A different approach?

[Data]

Many football rating methods are compiled on a match-by-match basis in real time, inasmuch a ratings is assessed for a game at hand, then using the result of said game, the rating is adjusted accordingly. The most recent match therefore has perhaps too much of an influence on how the new rating moves.

This approach presented here also uses past results to calculate the outcome, but it looks at ALL previous game results and treats them as equally important. Trends therefore are not picked up, whichever way a team's form is heading is totally ignored. A good or bad thing? . . . let's find out.

I'm using neural/genetic machine learning to process season-to-date league results to find the best fit of correct home, draw, away predictions to then apply to the next set of fixtures. In order allow the algorithms enough relevant information and data it is thought necessary to employ only where a league season is half completed.

Other caveats include no leagues of more than 20 teams so as not to overwhelm the net (and take far longer to process). No leagues with too few teams either where teams may play each other more than twice ~ nothing confuses artificial intelligence more than encountering the same situation with different outcomes, e.g. United v City showing one home win and one away win.

Having churned through a few leagues here's what it gives for this weekend; (odds are Oddsportal average odds from 50 odd bookies)

CHILE
18-Aug-18	16:00	San Luis	U. Espanola	A	2.57	3.28	2.59
18-Aug-18	19:00	Antofagasta	Union La Calera	nobet	1.94	3.54	3.57
18-Aug-18	21:30	U. De Chile	O'Higgins	A	1.64	4.01	4.51
19-Aug-18	00:00	Everton	Huachipato	H	2.04	3.27	3.52
19-Aug-18	16:00	Deportes Iquique	Colo Colo	A	3.19	3.51	2.05
19-Aug-18	19:00	U. De Concepcion	Deportes Temuco	H	1.83	3.43	3.99
19-Aug-18	21:30	U. Catolica	Curico Unido	H	1.72	3.57	4.45
SWEDEN
18-Aug-18	15:00	Malmo FF	Trelleborgs	H	1.15	7.89	17.16
19-Aug-18	14:00	Elfsborg	Norrkoping	X	2.87	3.44	2.36
19-Aug-18	14:00	Sundsvall	Hacken	H	3.26	3.51	2.13
19-Aug-18	16:30	Brommapojkarna	AIK	nobet	7.98	4.47	1.4
19-Aug-18	16:30	Djurgarden	Kalmar	H	1.65	3.63	5.53
20-Aug-18	18:00	Hammarby	Orebro	H	1.43	4.64	6.64
20-Aug-18	18:00	Goteborg	Ostersunds	A	2.93	3.25	2.41
JAPAN
19-Aug-18	05:00	Sapporo	FC Tokyo	X	2.79	3.27	2.45
19-Aug-18	10:00	Nagoya	Sagan Tosu	X	2.9	3.4	2.3
19-Aug-18	10:00	Shimizu	Urawa	H	3.54	3.38	2.03
19-Aug-18	10:00	V-Varen Nagasaki	C-Osaka	X	3.43	3.4	2.05
19-Aug-18	10:30	Kashima	Yokohama M.	A	1.92	3.77	3.53
19-Aug-18	11:00	Hiroshima	Kawasaki Frontale	A	2.88	3.41	2.32
19-Aug-18	11:00	Iwata	Kashiwa	A	2.74	3.17	2.55
19-Aug-18	11:00	Shonan	Kobe	X	3.23	3.42	2.13
19-Aug-18	11:00	Vegalta Sendai	G-Osaka	A	2.09	3.53	3.22
CHINA
18-Aug-18	12:35	Beijing Renhe	Beijing Guoan	A	7.35	5.74	1.31
18-Aug-18	12:35	Chongqing Lifan	Shanghai SIPG	A	7.94	5.83	1.29
18-Aug-18	12:35	Dalian Yifang F.C.	Henan Jianye	H	1.91	3.48	3.77
18-Aug-18	12:35	Guangzhou R&F	Changchun Yatai	A	2.14	3.62	3.01
18-Aug-18	12:35	Guizhou Zhicheng	Shandong Luneng	A	6.24	4.93	1.4
18-Aug-18	12:35	Tianjin Teda	Jiangsu Suning	nobet	2.43	3.38	2.71
19-Aug-18	12:35	Guangzhou Evergrande	Tianjin Quanjian	H	1.07	10.81	22.57
19-Aug-18	12:35	Hebei	Shanghai Shenhua	H	2.09	3.72	3.01

