Monday, 30 September 2013

BTTS, Win To Nil, And Clean Sheets

I had an email from Tony, who wrote:

It's always interesting to see your line on a game.
I'm assuming this comes from your ELO methods and certainly picks up on the fact Everton now haven't conceded a home league goal in 636 minutes.
My own line, which is 'not active' as its based on insufficient games this season has Everton at 1.77 and the under/over at 2.08/1.93
It made me think you might see value in the both teams to score market.
My line for that is 1.93 which is where the market is - your model has a higher Everton likely score so your chances of Newcastle being zero must be higher.
First of all, to clarify - it is Elo, not ELO. Elo is not an initialism. Elo ratings are named after Arpad Elo, a chess master who come up with the basic idea for the United States Chess Federation to improve upon the Harkness rating system they formerly employed.

The Both Teams To Score market is simply another derivative of the goal expectancy values. As I have write before, if your estimate of these numbers is more accurate than the market's, then there are multiple ways to profit. If your estimate is lower, then there is obviously going to be value in the Unders markets, and in all of them, not just the Under 1.5 and the Under 3.5, but also there will be value on the draw and on Both Teams Not To Score, as well as several Correct Scores and assorted other related bets.

In my opinion, too many people complicate their betting unnecessarily. If you expect less goals than the market, back the Under at whatever level your risk tolerance is best suited. If you like short losing runs, back the Under 3.5 or Under 4.5. If you're comfortable with longer losing runs, bet at the Under 2.5 level, and if you like the longer prices and can tolerate the associated even longer losing runs, back the Under 1.5 or even the 0-0. They all offer value, if your expectations are correct.
As well as the Both Teams To Score market (essentially a double bet - you are betting using tonight's game as an example, on Everton scoring [Yes or No] and on Newcastle United scoring [Yes or No]), there are also markets such as Everton Win To Nil, (another double - Everton have to score, Newcastle United have to not score), and Newcastle United Clean Sheet? (i.e. Everton not to score), but the point is that all these markets are determined by the goal expectancies.

On the subject of Doubles, I mentioned on September 9th that:
Skeeve maintains his first place, but worth mentioning is the bad luck he has had with his official double bets, having had six all so far come up empty, or rather come up with one winner and one loser. If your double is going to lose, at least the selections could have the decency of both losing or both winning!
In the long run, it makes no difference if you back selections in singles or doubles. Two selections can both win, both lose, or win one and lose the other. On a coin toss, you will win 3 units one time in four, and lose one unit the other three times and thus break even. You might as well just back them in singles so far as I am concerned, and doing just that, Skeeve would be up by approximately 1.57 points rather than down by 7.81. I say approximately because I use the Pinnacle closing prices while the Official bets are recorded earlier by Skeeve. While such a discrepancy is unlikely to last, (a couple of doubles coming in will soon reduce that difference), I'm not sure the irritation of one winner / one loser splits is a price worth paying for the rarer longer priced winning double. Again, it's all down to risk tolerance. If you can detach yourself from the short-term and see the big picture, then it makes no difference as I said, but if you prefer a smoother ride, then singles are the way to go.


I won't publish the latest FTL table until tonight's Everton v Newcastle United game is in the books, as a few people have selections. 

Premier Betting, Rubicon and Webbo all have selections, as do I with another Cassini Value Selection, although their 100% winning came to end at four selections, but the 'all selections end in 2-0' run continued as Getafe beat the value winning by that score at Espanyol.

Here's a quick run down of those entries who are done for the weekend. 

Emp's 23 selections included 10 winners, but a small loss of 1.93 points overall.

Skeeve found two winners from four selections, but made a small loss on the weekend, and Forza Fizzer continued his recent slump with two more losing selections. 

After two profitable weekends backing draws, Punters' Friend Neil came down to earth with a bump finding just one this weekend from 12 selections and a record 8.22 point loss. It rather highlights the challenge with backing draws - two weeks ago, it seemed like the Holy Grail when he found five draws from nine selections, but then you get weekends like this one where almost nothing goes your way. 

Football Elite found one winner from three selections, a lay of Monaco which wasn't enough to make a profit with the weekend resulting in a 0.67 loss. 

Fedslam made a little under 3 points on the back of Aston Villa's win over Manchester City, and Skeeve's Official bets made a 2 point loss. 

The two Bundeslayga selections came up with one winner for a small profit on the weekend, and Scatter Gun found three winners from his standard ten, but for a small loss.

And then there's The Football Analyst, for whom the good news was that he managed to improve on last week's loss of 5.31 points. The bad news was that it was by just one-hundredth of a point in a loss of 5.30 points. I'm sure 

I'll cover the 'live' entries later tonight or tomorrow. 

To the NFL now, and a much better weekend for John Walsh's selections up 4.26 points with seven winners and two losses.
Overall an excellent set of results, but that last loss hurt. With a lot of money looking to back the New England Patriots at 1.01 too early in my opinion, (as I often comment on), I gave up a lot of traded profit laying it. The Falcons came close to taking the game to overtime, but came up one touchdown short. It was high drama at Cassini HQ, and although I evened things up slightly and made £300 before commission, it felt like a loss. 
The regular baseball season ended yesterday, but there is plenty of drama still to come with a tiebreaker game tonight between the Tampa Bay Rays and the Texas Rangers. After that, there are wild-card games on Tuesday and Wednesday before the play-off series begin on Thursday (American League) and Friday (National League). Liquidity has been rather poor of late, well, all season really, but it is a good trading sport when there is some money around. Plenty of pauses and win probabilities for every game state based on well over a hundred years of data.

Finally, I mentioned BIRGing a few days ago, something we all do to some extent or another, but an odd choice of words from Ian Holloway after Crystal Palace's latest defeat shows that perhaps he is attempting to do quite the opposite:
"We need to dig in and find ourselves. I won't keep them [Palace] up if we keep making mistakes. We need to be more clinical and stop making errors."

Saturday, 28 September 2013

Suicide, Skipping And Tissues

A couple of comments on the last couple of posts to address. Emp is suicidal after failing to include West Bromwich Albion to win at Manchester United this afternoon, writing:

And I now feel like killing myself- Forgot to mention my last pick WBA to win against United, and they lead with 20 mins to go......
Indeed, they did, and that's how it stayed. I mentioned yesterday how your mind sees an event differently depending on where your money is, an similarly I'll bet there was part of Emp wishing that Manchester United would score. Anyone following a system will know that on occasion, it is simply not possible to get the bet placed. If the selection wins, it's annoying because you missed out. If the selection loses, you breathe a sigh of relief and thank fortune for getting the loser out of the way at no cost. This is understandable, but if your system has a positive edge, (and if it doesn't, why are you following it?), over time missing out on bets will cost you whatever the ROI is.

