Rugby Quarb System - The Punters Lounge

Wumpking · 31-08-2007, 13:28

This system will focus on identifying 'Quasi-arbitrages' or 'Quarbs' in fixed odds rugby union matches and exploiting these as hopefully profitable betting opportunities.

In essence, a Quarb occurs where one bookmaker is offering prices out of line with their competitors but not so far apart from the others as too cause an arbitrage. The theory of Quarbs was first outlined in the work of Paton and Vaughan Williams (2002) in 'Quarbs' and Efficiency in Spread Betting: can you beat the book? This theory was further developed and applied to fixed odds betting by Smith, Paton and Vaughan Williams (2005) in Information Efficiency in Financial and Betting Markets.

The system also draws upon other weak form information inefficiency theories, particularly the ‘favourite-longshot bias’, whereby low probability outcomes are overvalued and high probability outcomes are undervalued. Other forumites, particularly mcgin, have analysed the ‘favourite-longshot bias’ in rugby - see for example http://www.punterslounge.com/forum/showthread.php?t=38441 Indeed, this system owes much to mcgin’s work outlined in this thread and is based on his original paper/real life trial so many thanks mcgin -

The rules of the Rugby Quarb System are as follows:

Only home sides in rugby union matches, both domestic and international, will be considered for betting purposes. This is because there is a strong home bias in rugby (e.g. 68.5% of games played over the last 3 seasons in the Magners League were won by the home side). Furthermore, the draw outcome has limited impact (e.g. only 2.1% of games played again in the Magners League over the same period were drawn).
Only mean odds of 1.50 or less will be considered from an analysis of at least 8-10 bookmakers’ prices (normally using the Odds Comparison website). The underlying assumption of the system is that the consensus market price is a good indicator of ‘true’ probability. In this respect, the subjective probability of the home side winning a match will be at least 66.7%.
Where a bookmaker’s price is out of line with the mean by at least 5% or more (or 7% for Betfair to take into account commission), a Quarb or betting opportunity is considered to occur.
Stakes are based on the Kelly criteria with an added modification to limit risk and are calculated by the following formulae: Current Total Bank x % Difference between Mean Odds and Best Odds x Subjective Probability of Winning (as derived from the mean odds).

I tested this system on all Rugby Union games during Feb – May 2007 and had a strike rate of 85.4% and yield of 16.9%, hence this will be a real life system and will start with a bank of £1,000.

Would appreciate any input / suggestions for improvement from other forumites.

Wumpking · 31-08-2007, 13:53

Match No.	Date	Tournament	Home Team	Best Odds	Worst Odds	Av Odds	Av Odds +5%	Selection
1	31/08/2007	Magners League	Cardiff Blues	1.61	1.50	1.55	1.63	No Bet

mcgin · 31-08-2007, 17:41

Quote:

Originally Posted by Wumpking

The system also draws upon other weak form information inefficiency theories, particularly the ‘favourite-longshot bias’, whereby low probability outcomes are overvalued and high probability outcomes are undervalued. Other forumites, particularly mcgin, have analysed the ‘favourite-longshot bias’ in rugby - see for example http://www.punterslounge.com/forum/showthread.php?t=38441 Indeed, this system owes much to mcgin’s work outlined in this thread and is based on his original paper/real life trial so many thanks mcgin -

Cheers for the acknowledgment, although the original idea stemmed from from work done by Datapunter & the guy that runs football-data.co.uk

Good luck anyway, I'm sure you will be successful as I saw every reason there were huge inefficiencies with some of the books out there (keep an eye on Stan James he was always providing some far out odds)

A word of warning, you might want to be wary of anything > 1.30. I found those to be a bit dodgy when I was running this. Although that was more of a subjective thing rather then me tracking it and seeing the results.

Am looking forward to the WC myself. I believe there will be plenty of short favs that will fall into the system and was actually thinking of resurrecting my old thread for it. Mightn't have to now

Here's hoping you can beat my 33 in a row

trollerboy · 31-08-2007, 17:55

can you tell where to get infos about this for soccer?

