The Draw As Third Favourite

One, of many, curiosities about this Premier League season is that since 2000, every Home odds-on favourite in a match where the Draw was priced as either the third favourite or joint second favourite, had won. 

This isn't a scenario that shows up too often, so it wasn't anything to get too excited about, but it was interesting nonetheless.

Then along comes 2020-21 and the first two qualifiers of the season fail to win, the second being Manchester City v Liverpool this past Sunday. 

The Draw can of course be either the second or third favourite or it can be joint second favourite and I've written before about how backing the Draw when the Away team is the favourite is more profitable than when the Home team is favourite, but this was in the context of "Close" matches.

Blindly backing the Draw when the Away team is sole favourite will not be profitable, but if you stick to the "Close" matches, backing the Draw is far more profitable in these games than when the Home team is the sole favourite.

Of the seven types of match-odds, assuming nothing untoward is at work as sometimes seen in Italy, the lucrative match types for Draw backers are clear from the chart below. When there are fewer than 50 matches a neutral yellow is used:  

For the A-H-D category, some of you may be interested to know that for the "Toss Up" category, the ROI climbs to 26.65%.

On a day when my wit won me, and WON ME BY A LOT! in the words of Trump, the Internet, a betting related topic came up on my timeline with @MathewForth questioning how a tipster could guarantee three bets a day.

The answer is that he can't.  

It's a clear signal that the tipster has no clue about how betting profitably works because you can't set an arbitrary number of selections within any timeframe. 

You might have several value selections on any given day, or you might have none for several weeks. The result of forcing selections, or stopping on a day after a random number has been reached, is that you either bet at poor (negative value) or you miss out on positive expectancy bets. The three a day idea is nonsense. You bet when you have value, and you don't bet when you don't. The option of not betting should never be underestimated or looked down upon because it falls under the "discipline" heading. 

Meanwhile the Next President market on Betfair has still to be settled, with Biden out to 1.1 now. Presumably this is because Trump is refusing to concede but so far his lawyers have held a press conference outside the Four Seasons Total Landscaping business with the star witness a convicted sex offender who isn't even from Pennsylvania and otherwise have presented no credible evidence at all.

Claims that vans have been seen pulling up outside with people seen filling in ballots for Biden are all ridiculous but it's amazing what some people will believe. When you ask them why this fix didn't also get the Democrats a couple more Senate seats, and why they lost House seats, well - no explanation. The one person so far committing fraud was a Trump supporter who attempted to get a ballot for his dead mother, but was caught by the system and even if, and it will never happen, he gets the result in that state overturned, he still wouldn't have enough Electoral College votes to change the result. 

I've topped up at 1.11 as this is too good to be true. Pretty sure that with half a billion traded, Betfair will not be voiding this market! 

Dogs Stay On Track

For the sixth time in nine weeks this season, the Small Road 'Dogs were again profitable with the New Orleans Saints, Miami Dolphins and New York Giants all winners with two losers, the Detroit Lions and Denver Broncos. 

One weekend was a loss, one was a push and week five saw no selections and for the season, the record improves to 20-12. 

The three Divisional games went 2-1 this weekend for an 8-5 record overall this season. 

I'm still waiting, along with a few others, for the Presidential Election bets to be settled. In addition to the main event, I have smaller interests in the Arizona, Georgia and Pennsylvania state markets. 

The sum traded smashed, and continues to smash, the Betfair record set in this same market in 2016:

The Exchange 'Next President' market obliterated the record for the most money ever to be traded on any political and sporting event, with just under £200m traded

 

With three more US states (Louisiana, Maryland, South Dakota) voting this past week to legalise sports betting, bringing the total to 21, this 2020 record may well be beaten again in four years time. My prediction is that Biden will step aside and Kamala Harris will be the Democratic Party nominee, and John Kasich will be the Republican candidate.

No "Toss-up" Draw selections in the Premier League this weekend, and the one "Close" selection wasn't close at all with Crystal Palace scoring four goals against Leeds United for the first time ever, and in a match for the first time since the final day of the 2018-19 season.

Biden Beats Trump

The four year nightmare in the USA is entering its final days with the win of Joe Biden over Donald Trump which was finally called by the media yesterday, an announcement that was accompanied by spontaneous celebrations in cities around the country. The timing was interesting as it appears the networks waited until Trump was out on the golf course (yet again) before making the call. He wouldn't be the first right wing leader to learn of defeat while in a bunker.

Trump does at least set the record for most votes by a losing candidate, and how 70+ million people can vote for him given his record is really quite disturbing, as well as managing to lose the popular vote twice. 

The error yesterday by a Trump staffer who booked the Four Seasons Total Landscaping business for a press conference for his lawyers instead of the Four Seasons Hotel was a nice metaphor for the job the administration has done over the last four years. 

He promised to run the country like he ran his business. It ended with Rudy Giuliani outside Four Season Total Landscaping next to a dildo shop, after a staffer mistakenly thought the empty parking lot was a Four Seasons hotel.

Although at the time of writing, Trump is still whining that he won - and by a lot - he didn't, and there is no reason why the result in any state would be overturned, so if you have some money you don't mind tying up possibly for a few days, you can still get 1.05 on Biden.