Anonymous was unclear on the selections listed.

Are those all home wins, draws away wins?
I get the Swindon Town @ bit but the others aren't clear
When the selection is listed as Braintree Town v Alfreton for example, the selection is Braintree Town, with the versus Alfreton added for clarification. In other words, the selection is the team first mentioned  unless specified otherwise.

The Football Analyst had three winners from twelve selections, but until the prices are available I won't know whether these would have been profitable to follow or not. I'll summarise after the weekend is over. One person who did have West Bromwich Albion to win is Webbo, a result that will certainly help his numbers.

One oddity from Peter Nordsted's Account Bet selections this weekend was the selection of  Over 1.5 and Over 3.5 goals in the Hull City v West Ham United game. What puzzles me is why the Over 2.5 goals is omitted. If the expected goals exceeds the market's expectation, then every Over.5 goals market should be value. The logic of skipping the 2.5 goals market escapes me. In the event, Overs was not a winner with just one goal (a penalty) being scored. Incidentally, I had the Under 2.5 goals priced at 1.65 for this game, which leads me to a question I was posed by email by an XX Draws subscriber on a bonus selection I had made. 
Fascinating..just what I have been waiting have 'Selection' at 1.83 whereas Ian at FTS has them at 2.69! and has them as a lay bet at 1.78 accordingly.
I have queried his ratings before on individual ratings that seem blatantly wrong but he won't have it.
Any thoughts?
All I can say is that I have my football pricing spreadsheet, or model sounds better, and it is improving all the time as more and more data is entered. I can't speak for how Ian prices up matches, or indeed if he does at all. Many tipsters seem to offer a selection as value, yet fail to suggest how they come to this conclusion. No model is 'correct', but some are better than others, and looking at a small sample is meaningless. However, if Ian consistently prices a team at 2.69 and I have the same team at 1.68 (not 1.83 as stated in the email by the way), then clearly one model needs some adjustment.

In my opinion, the only way that you can make the statement that something is value is to price it up yourself, and be able to support your opinion with data. If you are consistently wrong, then you need to revisit your spreadsheet / model and make some changes. Ultimately all prices come down to the goal expectancies - if your expectation of goals is consistently too high for example, then you need to see why and fix it. Maybe it is too high for some leagues and not for others. Again, keeping records is essential.

For example, for tomorrow's Stoke City v Norwich City match, I have the following:
H: 2.11   D: 3.56   A: 4.07   U: 1.85   O: 2.18
As XX Draw subscribers can confirm, when I mention a team as value, I give the price that I think they should be priced at. The four selections so far have been at 1.7 (my price 1.6), 2.12 (1.82), 1.68 (1.52) and 2.03 (1.77). 
I could be wrong, and stand to be corrected, but the feeling I get is that among the FTL tipsters, only The Football Analyst keeps detailed statistics and prices up matches like I do. Skeeve uses his knowledge of the leagues he covers to form his opinions, while Matt at Football Elite looks for strengths or trends that may have been missed by the market (although this strategy may be less effective in recent seasons than it was once). Peter Nordsted looks at a more broader set of historical statistics such as how teams from one category have fared when playing teams from another.

There's no right or wrong approach - just whatever works for you. I lean towards objectivity and an emphasis on data, but the weakness here is that it doesn't account for 'soft' issues such as changes in manager or injuries or suspensions etc. Others lean towards studying line ups and a detailed knowledge of their league with merely a nod towards data. The overall total from the FTL shows how difficult it is to get ahead in football betting, and how people approach solving this puzzle is interesting. 

Late Selections

The Football Analyst's selections for this weekend are:

Braintree Town v Alfreton
FC Halifax Town v Chester
Hydea @ Lincoln City
Ipswich Town v Brighton and Hove Albion
Swindon Town @ Preston North End
AFC Wimbledon @ Cheltenham Town
Bury @ Dagenham and Redbridge
Yorquay United @ Newport County
Morecambe @ Northampton Town
Accrington Stanley @ Plymouth Argyle
Hibernian @ Inverness Caledonian Thistle
Partick Thistle @ St Johnstone
Emp had some late additions:
Back Arsenal @ Swansea City
Back Manchester City v Aston Villa
Back Cardiff City v Fulham
Back West Ham United v Hull City
Back Crystal Palace v Southampton
Back Liverpool vs Sunderland

Tips And Counterbets

The only profitable non-service in the FTL currently is Emp, and his selections, and there are a few of them, are here (additional punctuation on the Bundelsiga selections provided by Emp, not myself, and note that the Borussia Moenchengladbach bet already lost last night):
Lay Nottingham Forest v Derby County
Back Bolton Wanderers v Yeovil Town
Back Burnley v Charlton Athletic
Back Brighton and Hove Albion at Ipswich Town
Back Barnsley at Leicester City
Lay Millwall v Leeds United
Back Queens Park Rangers v Middlesborough
Back Reading v Birmingham City
Lay Sheffield Wednesday v Doncaster Rovers
Back Watford v Wigan Athletic
Back Borussia Moengladbach at Augsburg
Back Freiburg at Borussia Dortmund (!)
Lay Hertha Berlin v Mainz '05
Lay Hoffenheim v Schalke '04
Back Hannover '96 at Bayer Leverkusen (??!)
Back Eintracht Frankfurt v Hamburg
Lay Werder Bremen v Nuremberg
Nothing received from Fedslam or indeed The Football Analyst yet. There were a few selections from the FTL in play last night. Skeeve selected Kidderminster Harriers (v Aldershot Town) but the game ended 0-0.

Football Elite went for Augsburg (v Borussia Moenchengladbach) but that ended in a draw 2-2, as did the XX Draw selection from La Liga of Real Valladolid v Malaga. I'll take the win, but as I have written before, 2-2 draws are too fortuitous to take any pride in.

A few posts ago, I suggested that the better a trader, the closer their long-term average win size was to their average loss size. The discussion continues on the Betfair Forum, although disappointingly not too many people are either willing or able to offer their numbers.