I also processed Brazil Serie A, but they're not quite half way through their season yet, but here are the forecasts anyway;

BRAZIL					OddsPortal	average	odds
18-Aug-18	20:00	Santos	Sport Recife	H	1.51	4.01	6.46
18-Aug-18	23:00	Corinthians	Gremio	A	2.97	2.92	2.58
19-Aug-18	15:00	Atletico-PR	Flamengo RJ	H	2.74	3.12	2.62
19-Aug-18	15:00	Internacional	Parana	H	1.38	4.4	8.74
19-Aug-18	20:00	Botafogo RJ	Atletico-MG	X	2.36	3.31	2.92
19-Aug-18	20:00	Cruzeiro	Bahia	H	1.44	3.98	8.2
19-Aug-18	20:00	Vitoria	Palmeiras	X	3.42	3.4	2.08
19-Aug-18	23:00	America MG	Fluminense	H	2.31	3.09	3.19
19-Aug-18	23:00	Sao Paulo	Chapecoense-SC	H	1.39	4.31	8.88
21-Aug-18	00:00	Vasco	Ceara	A	1.75	3.36	4.91

 

Edited August 18, 2018 by Data
Info

[Data]

14 correct from the 38 picks (including the Brazil games) is not a start I would have wished for. -12.83 from a 33.8% strike rate and a -33.76% yield. But this thread is experimental and my thinking-in-progress, so all is far from doom & gloom.

My neural net classifier gives a binary decision on each game. For my first run, with last weekend's fixtures I merely selected the outcome with the highest network 'probability' for every match, with a no bet where two oucomes were too close to call. Better I think to select only the very best of the picks where the net is suggesting a near certainty. I'm not naive enough to think such games will be anywhere near the certainty suggested, but they will be the cream of the crop.

Elsewhere I've expressed my opinion that football draws are probably the hardest outcome to predict, and my weekend sample, although comparatively small, suggested 7 such outcomes - and every single one failed! Not a proof by any means, but something to keep in my mind.

So, the Brazil net newly trained with the latest results and bearing the above thoughts in mind, we have three midweek selections; ALL HOME WINS:-

BRAZIL A
23-Aug-18	01:00	Palmeiras	Botafogo RJ	1.44
23-Aug-18	01:00	Sport Recife	America MG	1.91
23-Aug-18	01:45	Fluminense	Corinthians	2.23

[Data]

Two from three for a profit is okay. But I've re-analysed and re-thought my methods.

I'm still using nothing more than season results to date to build the networks, no odds are used, no winning margins, just the season's win, draw & away records. I do have a very small sample now to help see how it is performing, and near to perfect P=1 is required to trigger a bet ~ but still work in progress.

This weeknend's matches are reprinted here in full showing all network P values for HOME,DRAW or AWAY.
My best bets are shown to the right hand side.