Trump says he plans to go to court on Monday, but reporting indicated Jared Kushner is trying to get through to him that it's a lost cause. Trump's defeat is made all the sweeter given that on Tuesday night, with his odds as short as 1.24, he almost certainly thought he had won. We can only imaging him waking up a few hours later, checking his phone, and the reality of his situation slowly dawning on him. With no immunity from his financial and legal troubles after January 20th, it's no surprise that he is crying so much, but come that day he is no longer President whether he concedes or not. Farewell to a dreadful human being, and Mexico never did pay for that wall. 

For a change of pace, if you like statistics and you like tennis history, you might enjoy the Netflix show Guilermo Vilas - Settling The Score which is about how Vilas was wrongly denied the No.1 world ranking in the 1970s. Great lockdown viewing.

Even though he won 16 ATP singles titles, including the French Open and the US Open and was the runner-up at the January edition of the Australian Open in 1977, Vilas was never ranked by the ATP as world No. 1 during 1977 which was due to the fact that the rankings at the time were based on the average of a player's results. He was instead year-end world No. 2, behind Jimmy Connors (who won the Masters and seven other titles and was the runner-up at Wimbledon and the US Open in 1977). Nevertheless "World Tennis" magazine listed Vilas as 1977 year-end world No. 1, and Borg No. 2.

Argentine journalist Eduardo Puppo and Romanian mathematician Marian Ciulpan investigated the 1973–78 period records, and delivered a detailed report with more than 1,200 pages in which they came to the conclusion that Vilas should have been ranked No. 1 for five weeks in 1975 as well as during the first two weeks of 1976 and handed over their research to the ATP at the end of 2014. Although the study was not refuted, in May 2015 the ATP announced it had decided not to make official the No. 1 position for Vilas because it happened in the interval between the publications of the official rankings.

This was also a topic covered by the New York Times in 2015 in which the ATP give more details about their reluctance to correct the record and recognise Vilas' achievement. 

When the Tie-Breaker is Tied

Thanks to @Bitcoinpari_, the flaw in my logic puzzling over why either Pennsylvania or Georgia was identified.

What I didn't know until today, was that in the event of a tie, while the deciding vote is cast by the House of Representatives, it is based on one vote per state and not one vote per representative, and with Republicans currently in the majority when measured by state, although maybe not in the new Congress, the deciding vote would go to Trump.

Some tried to claim that the flaw in my logic was due to correlation, but that argument never made sense. What I was pointing out was that if Trump's only paths to winning had a sum probability of around 2%, then why was his price around 10%?

As mentioned, Trump's probability was actually higher than that 2%, because a Biden win in Georgia wouldn't have been enough on it's own. 

Fortunately not only did I learn something new, although now I am wondering what would happen of the House was tied at 50-50**, but I got away with the mistake since Biden has since shortened to 1.08 / 1.09 with Trump out to 13.5 / 14. 

Pennsylvania is currently 1.08 for Biden, with Georgia around 1.34, Arizona 1.23 and Nevada 1.04.

If Biden wins Pennsylvania, it's over. If he wins Arizona AND Nevada, it's over. If he wins Georgia AND either Arizona or Nevada it's over, so the 1.08 for Next President still offers value, discounted presumably because of the risk of courts stepping in which seems less likely as the hours go by. There's probably also a few people who have a significant sum tied up and are looking to cash out.  

** I looked up the tie in the tie-breaker rules: Note that if a majority is not reached in the House vote (e.g., 25-25), that chamber needs to keep at it until the tie is broken. If the deadlock is still in place when the new term starts (noon, ET on Jan. 20), the vice president becomes acting president until such time as the House elects a president.

Trump's Roads to Victory

Unless I am missing something, which wouldn't be the first time, or there are some shenanigans in the courts, which seems less likely, as things stand, Trump has just three ways of getting to the 270 Electoral College votes he needs to win.

The five states still in-play are Pennsylvania (PA), Georgia (GA), North Carolina (NC), Nevada (NV) and Arizona (AZ) which has already been called a win for Biden by Fox News, much to Trump's annoyance and perhaps a little prematurely.

Certainly a fair amount of money was matched down to 1.01 in the AZ market for Biden, with the price now out to 1.2. The issue here is that pre-election day ballots are still being counted, and while generally the early / mail-in voting has favoured Biden, around 59% of the ballots to be counted are from areas favouring Trump which means the lead Biden currently has will close.

Anyway, the maths is that if Biden wins Pennsylvania or Georgia, he gets to 269, which with the deciding vote to be cast by the House of Representatives, again under Democrat control, would mean a win for Biden and on Betfair whose rules cover this possibility.

So for Trump to win he needs both PA and GA plus either:

NC and AZ or

NC and NV or 

AZ and NV.

Trump is currently the favourite in just one of these states, North Carolina (1.2) but using the current state prices, and they are moving all the time, these combinations add up to a 2% probability if my Pure Mathematics With Statistics 'A' Level skills are still with me. 

Biden is trading at 1.1 / 1.11 with Trump at 10 / 10.5 which means a big discount for the possibility of the courts intervening in a crucial state or on a state such as Michigan suddenly finding thousands of Trump ballots. 

Not in Kansas Anymore

Early in the 2009-10 EPL season, Betfair Forumite "The Magician" commented on the lack of Draws in that league to that point, and suggested that he was "thinking about backing every draw until the end of the season."