U.A. wrote:
You state that "When you watch enough games, you develop a sense for when the price is wrong". You also state that "I suspect that I cut my winners too short and let my losers run too long".
Why do you think that you can sense a price that is wrong when you place your original bet but not when you place your counterbet? Especially if you conducting a large number of trades on a game. Or do you feel that even though the price is wrong you just can't help but place the counterbet. I find this concept particularly interesting in light of the fact that you don't trade football and that you are referring to markets where there isn't a sudden massive price change. I.e. there shouldn't be the fear of something like a goal which would mean a big price change and consequently a big loss.
It's an interesting question, and my answer was that it's:
Because I am not totally detached when it comes to trading. When I am out of a market, with no exposure, it's relatively easy to identify a price that is wrong and enter the market, and if I enter with a small stake, it is easy to override the emotion of fear and exit at the right time.
What often happens is that in the heat of battle, there's no time to calculate Kelly, so I end up with a higher risk than I am comfortable with. As I mentioned in my rather amusing When Bets Go Wild post of June 2011, I know I'm in too deep when my breathing changes. It's a combination of risk and value, so a larger amount my not trigger the feeling of panic if the value is that much greater. In these circumstances, the overriding emotion is fear, so the tendency is to at least look to reduce the exposure somewhat, even if this is at the cost of giving up some value.
As I wrote in December 2011, [Prospect Theory], quoting from a study by Tversky and Khaneman, who "found that contrary to expected utility theory, people placed different weights on gains and losses and on different ranges of probability. They found that individuals are much more distressed by prospective losses than they are happy by equivalent gains. Some economists have concluded that investors typically consider the loss of $1 dollar twice as painful as the pleasure received from a $1 gain. They also found that individuals will respond differently to equivalent situations depending on whether it is presented in the context of losses or gains."
The last piece is interesting. Depending on where your money is, you see games very differently. If you want a team to score a basket with five seconds left, your thoughts during the time-out tend towards the improbability of it. "They're never going to score", but when you are on the other side of the same game, your mind keeps telling you that there is no way they can fail to score. That time-out gives too much time for thinking, and the cost of giving up a little value but guaranteeing a profit either way can often seem worth it. In a way, this trading out at a less than optimal price is adjusting the risk to a more Kelly like level, but there's a 'peace of mind' factor in play too.
Even though this is a hobby, I don't like to lose, and while I would like to think that I am a cool detached trader, I know that I am not, and giving up a little profit to ensure a less stressful life and smooth out the ups and downs of trading seems to be worth it. In my opinion, to be able to trade optimally, you need to remove emotions from the picture, but that's easier said than done.
And incidentally, price swings can be quite extreme in sports other than football, especially later in a game. The fear of giving up a late, game-changing score, is very much alive and well in many sports.

Thursday, 26 September 2013

2-0 Consistency

Last week, I decided to give away a couple of my top rated value selections from across the big five leagues to the XX Draw subscribers, and after both won, thought I would throw in some more from the midweek matches. I'm a generous chap. Another two winners midweek, but the remarkable thing about all four wins - two home, two away - is that they were all by 2-0. Boring is good. The odds on all four winning were 12.29 (Pinnacle odds) but on all four winning 2-0? High!

I'll add these into the FTL table from next week, but I won't make them eligible for prize money to give others a chance. Oh no - I'm turning into The Football Analyst! Seriously, I'm more interested in how they compare with the comparable Football Elite or Premier Betting account bets who pull from the same leagues as me.

Speaking of Football Elite, and they dropped a point as Granada went down 0-1 to Valencia. I had this as an XX Draw and was thus a little peeved to see the visitors win it in the 90+4'. Two other selections finished 2-0 (Nantes v Nice) and 1-1 (promoted Livorno v Cagliari). So three wins for the Under 2.5 selections, and Peter Nordsted's Drawmaster selection was also Livorno v Cagliari, so a little movement a the top.

Lifting Spirits

Anyone who has been a reader of this blog for the past year or so will know that I long ago highlighted the value in backing Barack Obama to win last November’s US Presidential Election. Nate Silver doesn’t get too much wrong, and some might say nor does Cassini, but there were doubters out there, Al springs to mind, who were of the opinion that Romney would win. Fortunately for the world, Al and his ilk were completely wrong, and fortunately for my trading balance, I was right.

The difference in prices between Betfair and Intrade was commented on at the time, a great arbitrage opportunity for those with an Intrade account, and one possible reason for the discrepancy is highlighted in the study reported below, with Intrade at the time being accessible to the US, while Betfair isn’t.

As the study concludes, there is a possibility that the trade was simply about "market manipulation for political purposes" and whoever lost the money really didn't care. If the losses on Intrade were matched or exceeded by profits elsewhere, then this would make sense. Lose $7 million by artificially boosting a price and attract donations of $8 million that would otherwise not have been made, and the strategy makes sense, and offers a big opportunity for those who can identify the strategy at the time. Clearly something strange was going on, and as Al showed during the days leading up to the election, any reduction in Obama’s implied probability of winning triggered (misplaced and temporary) elation.

The study suggests that “such dynamics are most likely to arise in electoral prediction markets and in sports betting”. When Crystal Palace and Watford played their £120 million play-off final last May, I wondered if the boards of both clubs had looked into hedging by having a big bet on the other team to win. It would certainly have made sense to have guaranteed something from the game (other than the relatively insignificant gate receipts which by tradition go to the losing club). A few million in the market pushing Palace’s price down to evens and the board can say to prospective season ticket holders or other investors “According to Betfair, we have a 50% chance of playing in the Premier League next season”. And with an artificially low price comes the opportunity to back or lay the hell out of it.

The study also mentions that “Some voters appear to have a preference for affiliation with a winning candidate”, BIRGing I believe it is called (Basking In Reflected Glory), and so it is with sports teams. How often this phenomenon offers a value opportunity in the betting markets these days is debatable, but occasionally it does appear that someone is willing to show their support at an unjustified level.

Here’s the article from Yahoo! in full:

Romney loss cost trader up to $7 million: Study

Mitt Romney may have taken the fall in the 2012 presidential election, but the trader who bet as much as $7 million on the Republican nominee was the one who took the hit.

An unidentified fan of the former Massachusetts governor flooded former online wagering site InTrade in the weeks before the election, wagering heavily that President Barack Obama was about to go down to defeat, according to an academic study released this month.

The consequences were twofold-an enormous loss for the wrong-way bettor, and an example of how such sites can be manipulated by one person with a clever betting strategy.

David M. Rothschild of Microsoft Research in New York and Rajiv Sethi at Barnard College believe the trader in question was using a strategy that not only sought personal gain but also looked to raise spirits.

Much of the wagering, they said, was done on Election Day "and at other critical moments of the campaign."

More plausibly, this trader could have been attempting to manipulate beliefs about the odds of victory in an attempt to boost fundraising, campaign morale, and turnout.
In all, the paper looked at a 15-day period during which 3.5 million contracts were traded between 3,200 accounts on Intrade-which has since been disbanded under regulatory pressure-and Betfair.

