Football Months (Betting School Article)

Another Betting School article. This time one I originally wrote in January 2010. It talks about the coming months in the season so should still be a good read and valuable.

Football Months

We are now in the new year and the fresh start gives us a chance to look ahead to the rest of the season. Most leagues they have passed the half way point in terms of matches but there is still plenty to come and lots to be decided. In January and February teams are still jostling for position and trying to move up the table but as the season nears the finish some games mean a lot and other games are meaningless.

The aim of this article is to look at the changes that occur over the course of a football season and see if there are any profitable angles

The data I use comes from www.football-data.co.uk and spans 5 seasons from 05-06 to 09-10 for the top leagues in Germany, England, France, Italy and Spain.

When looking at data over time charts come into their own and there is no better way to see patterns. The first areas to look at are goals scored and then shots taken as they go together.

Looking at the above you can see some similarities and obvious patterns. At the start of the season the goals are pretty low. However, the number of shots is not lower and so it may be that players take a while to get into the groove and so the early season dip in goals is to do with a bit of rustiness. Then come the months of January and February where there is a clear drop in both the number of goals and number of shots. I can think of no explanation for the drop in shots unless poor pitches and cold weather plays a part. The final part of the season sees a clear increase in both shots and goals possibly down to a lot of teams having little to play for. This brings a more carefree attitude and teams may concede and score more goals as there is less pressure.

As we can see a pattern here the best market to exploit it might be the over under 2.5 market. If we look at percentage of matches under 2.5 goals we hope to see the same pattern again. The big question is do the bookmakers also know that this pattern exists? By charting the average price of under 2.5 goals we should be able to see just that.

Although the under 2.5 graph is reversed you can see the same pattern exists in January and February with those 2 months having the highest averages of the season. If we look at the odds for the bookmakers (I have used BbAv<2.5 = Betbrain average under 2.5 goals) we can see a minor drop but nothing like we have seen in any of the other charts. The bookmakers have the climb at the end of the season spot on but there could be a bit of profit to be taken in the months of January and February.

The reason I chose the under market is that it is more profitable than the overs market. You can see this in the following table as well as a complete breakdown by month for over and under 2.5 results. Again this uses the Betbrain average over and under 2.5 goals price from football-data.co.uk.

Month	Matches	% Over 2.5	Over P/L	Over ROI	% Under 2.5	Under P/L	Under ROI
Aug	627	46.6%	-44.48	-7.1%	53.4%	-42.59	-6.8%
Sep	903	45.2%	-116.22	-12.9%	54.8%	-28.25	-3.1%
Oct	938	48.0%	-71.16	-7.6%	52.0%	-72.57	-7.7%
Nov	916	47.2%	-85.82	-9.4%	52.8%	-60.13	-6.6%
Dec	971	48.4%	-49.46	-5.1%	51.6%	-86.74	-8.9%
Jan	836	42.3%	-143.62	-17.2%	57.7%	5.91	0.7%
Feb	976	44.3%	-130.25	-13.3%	55.7%	-18.01	-1.8%
Mar	1018	45.6%	-102.08	-10.0%	54.4%	-51.02	-5.0%
Apr	1088	48.1%	-60.11	-5.5%	51.9%	-90.32	-8.3%
May	827	54.8%	-5.15	-0.6%	45.2%	-111.96	-13.5%
Total	9100	47.0%	-808.35	-8.9%	53.0%	-555.68	-6.1%

What stands out to me here immediately is that backing all games under 2.5 in January returned a profit and doing the same in February just returns a very small loss. If we had used the maximum odds then those two months combined return an 80pt profit and a ROI of 4.5%.

The other point is that even though the bookies expect more goals in May you can still do well backing overs. Looking at the maximum odds a profit is attainable for this month backing every game as over 2.5 goals.

I think it’s worth digging a little deeper into the January and February under 2.5 results to see if we can find more profitable angles. Maybe some leagues may be better than others so that is a starting point.