There was some debate about the soundness of this strategy, for example whether this was akin to backing tails after consecutive heads, but the OP was making the assumption that "football has not changed that much" and that the draw price would likely increase in response to their relative scarcity. 

He started his strategy on September 27th with the Sunderland v Wolves match. At this point in the season, there had been four draws in the opening 64 matches, well below the Premier League's 26.9% for draws since 1992.

As it turned out, The Magician was right. Nothing had fundamentally changed with the Premier League, and by the end of the season, the draws were at 25.26%, only one fewer than the previous season and backing the Draw in every remaining game at Bet365 prices would have generated a 7.1% profit.

This season has started off slowly for the Draw also at under 20% for the season, but the big story so far is in the percentage of matches being won by the Away team. Long time readers of this blog will know that I have predicted the decline of home advantage at the top level of the game, and with the implementation of first goal-line technology followed by VAR and now the added factor of no crowd, this 'advantage' has become a disadvantage. 

With one exception, the first season after the Second World War, Away wins in top division in England had never (going back to 1882-83) surpassed 30% until 2011-12. Five of the eight seasons since have seen that level beaten, and as of today the percentage is 45.6%. 

The problem is that unlike in 2009 when nothing had fundamentally changed in football, we can't say that this time. 

In the Bundesliga, Home teams currently have just a 35.2% winning record (a low since 1993 when my spreadsheet for the European leagues begins), while in La Liga home teams are also at an 'all-time' low (38.4%) with Away wins at a record high (35.6%). 

In Ligue 1 and Serie A, it's the Draws which are at an all-time low, with the Away at all-time highs.

In my formative years, the Home team in Division One won around 50% of matches, with 25% for the Draw and Away, but it's a different world now. Good luck if you fancy laying Away teams for the rest of the season. 

One final nibble at Biden's great value odds this morning. The odds will shorten as the day goes on, or at least that has been the pattern for some time now. Hopefully by this time tomorrow they will have shortened a lot more usual. 

All Souls Day Update

Another profitable Sunday for the NFL Small Road 'Dogs system which advances to a 17-10 record with the three wins and one loss yesterday. All four matches were Divisional games where the edge is historically bigger, and these are now 6-4 on the season. 

Tonight's game is not a qualifier. Since the league reorganised for the 2002 season, only 24 teams have been road favourites by more than 12 points, as are the Tampa Bay Buccaneers tonight for the first time in database history.

In non-Divisional games, the overall record ATS for favourites is a losing one (4-9-2) but the last loss was back in 2012. 

Both of tonight's EPL matches fall into the "Close" category but no luck with the one "Toss-Up" qualifier yesterday (Aston Villa v Southampton, coincidentally the second consecutive time this fixture has featured in a post!) 

The US Presidential Election market is closing in on £300 million matched on Betfair, a Betfair record, with Biden currently at 1.51 / 1.52. Once again last night, the price on Trump shortens overnight and then drifts during the day. 

A word of caution if you are arbing this market that the rules clearly state that "This market will be settled according to the candidate that has the most projected Electoral College votes won at the 2020 presidential election." In other words, the "winner" isn't necessarily the person who will be the next President. 

There is also a caveat in the Senate Majority market which specifically excludes the two Independent senators who both caucus with the Democrats. So if you are betting on the tie, you are betting on the Democrats getting 49 seats, and not the 48 seats which they need in practice. 

Football Outsiders

With few of my sports currently active, I have too much time on my hands, and during a trawl of the Betfair Forum I came across a thread titled Football outsiders - match odds market.

The OP Shanelee1966 wrote:

To back 4 teams every Friday/Saturday who are 5/1 or over. English leagues ideally.
I may venture into Europe, lets see how it goes.

Starting bank £20, fiver a throw.

Blindly backing outsiders in any sport is generally the fastest way to lose your money, and if this strategy had been applied to the Premier League since 2000, the ROI would have been around -14.08%. 

Given that in the early years of this century the over-round was sometimes as high as 113%, this number looks worse than it is, but what surprised me a little is that if you apply this strategy to Home teams only, then it is actually profitable with an ROI of 7.75% from 514 matches. 

If we look at the Pinnacle era (i.e. since 2012) when the over-round has averaged a more competitive 102.2%, the ROI on these Home outsiders increases to 20.7% from 270 matches. With one exception (Aston Villa v Southampton in April 2016), in every match the Away team has been one of the Big 6. In four matches the Home club was also "Big 6", and two of these won.

I took a look at this strategy for big Home 'Dogs in some of the other top leagues in the Pinnacle era to the end of last season, and in La Liga the ROI is 30.7%, while in Ligue 1, hardly any home teams are priced at such long odds, and all seven in the eight seasons lost. However, the return on Away sides priced at 5/1 or longer is 31.4%.  

As mentioned many times before, when it comes to European football, a one-size fits all approach never works as each league and division has its own personality. 

Blue: Dodgers, Moons and Waves

As it turned out, just the one World Series game was left with the Los Angeles Dodgers winning their first title in 32 years, but another profit took the final season total to £538.56 which, as I mentioned before, feels more of a win than it should. 

The bad news is that with the NBA, NHL and MLB seasons all ending fairly recently, it could be a while before the daily leagues return, which for the indoor sports is usually around this time of year. 