The sites were popular places for bettors to play the election, and the trends reflecting the probable results were cited frequently by the media:
The single largest trader was responsible for more than one-sixth of (double-counted) total volume, with more than 1.2 million contracts bought or sold in about 13,000 distinct transactions. The most frequent trader engaged in almost 34,000 transactions, accounting for more than 20 percent of (double-counted) total observations. The largest 32 traders (just 1% of the trading population) were responsible for 60% of volume, while the 32 most frequent traders accounted for 57% of transactions.
One of those traders clearly went all-in for Romney.
Trader A was responsible for one-third of the total money on Romney over the two weeks in our sample, and about a quarter over the entire cycle. The result was a loss of close to four million dollars over the two week period, and a likely loss of almost seven million overall. What could possibly have motivated this activity? Given that the trader bet on Romney and not Obama, we can rule out cross-market arbitrage with Betfair as a motivation.
This leaves three possibilities: (i) the trader was convinced that Romney was underpriced throughout the period and was expressing a price view, (ii) he was hedging an exposure held elsewhere, or (iii) he was attempting to distort prices in the market for some purpose.
Ultimately, the analysis "leaves open the possibility" that the trade was simply about "market manipulation for political purposes" and whoever lost the money really didn't care.
The trading losses, while hardly trivial, pale in comparison with the cost of contemporary political campaigns. Beliefs about the likelihood of victory are important determinants of fundraising as well as volunteer effort and morale.
Some voters appear to have a preference for affiliation with a winning candidate, and are prepared to abandon those seen as likely to lose. Turnout can also be affected by perceived candidate viability. Intrade was among the most closely watched indicators of campaign vitality, resulting in incentives for price manipulation to boost support, donations, effort and morale prior to the election and turnout while voting was in progress. The last swing state to close its polls was Colorado at 9pm ET (7 pm local time), and this is almost exactly when the floor in the Romney contract gave way.
The paper found the manipulation of the market successful at least to a point, with the trading strategy remaining intact even if Obama still won the election.

Similar strategies, Rothschild and Sethi said, could find their way into the stock market.
Such dynamics are most likely to arise in electoral prediction markets and in sports betting, but similar effects may well exist also for common stock. Especially in the case of consumer durables, attachment to products and the companies that make them is widespread. It would not be surprising if one were to find Apple or Samsung partisans among investors, just as one finds them among consumers.

Wednesday, 25 September 2013

Americas Cup - Round 19

Once every (usually) three, four or five years, depending on legal wrangling, I become slightly interested in sailing when the Americas Cup reaches its closing stages. This year's is proving somewhat remarkable. For the first time, the final is a best of 17, beating the previous high of best of 9 by some distance, and the USA started effectively on -2 after being caught cheating in March's warm-ups. With an 8-1 lead, New Zealand, not surprisingly, traded at 1.01 on the exchanges. At 8-2, the Kiwis were well clear in the 13th race but they failed to beat the time limit due to light winds in San Francisco Bay and the race was voided.

And now after seven straight wins, (talk about momentum), the USA team has tied the final, and set up a winner take all race today at around 9:15pm. The last time an America's Cup went to a similar outcome, was in 1983 - when the Australian challengers came from 1-3 down to 3-3 in a best of 7. I well remember laying in my sleeping bag on the floor of the living room of my granddad's North Devon bungalow watching the Australians clinch the win 4-3 and hand the USA its first loss in 132 years. It was quite exciting, but my Nan (also a Cassini) was buried the next day, and it didn't seem appropriate to talk about it too much over drinks and nibbles, but for some reason the memory of watching the sailing the night before is as strong as those of the funeral itself almost exactly 30 years on.  The USA are down to 1.31 to win, and as excited as they get about play-off game sevens, one can only imagine how crazy they must be going over a 'game' 17 (or is it 19?)

The Football Analyst picked up one winner from two last night as Braintree Town won at Wrexham. Nuneaton Town did not win the top of the table clash at Cambridge United though, and were hardly likely too with the hosts having a Crystal Palace star in the team, and one who couldn't have been Appiah to score...

No joy for anyone else (Forza Fizzer dropped on point, and Punters' Friend dropped two) and there are some matches in-play tonight for XX Draws and Football Elite. With form starting to settle down, I've started sending out a couple of bonus value selections with the XX Draws, and I probably shouldn't say this but they are currently three out of three. AS Roma, Atletico Madrid on Saturday, and Malaga last night. And one more in action tonight. No selections from Skeeve from the Conference games, and the poor start that most football tipsters are making to the season has brought these comments from my old friend Steve the Hamper Man - Warning - the below contains bad language. Do not read any further if you are under 18 or of a sensitive disposition.
Onto Soccer, what can I say. I don’t usually swear but FUCK it’s shit. The one positive if you can call it that was the fact that FB Elite didn’t make a loss. He made a whopping $650 profit. Combo FI lost $500. But it was the big two again that failed miserably. Let’s first talk about Skeeve, after just the other week saying he was going to down stakes he does quite well last week and the next week is back to full stakes and loses them all. A $4,800 loss and the excuses were inexcusable. Now he is saying that it is always the case that the first half of the season is terrible, if that’s the case then why did he not do this analysis before and realise he has no clue early in the season. We are now stuck once again with the hope of a good second half, but what happens when it ends up being as bad as the start? I don’t mind (well I do) losing the money, but I really hate the excuses. If they were valid then fair enough, but in my worthless opinion they are just crap. Really not sure what to do at the moment, but will continue to ride it out and see what happens, as always, will only stop if the bank goes to zero. Onto an even bigger loss for TFA, I have just combined all the models I follow into one as it was getting too time consuming to update them all each week. This weeks loss was over $10,000 for the second week in a row. At this rate the bank will be dead in the next month or so. Surprisingly I still have some confidence in TFA, only because I see the results and that we are betting value. Value is pointless if they all lose, but if it were for a bit of luck, I could easily be sitting on a small loss. I can see this model turning things around in quick fashion and really hope it does. If it doesn’t then I will be as harsh with TFA as I am with Skeeve. Lets hope that both perform so well over the coming months that they can tell me to shut the fuck up for writing such horrible things about them.
The life of a tipster is not an easy one. In many ways, you are only as good as your last couple of weeks. Doubts soon creep in as to whether the edge has gone, or the tipster somehow has lost his touch, however long a profitable record he may have had.

Steve is wrong with his suggestion that "value is pointless if they all lose". Value doesn't guarantee anything short-term, but if we don't have value, we are guaranteed to lose in the long run. It's someone who really doesn't understand betting who is happier with winning a fair coin toss at 1.7 than losing a coin toss at 2.3. Value IS the point.