Wumpking · 31-08-2007, 18:35

Thanks mcgin for the input and good wishes. Funny you should mention Stan James. I certainly agree that they tend to offer good odds in terms of rugby prices. I'm also currently 11 from 11 with them for rugby bets so enjoy betting with them!

Whilst I appreciate the warning in betting at odds > 1.30, I've found the system works well up to 1.50. I feel the risks are manageable, particularly as the staking plan is also linked to the mean odds.

Your run of 33 in a row was fantastic and sets the milestone -

. I was sorry to see your thread end at this point. I'm not confident this system will match that type of run as I expect it to lose around 10-15% of the time. I just hope it starts well

.

Wumpking · 01-09-2007, 10:33

Match No.	Date	Tournament	Home Team	Best Odds	Worst Odds	Av Odds	Av Odds +5%	Selection
2	01/09/2007	Magners League	Scarlets	1.37	1.33	1.35	1.42	No Bet

Whilst Scarlets potentially qualify as price < 1.50, currently it is a no bet as no value can be obtained. Will monitor Betfair prices up to kick off to see if minimum value price of 1.44 (i.e. average odds +7%) can be obtained.

Matthew · 01-09-2007, 11:51

Good luck with this Wumpking.

Looks very interesting.

Wumpking · 01-09-2007, 21:23

Quote:

Originally Posted by Matthew

Good luck with this Wumpking

Thanks Matthew for the good wishes.

Fortunately for the system, Scarlets price on Betfair never reached 1.44 as would have been a losing bet!

Wumpking · 03-09-2007, 22:05

Enough bookmakers have priced up Friday's opening match of the RWC to assess. France potentially qualify but are currently no value. Will monitor during week just in case any additional bookmaker’s prices are set out of line with the mean odds.

Match No.	Date	Tournament	Home Team	Best Odds	Worst Odds	Av Odds	Av Odds +5%	Selection
3	07/09/2007	RWC 2007	France	1.20	1.12	1.18	1.24	No Bet

Wumpking · 06-09-2007, 17:55

Unlikely that France's price will meet the system requirements for a bet.

Have decided to include all the RWC 2007 matches in the system with the favourite considered to be the 'Home' side. Assessment of the remaining weekend matches are as follows:

Match No.	Date	Tournament	Home Team	Best Odds	Worst Odds	Av Odds	Av Odds +5%	Selection
4	07/09/2007	Magners League	Dragons	2.37	2.00	2.17	2.28	No Bet
5	08/09/2009	RWC 2007	New Zealand	1.00	1.00	1.00	1.05	No Bet
6	08/09/2007	RWC 2007	Australia	1.00	1.00	1.00	1.05	No Bet
7	08/09/2007	RWC 2007	England	1.00	1.00	1.00	1.05	No Bet
8	09/09/2007	RWC 2007	Wales	1.03	1.01	1.02	1.07	No Bet
9	09/09/2007	RWC 2007	South Africa	1.02	1.00	1.01	1.06	No Bet
10	09/09/2007	RWC 2007	Scotland	1.00	1.00	1.00	1.05	No Bet
11	09/09/2007	RWC 2007	Ireland	1.00	1.00	1.00	1.05	No Bet

No bets on any games as no qualifiers.

Wumpking · 07-09-2007, 21:32

Quote:

Originally Posted by Wumpking

Unlikely that France's price will meet the system requirements for a bet

Thankfully it didn't as again would have been a losing bet!

Wumpking · 10-09-2007, 20:06

Have now assessed the matches for the 12th Sept as enough prices available:

Match No.	Date	Tournament	Home Team	Best Odds	Worst Odds	Av Odds	Av Odds +5%	Selection
12	11/09/2007	RWC 2007	Argentina	1.00	1.00	1.00	1.05	No Bet
13	11/09/2007	Magners League	Ospreys	1.40	1.28	1.34	1.40	Ospreys

Argentina game is a no bet as no value can be obtained. However, the Ospreys qualify for what is the first bet of the system.