For the record, in the MLB play-offs this season Unders had 28 wins to 25, (and 100% when the game time temperature was less than 55F as was the case in the Regular season), while backing the favourite was profitable on both the Money Line (+7.21 units / 9.1%) and the Run Line (+13.95 / 25.2%). Other than in 2017, this hasn't typically been a profitable strategy in recent seasons, but this season was by no means typical and I'm not sure we can read too much into it.

The NFL season appears not to be hugely impacted by the reduction in crowd size, sometimes to zero, with the Small Road Dogs System 14-9 for the season so far. Not a big surprise given that this simple system has had just two losing seasons since 2005.

The question for this season was always whether the market would continue to over-value the Home side even with no or fewer fans in attendance, and the answer so far appears to be yes.

In the Presidential Election, Biden has drifted slightly from 1.51 / 1.52 to 1.53 / 1.54, tremendous value unless you believe the polls are more wrong than they have ever been before.

The Economist has Joe Biden with a 96% probability of winning the Electoral College and to win Pennsylvania, around 93%. As Political Gambler Paul Motty noted today in this article, Biden at 1.53 (was 1.59 at the time of the article) looks like another solid bet. 

The race for the Senate is also interesting and having backed the Democrats at 2.84 early this month, the price has moved in the right direction. It's interesting to note that since 1900 there has never been a Congress where the House and President have been Democrat controlled while the Senate has been Republican controlled. The Economist actually gives the Democrats an 83% chance of controlling the Senate, although this includes the tie. 

The tight Senate races in Georgia (two), Iowa and South Carolina will decide this. 

Hopefully the theme of blue will continue next week. The Dodgers play in blue, there's a blue moon in America for the first time since March 2018, and the first for the whole country on Halloween since 1944, and a blue wave or tsunami to come on Tuesday.

Back to Baseball, Biden and Barrow

I'm back a couple of days earlier than originally scheduled, which is a good thing as it means everything went according to plan, and so with the World Series still in progress there are still another one, or maybe two, games for me to add to the season profit reported in the last post. Game Six is tonight with the Los Angles Dodgers 3-2 up in the series and favourites versus the Tampa Bay Rays. 

Draws in the EPL are also back after starting the new season with none in the first 20 matches, the pace has since picked up with 13 in the last 38. The "Toss-Up" matches are up by 3.71 points using Pinnacle's closing prices, but we've only had three selections. As a reminder, in these matches, backing the Draw in all matches has an ROI of 7.33% (adjusted for an overround of 103%) since 2000. The ROI in Big 6 games is now over 31% in that period following last weekend's perfect draw in the Manchester United v Chelsea game.

In politics, the US Elections are less than a week away, and Biden is still value at 1.51 / 1.52. To reach 270 Electoral College votes, Trump needs to secure all the Lean, Likely and Solid red states, the five "toss-up" states, plus three of the "Lean Democratic" states unless he can win Pennsylvania in which case one other state would do it. That seems a tough ask, and his poll numbers are tanking by the day with early voting at a record high with over 70 million people already having voted. For those who say "yeah, but the polls were wrong in 2016", one week prior to the election in 2016, Clinton was up by between 2.2 and 3.5 percentage points (and won the popular vote by around 3 million). Biden today leads by 7.4 to 9.1 points.  

Finally, congratulations to Barrow who tonight won their first Football League game in over 48 and a half years, a new record as even Brighton haven't gone that long between league wins.  

MLB 2020 - When Breaking Even Is A Win

Another work trip starts tomorrow, so I shall again be off-line, this time for two weeks. Rather unfortunate timing really, with the MLB play-offs in full swing, meaning I shall miss every game of the World Series, but my boss doesn't revolve his business around the MLB season apparently.

After a disastrous start to this unique season, I'm more than happy to finish the season with a small profit.

In a season where the several changes were implemented, some before the season and some during the season, it was always to be expected that some of the long-term profitable strategies might not be profitable.

Obviously not having fans in the stadium was likely to be impactful, but more significantly the rule changes implemented weren't exactly minor tweaks with some of the major adjustments being American league rules throughout, fixtures changing to play geographically close rivals, a shortened regular season (60 games versus the usual 162), changes to the rules for extra-innings with a runner starting on second base, and limiting double-header games to seven innings. Basically the 2020 season bears no resemblance to any previous season. 

My latest read is "The Psychology of Money" by Morgan Housel (highly recommended), and Chapter 13 is sub-titled "The most important part of every plan is planning on your plan not going according to plan".  Although this book wasn't published prior to the season, it's a philosophy I've adopted for some time now (sports betting provides some valuable lessons in life). Cut the losing systems, stake appropriately (based on Kelly), and keep evolving.

The biggest change in 2020 was the reversal of value of backing some Hot Favourites (1.5 or shorter), which after three years of increasing profits, hit a wall and lost 4.8%. The month of August saw a loss for the first time since 2011.

   

For the T-Bone System, which is also a favourites based system, here is how the season ended, with previous seasons included for perspective.

The Totals strategy overall was slightly profitable, although the seven innings rule for double header games reduced the number of qualifiers for Overs. 

Unders actually showed a small loss with the heavy lifting all being done by the Overs in games where both clubs were from the National League. Apparently the market underestimated their ability to adapt to the Designated Hitter rule, because this matchup was also sole reason why "Hot Favourites" were a losing proposition this season.