I'm glad I'm not the only one who finds Graeme's numerous systems and numbering a little confusing. Fortunately, the selections for the FTL are straightforward enough, and I can just about understand those.

One obvious problem tipsters have is that their selections reveal a lot about their strategy, and opens the door to copycats. Football Elite was the inspiration for me rating the top five leagues in Europe, the same five that Matt dips into for his selections. If Matt had a few seasons of profitably backing in Belgium, the Netherlands, Portugal, Scotland and Russia, my direction may have been different.

Emp commented that he wasn't only thinking in terms of smoothly transitioning promoted teams by ranking lower leagues, but also of being able to profit from them. He has a point. Skeeve finds value - at least after Xmas :) - in the fifth and sixth (south) levels of the English game, and as I have said before, but still done nothing about, this may well be a good league to keep ratings on and track true form. Time is a factor though. I spend long enough sitting in front of my PC updating my spreadsheet as it is. Hard to believe that the Conference is almost a quarter the way through the season as well.

Tuesday, 24 September 2013

Ranking Rabat Ajax

The Football Analyst lost his last bet of the weekend last night as Leyton Orient maintained their impressive 100% so the bottom of the FTL table now looks like this:
With a full slate of Conference games tonight, we may see Graeme dipping his feet in the currently icy waters again. 

There are also a full schedule of matches in Ligue 1, Serie A and La Liga so the XX Draws selections will be in action.

Emp, no doubt feeling rather happy right now, commented on my post on subjectivity / objectivity:
Couldn't agree more Cassini. If you wait till all is clear, there is no longer a price. I think the best opportunities are when other intelligent players feel that they don't want to be involved.
Two points to note:
a) If you were rating the lower league as well, there is no problem with promoted/relegated teams as you can have a seamless transition (unless of course 500M worth of incoming transfers off-season)
b) On transfers and things like that, I think it's just not possible to know precisely what impact new signings/departures/suspensions have, given that football is so team-oriented, that I don't know any particular way to predict whether loss of cohesion offsets improved quality, what the actual quality of transferred players is; etc. I am happy to defer to stats under most circumstances.
And another note, updated table? I have a sneaking suspicion I might have catapulted close to the lead.
The sneaking suspicion was correct. At one time, I did rate all the English teams down to the Level 6 Regional Conference Leagues, but replaced those with the top leagues in France, Germany, Italy and Spain. apart from taking a huge amount of time, the rankings are always from a season of games against lower division teams, and promoted teams usually make changes to prepare, and so in my opinion they would be of limited value, and certainly not worth the effort needed to maintain them. The Euro Club Index site ranks 714 teams, from Barcelona to Malta's Rabat Ajax (Barcelona being the top rated by the way).

Not a great weekend for John Walsh's NFL tips, with just two winners from nine selections:

Monday, 23 September 2013

Empire Strikes Back

FTL As Of 22.Sep.2013
A new leader after the weekend's results, with the Premier Betting Official account bets taking over at the top, but from a prize money perspective top place currently goes to Emp.

Peter Nordsted's Official bets fared better than his FTL qualifying bets due to multiple winning bets on various unders in the Liverpool v Southampton game as well as on the Half-Time score market.

Emp had  a mere 20 bets, with 13 winners and a profit of 7.11 points, while Skeeve slipped with just one winner. The XX Draws Under bets lost 0.58 points with six winners from twelve, while Forza Fizzer split two under bets in Scotland for a small loss.

No luck for the three Drawmaster selections, but Punters' Friend moved up six places with a 1.17 point profit. Football Elite made a small profit with one winner from two, while Fedslam lost 2.81 and Rubicon lost all four.

The Bundeslayga had one winner for three, for another overall loss, while Scatter Gun had his worst weekend yet with a 4.6 point loss.

Webbo had two winners from ten, and the XX Draws picked up 1.79 points with four winners from 12 including three perfect draws. A 90th minute missed penalty by Stuttgart helped the numbers and repaid us some luck.

And last, but possibly not least by the end of today, is The Football Analyst. 3 winners and 10 losers from the 13 matches played so far, but Graeme will likely move out of last place if Brentford (currently 2.09) can beat 100% Leyton Orient in the last game of the weekend. Currently his bounty liability is £325.

As always, any errors are unintentional - just let me know and I will confirm the numbers. Happy Autumn.

Sunday, 22 September 2013

Subjective Objectivity

The inclusion of Sassuolo v Internazionale in this weekend's XX Draw selections no doubt raised a few eyebrows, including my own, but an objective spreadsheet doesn't read facial expressions too well, only numbers. With the season only three games old, the numbers for promoted teams, including Sassuolo, are largely guesswork, and as mentioned by Scott in an email today:

Selections involving newly promoted teams seem intuitively flawed, although it would be interesting to see if the numbers support this.
While we won't know until a few more games have been played how bad Sassuolo are, with four losses (including today's 0-7) and one goal scored, 15 conceded, it's fair to say that they are not looking too good. They did win Serie B last season and only one Serie B champion in the last 10 years has been immediately relegated (which was Pescara last season - more on those little devils later). I see a new trend starting. With Sassuolo's home ground having a capacity of 4,000, they play at Reggiana's stadium this year so perhaps all games should be treated as away fixtures?

I had a look at last season's promoted teams and how they fared when selected, and interestingly the worst league was Serie A. Pesky Pescara were selected 13 times, and not once did they oblige with a draw. Sampdoria were also selected 13 times, drawing only twice, but Torino helped out with four draws from nine selections. Promoted teams in Serie A thus had a strike rate of 6 from 35 (17.1%) and implied odds of 5.83. Clearly we would have been better off in this league by excluding promoted teams.

At the other end of the spectrum is the English Premier league, where the promoted teams were a combined 12 from 31 - a much healthier 38.7% and implied odds of 2.58. Interesting, again one promoted team (West Ham United) failed to draw any of the eight times they were selected, although Southampton more than made up for it with 8 draws from 13.

Across all five leagues, promoted teams were selected 154 times, and 41 ended in draws (26.6%, implied odds 3.75). Based on this evidence, promoted teams probably should be excluded, but the evidence is from just one season, and I'll continue to include them this season. Perhaps with a disclaimer something like "Warning: The inclusion of newly promoted teams can damage your wealth".

