angeles Dodgers by one game (and hold the tiebreaker) in the race for home field advantage in the World Series should both teams advance that far for a repeat of the 2017 finale.


Doubleheader Delay

For those of you hoping to cash in on the September Sweep System yesterday, you're going to have to wait for 24 hours after both games between the Detroit Tigers and Chicago White Sox were washed out.

Game One was called in the fourth inning with the Tigers 4-2 ahead, but a game needs at least five innings to count.  

With only two days remaining in the regular season, and four games still to play, MLB have decided that we will have only the one traditional doubleheader today, and both teams will end the season on 161 games rather than the full 162. 

Since neither team can move up or down in the AL Central Division, it's understandable, but two doubleheaders each day would have been fun. 

We did have a couple of T-Bone winners yesterday from three selections, and the only totals selection was a comfortable winner with 21 runs scored. 

At the start of this season, there was an article published by the Action Network which included this system:

This is the simplest system you’ll find for betting baseball. Take a bad team (won 40% or less of games) after a win. Why does this work? Recreational bettors think it’s unlikely that a poor team will win two games in a row, which leads to inflated lines.
The problem is that it actually doesn't work these days. This season's Money Line ROI is -16.4% following last season's -8.9%.

Another system was to back the Under when two winning teams play each other: 
Winning teams score runs, and the public likes to bet the over. So, if you have a pair of winning teams on the field, the oddsmakers will inflate the total expecting casual bettors to get down on the over.
Again, this would have not been a profitable strategy, with the loss at 3.8% this season although to be fair it was profitable in 2018 at 1.5%.  

Perhaps too many people read the article and took all the value out of the systems, but they don't seem to have identified market inefficiencies. I do find such articles interesting though, as they sometimes trigger ideas that do lead to finding inefficiencies, which after all, is what it is all about.

Friday 27 September 2019

MLB Regular Season Finale

There are now only three more days of the regular MLB season remaining, and once again, it has been a very profitable one. I'll update the final numbers next week, when the focus will be on the post-season.

The September Sweep System came up with another winner on Tuesday when the Washington Nationals swept the Philadelphia Phillies with the second win at -220 (1.45).  

The Houston Astros also won on Tuesday beating the Seattle Mariners at -480 (1.21) and another shorty won at the same price on Wednesday when the New York Mets defeated the Miami Marlins. Neither winner game up a run. 

In the NFL, things are going rather nicely too. After seven winners from seven selections on Sunday, the Philadelphia Eagles made it eight from eight for the Small Road 'Dogs System beating the Green Bay Packers last night. Another five likely selections from the remaining Week 4 matches, two of which are Divisional games.

The 2019 ROI is meaningless after so few matches, but an ROI of 7.5% over the past five seasons (plus the current one) should draw some attention.
The biggest favourites of the weekend are likely to be the Los Angeles Chargers (-16.5) who visit the Miami Dolphins who have lost their first three games by a total of 117 points. This is the shortest price the Chargers have ever been in an away game, and historically road teams against the spread are 16-11-1 when favoured by two touchdowns or more.

Currently eight games are in scope for the College Football equivalent, which currently boasts a 13-9-2 record. 

Tuesday 24 September 2019

Cole On Fire

For anyone still backing red hot MLB favourites, there's yet another sub -400 selection today, once again the Houston Astros and starting pitcher Gerrit Cole's fourth such outing having won all three previous opportunities and who recently got his 300th K of the season. Justin Verlander is just 12 strikeouts short of 300, and if he does make that number, it'll be only the second time ever that two pitchers from one club have reached this in the same season. 

For some historical perspective and to show how times are changing, from 2004 to 2015 there were just three favourites in this range. In 2016 there was one, and in both 2017 and 2018 there were three in each season. 

Tonight's Houston Astros v Seattle Mariners game is the 13th of the 2019 season, ten of which have been the Astros. Of the 22 matches, only 17 have been won for a loss of 5.1% while backing the favourite on the Run Line has a disastrous return of -21.5%

The edge in these games would appear to favour the underdog, but we're looking at a very small sample size. Tonight's game is a Divisional one, and here favourites do have an edge.  

The Extra Innings system came up with a winner last night (St Louis Cardinals) with another selection today, which along with four probable T-Bone selections, one Overs and the doubleheader game makes for a busy day.

While I was researching the doubleheader sweep / split post, I came across this related idea:
If a team has lost back-to-back games as a road favorite, there is a good chance they will break through with a win if they are favored in the next game. For example, if the Chicago Cubs are visiting the Cincinnati Reds and they drop the first two of a three-game set, there is a good chance they will bounce back and avoid the sweep. This system is harder to execute than the previous one because there are fewer instances where a road team loses twice in a row as the favorite and then enters the third game as a favorite as well. However, there are situations where it happens so it makes sense to track them and capitalize with the road favorites split.
This is certainly profitable with returns of 3.6% / 1.2% (ML/RL) but it's excellent in American League Divisional night games with returns of 23.9% and 29.8% respectively.   