Usually the winter is the time for looking at the previous season's results, seeing what edges may be vanishing and where new ones might be appearing, but I don't see any point in going over the 2020 season in much depth.

Some of the rule changes may become permanent, but a 60 game regular season played in empty stadiums against geographically selected opponents is hopefully unique rendering the data from this season almost useless, but this pandemic is nowhere near over, so perhaps not.  

New Balls Please

I have been critical in the past about the quality of some of Pinnacle's articles which are frankly sometimes embarrassing, but an excellent article and a great example of creative thinking comes in this article about the impact of ageing balls in tennis. 

It's probably fortunate that author Jonathon Brycki's interest is in sports and not medicine, or the article might have been quite a distressing read for men of a certain age, but for anyone who is a tennis in-play "between games" bettor, this is a must-read article.

I say "between games" because as I have written before, court-siders will always an advantage in-play, but between games when the market has settled, an edge such as the one proposed by Jonathon might be significant. 

The data from eight years of Grand Slams indicates that ageing balls do have an affect on serve and hold percentages. The magnitude is somewhere in the order of 1-2 percentage points for both. While this may seem relatively insignificant, in betting markets that are becoming increasingly efficient, a percent or two will often be the difference between getting your money in behind and having an edge.

The margin might still be too large to overcome, as Jonathon makes clear here:

It’s unclear whether bookmaker’s live tennis models incorporate the age of the balls, and even if they don’t, the inefficiency may not be large enough to overcome their margin. In any case, incorporating potentially unique data and insights into your model will drive it in the right direction, and improve the accuracy of your predictions.

As I have mentioned before, in a binary market, grandma should be able to pick winners 50% of the time, so in a market with a tight margin, you don't need a huge advantage to be into profit.

But back to the article, and kudos to Jonathon for the idea, the research and for sharing his findings, although perhaps the fact that his findings have been shared is proof that the edge isn't of any practical use. 

Regardless, it's a great example of the process we should all be following in our never ending quest for an edge. I wonder if any such study has been done in cricket regarding the impact of an ageing ball?  

COVID-19 Suspends Play

For the last few days I have been away on my first work trip since January, and as a result of a positive COVID-19 test at one of the sites I visited, am now in quarantine. The good news is that you will likely see an increase in posts, and if this is not the case, then it's likely bad news and it was good knowing you!

The big news while I was away was that several of the US Presidential Election markets are currently suspended on Betfair due to the President testing positive. There's some confusion about why Betfair would do this given that in similar, although arguably less serious previous instances (e.g. Clinton 2016 pneumonia, Sanders heart attack) the market has not been suspended. The last price traded for Biden to win was 1.52, so hopefully many of you are in good shape, unlike Trump.

With 35 Senate seats up for grabs, the market here is also interesting, with the Democrats now favourites to have a Senate Majority at 2.8 in a three-way race with the tie very much a possibility. Republicans currently have 53 of the 100 seats and 23 of the 35 seats in this election are held by Republicans.

The Democrats will probably lose their senator in Alabama (Doug Jones) who won a special election in 2017, which means that Democrats need to win five of the 23 seats for control. Four would give them a tie, and with the Vice-President likely to be Kamala Harris, in effect that would be a win, but for betting purposes, only a tie.

Of the 23 seats, the Democrats are currently projected to win those of Martha McSally (Arizona), Cory Gardner (Colorado) and Susan Collins (Maine) which means they need one to tie and two to win from the toss-up seats of Joni Ernst (Iowa), Steve Daines (Montana) and Thom Tillis (North Carolina). In addition, four seats are in the 'lean' category and could yet move to toss-up. One seat to keep an eye on here is Lindsay Graham's in South Carolina where polling shows a close race, the incumbent being out raised financially, and as chairman of the Senate Judicial Committee will be spending a lot of time in Washington on the Supreme Court nomination. 

On the more traditional sporting events, the NFL Small Road 'Dogs System continues to produce winners. Week Three saw nine selections, with six winners taking the season record to 10-6. It's looking like we only have two selections today.

My father-in-law is currently in shock following the San Diego Padres first playoff series win in 22 years last week as they advanced to the NLDS where they will play the Los Angeles Dodgers. One piece of history from their win in Game Three against St Louis Cardinals is that the Padres were the first team in major league history, regular season or playoffs, to complete a nine-inning shutout by using at least nine pitchers.

So far in the MLB playoffs, as was the case with the NBA (41-35) and NHL (66-52), Unders is the play on totals with an 11-7 record. 

With the increase in goals per game in the EPL, the Draw continues to be an elusive outcome, but more on that later once we have the official prices for the fourth round of games from Mr. Buchdahl. 

Mother-In-Law Issues at the World Series

The Stanley Cup Finals are now set, and will see the Dallas Stars play the Tampa Bay Lightning. As in all play-off games in the NHL, Under is where to start looking for value, but in the Stanley Cup Finals this is even more true with 63% of games going Under in the last ten seasons. 

In a low scoring sport, we can't assume we will get 1.952 but Unders is a good place to start. In the past three seasons, this hasn't been a profitable strategy with every Game 1 going Over but with no crowd or home advantage, this season's playoffs have seen Unders win 56.5% of the time. 