Why should these fourteen or fifteen promoted teams by singled out for special treatment? The reason is that while established teams have a solid rating after at least a year and 34/38 games of adjustment, the initial point allocation is necessarily partially subjective. I wrote an article on this subject a little over a year ago for Betting Expert. This season, I had to break my rules for Monaco and make a manual adjustment, as they were clearly not a typical promoted team. At the time of writing, this decision is somewhat vindicated by the fact that Monaco sit atop the league. Given time, Monaco's ratings would adjust, but to knowingly go several matches with a clearly false rating would have been ridiculous. The problem is where to draw the line. Do I adjust Tottenham Hotspur's rating after Gareth Bale is sold, or temporarily adjust Liverpool's rating while Luis Suarez sits out his latest ban? Clearly these are factors to be taken into account before you place serious money on any game, with the spreadsheet numbers just a starting point, but for the weekend warrior, there is only so much we can do.

Several tipsters seem to be struggling this season. Some say that it is still too early, and caution needs to be paid, and there's nothing wrong with that, but the problem is that if you eliminate too much of the early season, and then start wrapping things up in February because it's near the end of the season, you end up with a couple of weeks in late December or early January when the stars are aligned and it is an auspicious time to bet. I exaggerate of course, but one downside of subjectivity is that there are too many reasons not to make a selection.  

Saturday, 21 September 2013

Five Year Anniversary

Happy Premium Charge
I realised today that it is almost exactly five years since Betfair introduced their first Premium Charge - I believe Monday 22nd September 2008 was that inauspicious day. According to my spreadsheet, that was day 996 of my taking Betfair seriously, and my first charge was £260.21 the next week.

Tomorrow is day 2,822 of my Betfair records, so the Premium Charge has been around in some form or another for the great majority of my time on there. Looking back, that opening salvo of 20% from Betfair was quite mild. From those good old days:
Although no figures are available as to how many of Betfair's customers will be subject to the new charge, with the exchange claiming to have up to 500,000 active users, it is not expected to exceed 2,500 account holders.
I have no idea what the demographics are for Betfair's active users today, but there must be a fair, and growing, percentage of customers for whom the Premium Charge has always been there. The good news for anyone new is of course that you can, to a certain extent, design your betting so that you do not face the Premium Charge in the future, an advantage that customers at the time were not afforded. I say 'to a certain extent' because as Betfair did with the Super Premium Charge implementation, it's their game, and they can change the rules as, when and how they like, so you can never be sure. From 2008 again:
One of Elitebet shop's customers, who wished to remain anonymous, added: “The calculations are fundamentally flawed and the only resolution I can see is a mass exodus of clients to another betting exchange".
Well I'm not sure the anticipated exodus was of Biblical proportions, and certainly for many, Betfair remains the only place in town to apply profitable strategies, with many events, even if offered in play elsewhere, having close to zero in liquidity.

Ultimately this is a hobby, and it is still worth my time to make 50% of something, although 80% or 95% were preferable. As I have said before, it's the people who gave up decent jobs to trade full-time who have my sympathy. With the trading climate unlikely to improve, I imagine many are second-guessing that decision or have already returned to the workforce. Those National Insurance contributions might come in handy somewhere down the road.

My profits these days are almost exactly half of what they were pre-Premium Charge, 51.4% to be precise, but while I don't keep track of my hours spent trading, it is certainly less than it was before, and part of the decrease is for that reason.

Premature e-Publication

I published the The Football Analyst's selections yesterday upon receiving them, much to the chagrin of Anonymous, who wrote:

As you know Graeme has closed at 100 members to protect prices.
But I understand that from now on every one can put pressure on these prices only a few hours after they are released to paying subscribers.
But then again, you have a chance of winning £25!
I actually wasn't aware that Graeme limits his membership to 100, nor am I aware of when the selections are released to paying subscribers. As Graeme says in his comment on the comment:
Cassini had no way of knowing I didn't release the bets until Friday night and it was my fault for not saying to Cassini in the email that I didn't want these published until Saturday morning.
I have no issue with Cassini posting the bets as long as it's 24 hours later after I release the bets. I should have mentioned it in the email to Cassini.
Now that I am aware of the sensitive nature of these selections, and the fact that the selections for the FTL overlap those sent to subscribers, I shall delay publication. Anonymous is right about one thing - after paying over £1,600 in Premium Charges last week, I certainly do need that £25!
I actually didn't look at backing Graeme's selections until this morning, and although the Betfair graphs leave much to be desired, it did appear that there was generally a recent dip, before the price recovered. Graeme did continue:
I do feel people need to cut me a little slack at times. I limit members, I quote below market prices for proofing/recording purposes and now someone isn't happy that some bets are appearing on another blog. For two seasons, I posted every bet on my own blog beforehand and gave away the bets for free.
I do understand that having paid for something, you don't want to see it being given away for free, and hopefully Anonymous understands this was a miscommunication / misunderstanding that won't be repeated

After all that though, another down day for Graeme today with his 12 Saturday selections just wrapping up. Three winners and nine losses.

Skeeve had a poor day too - five FTL selections and just the one winner.

The XX Draws have so far had two winners snatched away late in the game - Everton's 85' winner at West ham United and Hannover '96 scored an 89' penalty to beat Augsburg - but one man's pain is another man's joy, and Matt at Football Elite will have been turning cartwheels having selected Hannover '96 to win.

Peter Nordsted will also be a happy man today, with his Account Bets already guaranteed to be in profit for the weekend. He predicted a low scoring Liverpool v Southampton game, a game which caught out several entries, as did the Norwich City v Aston Villa game which many, including myself, predicted the draw.

Friday, 20 September 2013

FTL Selections September 21-23

The Football Analyst has his selections in for this weekend, and twice as many as last week. Here they are for anyone following:
Lincoln City @ Luton Town
Tranmere @ Notts County
Oldham Athletic v Crewe Alexandra
Stevenage v Carlisle United
Rotherham @ Walsall
Blackburn Rovers v Huddersfield Town
Middlesbrough v AFC Bournemouth
Hartlepool @ Bristol Rovers
Newport County @ Exeter City
Fleetwood Town @ Portsmouth
Torquay United v Cheltenham Town
York City @ Wycombe Wanderers
Ipswich Town @ Wigan Athletic (Sunday)
Brentford v Leyton Orient (Monday)
I typically only publish the selections from those in profit (and not subscription services obviously) but since everyone stands to make £25 off Graeme, we can all hedge by backing his selections if we want to! As you know, Graeme had a poor week with seven losers last week, and currently sits in 16th place with a £300 liability to the 12 qualifying entries above him, but Mark has faith in him, commenting:
Followed Graeme's soccer bets for a long time and I would take 1.5 on him topping this league at the end of season.Any takers?
I can see a whole new derivatives market springing up around the FTL placings.

