Dispersion Of Skill

One of the longer running blogs linked to by the longest running of all sports investing blogs is that of Simple Soccer Stats where Leon Pidgeon has this week decided to hang up his keyboard after ten and a half years and return to his loft:

So much has changed in those years but right now I don’t have the time or dedication to continue with the site. I was a keen bettor with plenty of time on my hands staying up late and studying everything. I’m now 10 years older with a family and no spare time on my hands. I also don’t even bet on matches anymore so I am keeping the site going more out of a duty to the people who use it.

Leon certainly published a lot of stats, but as I have written previously, you're not going to find it very easy gaining an edge over the big players when trading top flight football. In his defence, he does refer to them as 'simple' stats, and looking at an edge from early goals, bogey teams, half-time scores etc. is unlikely to lead to riches. Leon doesn't mention how much he made in those ten years, but I hope it was worth the time invested and I wish him well in the future.

I'll leave the link up for a few months, before taking it down, but to compensate, I've added a couple of well-written and researched horse racing blogs from Gareth, one focused on the Grand National and one on Trainer Profiles. If horse racing is your thing, then check them out. 

I'm traveling to Texas this weekend hoping to catch a Golden State Warriors game there on Monday night, and there's an interesting article on the progress of sports betting legalisation in the US at Axios. Actually, it's not really an article but more of a list of bullet points, but a couple are interesting, if not news to readers of this blog.

The writer, Kendall Baker, writes that:

...too many of us equate sports betting to playing roulette (a game of luck), when in reality, it's much more like investing in a stock (a game of skill).

As I've written before, when someone writes or talks about 'bashing the bookie', they are showing their ignorance about how the industry works.

"Bookmakers operate in a manner strikingly similar to that of stockbrokers.The customers of each try to predict the outcome of events. The sports bettor thinks, feels, or relies on someone else's opinion that a specific team will win a game. An investor in securities also thinks, feels, or relies on someone else's opinion that a specific stock's price will increase or decrease. As more money is bet on a team than its opponent, the money odds, or point spread, will increase. As the demand for a certain stock increases, the price of the stock increases." - Jack Moore

Mr. Baker also quotes Ted Leonsis, owner of Washington D.C.'s Capital One Arena as well as the four teams who play there (Wizards, Capitals, Mystics, Valor), about this disconnect between how sports betting and stock trading are viewed:

"It's a generational thing. When someone my age closes their eyes and thinks of sports betting, they imagine some back room with cigarette smoke and cocktail waitresses and a mafia guy walking in with a paper bag full of cash."

"For the younger generation. what conjures up in their minds is playing daily fantasy sports. Rather than thinking of sports betting as this criminal activity, they associate it with the idea of 'how much smarter am I than you because I did my research.'"

Fortunately for those of us with an at least somewhat functioning brain, sports betting attracts not only those who understand what is required to gain an edge, but seemingly many more, almost always uneducated and almost illiterate, who think it is an easy way to get rich. It isn't, despite what you might read on social media, but this dispersion of skill is what we need to be profitable. 

Back in the US and the NHL regular season has just 23 days to go, with one team already guaranteed a play-off place. I'll update the numbers next month, but the basic system currently has an ROI of 12.2% while the more exclusive selections (only 36 bets) have an ROI of 22.1%. The Vegas Golden Knights play tonight and I'm hoping for 12 wins from the last 14. 

The NBA is also in the final stages of the regular season, and March has so far not been kind to the Overs on High Totals backers. It all depends on your entry point of course, and which sportsbook you use for your line, but if you are backing when the line is 225 and over, March has probably not been a pleasant time so far.

Some perspective though - the last full month when backing Overs in the regular season when the Total was 225 and higher was under 50% was December 2009.

Finally, MLB is poised to return. Will the famous T-Bone System be profitable for the 8th time in 9 seasons?  Last season certainly showed no signs of a slowdown and weaknesses in American sports markets certainly seem to persist a lot longer than in say European top-class football. 

Gone Dancing, Madness

One Draw from the ten English Premier League games this weekend, and it wasn't the Big 6 match, means that the percentage of Draws this season drops still further to 19.06% and is challenging 1929-30 for the third fewest in the past 100 years: 

If there should be no Draws in the five Premier league games next weekend (the big teams such as Crystal Palace are all engaged in F A Cup Quarter-Final action) then third place it will be. 

And if there are 10 or fewer Draws from the remaining 81 matches, the 1931-32 record is beaten. The all-time record 12.12% of 1890-91 is safe however, at least for another season.

Moving on to bigger balls, and it's March Madness time in College's version of basketball, and an event I have written about in the past both here and elsewhere. It's like the FA Cup from the third round, with four second round replays pending, played over three weeks on neutral courts, and it's hugely popular in the US.

The 'big dance' is preceded by Conference Tournaments, which for the major conferences, take place this week, but be wary that trends seen in regular season conference play do not often carry over.

Having lived for several years in SEC country, that's the conference I focus on, and while in regular season conference play, the money is on Unders in narrow spread games. The conference expanded in 2012, and over the last five seasons, this system has a 55.2% record in regular season games, but in the conference tournament, the edge on these games moves to the Overs with a 56.2% strike rate. 

With thirteen national champions this century between them, the ACC (7) and the Big East (6) are also worth following. 

The ACC expanded in 2014, and in the last five years, in conference tournament games Overs has a 65.6% winning record, while in 'league' games Unders has the edge at 52.7%.

In the Big East, since the conference was reconfigured in 2013, there have only been 16 qualifying matches in tournament play, far too small a number to be useful, but the regular season offers value on spread bets in close games. 

Stop The Press: Spurs Draw

So Spurs finally draw a league game, ending the streak at 32, their first 'Big 6' Draw in 11 months.

Qualifying Big 6 matches are now +3.62 points from 14 matches, with six games remaining this season.  

Before I departed for another work trip last month, overall the Draw in the Premier League was striking this season at exactly 20%, but since then has declined still further, currently at 19.4% which is the third lowest in top English division since the Second World War, and the 5th lowest in 100 years, in fact since the game-changing 1925 change in the offside law. 

For those who can't remember that time, the impact on goals per game, and thus on the probability of a Draw, was dramatic, In the 1923-24 season, the goals per game average was the lowest it had ever been at 2.474 with no less than 55 goalless draws. 

As attendances dropped, the FA looked at changing the offside law and the current rules, albeit with a few subsequent tweaks, were introduced in the summer of 1925. The result was that the 1925-26 season saw the average goals per game jump to 3.686, and goalless draws drop to just 15.

Averages (mean) can be a bit 'mean'ingless when there are outlier results such as 5-5, 8-3 and 11-2, but a more useful indicator perhaps is that the median goals per game doubled, from 2 to 4.

As readers of this blog and of my articles on the Draw elsewhere will know, the strike rate of the Draw has varied over the years. In the formative first eight seasons of the league, fewer than 16% of matches were Draws. As the number of goals dropped, so the number of draws increased, averaging 22% in the remaining seasons before the Great War stopped play. 

Draws increased after the war until the Offside Rules changed, and the higher number of goals of course meant fewer Draws.

Draws occurred in 29% of matches in 1920-21, a frequency not beaten until 1946-47, and coincidentally both seasons were just after a major war (something to make a note of perhaps, in the event that a Third World War takes place). That record was surpassed in the early days of the Premier League, with 30.74% in 1993-94 and the all-time high of 31.32% in 1996-97.

The first five seasons of the EPL saw Draws in 29% of matches, generally declining to this seasons EPL record low.

Backing the Draw in every EPL game this season (never a good idea) and you'd be down 74.43 points, a -26% ROI, but backing it in games where the difference in 'true' prices is less than 10% and you're again in profit (see above). This strategy is also profitable in every season in the Pinnacle era with an ROI of 20.6% from 159 matches.

Looking at the pre-Pinnacle years, using prices adjusted for the challenging over-rounds of those days (up to 112%), and it's interesting that when the Home team was the slight favourite, the Draw wasn't a value bet, but since Pinnacle came along, it is actually slightly more value then when the Away team is slightly favoured.

When the prices on both teams are exactly the same, the ROI is a rather impressive 96.6% in the Pinnacle era.   

As others have previously noted, in matches between evenly matched teams, backing the Draw is a solid strategy and the overall low strike rate for Draws this season shouldn't be of undue concern.

More interesting than that the Draws have dried up, is the question of where have they gone to? In a word, they have gone Away, with Away wins currently at 33.22% of matches which will be a record for the top division in England if it remains that way. 

25/25

It's been a while since we looked at the Draw in the Premier League, in part because there haven't been that many this season. With 250 games played, exactly 20% have ended as Draws, an all-time low for the Premier League. As always with football, using Pinnacle's closing odds courtesy of Joseph Buchdahl's Football-Data web site, the outcome of backing the Draw in every game is, as usual (2010-11 was an exception), a disaster.

Currently you'd be down 56.97 points from 250 bets, a negative ROI of 22.8%, and you'd be forgiven for thinking the Draw offers no value, but as I wrote back in November:

A more selective rule of less than 25% and this climbs to 41.78 points from 155 matches, an ROI of 27%".

The rule being that of backing the Draw in "Big 6" matches when the difference between the 'true' win probabilities for both teams is in that 25% to -25% range. 

With only seven qualifying matches so far, it's not a big sample (to put it mildly), but these close contests have produced three winners and a 3.58 point profit. 

What about other match types in the 25% to -25% range - the Little 14 and David v Goliath matches you may be asking? Backing the Draw in both these categories is profitable, no surprise there as they are by definition 'close' matches, but what is a surprise is how much profits are improved by ignoring Little 14 matches where the market has the Home team as favourite. 

Overall, the 25/25 system since 2000 is up 73.10 points from 1,456 matches, but eliminate those Little 14 matches where the market makes the Home team favourite, and the profit jumps to 129.91 points, while the number of matches drops to 664. The difference in ROI is huge, almost quadrupling from a decent 5% to a very decent 19.6%.

The outstanding value in the EPL this weekend would appear to be Tottenham Hotspur at a very generous 2.96 with Pinnacle, as advertised in their Tweet yesterday (above). Good luck collecting though. Clearly a palpable error will be claimed as there is no such team as Leicetser City. 

Bayes Laughs, Others Miss The Joke

It was't the most exciting end to the 2018 NFL season with the lowest scoring Super Bowl ever, but from an investment perspective, it was another profitable one.

For those following the Small Road 'Dogs system this season, the regular season ROI was 7.8%, and for those saving themselves for Divisonal games only, the ROI was 9.9%. In the playoffs, the ROI was 30.1%, and for all categories combined the ROI was 9.6%. 

As readers will be aware, the Small Road 'Dogs strategy isn't new. Since the 2002 season, the numbers for the Regular season are:

Those are the baseline numbers. If you are a little more selective, those percentages can all be increased and only one season since 2005 has seen the strike rate below 50%:

It may not be the most exciting way to make money, but 56.9% is fantastic over 17 years. Ignore the claims of those who say they can win coin-toss bets at a strike rate anywhere close to 70% as one Twitter handle (now deleted) claimed:

The esteemed Joseph Buchdahl (@12xpert) has a few tweets on this conman, and I liked this comment from Andrew Mack after the claimed 70% strike rate went closer to 40%:

No one, not even Mel (his claimed strike rate was 50% to 60% at 2.8), can beat the odds long-term by this kind of a margin.

To paraphrase Steve, for profitability you need an edge, discipline, patience and perseverance. A sense of humour might be helpful too:

I may have added at least six years to mine.

Undesirable Behaviour

In my last post, I mentioned a comment made by spanky, whose bio on Twitter describes him as a Professional Sports Bettor. Some of you follow him, and would have seen that a few days ago he posted some videos of himself being restricted at some sportsbooks.

At the time, my thoughts were that this might not be the smartest move to make. Not only do some of the videos show sportsbook employees, but his own face is clearly identifiable, and I'm not sure why someone already having trouble getting on would draw attention to themselves in that way. Sportsbooks in the US have strict rules about filming and clearly his actions with a hidden camera breached those rules. 

Spanky Tweeted on January 14th:

Yesterday, he gripes on Twitter that:

"I just got a certified letter dated January 16th banning me. My “Undesirable Behavior” is winning too much!?! I never even stepped foot in the joint! Regulated Bookmaking in America sucks!"

As quite an expert on undesirable behaviour myself, I'm not convinced that winning belongs in this category. My suspicion was that it was actually his filming in casinos which had brought about the ban, and had nothing to do with "winning too much" but this justification drew the sympathy of many followers judging by the comments. 

But sure enough, there in the comments several hours later was a credible reply explaining that:

Presumably the casino made clear the reasons for the eviction at the time, and leaving out this rather important detail, and promoting one more to his liking is poor form. I wonder why he might do that?

It's probably also not in his favour that he's previously been in legal trouble, with a criminal record, agreeing to a plea deal in Queens, NY, to stay out of jail after being charged by the FBI with:

"enterprise corruption; fourth-, third-, second-, and first-degree money laundering; first-degree promoting gambling; fourth-degree conspiracy; and fifth-degree conspiracy."

That's a lot of degrees! If spanky really has an edge, he should bet on the exchanges (now legal in New Jersey I am told) and keep a much lower profile than has been the case of late.  

The Super Bowl Trap

According to the NFL Players Association, the average career length is about 3.3 years, which makes it all the more remarkable that of the 18 Super Bowls since 2001, the New England Patriots will have played in exactly half of them. Since 1986, they will have played in 11, with a 5-5 record to date. Their opponents next Sunday are the Los Angeles Rams.

Seventeen seasons of data isn't much, but there are a couple of trends if you feel the need to bet on the biggest event of the year, the risks of which I have written about before. 

These big sporting events are seldom good for the serious investor. With so much interest, value is impossible to find. The Derby might be a big deal for the owners, jockeys, trainers, but for the sports investor a winner is a winner whether it be in a maiden race at Cartmel (I've been there too) or the Prix de l'Arc de Triomphe. It might not sound so good in the pub that you didn't have a bet on the Derby but that you had the winner (Breeze Out) in the 2:10 at Hexham, but then why bother? No one except you really cares anyway!

Here's Spanky's opinion on this topic:

The argument against this of course, is that because it's such a huge game, the ratio of sharp to mug money is at its lowest.

You have been warned. This season's game has (currently) the highest total since 2001 at 58.0 points. Of the four previous Super Bowls with a total at 50 or higher, three have gone Under, and the one Over outcome needed an overtime. Overs is historically the play when the line is 50 or less, with a 61.5% record, but the total won't be anywhere near this. But before you lump on Unders, read on.

For the line, Favourites have a poor record, covering the spread in just 5 of the 17 games. However in the five games with the narrowest spreads, Favourites have a winning record. 

The line (currently) for next week has the New England Patriots at -2.5 / -3.0, and a narrow spread tips the scales in favour of Overs, with 7 of the top 8 games ending Over. 

The Rams actually opened as one point favourites, but William Hill reported that 94% of money that came in at that number was on the Patriots, and the line moved. The amount of money bet is useful data - Pinnacle's 'Betshare' is not.