Not as rare as the tripleheader, the last one being in 1920, the doubleheader is a relatively rare feature of MLB and is traditionally defined as two games played between the same teams on the same day in front of the same crowd, but which is more generally used these days to refer to two games played by a team in a single day, but in front of different crowds and not in immediate succession.

They typically occur later in the season, as they are no longer scheduled, but are the result of games being called off, most often due to bad weather.

In 1969, Michael L Goodman published an article in The American Statistician titled On the Incidence of Swept Double-Headers in which the author claimed that contrary to popular opinion, doubleheaders are more often swept than split: 
1969 was a long time ago, but over the last ten seasons, including the current one, I make it pretty much 50/50, with 130 sweeps to 134 splits, but there's an interesting, and perhaps not totally unexpected, tendency for teams to sweep in the later months of the regular season.

If the public perception is that it's difficult for a team to sweep, you might ask yourself how does backing the winner of doubleheader game one in game two work out?

Turns out overall, not too well, with a negative ROI of around 3.4%.

In September and October though, when for many teams the season is effectively over, and for others, they are fighting for a post-season place, it's a different story. 

Here, backing the game one winner in game two has an ROI of 25.7% (ML) and 24.8% (RL). 

When the team is the underdog, the returns increase to 46.8% and 27%, and when the 'dog has fewer wins in the season than their opponent, we're up to 82.9% and 40.5%. 

And when the 'dog is at home, the ROIs are a ridiculous 151.9% and 82.2% respectively, but we're also down to a very small sample size! 

There have been 31 doubleheaders this season so far, with another one today (Washington Nationals v Philadelphia Phillies) and one on Friday (Chicago White Sox v Detroit Tigers).  

Monday 23 September 2019

Seven Heaven

A perfect Sunday for followers of the NFL Small Road 'Dogs System, with all seven selections winning. 

Three of the seven covered by half a point - the Cincinnati Bengals and Pittsburgh Steelers were both getting 4.5 and both lost by 4, while the Baltimore Ravens were getting 5.5 and lost by 5 (after being down by 17 at one point). 

The New York Giants, Houston Texans, Detroit Lions and New Orleans Saints all won straight up to complete the sweep. 

The season record is now 12-3, with the record in Divisional games 3-1. 

Somehow I suspect the 56.2% ROI number may slip a little as the season continues, but hopefully at least some of you made a decent amount this weekend.  

Not quite such a good day in MLB, with the Extra Innings idea losing two bets from three but a winner on the T-Bone System (New York Mets). No action for either of the Totals systems.  

After six rounds, the EPL is starting to take shape, and a new Draw system I am following is so far looking very good, although I would have preferred a loss on yesterday's winner:
Only one losing round out of the six, but another ROI (62%) that is unsustainable.  

Sunday 22 September 2019

Extra Innings and More

Back from a week in humid Orlando, home of the University of Central Florida (more on that later), and to a rare, but much appreciated, donation to my retirement fund. Thank you Justin. Putting a monetary value on my invaluable advice is not an easy task of course, but either my retirement has been brought forward by at least 15 minutes, or I may take the donor up on his suggestion to have a few beers this weekend instead. 

Another area where the markets might be inefficient came to my attention  during the week. 

As always, these systems work because the public (and thus the market) reacts inappropriately to recent events, either under-reacting or over-reacting based on long established biases.

With the MLB season almost over, there's probably not much point in going into too much detail at this time, but the off-season will allow time to look into it further. 

A basic implementation of the idea, which is based on the premise that the impact of extra innings isn't always accurately accounted for, looks promising with the following results from the past three seasons:

The ROIs are even better at 21.0% and 17.6% over the same period when being selective about the type of match, and the one qualifier today is the Washington Nationals, with the Tampa Bay Rays and Baltimore Orioles worth considering. 

The more established T-Bone System continues to be a winner in September, not usually its best month, adding another 22.65 points to the season total, which is now the highest ever.
It's hard for me to understand why such an inefficiency persists in the market for so long, but I'm not complaining. There's a 1 in 3452 probability that these results are by chance but anyone following will be doing very nicely.

The Totals systems have had an unexciting month so far, with Overs having a 15-14-3 record, and Unders a 3-2 record.

In College Football, the perennially successful Small Road 'Dogs system (no losing records this millennium) went 5-1-1 yesterday and 13-9-2 for the season. 
Not everyone had a good Saturday:

Pittsburgh led 21-0 before UCF scored 31 points straight, before giving up a last minute touchdown and losing 34-35. While it is true that Central Florida has a great record ATS overall as a favourite, (22-9-2 since 2015) I'm not sure 1.21 was the best value.    

Early days still in the NFL, but the similar Small Road 'Dogs system here is 5-3 after two weeks, with five possible selections today. The record in Divisional matches is 3-1 but there are no qualifiers today unless the line on the Patriots v Jets game shortens considerably from its current 21 points! 

Lines of 20 points and over are rare occurrences in the NFL with just six such instances since the league re-organised in 2002, but today may well see two more with the Miami Dolphins getting 22.5 points at the Dallas Cowboys. 