The Drawmaster selections from Peter Nordsted, with Pinnacle prices, are:
Liverpool v Southampton @ 4.11
Cardiff v Spurs @ 3.64
AC Milan v Napoli @ 3.44
Rubicon is less than a point in the red, so here are his selections for this weekend:
Back Liverpool v Southampton
Lay Manchester City v Manchester United
Draw West Ham United v Everton
Draw Norwich City v Aston Villa

Wednesday, 18 September 2013

Wipeouts And Winners

FTL as of 17.Sep.2013
Not the start to the season that The Football Analyst was hoping for I'm sure, or is Graeme Dand just giving some of us mere mortals a head-start? All seven of his Saturday selections lost, six by one goal, and AFC Wimbledon by two goals, and if he keeps this up for too much longer, there will be a flood of entries into the FTL eager to collect the bounty on offer for finishing ahead of Graeme.

Graeme also had seven selections from last night's lower league schedule, and found two winners for a small loss. With his offer of £25 to anyone finishing above him, Graeme is currently on the hook for some £300, and as entry to the FTL remains open, if you think you can stay ahead of him with a 7.16 point start, then you have a free play. Win prize money or at least get your money back.

In stark contrast to TFA, Skeeve had a good weekend, despite the Friday night loss. In the FTL qualifiying bets, his selections of Nuneaton, Forest Green Rovers, Kidderminster Harriers and Salisbury City were all winners. Surprisingly, no more selections from the midweek games. Skeeve leads by 2.2 points.

A solid weekend too for Peter Nordsted with his Premier Betting selections profitable overall, and one winner from two for his Drawmaster selections, which move up to fourth. Premier Betting moves up to second in the table.

The biggest mover of the prize eligible entries was yours truly and the XX Draw Under 2.5 selections had a good week with 8 winners from 13, and are now at -0.27. Hard to gloat too much though when the flagship of the service - the XX Draws - still languish in last place although they did show a profit of 3.68 points with 5 winners from 13 this weekend.

Just outside the prize money is Rubicon who made a small profit this weekend, followed by Fedslam who made a small profit.

Forza Fizzer fizzled this weekend with selection St Johnstone losing 1-2 to Hibernian despite scoring first.

Emp reduced his opening week's losses by finding 14 winners from 23 selections for 1.02 points profit.

The Scatter Gun spluttered a little, with a loss of 3.2 points on the weekend, and then comes Football Elite (the only service with an 'official' entry but not a "prize-eligible" entry). FE went for Elche on Monday night, but to no avail. Football Elite also has the dubious honour of the lowest ROI at -60% but he has had only six selections so a couple of winners will change that number fast. Strangely five of the six selections have failed to score, never mind failed to win.

Webbo had one winner from five for a small loss and then we come to Punters' Friend Neil, who had a great weekend switching to picking draws and found five from nine. I say great 'weekend' because unfortunately for Neil, his five midweek draw selections all ended as home wins so Neil only moves up by a couple of spaces overall.

The Bundeslayga had just one selection, and a losing one at that.

As always, it is possible that I have erred in some way, so please let me know if you think this is the case.

NFL update now, and the smallest of profits (0.16 points) for John Walsh.
John's season total is 2.31 points, comprised of Totals +3.4, Moneyline +1.15, Handicaps -2.24. Interestingly (to me) the average margin of victory in both Weeks 1 and 2 is 7.63 points with 22 of the 32 games so far being decided by a touchdown (and point after) or less.

Back to real football, and the topic of the lack of goals in the EPL so far this season has received much attention, including from the always very readable Mark Taylor at The Power of Goals. A Mark (may or not be the same one) did drop a comment here, which was:
Perhaps one significant factor could be the change in managers at several top clubs. Man Utd, Chelsea, Man City and Everton all have new managers and at least two of them are not renowned for attacking football. I haven't done any analysis on this theory, but it could be a possible 'intrinsic' reason.

Sunday, 15 September 2013

Dry First Half

On the topic of the low-scoring start to the English Premier League this season, I wrote yesterday my opinion that 

"Unless there is some intrinsic reason for the drop off in goals, a change in the offside law that I missed or a reduction in the size of the goals, the goals will pick up soon enough"
Emp comments that:
It's too early to know if this is a trend or just variance, but to me an "intrinsic reason" for the drop in goals seems quite plausible.
10% of a season is indeed far too early to be drawing any firm conclusions from, but if there is an 'intrinsic reason' for the drop in goals, I do not know what that might be. The only change to rules or officiating is the introduction of goal-line technology, and unless something like nine goals a week are now correctly being disallowed, this is unlikely to be a significant factor.

Has the quality of goalkeeping improved dramatically, or that of strikers declined? No team is averaging above two goals scored per game, and no team is averaging more than two goals conceded per game, so this is possible.

Extreme weather can be ruled out, but there is one external factor that may possibly explain why games are a little tighter to date.

With the TV rights up by 70.2% on last season, the value of Premier League membership is more valuable than ever. Just a thought, but one that I don't seriously think explains the decline. I think it will turn out to be variance, but one statistic is particularly interesting, which is that the decline in goals is biggest in the first half.

Last season at this stage there were 52 first half goals. This season just 29. Since the opening weekend, only one team has scored two or more first half goals, and only one of the last 28 matches (and six all season) have been over 1.5 first half goals.

16 of 39 matches have been 0-0 at half-time, and 17 have been 1-0 or 0-1.

Hold the front page for a Swansea City v Liverpool goal-fest tomorrow.     

Saturday, 14 September 2013

Below Average

With 'only' 3.8 rounds of matches completed in the EPL, any worries from 'fans, pundits and punters' over the drop in goals this season might seems a little premature, but 38 matches represents exactly 10% of the season. Right now, the average goals per game is 1.92, while the average over the last three full seasons is 2.8.

Peter Webb discusses this in his post today and while the current rate is certainly below average, it should be noted that the last three seasons have seen more goals than any other period in Premier League (20 team) history.

Over 18 completed seasons, the average is 2.64, ranging from a low of 2.45 (2006-07) to a high of 2.81 (2011-12), so 1.92 is well out of this range with exactly 10% of the season played.

Strangely the percentage of draws is lower than might be expected with a lot of binary scores, currently at 23.7%. Only six teams have seen a team score more than two goals, Arsenal (twice), plus Aston Villa, Cardiff City, Crystal Palace, plus both Manchester teams, and only five teams have an average goals (total) per game exceeding the long-term average - Arsenal, Aston Villa, Manchester City, Sunderland and Swansea City. It won't last, but will the average of recent years be reached? Probably not.

The remaining 342 matches would need to see an average per game of 2.89, a total not reached in a full top division season since 1967-68 when the average was 3.03.

