Sunday 22 November 2015

Under Taker

Gratifying that at least a couple of people have had their interest in NBA betting piqued by my recent observations on the effect of big spreads on Totals betting.

bossman megarain (possibly not his real name) commented on Spread The Word:
Very interesting post, involving a lot of data analysis/research.
I rarely bet on basketball, (which is probably a mistake), but was fascinated by the book 'Gaming the Game', which, as you are probably aware, largely recounts the tale, of a bent basketball ref, shaving points on games, he had bet on.
In trades across many sports, 'mug' punters want to see action, and many 'Mr Unders' type betting strategies will be profitable, as it seems the lines are inflated, to account for this public need.
Keep up the good work.
Whether or not rarely betting on basketball is a mistake rather depends on whether your betting would be profitable in the long-term.

The book Bossman references, "Gaming The Game" is about the 2006-07 scandal involving NBA referee Tim Donaghy and professional gambler Jimmy Battista.

The suggestion that Unders betting strategies might be profitable because mug punters like to see scoring is one that has been mooted before. While Unders is the preferred long-term strategy, I'm not convinced it is because of the 'mug' punters.

Across the major US sports, Unders wins more than 50% of the time (excluding Pushes) in every sport, but only in the NHL would blindly backing it be a profitable strategy. Since the start of the 2013 season, Unders are hitting at 52.34%, which at Pinnacle Sports' 1.952 represents a 72.62 point profit from 3,749 bets, an ROI of 1.94%.

In Canada, the Canadian Football League from 2013 on has an even better ROI of 3.72%, but from just 321 bets. Only three more games to play this season with the Division Finals today, and the Grey Cup Final next week. In the NFL, last week was a big week for Unders with 10 wins from 13 (two Pushes).
Incidentally, another winner (2.1 on Betfair) for the BLUnder System last night as the Miami Heat and (still winless) Philadelphia 76ers played out a 187 point game. The Sixers led by 17 points at one stage, so this loss won't do anything for their confidence! The BLUnder ROI moves up to 23.8% on the season after 15 selections.

Friday 20 November 2015

Stewing It Under

Football Investor Stewboss asks:

Daft question but are you advocating betting in the Overs / Unders market or the handicap market for these?
I was told years ago in a training class that "there are no stupid questions", but the trainer had apparently never met Stewboss :)  I'm just kidding. It's good practice to look at other possibilities.

On double digit handicap spreads, there is no obvious edge with the split over the last three full seasons (and 2015) 51.6% to 48.4%. Going back to 1995 onward, and it's even closer at 49.2% to 50.8% over 3,929 events.

On the rare 'very' large spreads, those of -17.5 or more, there is an edge in backing the big favourite to cover the spread, but since 1995 there have been just 64 such opportunities resulting in 38 wins, 26 losses.Since 2005, these are 16-11.

Strangely in the five seasons from 2008 to 2012, only 2010 saw any selections, so you would need some patience to follow this, although last season (2014) did throw up a whopping nine selections going 4-5. The Golden State Warriors were the big favourite in six of these (2-4), and have also been the selection twice this season already (0-2).

So my suggestion is to look at backing the Unders in the Totals markets on games where the handicap is large.

Wednesday 18 November 2015

Spread The Word

Jack Sparrow, possibly not his real name, commented on yesterday’s Blunderful post:

The spread was -12, In your previous post you mentioned -16.5, can you clarify this please?
"In the last 20 NBA seasons, only one team favoured by 16.5 points or more has lost the game straight up."
Are we to use all double digit spreads or is there a cut off?
The story about the straight up success of teams with a spread of -16.5+ was meant as merely an introduction to the world of big NBA spreads in general. Perhaps no one else found it interesting, but that the Brooklyn Nets came within a missed lay-up of ending a sequence dating back to 1998 seemed to me worthy of mention.

The line that I use for the Blunder System is shorter than the -16.5 of that story. Games of this spread are relatively rare, and don’t generate enough bets each season to make it worthwhile.

