Saturday 25 September 2010

Drawing Encouragement


A busy football weekend got off to a good start with a draw between Koln and Hoffenheim backed at 3.5. The Elo ratings are particularly good at finding value on the draw, with a strike rate last year in the Premier League of 38.67%, although that figure includes cup matches between teams, something I am not tracking moving forward. I'm now backing the draw in the top European leagues, and only eight games in, but so far so good. Fifteen more selections this weekend. I like hunting the draw. Except for the occasional end of season Italian game, the price is always over 3.0, which means of course that if you can hit a draw every three matches, a healthy profit ensues.

Pete Nordsted is having some success with a similar strategy in the Premier League, and has so far found at least one winner from three draw selections each weekend of the season. Of his three selections this weekend, two match mine (Blackpool v Blackburn Rovers and Fulham v Everton) but he has Newcastle United v Stoke City whereas my third selection is Wolverhampton Wanderers v Aston Villa. I should add that I don't look for three draws - that's just how it worked out this weekend. Some weeks I might have none, other weeks 10. Not likely, but I think it's a mistake to set yourself a specific number of bets. In France I have 5 (of 10) selections, in Spain 3 (of 10), in Italy 3 (of 10) and in Germany (including last night) 2 of 9.

Back to Peter's strategy, and the article says

The strategy uses a past history of team’s results to provide a series of matches likely to draw from the English Premier League. Each of the teams in the Premier League has been graded from A-D with A containing the Champions League hopefuls and D being the relegation fighters. Grouping the teams in this way allows the strategy to reduce the number of variables and account for promotion/relegation without interrupting the collection of his vital statistics.

Having grouped the teams into these categories, he will look for patterns and previous results from similar graded teams when it comes to the upcoming fixtures.
Although I don't take past results directly into account (they are in there as an adjustment to the ratings of course), the strategy sounds very similar to something I wrote about almost exactly one year ago!

9 comments:

Anonymous said...

Do you know wow would your record on draws in the PL would have looked last season had you started at the beginning of the season during the dearth of draws?

As far as I can tell, both yours and PN's method of choosing draws simply involves finding two vaguely similarly matched teams and betting the draw at any price? You say your system is good at finding value on the draws - but what are you defining as value? You simply appear to bet them whatever the price when they meet your loose criteria.

I wouldn't be particularly pleased with myself if I came up with the same bets as PN. I've looked at one of his books and it's extremely flakey with some horrible advice. Him and his partner bang on about value seeking and then base selections on what they THINK will happen. And claim they get value prices. e.g. They claim they get 6-5 on an evens shot by removing a scoreline they THINK won't happen and dutching the other scorelines. Nonsense obviously. And there are so many other mistakes, fallacies and contradictary advice it's unreal. e.g. They seem to think you can price up a game based on about 4 recent games.

Anonymous said...

Wow should obviously read "how" in the first sentence before you jump on a spelling error. There are probably numerous others too.

Anonymous said...

I would like to have a free copy of the book. My email is s.khanna143@gmail.com

Anonymous said...

I have read their books and the comment they price up the game based on 4 recent games is completely wrong

They base there ratings on rolling 30 game spread.

Also obtaining a better price on short price shots by eliminating certain factors isnt nonsense

Every bet you make, you're eliminating the other outcomes from your thought process aren't you to make the bet in the first place?

So to me if you can get 6/5 dutching a winning margin rather than taking 4/6 on a team just winning the match, then I am all for it.

Anonymous said...

What you say is true. However, the way they present their "dutching" is such that they suggest the "dutched price" is value and the straight win price isn't. Which is normally nonsense.

Anyway, they bang on about value througout their writings yet clearly have little idea as to what represents value.

And just about all bets I have seen them tip up this season are based simply on last season's stats for each particular team with no consideration for numerous other factors. I don't have anything to hand but they spout crap like Utd scored lots of second half goals last season, therefore back them on the Draw-Home HT/FT forecast. With no consideration of price.

If I had the book to hand I could rip it to pieces. So much plain wrong, or totally misleading, analysis.

Anonymous said...

You see that is you problem, you're making assumptions about what they do without having the hard evidence of facts. I every single bet all 29 of their account bets this season they have given out and not one of them is the bet you have mentioned above. I also have all the bets they have given out during the world cup, once again they have never given out a Draw-Home HT/FT forcast. The value they speak about is to do with their ratings which is totally free to view on their site.

What they then explain based on their ratings (price) is a way of ramping up the price by eliminating other factors out of the match.

As for my thoughts on value, there is no right or wrong answer as its all down to the individual and only becomes value because you feel that percentage of a certain outcome is better than the bookies do

Anonymous said...

Take a look at pages 353-360 in Premier Football Betting handbook 2010/2011 for a whole host of terrible advice.

There are something like 50 suggestions for general bets involving specific teams this season. All of which are based purely on last season's stats. And just about all of which make no mention whatsoever even of what price to look for.

e.g.:

"Back under 2.5 goals in Stoke's home and away games"

Yep - that's it. No price, no consideration of who are the opposition are. No clue.

Anonymous said...

It says "14 of 38 matches finished with 3 or more goals (36.84%), backing under 2.5 at above 1.58 would have yielded a level stakes profit".


Sounds fairly simple to me, if you had backed unders above 1.58 in all Stokes games your winning.

If you dont like that way of reading past data and using it for future picks thats fine use your own system for finding selections.

Anonymous said...

So you think it makes sense, based on results over just one season, to back unders in all Stoke's game this season if you get 1.58 or better? And you think betting the unders in every game last season at 1.58 or better was a good bet? You need to have a rethink

The logic (incorrect logic) implies that, for example, you should be backing under 2.5 in Chelsea v Stoke at Stamford Bridge at 1.58? That would be absolute madness - It would clearly be an exceptionally non-value bet.
Think about it.

The logic applied to the suggested "system" of backing Stoke unders at 1.58 or better, regardless of opposition, venue or any other factors, is completely backward thinking. There is no argument that can possibly change that simple fact. It may have proved a profitable investment last season but that, my friend, doesn't make it correct. For each team in the Premier League, it would be very simple to come up with a similar piece of highly incorrect advice. Indeed, that's what they have done.

Here's an example to explain why it's so wrong:

Remember the season Arsenal went unbeaten?

If we wound the clock back to the end of that season then Pete Nordstedt's book for the next season would come up with the stat "Arsenal never lost last season, laying Arsenal's opposition to Arsenal at any price was a profitable system with a profit of X and a ROI of Y%. Advice: Lay Arsenal's opposition blind every game"

I suppose you think this would have been correct?

Essentially, all they are doing is using hindsight to come up with some parameters which would have been profitable for a given subset of results within the Premier League. In terms of foresight, well, I would think they're fairly clueless.