country date KO home team away team OddsH OddsD OddsA Phome Pdraw Paway My Pick BRAZIL 23/08/2018 23:30 Flamengo RJ Vitoria 1.36 5.28 13.93 0.9350 0.0170 0.0480   BELARUS 24/08/2018 16:45 FC Minsk Zhodino 3.78 3.10 1.93 0.0000 0.9997 0.0003   BRAZIL 24/08/2018 00:00 Atletico-MG Vasco 1.58 4.32 6.50 0.9981 0.0001 0.0018   CHINA 24/08/2018 12:35 Tianjin Quanjian Guizhou Zhicheng 1.92 3.80 4.03 0.9993 0.0007 0.0000 HOME JAPAN 24/08/2018 11:00 Kashima Iwata 2.00 3.75 4.33 0.3913 0.5997 0.0090   BELARUS 25/08/2018 13:45 Gorodeja Isloch Minsk 2.42 2.73 3.08 0.0000 0.0232 0.9768   BELARUS 25/08/2018 15:45 BATE Vitebsk 1.40 3.84 7.41 0.5029 0.4971 0.0000   BELARUS 25/08/2018 17:45 Neman Torpedo Minsk 1.41 3.70 7.99 0.1481 0.1408 0.7111   CHILE 25/08/2018 00:00 Curico Unido U. De Concepcion 2.61 3.38 2.80 0.0000 0.0135 0.9865   CHILE 25/08/2018 16:00 Colo Colo U. De Chile 2.10 3.60 3.80 0.0000 0.0000 1.0000 AWAY CHILE 25/08/2018 19:00 Deportes Temuco Everton 2.31 3.44 3.30 0.9981 0.0018 0.0002   CHILE 25/08/2018 21:30 Huachipato Antofagasta 2.51 3.40 3.03 0.0006 0.5082 0.4912   CHINA 25/08/2018 12:35 Chongqing Lifan Dalian Yifang F.C. 2.32 3.66 3.04 0.0000 0.0000 1.0000 AWAY CHINA 25/08/2018 12:35 Shandong Luneng Guangzhou Evergrande 3.12 3.74 2.21 0.4912 0.5088 0.0000   CHINA 25/08/2018 12:35 Shanghai SIPG Henan Jianye 1.12 9.70 27.00 0.9935 0.0000 0.0065   CHINA 25/08/2018 12:35 Tianjin Teda Beijing Guoan 5.45 4.30 1.62 0.0000 0.0004 0.9995   ICELAND 25/08/2018 18:00 Stjarnan Breidablik 1.97 3.76 4.00 0.0001 0.7302 0.2697   ICELAND 25/08/2018 18:00 Vikingur Reykjavik KA Akureyri 2.64 3.60 2.76 0.0000 0.0132 0.9868   ICELAND 25/08/2018 21:00 Valur Fjolnir 1.46 4.97 8.00 0.5246 0.4754 0.0000   JAPAN 25/08/2018 11:00 C-Osaka Hiroshima 2.48 3.62 3.00 0.0034 0.2566 0.7399   JAPAN 25/08/2018 11:00 Kashiwa V-Varen Nagasaki 1.80 3.92 5.25 0.8272 0.0001 0.1727   JAPAN 25/08/2018 11:00 Kawasaki Frontale Vegalta Sendai 1.61 4.50 6.50 0.0103 0.0017 0.9881   JAPAN 25/08/2018 11:00 Shimizu Sapporo 2.52 3.52 3.00 0.0000 0.4876 0.5124   NORWAY 25/08/2018 17:00 Haugesund Stabaek 1.82 3.91 4.83 0.9746 0.0253 0.0001   NORWAY 25/08/2018 17:00 Molde Ranheim 1.44 5.25 9.17 0.7594 0.0002 0.2404   NORWAY 25/08/2018 17:00 Sandefjord Lillestrom 3.10 3.50 2.53 0.5435 0.0003 0.4562   NORWAY 25/08/2018 17:00 Sarpsborg 08 Kristiansund 1.83 4.00 4.65 0.9991 0.0009 0.0000 HOME NORWAY 25/08/2018 17:00 Tromso Bodo/Glimt 2.02 3.75 3.95 0.0000 0.9906 0.0094   BELARUS 26/08/2018 13:45 Smolevichi STI Brest 8.10 4.36 1.33 0.0001 