I mentioned earlier in the week that there were three qualifiers for the NFL Small Road Dog System, but the line moved on the Chicago Bears game so 'officially' it wasn't a winner. We did have a winner on Thursday though, so 'officially' we are back up to 2-1 with possibly another four selections tomorrow.

Still no draws in the Premier League this season with plenty of goals reducing the probability. Newly promoted clubs have conceded 22 goals combined after just two games.

I mentioned that the World Series would be held in one stadium this year (Texas Rangers) but a little research revealed that this isn't a first.

In 1944, the St Louis Browns played the St Louis Cardinals in what was called a "Streetcar Series" with both clubs sharing the imaginatively named Sportsman Park.

There's an amusing story about the logistics of this series:

Since there was a shortage of housing during the war, and the teams were never at home at the same time during the regular season, the two managers — Luke Sewell of the Browns and Billy Southworth of the Cardinals — shared an apartment.

Of course, they’d both have to be in St. Louis for the World Series.

“However admirable this display of interleague cooperation might have appeared during the season, it would never do for the opposing managers to sit in the same living room after a World Series game, sipping bourbon and chatting politely with their wives,” wrote William B. Mead in his book “Even The Browns: The Zany, True Story of Baseball in the Early Forties.’”

“Besides, Sewell wanted to invite his mother, and Mrs. Southworth could hardly be expected to put up with a mother-in-law from the wrong family and, indeed, the wrong league.”

Fortunately, according to Mead, another resident of the apartment building was out of town for the month. The Southworths were able to move into the unoccupied unit during the series, snuffing out that potential crisis.

The Cardinals came from 1-2 down to win the Series 4-2 and the Browns later moved to Baltimore becoming the Orioles.

All World Series games in the same stadium also happened in 1921 and 1922 when the New York Giants twice beat the New York Yankees while sharing the Polo Grounds, which was the original home of the New York Mets. The Giants moved to San Francisco and I believe the Yankees are still in New York...

Playoff Updates: NBA, MLB, NHL

Teams arriving at a Game 7 after losing a 3-1 series lead have a losing record and in this series, four games have been Under so far with two Pushes.

Two of the easier wins last night in the NBA as the Denver Nuggets won comfortably, not needing the 7.5 points they were getting, and the Under also an easy win (by 14.5 points) for the ninth consecutive time in a Game 7. Hopefully some of you took note of the observation in my last post.

This win was the second time the Nuggets have come from 1-3 down in a series to win, following their First Round comeback against the Utah Jazz. 

The Conference Finals started in the East last night with an overtime win for the Miami Heat as 'dogs versus the Boston Celtics, while in the West the Nuggets will play the Los Angeles Lakers.

MLB announced a unique play-off format yesterday, with all Division Series, Championship Series and World Series games scheduled for a neutral venue in either Texas or Southern California, and importantly with no rest days scheduled within a series meaning some tough starting pitching decisions ahead.  

The host stadium for the World Series will be Arlington, Texas, and with the home town team currently the second worst in baseball and nine games out of a Wild-Card place, the chances of them having home advantage here are somewhat slim.

Other clubs hosting games are the National League's San Diego Padres and Los Angeles Dodgers, both staging American League games, and Houston, whose Astros play in the American League and will host National League games.

The expanded Wild Card games are all best of three games, and the higher seeded team will have home advantage. 

In case you need any more evidence for how unusual this season has been, the five clubs with the longest playoff drought are all in with a chance of ending their disappointment this year, including my wife's hometown club. 

They are, in order of woefulness, Seattle Mariners (last playoff appearance 2001), Miami Marlins (2003), San Diego Padres (2006), Chicago White Sox (2008), and the Philadelphia Phillies (2011). Every dog has its day.

In the NHL, the Dallas Stars have reached the Stanley Cup Finals thanks in large part to two overtime wins. I mentioned the edge on backing Unders in play-off games earlier this month, and since then we have had six Unders and two Overs. Two matches resulted in a Push. 

Patience and Discipline

Yet another Game 7 in the NBA Play-offs this week, after the Denver Nuggets overcame a 19 point deficit to beat the Los Angeles Clippers by 13 points on Sunday. 

Teams arriving at a Game 7 after losing a 3-1 series lead have a losing record and in this series, four games have been Under so far with two Pushes. 

The NFL is back, and the Small Road Dogs System had three selections in Week One, namely the Chicago Bears, Tampa Bay Buccaneers and the Green Bay Packers and I'll take the 2-1 outcome any day of the week. 

Although none of the opening eight matches ended as a Draw, with the 2020-21 EPL season underway, a timely reminder that backing the Draw in close matches is generally profitable in this league. As David Sumpter wrote back in 2016 or so:

It turns out that when two well-matched teams meet (i.e. the probability of a home win is only slightly bigger than the probability of away win) then draws are under-priced. When matches are skewed so there is a strong a favourite (i.e. the probability of one team or the other winning is larger than the other) then draws are over-priced.

Somewhat arbitrarily I define 'close' matches as those where the difference between the fair win probabilities of the two teams is less than 25%, and 'toss up' when the difference is less than 10%.

In the eight seasons of the "Pinnacle era", with the overround adjusted to 103% for consistency, a number to which Pinnacle unfortunately appear to be headed quite rapidly, here are the results:

As a baseline, backing the Draw in all matches would have resulted in a 208.49 point loss, an ROI of -6.9%.