Jack asks “Are we to use all double digit spreads or is there a cut off?

First, there is no “we” here – no obligation for anyone else to act on these findings at all. I’m simply sharing some findings with you all, free of charge I might add, for you to completely ignore or for you to modify and take the idea further.

Second, it’s important to understand that there is no single official line for spreads or totals on these games - witness Betfair’s habit of including several spreads for each game, or the differences between various sportsbooks in the US.  The Spurs v Trail Blazers game of Monday night is a case in point. Jack says the spread was -12, the source I use had it at -12.5. You need to make your own choice as to what data to use.

Third, yes there is a cut off that I use, based on my choice of spread source, historical results and number of bets to be generated per season (for me, around 100 or so). Anyone else wanting to follow this idea would need to reach their own decision on where to draw the line (pun intended), preferably after looking at historical results generated from their selected data source.

Pegguy Wort, again possibly not his real name, commented on the initial post on this topic, asking:
What size of handicap do you look for when you're selecting games for this system?  
As the table above shows, backing unders on any spread of -6.5 or more would be profitable over the past ten full seasons, so it's a personal choice of where to draw your line. If you want 100 bets a season, you'd be looking at around 11 for your spread. If you want two bets a season with a bigger ROI%, you might look at -17.5 and greater!

The San Antonio Spurs are again a selection tonight, with a spread of -14 v the Denver Nuggets and a Total Line of 195.5. The Spurs are 3-0 for us so far this season on Unders.

Tuesday 17 November 2015


Another very comfortable winner last night for the newly christened BLUnder System - Big Line Under - as the San Antonio Spurs and Portland Trail Blazers played out a 64 point first half in a game where the Totals line was set at 201 points. Although calculated to 1.952, the Unders (200.5) traded pre-game at 2.08 on Betfair.

The final points total was 173, and the win by 28 points was the easiest since Feb 9th, 2015.

Yet another International break this past weekend provides a good opportunity to review the Bundeslayga system which is also showing another nice profit on the season to date.
Profits from strategies like these might not be very sexy compared with the thrill of trading, but they are a lot less time consuming, and a lot less stressful.

Sunday 15 November 2015

Widespread Unders

In the last 20 NBA seasons, only one team favoured by 16.5 points or more has lost the game straight up.

March 17th, 1998 was that famous occasion, when the Washington Wizards favoured by 17 points lost by one point at home to the Denver Nuggets some 6,451 days ago!

The reason I tell you this is that last night saw the closest we have yet come to a repeat, with the Golden State Warriors (-17.5) needing a late 3-pointer to tie and take the game into overtime to beat the Brooklyn Nets.

The all-time record of teams at -16.5 or more is now 98-1. Perhaps the Warriors pre-game price of 1.03 / 1.04 was value, but a few backers would have been sweating had they been watching the game.

The reason I happened upon this little fact is that during the Summer, I was asked my opinion on NBA Totals betting in games where one side was heavily favoured. My gut response was that Unders was more likely to offer value, my thinking being that a team with a comfortable lead has a tendency to psychologically ease up, rest starters, and in addition the valuable extra points from the fouling towards the end of a close contest are absent.

I looked at the data for the last twenty seasons to see if this idea had any merit, and found that while the older data from 1995-2004 showed more equality between Overs (474) and Unders (480), the numbers since 2005 showed more of a bias (54.8%) towards Unders (487 - 402) and since 2010, even more of a bias at 56.5% (247 - 190).

As a result, we are now busily backing Unders on qualifying games, although 'busily' isn't really the appropriate adjective to use here, since the number of qualifiers is a little under 100 per season, a very manageable number for a part-timer.

Similar to the UMPO System, this is not a "get rich quick" strategy, but it is useful for generating slow and steady returns, for mitigating current and future Premium Charges if you are on Befair, and with just two losing (both small) seasons since 2004, it's another relatively low-risk system that doesn't require hours of precious time.