Selection is Ospreys - £35.00 win @ 1.40 with Blue Square -

.

Wumpking · 10-09-2007, 21:37

Enough prices available to assess the Tonga v USA match on the 13th Sept:

Match No.	Date	Tournament	Home Team	Best Odds	Worst Odds	Av Odds	Av Odds +5%	Selection
14	12/09/2007	RWC 2007	Tonga	1.40	1.22	1.29	1.36	Tonga

Selection is Tonga - £63.00 win @ 1.40 with Blue Square -

.

F. · 10-09-2007, 22:41

Hello,

First of all, rugby is not my sport... I barely know the rules

but I regard it as one of the best sports to watch once you get the hang of it...

Anyway, I got interested in your system because you are basing it in sound mathematical theories:

Quote:

Originally Posted by Wumpking

The theory of Quarbs was first outlined in the work of Paton and Vaughan Williams (2002) in 'Quarbs' and Efficiency in Spread Betting: can you beat the book? This theory was further developed and applied to fixed odds betting by Smith, Paton and Vaughan Williams (2005) in Information Efficiency in Financial and Betting Markets.

The system also draws upon other weak form information inefficiency theories, particularly the ‘favourite-longshot bias’, whereby low probability outcomes are overvalued and high probability outcomes are undervalued.

Not only your selection criteria seems solid, you are basing your staking plan in Kelly's criterion, also very solid (mathematically speaking)

Quote:

Originally Posted by Wumpking

Stakes are based on the Kelly criteria with an added modification to limit risk and are calculated by the following formulae: Current Total Bank x % Difference between Mean Odds and Best Odds x Subjective Probability of Winning (as derived from the mean odds).

As I'm sure you already know the main downside to Kelly's staking plan is that your bank will be on a "rush mountain ride", but your risk limitation variant should smooth it out...

I'll be following with interest. Best of lucks to you...

F.

Wumpking · 11-09-2007, 20:46

Quote:

Originally Posted by F.

I'll be following with interest. Best of lucks to you...

F.

Thanks F. for the interest and good wishes, it is much appreciated.

Sadly the system has got off to the worse possible start with the Ospreys losing at home to the Scarlets

. As such, the current status is:

Starting Bank	£1,000.00
Current Bank	£965.00
Total Staked	£35.00
Number of Bets	1
Winning Bets	0
Losing Bets	1
Strike Rate	0.0%
Yield	-100%
Profit / Loss	-£35.00

F. · 11-09-2007, 22:33

It is still too early to tell. Don't despair !!

F.

Wumpking · 11-09-2007, 23:26

Early days indeed F.

It is Sod's Law however, that the first bet of the new rugby season would lose. Also particularly galling as the Ospreys haven't lost at home in the Magner's League since May 2006!

As you state, Patience is the key, Discipline is the way...

Wumpking · 12-09-2007, 12:14

Assessment of matches later today are as follows:

Match No.