I'd actually previously not noticed the distinction in David Sumpter's quote above that he's looking at matches where the probability of a Home win is only slightly bigger than that of an Away win.

Close matches can also include matches where the probability of the Away team is slightly higher, and of course the two probabilities can be the same.

I'm not sure that Sumpter meant to make this distinction, but as I've mentioned before, the ROI on the Draw is actually slightly higher (16.8%) when the probabilities are the same or that of the Away team is slightly higher. Perhaps that wasn't the case in the Big 6 matches during Sumpter's period of study.

Finally, some words of wisdom from the Betfair Forum on a thread asking "Why do so many people lose at gambling?"

Zealot suggested that:

Most people lose because they are too greedy and 90% of them have no idea what a good return would be. Say, for example, you backed 500 horses over a 12 month period and your total outlay was £10k (£20 per bet). 

How many Joe Punters would be happy with a profit of £800? That's a very good 8% return. I would say less than 1%. If you ask a betting shop punter this question, most will say anything up to "Doubling It". Most have no clue of what it takes and how much work you have to put in. 

First step should be breaking even, and then with a little bit of tweaking, you will be well on your way. Patience is the key word in my opinion.

Those last two sentences are very sagacious. In binary betting, a blind squirrel will be able to win 50% of the time and you only need to convert 1.02% of those losers into winners if you are on Betfair's Basic Plan, or 2.46% at Pinnacle's Dime Line (1.952)  

Patience to wait for value opportunities is key, as is the discipline to stake appropriately when such opportunities come along.    

G7 Home Dog and MLB Woes

We have another Game 7 in the NBA tonight, in the Boston Celtics v Toronto Raptors series, where five of the six games so far have gone Under, with the exception being the double-overtime Game 6 where the 51 extra-time points took the total comfortably over. With the Toronto Raptors getting 2.5 points, this is only the third time in database history (since 2002) that the Home team has not been favourites, although this season all games are being played on a neutral court so it's not a huge surprise. Both of those previous games went Under.

Baseball continues to be an outlier of a season, with hot favourites unprofitable for the first time since 2014:

With some perspective, the results this season don't look so terrible with some nice profits in the bank from previous seasons, but it's tough to keep backing during a downturn, especially when the schedules and rules have been changed so dramatically.

Backing weaker favourites this season would have been profitable, but typically these are losing propositions. Clearly all the changes MLB has made this season have affected the markets. 

The Totals systems have also been messed up with the mid-season decision to make all double header games seven innings. August saw Unders as profitable in these games, but so far in September the value has been on the Overs. 

Pump It Up

The NFL season is here, and as with most other sports at the moment, it will be one with no fans at some stadiums, but fans at others, and thus a season with no precedent.

At least some Home Field Advantage (HFA) has long been held to come from the thousands of screaming fans in the stands who intentionally make more noise when the visiting team is driving and hard for them to hear play calls.

This season, some teams will have fans from the start, e.g. Miami Dolphins. Others hope to have fans as the season unfolds, e.g. the New Orleans Saints and Green Bay Packers, while others, e.g. the New York Giants have ruled out fans completely, at least for the time being.

Last week, NFL Commissioner Roger Goodell told reporters on a conference call that there will be no advantage for teams with fans in the stands.

How can this be? Here's how - on the same call, it was announced that:

"For teams with no fans in the stands, we’ve created audio that will be played in all stadiums"

Apparently details are still being finalized, most notably about the decibel level to be allowed for the pumped-in noise, but it certainly appears to be a case of an unfair playing field.

Readers will be aware that the strategy of backing away teams getting a handful of points (the Small Road 'Dogs System) in the regular season has been profitable in most recent seasons:

That ROI is 10% if you restrict your bets to Divisional games only. Unlike MLB, the NFL have not changed their scheduling, nor made any COVID-19 specific rule changes, so I am optimistic that the season will play out in a fairly typical way. One big concern is that the market may see the HFA as reduced or non-existent this season, and the points Road teams are getting will be reduced. 

Three road teams opened as Picks for Week One, the Las Vegas Raiders, (get used to that name), Seattle Seahawks and Chicago Bears, although the Raiders are now 3 point favourites in some books while the Bears are now getting 3 points (and are currently a Small Road Dog pick) along with Green Bay Packers and Tampa Bay Buccaneers.

A quiet opening week might not be a bad thing. The reigning champions Kansas City Chiefs host Houston Texans on Thursday night and the Washington team still have no name and will start the season with the temporary, if unimaginative, name Washington Football Team.

Although not tied to the pandemic, the NFL has changed its play-off format this season with 14 teams (of the 32) now qualifying, and only the top seed in each conference now receiving a Bye in the first round.   

Finally, as I've mentioned before, I occasionally peruse the Betfair Forum, though more for amusement than in any expectation of learning anything useful.

One recent thread on the Football Forum was started with this comment:

It is nice for b/f to provide us live view...but until one team scores how are we to know which team is which?

Team colours would be nice.

Good grief. As one of the replies said:

If you are researching teams to bet on and don't know what colour shirts they play in maybe time to call it a day. Failing that a 10 second search on Google.

NBA: Evolving Playoffs and Game 7 Totals

After the two NBA Game 7s this week, here are some more details about how the Totals in a long series tend to play out.

First of all, some history. Since 2014, NBA play-off series are all Best-of-Seven, played in a 2-2-1-1-1 format, with the highest seed (better team based on regular season) having Home advantage in Games 1, 2, 5 and 7, although sometimes Game 5 and often Game 7 are not required,

Prior to 2014, the series were played in a 2-3-2 format, and the format has been tinkered with on several occasions, which means we have to be careful when comparing data across seasons, especially in Games 5 and 6.

The format was changed in 2003, when the First Round became a Best-of Seven series, reducing the chances of a higher seed being eliminated, (in 2007, the Golden State Warriors became the first team to defeat a number one seen in a best-of-7 First Round series), but also increasing the probability of a team sitting around for a few days waiting for their opponent who may be playing three games more.

In 2005 the two conferences (Eastern and Western) were realigned into three divisions each, with each division winner qualifying for a top-three seed regardless of their record.

This proved unpopular, and in 2007 the rules were changed again so that the division winners are now only guaranteed a top-four seed. If the top two teams in a conference are in the same division, this change means they won't meet until the conference Final, whereas before they could meet in the second round (the conference semi-finals).

Eight teams from each Conference make the playoffs, with the winner of the Eastern Conference playing the winner of the Western Conference for the NBA Championship.

So each season there are 15 playoff series, 7 between Eastern Conference teams, 7 between Western Conference teams plus the Finals Series. 

Teams need to win 16 games to be champions, and 71 of the 73 winners to date have been seeded either number 1 (49 wins), 2 (16), or 3 (6).

My suspicion is that there is more interest from the general public in playoff games, and that interest increases within each series peaking with Game 7s and also as the playoff rounds progress peaking with the Championship Series, and the public like to see points which may offer value on the Unders. 

Does the data support this idea?

Since 2002, when we first have data, Unders has been the result 51.5% of the time which is not a huge edge, but at the standard (Pinnacle) 1.952 (US -105) odds, you only need to win more than 51.23% of the time to be profitable.

In the 'If needed' games, i.e. Games 5, 6 and 7, the Unders record all-time is impressive, and notably more impressive if the game was preceded by one or two Unders also.

Unfortunately since the latest change in format in 2014, the trend is not so strong. One reason may be that we have fewer matches to look at, but also it seems reasonable to assume that markets are more efficient these days than they were in the 2000s.

There's also a noticeable divergence in results between Conferences since 2014. 

In the NHL play-offs, where 58.6% of Game 7s go Under, there was another one last night between the Philadelphia Flyers (nominally the Home team) and the New York Islanders. With four goals in total, this was another win for Unders.

With baseball not playing out this season anything like it typically does, systems that have performed well for many years are struggling. With different rules in place, it seems likely I should exclude these matches from analysis in future, but Signora Cassini is happy with her hometown San Diego Padres looking likely to end their thirteen year play-off drought with the third best record in the National League West, albeit playing under American League rules. Her father was a minor league player with the Angels (then California Angels) organisation, so he's not quite so happy. 

Trump Support Swamped by Blue Wave

Some proof that there is a god came yesterday when five* boats in a 'Boaters For Trump' Labor Day weekend parade on Lake Travis near Austin, Texas, sank in 'choppy water'. 

Metaphor jokes have been deleted - editorial ruling: too obvious. 

Poseidon, traditionally known as the god of the sea, but also god of water in general, gets the win.

Some unconfirmed reports claimed that when boaters radioed the local Sheriff's Office saying that they needed help and were sinking, the response was "All boats matter". 

I found this first-hand report from one of the boat (the SS Magaritaville) owners. It's a tough watch, and to rub 'salt' into the wound, apparently Jimmy Buffett, no relation to Warren, has asked that the name not be used again on any replacement boat. 

In other Trump news, while his odds on Betfair haven't shifted much, I don't think the reporting of Trump's contempt for the military and religion (he may have a point regarding religion to be honest) this weekend isn't going to win him any new voters and polling next week might well reflect this.

Biden is still available at 1.97, and his polling numbers are looking much stronger than that price would imply when looking at the individual states he needs to win.

With 249 Electoral College votes if there are no upsets on the blue side down to Michigan (which went to Trump narrowly in 2016), then Biden needs to pick up either Pennsylvania (to tie), or Florida for the win. If he loses these two states, he can still win by taking Wisconsin and Arizona. North Carolina is a toss-up state also, and with 15 EC votes may be crucial.

From the Trump perspective, he will have 219 EC votes if he wins all the states where he is currently ahead, meaning he needs 51 more votes to win. Where are they coming from? If we concede Florida to him, he still needs another 22 votes, i.e. Michigan or Pennsylvania plus one of Arizona, Wisconsin, or Minnesota. Or those last three states combined. It's really not looking likely right now, but Trump is available at 2.1 which looks a terrible price to me, but £3,109,940 has already been traded at this level, with another £2 million matched at 2.12. 

There are still two months to go until the election, and an October surprise may yet come along, but it's not clear to me how Trump is going to make up the deficit seen in the fairly consistent poll numbers. Surprises can work both ways, but although Trump's base has been pretty consistent and might not drop much in the event of yet another Trump scandal, I'm not sure what kind of Biden story would win Trump enough new voters to overcome his current position. 

* Post updated to reflect that in fact five boats, not four, sank during this parade.