Away Team Experiment - Bookies

Culchie wrote:

My mate 'Statto' seems pretty sure it will be ahead at the end of the season.

On what reasoning does he/she base that on?
Has he done any back testing? any testing at all?
Is it just a hunch?

As I said before, the Away-win "system" will not end up in profit, at least not statistically. It may happen to be slightly in profit this season, but it may not be the case next season or the season after that. Over time, if you place a bet for an Away Win on every match, you will lose. Statistically, it is a loss-making system.

If he has a mathematical/statisical reasoning as to why he thinks it would be a profitable system, then get him to write it out and you can post it here. I dont think others should waste their time trying to follow and record this system.


Meanwhile, the suggested Selected Draw system may have something to it statistically but it will need to be investigated. It has drawbacks, such as ensuring that the correct teams are selected who are most likely to have 0-0 or 1-1 draws.

I dont have the stats from last season or other seasons on how many 0-0 or 1-1 draws occurred for the likely "selected" mid-table teams. The table from last season and the season before would suggest that there are many draws in the middle of the table. Only many seasons of testing would fully prove that this system works. In the meantime, a mathematical calculation could be done on its effectiveness on previous seasons if all the results are known. If someone has the data, it could be tested whether this system would be able to beat the market.

Lets say on average that there will be 3 games each week and that the odds offerred will be 15/2 for a 0-0 and 5/1 for a 1-1. For this system to break-even, it is necessary that out of the 228 bets, 32 are winners with 14 0-0's and 18 1-1's, approx. Below is a table that shows profit levels depending on the frequency of the 0-0 or 1-1 results:

Odds Odds
0-0 1-1 #Bets Euro 0-0 1-1 Profit Profit% WinningBets
8.5 6 228 228 10 12 -71.0 -31% 9.6%
8.5 6 228 228 11 13 -56.5 -25% 10.5%
8.5 6 228 228 12 14 -42.0 -18% 11.4%
8.5 6 228 228 13 15 -27.5 -12% 12.3%
8.5 6 228 228 13 17 -15.5 -7% 13.2%
8.5 6 228 228 14 18 -1.0 0% 14.0%
8.5 6 228 228 15 19 13.5 6% 14.9%
8.5 6 228 228 16 20 28.0 12% 15.8%
8.5 6 228 228 17 21 42.5 19% 16.7%
8.5 6 228 228 18 22 57.0 25% 17.5%
8.5 6 228 228 18 23 63.0 28% 18.0%
8.5 6 228 228 19 24 77.5 34% 18.9%
8.5 6 228 228 20 25 92.0 40% 19.7%
8.5 6 228 228 21 26 106.5 47% 20.6%
8.5 6 228 228 22 27 121.0 53% 21.5%
8.5 6 228 228 23 28 135.5 59% 22.4%
8.5 6 228 228 23 29 141.5 62% 22.8%
8.5 6 228 228 24 30 156.0 68% 23.7%
8.5 6 228 228 25 31 170.5 75% 24.6%
8.5 6 228 228 26 32 185.0 81% 25.4%
8.5 6 228 228 27 33 199.5 88% 26.3%
8.5 6 228 228 28 34 214.0 94% 27.2%
8.5 6 228 228 28 35 220.0 96% 27.6%

Note: 6 = 5/1 plus return of original bet, no tax deducted !
Note: 8.5 = 15/2 plus return of original bet, no tax deducted !

redspider

masterK wrote:

I have done some analysis on last years results on 0-0 and 1-1 draws. From last years Premiership if we put the qualifying list as:
... Portsmouth, Southampton, West Brom
There were 11 0-0 draws, 25 1-1 draws .. a profit of €235
If we had only backed 1-1 draws the profit would have been €400 on an outlay of €1,100 which would have been 36% return on investment

Good analysis MasterK. That shows that the system with one season of stats would prove profitable. One question though, at the start of the season, would someone betting have selected Portsmouth, Southampton and West Brom? Anf if these teams were left out, would that affect the result. One of the difficulties with the system is selecting the team, is theri criterion to do that or is it based on hunch. I understand where you get your hunch from from a footballing sense, but for a system to be fool-proof, it needs objectivity.

In terms of the analysis, "one Swallow does not a summer make", hence more season tests are needed. You've done your bit MasterK, mayb someone else wants to volunteer to do another season. I'd say if we have analysed 10 seasons, it may show some sort of pattern. ie: if every season is in profit, then it is unlikely to be an anomaly, although that can never be ruled out.

I wonder are 1-1 bets on their own a winning system?

redspider

Slightly of topic but one bet that I like to do now and again is the double result. Say Chelsea @ H/T and Man Utd @ F/T. No idea what the stats are on this but I usually do either way e.g
Man Utd/Chelsea
Chelsea/Man Utd.
Odds are usually about 30/1, vary depending on the difference in how good each team is. If you select your matches, pick teams that are close to each other in terms of how good they are, I reckon that you could make a nice bit of profit doing this. The last one that I done I lost, it was the Champions League Final and done it:
AC/Liverpool
Liverpool/AC
You could say I won it but as Liverpool won on penalties I lost. I have won before on a Man Utd vs Newcastle match.
Birmingham v Middlesbrough
Portsmouth v Aston Villa
Sunderland v Man City
Arsenal v Fulham
Blackburn v Tottenham
Bolton v Newcastle
Chelsea v West Brom
Not sure how this works out statistically, last weekend it wouldn't have worked at all, but just for expermint on the midweek matches I have highlighted the ones that I think have a chance.
So say you put a E1 on each one, this would be 8 different bets, if one of them comes through you would make a profit of about 3/1 or 4/1 depending on the results or if 2 come through then you are laughing .
Of course the odds are that none will work out but worth a try I think.
I can't get access to betting websites at work so if someone could check the odds please.

I do a similar bet occasionally, I limit it to when Man U, Arsenal, Chelsea and Liverpool play one and other, I do one side to be leading at half time and the match to finish a draw, the odds are usually around 20/1.

masterK wrote:

I think a more selective approach than just backing every draw would yield more of a profit. As teams go through peaks and troughs they may be selected or left out, let's say for example West Brom suddenly go on an awful run and are conceding goals left right and centre then it might be worth leaving them out until their form stabilizes.

Perhaps. However, wherever judgement is used in a system there is a need for an objective formula to determine the selection. I agree that it would make sense to drop teams from the selection that are generally losing and heading towards relegation. This system would seem to work better for mid-table teams, but I dont have the stats to back that up.

No doubt as the system (does it have a name: MidTable01draw's?) is fine-tuned the selection formulas may be long, but if selection is fully objective, selections can be programmed and hence produced instantly.

redspider

im still trying to keep the s/sheet goiong, but im way too hungover to get the odds today, could somebody post em and i'll sort it out this evening?
thanks

masterK wrote:

Maybe we'll call it the "Bore Draw" system.
West Brom V Birmingham 0-0 7/1, 1-1 9/2
Aston Villa V Blackburn 0-0 7/1, 1-1 9/2
Fulham V Everton 0-0 8/1, 1-1 9/2
Man City V Portsmouth 0-0 9/1, 1-1 11/2
Wigan V Sunderland 0-0 7/1, 1-1 9/2
Middlesborough V Charlton 0-0 7/1, 1-1 5/1

Well, the 1st 6 of these have been played and none resulted in 0-0 or a 1-1 draw. Mid v Charlton to go, so lets see.

But one bad week doesnt mean the "Bore Draw" system doesnt work, it just means statistcally that it was a poor return. Things may change next week, or a month from now. Maybe the "0011" system is a better name, more apt, and apart from that, many of these games can be "exciting".

redspider

An awful weekend for the bore draw or 0011 system.

West Brom V Birmingham
Aston Villa V Blackburn
Fulham V Everton
Man City V Portsmouth
Wigan V Sunderland
Middlesborough V Charlton

Not one single draw, looking at the fixtures who'd have predicted that. That's a loss of €120 for the week and an overal loss of €110 for the season. Hopefully it will pick up with the next set of matches.

well we are at a loss of 4.32 at the moment, i really dont see it working but we will see..

redspider wrote:

As there are international games on, does anyone have a system for those or is it better just to ignore these for betting?

In my experience, the favourites slip up much less often than in domestic club competition, because the groups are shorter, and the margin for error is a lot less than in a 38 game season. Home advantage can of course also be much more of a factor than in club competition. System guaranteed to win money? Not sure about that one!

To be honest I totally forgot about it this weekend. I think the best course of action may be to wait until 10 games have been played and then select the list of teams that qualify for the system. The qualifiers I selected were based on last seasons results and since the makeup of most teams has changed(in some cases dramatically) it might be worth letting the early season for settle down before having another bash at it.