0.9785 0.0214   BELARUS 26/08/2018 15:45 Shakhtyor Soligorsk Dnepr Mogilev 1.22 4.98 11.56 0.5000 0.5000 0.0000   BELARUS 26/08/2018 17:45 FC Gomel Din. Minsk 6.42 3.49 1.51 0.0000 0.5000 0.5000   CHILE 26/08/2018 00:00 O'Higgins U. Espanola 2.32 3.40 3.36 0.1187 0.0077 0.8736   CHILE 26/08/2018 16:00 Palestino San Luis 1.82 3.70 4.90 0.9999 0.0001 0.0000 HOME CHILE 26/08/2018 21:30 Union La Calera Deportes Iquique 2.16 3.44 3.70 0.0000 0.4826 0.5174   CHINA 26/08/2018 11:00 Changchun Yatai Hebei 2.86 3.70 2.41 0.4842 0.0318 0.4841   CHINA 26/08/2018 12:35 Jiangsu Suning Guangzhou R&F 1.92 3.68 4.14 0.6025 0.3969 0.0006   CHINA 26/08/2018 12:35 Shanghai Shenhua Beijing Renhe 1.72 3.80 5.30 0.0011 0.9988 0.0000   ICELAND 26/08/2018 15:00 KR Reykjavik Vestmannaeyjar 1.62 4.09 6.55 0.5304 0.4696 0.0000   ICELAND 26/08/2018 19:00 Keflavik Hafnarfjordur 9.50 5.80 1.36 0.0000 0.4922 0.5078   JAPAN 26/08/2018 10:00 Kobe Yokohama M. 2.09 3.92 3.75 0.9973 0.0026 0.0001   JAPAN 26/08/2018 10:00 Nagoya Urawa 3.60 3.90 2.15 0.0002 0.9207 0.0791   JAPAN 26/08/2018 11:00 Sagan Tosu G-Osaka 2.32 3.48 3.60 0.9531 0.0469 0.0000   JAPAN 26/08/2018 11:00 Shonan FC Tokyo 3.79 3.60 2.10 0.0028 0.4840 0.5132   NORWAY 26/08/2018 17:00 Rosenborg Stromsgodset 1.56 4.50 7.00 0.4876 0.5014 0.0110   NORWAY 26/08/2018 19:00 Valerenga Brann 2.80 3.55 2.74 0.0000 0.4651 0.5349   SWEDEN 26/08/2018 14:00 Hacken Brommapojkarna 1.26 7.00 14.50 0.9854 0.0146 0.0000   SWEDEN 26/08/2018 14:00 Malmo FF Sirius 1.33 6.10 10.75 0.8920 0.1078 0.0003   SWEDEN 26/08/2018 14:00 Ostersunds Sundsvall 1.80 4.00 5.00 0.8661 0.1325 0.0015   SWEDEN 26/08/2018 16:30 AIK Trelleborgs 1.20 8.00 27.50 0.5003 0.4997 0.0000   SWEDEN 26/08/2018 16:30 Djurgarden Elfsborg 1.79 3.90 5.05 0.8053 0.0000 0.1947   BELARUS 27/08/2018 17:00 Slutsk Luch Minsk 1.31 4.46 8.63 0.9142 0.0005 0.0852   CHILE 27/08/2018 00:00 A. Italiano U. Catolica 3.22 3.50 2.37 0.0000 0.4993 0.5007   ICELAND 27/08/2018 19:00 Fylkir Grindavik 2.22 3.54 3.46 0.9985 0.0000 0.0015   NORWAY 27/08/2018 18:00 Start Odd 2.75 3.60 2.85 0.2554 0.3096 0.4350   SWEDEN 27/08/2018 18:00 Dalkurd Hammarby 5.83 4.20 1.66 0.3092 0.3995 0.2913   SWEDEN 27/08/2018 18:00 Kalmar Orebro 1.83 4.05 4.55 0.3716 0.6284 0.0000   SWEDEN 27/08/2018 18:00 Norrkoping Goteborg 1.65 4.20 6.00 0.4999 0.0004 0.4998

Edited August 23, 2018 by Data
text size

[Data]

A couple of Brazil Serie A best-bet picks from this method to add to the above;
(one result in so far, win at 1.92)

26-Aug-18 | 20:00 | America MG | