Simple EPL Lay System

[crazybadger]

Recently I tested a simple EPL system and found it was looking sucessful in the early days but I didnt have the time to follow each match. Instead I have pulled out all matches from the 08/09 season (and betfair prices) from my database and it still looks good so I thought I would share for anyone interested...

Theory
The Favourite in the Full-Time match odds market won 53.95% in 08/09 (57% in 07/08, 52% in 06/07, 53% in 05/06, 52% in 04/05). If you take these selections into the Half-Time market then in 08/09 they only won 37.37% (43% in 07/08, 38% in 06/07, 42% in 05/06, and 36% in 04/05). Based on this I thought there may be good value in Laying the Full-Time favourite in the Half-Time market. I did say it was simple...

Stats (08/09)
To analyse these I've pulled out the absolute highest and lowest prices that each runner was matched at prior to kick-off. I then also took the middle of these prices. I would prefer to get an average price (or even a mode price) but the data in the Betfair CSVs wasn't organised too well and I will need to re-write my extraction tool, but for now these will give a rough picture. It also is just using a flat staking plan (as I havent figured out when/if there is a good frequency to update stakes - eg daily would not work but would weekly/monthly?). And I've also factored in 5% commission too so if you have a lower rate then the results would be obviously better.

Selections: 378 (Betfair data was missing 2 games from the season)
Wins: 236 (62.43%)

At Lowest Price
Risked: 594 units
P/L: +23.4 units
ROI: 3.94%

At Highest Price
Risked: 682 units
P/L: -4 units
ROI: -0.60%

At Mid Price
Risked: 639 units
P/L: +9.5 units
ROI: 1.48%


So that's alright but not great. You wouldn't get the worst lay price every single time (and that is only just a slight loss) and would expect to be more around the middle price so you should be able at least to get some profit. However the interesting thing I found was when I added a rule to only lay those that were 3.0 or more in the Half-Time market. That had the following stats

Stats (08/09) - minimum price of 3.0
Selections: 150
Wins: 109 (72.67%)

At Lowest Price
Risked: 317 units
P/L: +18.7 units
ROI: 5.90%

At Highest Price
Risked: 365 units
P/L: +8 units
ROI: 2.19%

At Mid Price
Risked: 342 units
P/L: +13.2 units
ROI: 3.87%

That is much better. Also because these are "favourites" then they prices never get above 4.0. As such, with a strike-rate around 70%, the staking plan could be a little more aggressive than 1% as there are no large prices winners that really hurt.

Thought/questions/comments?

[MattR]

Interesting CB. Looks promising. Are you extracting the team names with the data? Just wondering what the figures would be like if you removed the big 4 from the equation and were to lay all the other qualifiers?

[crazybadger]

Interesting idea. I also thought of exlcuding Home teams...so only laying the Away team favourites as they might be less likely to get up at half time when on the road...

Exclude Big 4 Qualifiers:
Selections: 238
Wins: 157 (65.97%)

P/L @ Lowest price: +12.1 Units (ROI = 2.78%)
P/L @ Highest price: -4.5 Units (ROI = -0.91%)
P/L @ Middle price: +3.7 Units (ROI = 0.79%)


Exclude Big 4 Qualifiers AND exclude any under 3.0:
Selections: 133
Wins: 97 (72.93%)

P/L @ Lowest price: +18.1 Units (ROI = 6.43%)
P/L @ Highest price: +9.1 Units (ROI = 2.81%)
P/L @ Middle price: +13.4 Units (ROI = 4.45%)


Exclude Home Team Qualifiers:
Selections: 105
Wins: 65 (64.76%)

P/L @ Lowest price: +11.8 Units (ROI = 6.94%)
P/L @ Highest price: +3.8 Units (ROI = 1.95%)
P/L @ Middle price: +7.8 Units (ROI = 4.24%)

[MattR]

Thanks CB.

That exclude big 4 and all else under 3.00 appears very good as does your away teams idea. Out of interest how does it work out if you were to go totally opposite and lay just the big 4?

This seems a really good idea in principle, the first half has less goals on average (I think it's around 43% of goals in 1st half) so it makes good sense.

I wonder what a lower scoring league like Serie A or the French league would come out like?


Is there a correlation to the ht odds compared to the match odds (I would assume so) if so I wonder how it would work out if you were to lay the haf time on just those teams who were 2.00 or above for the match odds.

[crazybadger]

That exclude big 4 and all else under 3.00 appears very good as does your away teams idea. Out of interest how does it work out if you were to go totally opposite and lay just the big 4?
It's alright but nowhere near as good as the others.


I wonder what a lower scoring league like Serie A or the French league would come out like? I also tested Serie A, German Bundesliga, French Ligue 1, and Spanish Primiera and no luck. A couple were positive for the "Best odds" price but most results were negative and the extra filters didn't add anything. EPL is the standout performer so far.



Is there a correlation to the ht odds compared to the match odds (I would assume so) if so I wonder how it would work out if you were to lay the haf time on just those teams who were 2.00 or above for the match odds.
A short priced FT fav is genrally the HT fav too. However where the FT fav is not as clear cut then often the HT fav option is the draw. I like the idea of laying those at 2.0 or more in the FT market. I will have a look at that...

[MattR]

That's interesting that just the Premier is a standout one. So it doesn't seem to have a correlation to the average goals of a league which I thought it might do. I thought the spanish and german being higher scoring may not be so good, but the french or italian may have been better but it seems not then. What about the Greeks - they seem to have a fair share of 0-0,1-0,1-1 games. Espeically if the likes of Olympiacos, AEK and Panathinaikos were removed. Having said that AEK have let me down a few times so they might be prime candidates for this!

Look forward to seeing what else you discover on this CB.