Notable that on all six previous occasions, the 'dog covered the spread, but also notable that the Dolphins have lost their opening two games by 49 and 43 points.

Tuesday 10 September 2019

Dogs and Hotties by the Numbers

The (American) Football seasons are underway with a winner in the opening game of the 100th NFL season on Thursday night as the Green Pay Packers easily covered the spread against the Chicago Bears, and as it was also a Divisional game.

Selections for the basic Road 'Dogs system today were the Tennessee Titans, Atlanta Falcons and Buffalo Bills (in a Divisional game).

The Week One 'Dogs System was also profitable with an 8-6-1 record.

We had our first Division 1A Conference game in the College version of the sport on Saturday, although it was a loss with Stanford being thrashed at USC. 

In Conference games overall the Small 'Dogs system is 3-2 to the good after two weeks, and in all NCAAF games is 9-6-2. 

Overall a small profit on the weekend, but they all add up.

In my last post, I wrote about the over-reaction to a high score in baseball:
A similar over-reaction by the less sophisticated is seen when a team gives up a high (15 or more) number of runs. If the next game is a day game, the Money Line ROI is 14.5%.
Some of you are no doubt wondering if this inefficiency applies to the NFL, and are busily poring over the records. Well, more likely none of you are, and it's never crossed your minds, but I'll here to tell you the numbers anyway. 

When a team covers by 14 points (i.e. two converted touchdowns) or more in Week One, they are 18-29 in their next game ATS, and we have four qualifiers for next week.

As with baseball, in recent season it pays to back big favourites in the NFL, and while there were no double digit favourites in Week One, there are likely to be two next weekend. Over the last ten seasons, such favourites are 164-130-4 ATS. For teams favoured by more than two converted touchdown, the win percentage is 62.9%, and the New England Patriots are 7-1 in these games, with a loss in 2011. They could be the biggest favourites in a game since 2013 after opening at -17.  

The MLB regular season continues to wind down, and those who said the T-Bone was overdone are being treated to an unexpected bonus with a 6-0 September record.

Unfortunately Overs is just 2-4 while Unders is 1-2 this month. 

Two more wins for the Houston Astros this weekend at -430 (a narrow 2-1 win) and at -450 (a slightly wider 21-1 win), and backing at -300 or hotter in divisional games and you'd be up 12.4% this season on the Money Line, 15.9% on the Run Line.   

Tuesday 3 September 2019

Rare and Publicised

By some measures, Houston Astro's Justin Verlander hasn't had the best of seasons, managing to lose two games in the space of eleven days as a -430 or shorter favourite, but he did come through on Sunday with the third no-hitter of his career, and the second one in Toronto while playing for the road team.

Only five other pitchers in baseball history (since 1876) have managed to throw three or more although Verlander has some way to go to match Nolan Ryan and his seven.

Well publicised and rare events do offer an opportunity to the more sophisticated bettor, as the public overreacts to the news and goes into a collective recency reaction. 

For example, when a team holds an opponent hit-less, in the following game they have a 28.8% ROI on the Money Line when playing the same team again. 

If the no-hitter was the last game of a series, then the play next day is to fade them, currently with an ROI of 34.1% but a losing bet yesterday, albeit after extra innings.

A similar over-reaction by the less sophisticated is seen when a team gives up a high (15 or more) number of runs. If the next game is a day game, the Money Line ROI is 14.5%

I mentioned here the possible new record for the line in baseball on the 22nd, and sure enough a new record was set with a line at -530, Gerrit Cole becoming the third Astros pitcher of the season to start a game at -400 or shorter in August. Not surprising that with a starting rotation this strong, the Astros are as short as 3.15 to win the World Series this year. Currently they are tied with the New York Yankees for the best record in MLB and the American League, with the National League's Los Angeles Dodgers close behind, A repeat of the 2017 World Series looks possible. 

For some perspective on the history of red-hot favourites in baseball, since 2004 there have been just 16 matches where the line was sub -400 and six of those occurred last month, with five being with Astros pitchers. 

“Here's the essence of risk management: Risk no more than you can afford to lose, and also risk enough so that a win is meaningful. If there is no such amount, don't play.” – Ed Seykota

Sunday 1 September 2019

T-Bone Overdone

September's arrival means that we have just one more month of the regular MLB season remaining.

For those playing the T-Bone, official results for August are Money Line +4.75 points (ROI 8.0%) while the Run Line was up +2.25 points (ROI 5.7%).

Season overall: Money Line +27.45 points, (ROI 11.1%), Run Line +17.10 points (ROI 10.5%).

Historically, the month of September sees the most selections for this system, but is the least profitable (especially on the Run Line), so be careful if you are following along. 'Least profitable' is a charitable way of saying 'costly', especially in the opening game of a series. If the opponent won their previous game, you're better off fading the T-Bone pick.  

Ditching the 9 run total at the All-Star Break ensured the Overs system was profitable in August with a 27-18-1 record, while the Unders had only eight selections, but five winners for another profitable month.