Month	Country	Matches	% Under 2.5	Under P/L	Under ROI
Jan	Germany	61	47.5%	-3.32	-5.4%
Feb	Germany	189	48.7%	-4.59	-2.4%
Jan	England	183	57.4%	5.34	2.9%
Feb	England	170	58.2%	3.45	2.0%
Jan	France	172	69.2%	15.14	8.8%
Feb	France	192	58.9%	-14.89	-7.8%
Jan	Italy	207	53.1%	-12.21	-5.9%
Feb	Italy	225	58.2%	6.65	3.0%
Jan	Spain	213	55.9%	0.96	0.5%
Feb	Spain	200	54.5%	-8.63	-4.3%

Very interesting that the Premier League is the only league to show a profit for both months. Serie A and the Bundesliga are the only two leagues not showing a profit in January and that does look a better month to concentrate on. In February, aside from England, Italy is the only country that shows a profit. If we stick with the Premier League in January we’ll see if it’s possible to improve the figures using a rule. If we exclude any match featuring one of the big 3 of Arsenal, Chelsea and Manchester United we get these results

Month	Matches	% Under 2.5	Under P/L	Under ROI
Jan	131	61.1%	16.73	12.8%
Feb	121	59.5%	7.64	6.3%

Normally I would include Liverpool in the big 4 but with their current situation and the poor time they had last season I thought it sensible to stick with a consistent 3 and they make little difference to the results either way. So a simple system of backing under 2.5 goals for Premier League matches in January when one of the big 3 is not playing returns a very nice profit.

It makes sense to look at some other statistics now and see if patterns exist. There is data for corners and cards but sadly no data for the odds so we cannot search for profitable angles but merely speculate.

January again features as the lowest month along with October. I am not sure whay October appears but the drop in January would fit with what we found earlier with less shots. I would assume less shots would contribute to a lower corner count. With corners not being a totally mainstream market there could well be opportunities going lower in January. As in all the charts we see the peak towards the end of the season.

The last area to look at is cards.

With the yellow cards there is a clear pattern at the start and end of the season of less cards but from October to March it all looks pretty steady. Refs probably start the season trying to be more lenient or maybe firmer pitches then and at the end of the season play a part. Again I would imagine the bookies may be aware of the lower card counts at the end of the season but the drop is quite considerable and they may not judge it correctly. In terms of red cards there is a drop in November which seems a bit out of character and if we ignore that then the red cards would show a similar pattern to the yellows.

Overall it looks like there are a few distinct sections to a season and we are in the midst of one now in January and February to come. The league may not be decided in these months but the games are very important as everyone still has something to play for. We can expect less goals and less corners it seems as well. Then when we get to May we should see more patterns emerging like more shots, more goals, more corners and less cards. Further research may be needed to see how the end of season stats are affected by teams playing for nothing or those at the top and bottom of the table.

Leon Pidgeon

Hit or Myth: Does a new manager make his point or points! (Betting School Article)

This is an article I originally wrote for the Betting School in November 2011. Some of the content should be worthwhile despite some of the data being slightly old.

Hit or Myth: Does a new manager make his point or points!

You have to love some of the football myths. I remember being asked which player had won the league title with Arsenal and been involved in the Munich air disaster. The answer I was informed was John Lukic as his mum had been aboard the plane when pregnant and survived the disaster. This is absolute tosh as he was born nearly 3 years after the air crash and was based on the fact that Harry Gregg saved a woman called Mrs Lukic during the disaster who was pregnant.

Thankfully the advent of the internet means it is a lot easier to check this kind of thing and not embarrass yourself with some made up trivia which I have been guilty in the past. My brother in law likes to remind me of some dodgy trivia question I asked him about Imre Varadi and the teams he had played for which he was able to expose as rubbish and declare my knowledge as ‘bobbins’.

This month I want to look at how teams perform with a new manager. Often when a new manager arrives the perception is that the results will pick up as players try to impress the manager. This is the theory I will try to test this by looking at how teams perform in the games before and after the manager arrives to see if there are any profits to be had. I have decided to look at 4 games either side of the appointment. The reason I have chosen 4 games is then we should get two home and two away matches for each team to give a balanced result set.

I am looking at the English Leagues and have taken the information about managers leaving their jobs from the League Managers Association website which has a lot of data for this.

According the league managers association there have been, on average, just over 50 managerial changes in each of the last 5 seasons as shown below.

League	2006-07	2007-08	2008-09	2009-10	2010-11
Barclays Premier League	12	12	14	6	10
npower Championship	11	13	13	19	15
npower League One	13	12	14	12	17
npower League Two	13	9	9	13	16
Total	49	46	50	50	58

I wanted to be able to exclude any changes that happened in the off-season so to do this I only included matches where the next match after hiring was within 14 days and the last match before a firing was within 14 days. Also I have excluded any changes where the new manager was not in place within 30 days of the initial manager leaving. This leaves us with 129 managerial changes over the 5 years.

I have seen some analysis of this kind of thing before and I was expecting to confirm this as a myth but the results are quite surprising. The first thing to look at us how many matches the teams won in the 4 games before and after the managerial change.

	Wins	Games	Win %
4 games before	37	129	29%
3 games before	37	129	29%
2 games before	26	129	20%
1 game before	17	129	13%
1 game after	34	129	26%
2 games after	52	129	40%
3 games after	45	129	35%
4 games after	45	129	35%

You can see the win percentage fall as we get closer to the dismissal and then it picks up in the first game but then makes much bigger jumps in the 3 games after. Taking all 4 games together this equates to a 23% win percentage in the period leading up to the change and then a 34% win percentage afterwards.

Another thing to notice is that it is invariably a bad result that triggers the sacking as there is just a 13% win percentage in the game before a sacking. This is not rocket science but with the number of markets available to bet on for managers being sacked it could help.

After seeing the results the next question that needs to be asked is can we profit from this and the answer looks to be yes if we look past the opening match.

	Win %	P/L
4 games before	29%	-33.99
3 games before	29%	-31.69
2 games before	20%	-53.66
1 game before	13%	-76.61
1 game after	26%	-15.75
2 games after	40%	24.54
3 games after	35%	8.12
4 games after	35%	-0.28

The second and third game after the managerial change look to be the key games for getting a profit and there is some logic to this. In the first game the manager cannot change very much and sometimes is even just watching from the stands. I also believe in this game the prices reflect the fact that bookmakers expect some improvement. In the second and third game the manager has been able to work with the players and get across any new methods/tactics etc. In the fourth game the win percentage is the same as the third game but no doubt the fact that the team has improved will mean the bookmakers will adjust their prices to be lower and so the profit is eradicated.

I must that admit when I chose this choice of topic I firmly had the idea in my head it was going to be a myth but now it is time to dig deeper and see if we can find any more profitable angles.

Looking at the 129 changes the only manager that was able to produce 4 wins out of 4 was Guus Hiddink when he took over from Big Phil at Chelsea. Maybe there could be some scope in looking at the results by division.

	Profit/Loss
	Barclays Premier League	npower Championship	npower League One	npower League Two	Conference
4 games before	-10.66	-17.47	-6.26	-1.15	1.55
3 games before	-10.76	-19.42	0.09	-0.17	1.57
2 games before	-21.88	-19.84	-10.74	-8.95	-3
1 game before	-23.86	-26.19	-9.62	-16.91	-0.37
1 game after	-2.31	6.54	-10.66	-4.97	-1.35
2 games after	-2.49	13.89	1.52	12.17	2.45
3 games after	-11.11	0.69	15.82	5.6	0.12
4 games after	7.82	-3.68	-8.83	4.85	2.56

This table above shows the raw profit and loss by division. The Premier League still shows a loss for the 4 games after a managerial change. Of the 4 games it is only the fourth one that shows any profit. If we look at the Championship it is the reverse and only the fourth game does not make a profit and overall there are some decent results. League One shows an overall loss for the games after managerial change but the big loss is in the first game after the change which we discussed earlier as probably being no surprise. League Two shows a decent profit for all but the first game in and is the most profitable of the lot. There are not many results for the Conference and it is hard to take anything positive or negative from there. I will remove the Conference from the rest of the results.

If we look at the 4 games post change as a whole in the table below we can now more clearly see the loss in the Premier League and the fact the other Leagues are all quite consistent in terms of win percentage. Why is the Premier League different? My personal take is that in the Premier League has huge gulfs in class and that a poor team is a poor team and no amount of managerial magic can change that. Yes, some teams over perform but by and large most teams play to their own level. In the other divisions maybe the improvement brought about by a manager can make a very big difference.