Last night was a busy one though, with two qualifying games - the previously mentioned Golden State Warriors v Brooklyn Nets game which finished Under despite going to overtime, and the San Antonio Spurs v Philadelphia '76ers game which ended with a total of 175 points, i.e about as comfortably Under as they come.

The numbers above are calculated to a price of 1.952, which is the usual Pinnacle Sports price, and are intended as a guide.

Saturday 14 November 2015

Royals Killing, Car Pimping And Hairdressing

It has been a few days since I posted - unfortunately my 88 year old father was admitted to hospital a few days ago, and remains there in a precarious state of health, so I have been busy ferrying my Mum to and from the hospital and providing support as best I can, which is not my forte I have to admit. There can be few places as depressing as a hospital ward full of confused old people whose lives are all but over.

Since my last post, I had the pleasure of attending the Rugby World Cup Final at Twickenham and although at first sight it appeared that Steve M (currently in the midst of a massive winning run) and myself were sat on each others laps, on further review it appears we were actually diagonally opposite each other (Steve's view on top, my view below):

It seems a long time ago now, but a memorable day out.

Also rapidly fading into the past is the 2015 MLB season, which saw the Kansas City Royals win the World Series versus the New York Mets and single-handedly see the UMPO system return a handy profit. Here's baseball expert Tony's thoughts on the subject:
Enjoyed your October blogs, as always, and good to see the UMPO end up with another good year thanks largely to those Royals.

You will have the precise figures but it looks like UMPO made about 9.5% in the end, going 17-19. Within that the Royals went 9-4 with over 53% ROI.

Reminiscent of last year with UMPO going 18-14 and, within that, the Royals going 7-3 (+54.5% ROI).

The general public really doesn't seem to like the Royals, which has been great for UMPO followers. I think it's because the Royals are unspectacularly good. The general market jumping on board playoff games seems to go for the big name pitchers followed by power batters. The Royals really don't have either. What they do have is the ability not to strikeout, to run the bases well putting pressure on the other team, and to play great defence. None of these qualities drive the public imagination when it comes to betting.

This article in the Pinnacle betting articles archive touches on that from last year. In the example used, only 11% of bettors were on the Royals - but these were the heavy hitters.

There has been a lot of criticism on twitter/blogs etc today about Eric Hosmer making that game tying run in the 9th, saying he should never have attempted it and he just got lucky. What do the Royals have to do to get the praise due to them? Really pleased to see Joe Peta (12-2 in October by the way, +10.77 units) come out with this response today looking at the true value decision of running vs not running.

Finally, 2016 World Series odds came out today - still no love for the Royals, they are 11th Favourite at 18/1, the Mets are 2nd favourites at 10/1. Give me strength! All UMPO followers will be hoping for the Royals, and other underrated teams like them (Giants?) to be back in the playoffs next year.
I have to admit that with everything else going on, my interest this year was limited to putting on bets rather than watching the games. The actual 'official' figures (although in practice these were beaten quite handily) for UMPO 2015 after the World Series were:
As Tony noted, wins on the Royals contributed all the profit - 5.84 points to be precise, as they went 8-4 during the post-season. The Royals are not a fashionable or big market team, which is what casual bettors tend to look for.  The Pinnacle Sports article Tony provides the link to above is well worth checking out. It includes these lines:
Three times, the percentage of bets on KC was in single digits - In the months we had been running BetShares, we hadn’t seen this many single-digit occurrences in any major markets -and in those three games, the lines barely moved against the Royals, if at all.
Money moves our markets, so if the market hasn’t moved, the amount of money on each side should be fairly even. In a case where only 5% of bets are on one side and 95% on the other, the total amounts wagered on either side are very likely similar, in which case the average bet on the 5% side is nothing short of massive. It’s well-known that one habit of sharp (big) bettors is to fade the public; when you recognize that Pinnacle Sports offered the best price in the market on the Royals’ opponents in all three games, you understand why the public was betting against Kansas City, and what those sharps saw to keep the market in check.
We asked our lead MLB trader how much the line would move if all those bets were of the same size. He estimated it would represent a probability shift of 30-40%, so 60-80 cents.
Those bets on the Royals were likely very, very big.
I probably owe Leon Pidgeon of Simple Soccer Stats something of an apology, having lambasted his corner betting analysis in this post last month. Mr Pidgeon (perhaps not his real name)
I normally enjoy reading your blog as it's regularly updated, interesting and intelligent. I also have a chuckle at your sharpness but not this one as I see my site is on the receiving end. The one plus point is I know someone actually reads what I post on the blog.

Yes, 4 games is a ridiculously small sample size but all these stats are taken from pages on the website and they provide these stats whether a team has played 1 match or 45 matches and it's automatic. For me the important phrase in my post was "The price is just a guide to convert past results into a decimal price."
That is the reason I provide these equivalent odds on my site so a user can see what a percentage is in terms of odds. If something has happened 57% of times then I think it is useful to know that equivalent odds for that would be 1.75. It probably helps when making decisions on whether to bet or not and saves time for site users. None of the stats are recommended bets and I did point that out to the user. I don't see how that is irresponsible.
Back to enjoying your other posts now.
I can be a little acerbic at times, part of my charm, and perhaps I was a little harsh on Leon. I do still have an issue with the comment "if something has happened 57% of times" in this context. For example, that a team has been awarded 10 or more corners 57 times in their last 100 matches might be true, but those 100 matches were all unique events, and lumping them together for the sake of analysis isn't logical and any price generated from such an approach would be nonsense. A team might well be averaging three goals a game over whatever period, but it makes a big difference to their expected goal figure if their next opponent is Manchester United or Yeovil Town.

Finally James Butler, author of Programming for Betfair: A Guide to Creating Sports Trading Applications with API-NG commented:
I am guessing that we are of a similar age, considering your musical taste and that you are a fellow reader of The Punter's Revenge (TPR).
Great content, awful binding. The book sits on my shelf in pieces. My first degree was in astrophysics during the 1980s. One day the computing teacher presented a lecture on AI in which he discussed BEAGLE, which is mentioned by its creators in TPR.
I got my hands on a copy of TPR and attempted to replicate Forsyth's BBC Micro code (unless you have a different copy of the book) with not much success on my ZX Spectrum. Forsyth was the programmer whilst Drapkin was the "punter". I still communicate, now and again, with Tony who still produces HOOF ratings from his home in France.
The lecture on AI made me realise that I was in the wrong game (at the time - now I wish I was an astrophysicist rather than a code jockey) and I dropped out of the degree. I was a a full-time amateur cycle racer at the time and got a few fill-in jobs as a BASIC programmer and mini-computer operator whilst seeing if I had it in me to be the next Robert Millar. I was a good climber but not able to keep up with the peloton on the mostly flat courses.
I gave up cycle racing and went back to study when I was 25 years old. This time choosing Essex University, which, if anyone is interested, has a damn fine computing department. I have done post grad at Oxford but I still think Essex is better at computing than Oxford. It's only negative is that Essex not a Russell Group university, which probably says more about the Russell Group than it does Essex. Even though I jokingly add, after someone asks "Where did you study?" and I answer "Essex. I have an honours degree in car pimping and hairdressing." I shouldn't really put down Essex or myself like that.
TPR is a good read, even today. You can get a free copy in PDF format from Tony's HOOF ratings website. Link can be Googled as my eyes are just about buggered for today.
Aside from our ages, it does appear that James and myself have at least a few common links, although my copy (1987 first edition) of The Punter's Revenge is actually still bound together quite nicely, and is still dipped into from time to time.

We both had brief cycling careers and links with the University of Essex which certainly suited my son who graduated with a First and is well on his way to a successful career in Finance and Accounting.

That's all for now. Posts may be few and far between for a while, but there is a wealth of valuable and interesting content in the archives. Even with no new posts, the average hit count over the past two weeks was close to 500 and surely
people can't be wrong!