I have no official prices yet for today's games, but backing the Under 2.5 goals over the first 30 games this season would have seen a profit of  11.5 points (using average prices), with 27 Unders (71.1%). As a comparison, last season saw just 44.2% of games stay Under 2.5 goals.

Unfortunately knowing what Unders backing would have returned is of academic interest, and turning this data into money is quite another matter. Peter generously suggests that:
At this stage we’d cautiously suggest that ‘unders’ represents slightly greater value for the next few weeks. However, there could be real value in ‘overs’ from weeks 11-20, as the bookies might over-react to the low scoring start. This is also historically the most profitable quarter of the season for backing ‘overs’ (there tends to be a slight drop in goals in the third quarter, while ‘overs’ gets priced in during the fourth quarter).
I somehow doubt that the bookmakers will overreact to a low scoring start. Unless there is some intrinsic reason for the drop off in goals, a change in the offside law that I missed or a reduction in the size of the goals, the goals will pick up soon enough, although as I said earlier, probably not by enough to match the averages of the last three seasons.

Friday, 13 September 2013

Something For The Weekend

The Football Analyst has his selections in for the weekend, and they are:

Home Wins for Macclesfield Town v Alfreton Town and Newport County v Morecambe

Away Wins for Braintree Town @ Grimsby Town, Oldham Athletic @ Rotherham United, Swindon Town @ Wolverhampton Wanderers, Torquay United @ Rochdale and AFC Wimbledon @ Chesterfield.

Skeeve had one selection in action tonight, but Luton Town went down at Wrexham which means the slenderest of leads as we head into the weekend proper. Lucky for me that the selection arrived too late for me to put any of my hard-earned money on it - note to Skeeve - some people need more than 90 minutes or so notice!

There were a couple of excellent comments on my post earlier regarding risk from Matthew, and I shall return to this topic and the comments later in the weekend, but they have been published and are worth checking out.

Ratios And Risk

I posted the question of loss size : win size ratios on the Betfair Forum, and have had a few more replies.

One Forum member Darlo Bantam felt that to compare the numbers "properly" (whatever that means), the strike rate was needed, but for me the question was not how much people make, just whether my theory that an ace trader would have a close to 1:1 ratio (loss size : win size) and a less expert trader have a higher loss size to win size ratio.

I suspected that better traders would be close to 1.0 on the ratio with their strike rate being somewhere north of 50%.

Mardock posted that after 3,000 markets, his ratio is 0.82, with a strike rate of 72%.

Bayes, a much respected Forumite, posted that:

In tennis, my main trading sport, I win in about 60-65% of markets and my average win to average loss ratio is very close to 1. Four or five years ago this ratio would have been nearer 0.8 but I am much better nowadays at staying 'at market'.
His numbers for a couple of bots were reported at:
Markets 1513, Win rate 50.9%, Win/Loss Ratio 0.815
Markets 2724, Win rate 57.8%, Win/Loss Ratio 1.115
Then for tennis, and manual trading, he came up with:
All-time: Markets: 5374 Win rate 62.4% Win/Loss ratio 0.998
Pre-2010: Markets 2570 Win rate 62.6% Win/Loss ratio 0.961
TheInvestor2, another well respected Forumite, reported from the world of football:
For the first quarter of 2013, my average winning match gives me about 88% of the average losing match, and I won on 75% of matches (counting manual trading only). That was a pretty good quarter though.
Pretty close to Bayes, (and myself), and then jptrader chipped in with:
For a large number of mainly in-play football markets my average win size is about 88% of the average loss size, remarkably close to both OP and Investor.
So most traders do appear to have larger losses than wins, (wins typically ~80% of losses) and the better traders do tend to be closer to a ratio of 1:1.

There was one comment on Mark's numbers, which was posted anonymously, and a little sarcastically it has to be said:
Is that the "Professional Sports Analyst" Mark Iverson? So over 7 years he's made £200K, just have to hope his football app is bringing in the big readies cos his trading isn't.
Marks's contribution to this debate was much appreciated, and it should be made clear that Mark's numbers refer to just two sports, cricket and the NFL, and to trading activities on those two sports only.

And speaking of Mark, is he finally rethinking his strategy of easing up at the end of a calendar month? It appears so. My Casssini v Iverson post of just over a year ago (July 2012) is the ninth most popular all-time. Maybe not 'popular' but the ninth most read. What took Mark so long?

Mark is introducing a 'Market Risk Level' indicator, although what determines high, medium or low risk isn't explained. It could be volatility - certainly football would fall into the 'high volatility' category with a goal having a big impact on the outcome, but other sports like cricket or basketball have lower volatility because, (except near the end of course), any one score has less of an impact on the result. Mark wrote:
...if I’m trading a low risk event my mindset is set to be aggressive, whilst if it’s high risk I’m keen to be more cautious.
For me, distinguishing between a low, medium or high risk event is tough - at least before the event starts! If the definition of 'high risk'is an event that I am likely to lose at, then I skip it, or play with small stakes.

Play-offs certainly seem to traditionally be a challenge for me, but generally speaking I have my good sports, and my bad not so good sports, and I have no idea which games within a season are going to be good for me, and which are going to be bad for me. If I knew that...

While Mark's screenshot is the teeniest of sample, his three biggest wins come in the two low risk and one medium risk game, with the high risk games producing the smaller wins (and one small loss).  

Peter Nordsted did stop by to clarify the John Walsh selections question from Paul, explaining:
John Walsh charges me a fee for his NFL picks and I then provide these picks to readers of my newsletter which is a £15 per month subscription.
He also provides picks during the NHL season
Nick from GolfBetting sent me a message, and while golf is no longer one of my top sports, (far too much time relative to returns for me personally) he has a site full of links that some of you golf people may find useful:
Hi Cassini, Hope you don't mind but I couldn't find an email on your blog so thought I'd try this one. I thought on putting a comment on a post but thought it might be a bit spammy. I've just started a website on Golf form and have put a link on there to your blog. I know Golf isn't one of your main sports (although I seem to remember the occasional post on it). But I thought I should let you know the links there. If you have a spare minute to have a look at the site and have any thoughts, that would be great -
One thing puzzled me about the site though, was the description of this blog:
Green All Over – possibly the best betting related blog around. Mainly taken from the angle of Betting Exchanges and focusing on Football and US Team sports. Well written (and sometimes combative in style!), there is plenty that can be gleamed from the blog if you want to bet successfully.
Possibly? Combative? And I think Nick means 'gleaned' although as gleam is defined as 'A brief beam or flash of light' perhaps gleamed is just as appropriate in this case. This blog can certainly be most illuminating.

Green All Over - the betting blog that shines.