4 games post managerial change	Barclays Premier League	npower Championship	npower League One	npower League Two
Won	34	56	45	33
Total	112	164	132	96
Win %	30%	34%	34%	34%
Profit/Loss	-8.09	17.44	-2.15	17.65

So now let’s pull all this together and see if we can find a profitable angle. The Premier League looks worth avoiding and also we have seen that the first game does not seem to yield decent results and that by the fourth game any improvement has been spotted by the bookmakers. The results for the second and third game a new manager has in charge of a team are shown below by the 3 lower divisions.

	Won	Total	Win %	Profit/Loss	ROI
npower Championship	29	82	35%	14.58	18%
npower League One	28	66	42%	17.34	26%
npower League Two	19	48	40%	17.77	37%
All	76	196	39%	49.69	25%

Some very nice returns in there and with a total of nearly 200 games spanning 5 seasons there is an average of around 40 games per year that would qualify. Something worth looking at for sure and also on the flipside you would not want to oppose a team with a new manager in these games.

The new broom does look to sweep clean in the lower divisions and I would put this one down as a hit for them. In terms of the Premier League I do not think it makes much difference and would not be too concerned with this factor.

If you have any suggestions for any myths to investigate please contact me.

Leon Pidgeon

Goals Galore Article (Betting Insiders April 2011)

This is an article I originally wrote for the Betting School in April 2011. Some of the content should be worthwhile despite some of the data being slightly old.

When both teams score …

In the last few months I have often come across people talking about the Goals Galore coupon with Betfred and the equivalents that other bookmakers offer. For those who are unfamiliar with it the aim is to find matches where both teams score and Betfred pay 9/4 on 2 correct, 9/2 on 3 correct all the way up to 1150/1 for 12 correct for those keen to lose money.

Working with Betting School we have come up with a console that will hopefully help to pinpoint games that are more likely to have both teams score and that is already available to members.

The idea with this article is to search round the both teams to score angle to see what can be found and how the console can be used with these ides. In this article the abbreviation BTS will be used for both teams score/scored/scoring. The data used is from the English Premiership, Championship, Leagues 1 and 2, the Conference and the top league in Scotland, France, Italy and Spain. I have included the results for Germany in a table below and we see why Betfred don’t include it on their coupon.

Some readers may think this is a bit late in the season to be looking at this type of bet but as I showed in the January article as we get to the end of the season the number of goals increases and therefore so does the number of games where BTS. The table below shows the percentage of games where BTS by month.

Month	Matches	BTS	No BTS	% Games where BTS
August	1742	905	837	52%
September	2154	1126	1028	52%
October	2018	1080	938	54%
November	1753	863	890	49%
December	2214	1110	1104	50%
January	1922	957	965	50%
February	2040	977	1063	48%
March	2376	1167	1209	49%
April	2417	1226	1191	51%
May	966	521	445	54%

You can see that the start and end of the seasons look the best times for attacking this angle and overall in 50.7% of matches BTS. Therefore the chance of getting a double with two random games is 1 in 4 which equates to 3/1 and so Betfred only offering 9/4 mean we need to find more angles. To get to a break even situation we need to find games where the chance of BTS is between 55-56%. If we find games where the chance is 56% the decimal odds should be 3.19 on the double but Betfred are offering 3.25 (9/4) so we get value.

If we break the statistics down by the individual leagues we can see if some are better for concentrating on than others.

Div	Matches	BTS	No BTS	% Games where BTS
Ger Bund	1530	853	677	56%
Eng Champ	2760	1457	1303	53%
Ita Serie A	1900	998	902	53%
Eng Lge 2	2760	1436	1324	52%
Eng Lge 1	2760	1428	1332	52%
Eng Conference	2624	1343	1281	51%
Spa Primera	1900	966	934	51%
Scot Prem	1140	568	572	50%
Eng Prem	1900	903	997	48%
Fra Ligue 1	1900	859	1041	45%

The German league comes out on top and it’s no surprise that this does not feature on the Betfred coupon. The French league is considerably lower than the others and it’s a bit of a surprise to see the Italian league so high up. The 4 English lower leagues also fair well but in any analysis like this I like to look at results with and without the big teams to see if this makes a difference to the results.