Date

Tournament

Home Team

Best Odds

Worst Odds

31-08-2007, 13:28	#1 (permalink)
Wumpking Seasoned Punter Join Date: 09 Jan 2007 Posts: 464	Rugby Quarb System This system will focus on identifying 'Quasi-arbitrages' or 'Quarbs' in fixed odds rugby union matches and exploiting these as hopefully profitable betting opportunities. In essence, a Quarb occurs where one bookmaker is offering prices out of line with their competitors but not so far apart from the others as too cause an arbitrage. The theory of Quarbs was first outlined in the work of Paton and Vaughan Williams (2002) in 'Quarbs' and Efficiency in Spread Betting: can you beat the book? This theory was further developed and applied to fixed odds betting by Smith, Paton and Vaughan Williams (2005) in Information Efficiency in Financial and Betting Markets. The system also draws upon other weak form information inefficiency theories, particularly the ‘favourite-longshot bias’, whereby low probability outcomes are overvalued and high probability outcomes are undervalued. Other forumites, particularly mcgin, have analysed the ‘favourite-longshot bias’ in rugby - see for example http://www.punterslounge.com/forum/showthread.php?t=38441 Indeed, this system owes much to mcgin’s work outlined in this thread and is based on his original paper/real life trial so many thanks mcgin - The rules of the Rugby Quarb System are as follows: Only home sides in rugby union matches, both domestic and international, will be considered for betting purposes. This is because there is a strong home bias in rugby (e.g. 68.5% of games played over the last 3 seasons in the Magners League were won by the home side). Furthermore, the draw outcome has limited impact (e.g. only 2.1% of games played again in the Magners League over the same period were drawn). Only mean odds of 1.50 or less will be considered from an analysis of at least 8-10 bookmakers’ prices (normally using the Odds Comparison website). The underlying assumption of the system is that the consensus market price is a good indicator of ‘true’ probability. In this respect, the subjective probability of the home side winning a match will be at least 66.7%. Where a bookmaker’s price is out of line with the mean by at least 5% or more (or 7% for Betfair to take into account commission), a Quarb or betting opportunity is considered to occur. Stakes are based on the Kelly criteria with an added modification to limit risk and are calculated by the following formulae: Current Total Bank x % Difference between Mean Odds and Best Odds x Subjective Probability of Winning (as derived from the mean odds). I tested this system on all Rugby Union games during Feb – May 2007 and had a strike rate of 85.4% and yield of 16.9%, hence this will be a real life system and will start with a bank of £1,000. Would appreciate any input / suggestions for improvement from other forumites.

31-08-2007, 17:55	#4 (permalink)
trollerboy Junior Punter Join Date: 19 Nov 2005 Posts: 55	Re: Rugby Quarb System can you tell where to get infos about this for soccer?

31-08-2007, 18:35	#5 (permalink)
Wumpking Seasoned Punter Join Date: 09 Jan 2007 Posts: 464	Re: Rugby Quarb System Thanks mcgin for the input and good wishes. Funny you should mention Stan James. I certainly agree that they tend to offer good odds in terms of rugby prices. I'm also currently 11 from 11 with them for rugby bets so enjoy betting with them! Whilst I appreciate the warning in betting at odds > 1.30, I've found the system works well up to 1.50. I feel the risks are manageable, particularly as the staking plan is also linked to the mean odds. Your run of 33 in a row was fantastic and sets the milestone - . I was sorry to see your thread end at this point. I'm not confident this system will match that type of run as I expect it to lose around 10-15% of the time. I just hope it starts well .

01-09-2007, 11:51	#7 (permalink)
Matthew aka plopplop Join Date: 08 Aug 2006 Location: Nottingham Age: 28 Posts: 547	Re: Rugby Quarb System Good luck with this Wumpking. Looks very interesting. __________________ http://www.punterslounge.com/forum/s...d.php?p=855832 CARE TO SHARE? http://www.punterslounge.com/forum/s...ad.php?t=55571

11-09-2007, 22:33	#16 (permalink)
F. Unbelievably lucky Join Date: 30 Jun 2006 Location: Mexico City, MX Age: 33 Posts: 196	Re: Rugby Quarb System It is still too early to tell. Don't despair !! F. __________________ Probability is nothing but common sense reduced to calculation.

11-09-2007, 23:26	#17 (permalink)
Wumpking Seasoned Punter Join Date: 09 Jan 2007 Posts: 464	Re: Rugby Quarb System Early days indeed F. It is Sod's Law however, that the first bet of the new rugby season would lose. Also particularly galling as the Ospreys haven't lost at home in the Magner's League since May 2006! As you state, Patience is the key, Discipline is the way...