The leagues I am using and the clubs I am classifying as big are as follows:

England – Arsenal, Chelsea, Liverpool and Man United

France – Bordeaux, Lyon and Marseille

Italy – Fiorentina. Inter, Juventus, Milan and Roma

Scotland – Rangers and Celtic

Spain – Atletico Madrid, Barcelona, Real Madrid, Sevilla and Villarreal

If we look at matches involving big teams only we some interesting results.

Big Team v Big Team	Matches	BTS	No BTS	% Games where BTS
Eng Prem	60	27	33	45%
Fra Ligue 1	30	22	8	73%
Ita Serie A	92	58	34	63%
Scot Prem	20	8	12	40%
Spa Primera	150	86	64	57%

The two British leagues are the odd ones out in this as all the other leagues have a considerably higher percentage of BTS when it’s just the big teams playing. It is very high in France, and although the sample is very small this would be a pattern I am happy to follow. The same has to be said for Italy where the 63% BTS is well above the 56% threshold we are after.

Normally I would look at games with big teams against non-big teams overall but in this I think it makes sense to break it down depending on who is at home. If we start with games where the big club is at home against a non-big club we get the results below.

Big Team at Home	Matches	BTS	No BTS	% Games where BTS
Eng Prem	324	126	198	39%
Fra Ligue 1	255	123	132	48%
Ita Serie A	364	185	179	51%
Scot Prem	170	75	95	44%
Spa Primera	420	206	214	49%

France is the only league where these results are above the average for all games. I would expect this figure to be lower as you would think these teams are more likely to keep clean sheets. I expect most people would not use these games anyway in their coupon but they can be useful when you look back over a team’s form as you may see a couple of games where both teams did not score, but if this was against a big side it may be worth ignoring it. For example on the console I looked at a Bolton away match and both teams had scored in a healthy 10/15 matches. However, the games when they didn’t were against Manchester United, Manchester City, Chelsea, Sunderland and Stoke. The first 3 teams, along with Arsenal fill the top 4 places in the league and have done most of the season so if we ignore games at the top 4 clubs their record becomes an even healthier 9/11. I would imagine these results also work for lower divisions based on the top sides each season but without testing this cannot say so for sure.

The table below shows when the big team is away.

Big Team Away	Matches	BTS	No BTS	% Games where BTS
Eng Prem	324	144	180	44%
Fra Ligue 1	255	112	143	44%
Ita Serie A	364	188	176	52%
Scot Prem	170	103	67	61%
Spa Primera	420	217	203	52%

Aside from France the results go up compared to the table above. The Scottish result is a particularly high one and any games with an old firm team away could warrant a second look.

Finally we look at games where none of the big teams are playing below.

No Big Team	Matches	BTS	No BTS	% Games where BTS
Eng Prem	1230	623	607	51%
Fra Ligue 1	1360	602	758	44%
Ita Serie A	1080	567	513	53%
Scot Prem	780	382	398	49%
Spa Primera	910	457	453	50%

The Premiership is again an exception here but for all the other leagues we can see that the results for this are more or less equal to the average for all games. Essentially in these games there is nothing outside of norm so we would need to look for patterns with particular teams.

With a bet like this you want to look for teams that follow patterns as they are likely to keep them up. I believe Peterborough and Crewe are often taken off the Goals Galore coupon as they were having BTS so much. Whist writing this article I used the console and saw that Southend host Aldershot on 2^nd April and 14/19 Southend home games and 13/19 Aldershot away games had BTS. Both of these teams have a good pattern and would be a decent selection if featured on the coupon.

I looked at individual team patterns a bit more and found some startling results for certain teams over the course of a season. The table below shows 5 teams on either end of the extremes.

Season	Div	Home Team	Matches	Both Scored	One or None	% Games where BTS
06-07	Eng Prem	Man City	19	3	16	16%
07-08	Fra Ligue 1	St Etienne	19	3	16	16%
08-09	Eng Champ	Doncaster	23	5	18	22%
09-10	Eng Conference	AFC Wimbledon	22	5	17	23%
08-09	Eng Lge 1	Southend	23	6	17	26%
09-10	Scot Prem	St Johnstone	19	14	5	74%
07-08	Eng Champ	Colchester	23	18	5	78%
07-08	Eng Conference	Forest Green	23	18	5	78%
08-09	Eng Lge 2	Luton	23	18	5	78%
06-07	Ita Serie A	Atalanta	19	15	4	79%

Some of these values are outstanding but show how some teams get a pattern and stick with it all season. That’s not to say you don’t need to be aware of a change in pattern but hopefully the console will help with that. An example of a change in pattern would be Rotherham this season. At home 8 of their first 10 games had BTS but since then just 2 of the last 8 games have had BTS and what’s interesting is that they have kept scoring but have started to keep clean sheets and so those 6 games in the last 8 were all where they kept a clean sheet.

I hope this will all help with anyone interested in trying to make money from the Goals Galore coupon or similar using the console. The other two sections the console has show over/under goal statistics and also the average shots and shots on target. The over/under section just helps to give an idea of how many goals are being scored by these teams and it is all relevant to the both teams scoring market. In the nearly 20,000 games I have been through the games with both teams scoring average 1.4 shots more than those without. 0.95 of that difference is down to the visiting teams having more shots so finding teams with a better than average number of shots, particularly when away can be another factor that will swing whether to bet or not.

I can be contacted with ideas for improvements for the console, problems or any questions to do with it or anything in my articles.

Leon Pidgeon

Betting School Article

Once a month I write an article for Betting School and this month is no exception. I am doing something based on shots and the idea coming out is that is could be a decent indicator of success. So to test it this weekend I have a few selections that I am putting up. I am putting them up for reference as they are in the article and I want to post them

Bundesliga – Hannover look to be a bit lucky being so high up with so few shots but they are at home to Borussia Monchengladbach who also are poor. Freiburg have had a lot of shots and so have Leverkusen and Hoffenheim could be better off than they are. Sadly Leverkusen and Hoffenheim are away from home and the only possible looks to be Freiburg.

Eredivisie – Roda look to be in a false position as do Waalwijk. Vitesse and Hercales both play at home and look to have less points than their shot ranking deserves. I like them and also Venlo who have just 1 point at home but are ranked in the top half of average shots.

Premier League – West Brom and West Ham look to have been lucky and again Liverpool look to be unlucky! Swansea, Spurs, Arsenal and Villa all look to have less points than their shots deserve. Of these only Villa are at home.

Predictions (Best odds available at 18.05 (CET) on Thursday 25^th October)

Freiburg to bt Dortmund @ 6.0 with WH

Vitesse to bt AZ @ 2.18 with Pinnacle

Heracles to bt Heerenveen @ 2.04 with Pinnacle

Venlo to bt NAC Breda @ 2.28 with Pinnacle

Aston Villa to bt Norwaich @ 1.96 with Pinnacle

Recent Posts
April 2024

M T W T F S S

« Mar

1 2 3 4 5 6 7

8 9 10 11 12 13 14

15 16 17 18 19 20 21

22 23 24 25 26 27 28

29 30
Search by category
- Articles (28)
- Betting School (9)
- Both teams to score (4)
- Brier Score (1)
- Champs League (42)
- Corners (42)
- Data Tables (9)
- Euro Championships (3)
- Europa League (13)
- Guest Article (15)
- highcharts (2)
- In collaboration with NS (1)
- Infographics (4)
- London Clubs (1)
- Manager profiles (1)
- Premier League (12)
- Previews (3)
- Promotions (7)
- Pythagorean Expectation (3)
- Ratings (2)
- Recommendations (13)
- Roundups (87)
- Secret Betting Club (10)
- Simple Weekend (21)
- Site information (21)
- Spanish Football (7)
- Sporting Gijon (2)
- Stat ideas (24)
- Status Updates (30)
- Tips (129)
- Twitter Bets (26)
- Uncategorised (1)
- User questions (5)
- World Cup (34)
- World Cup 2018 (3)
Search

Search for: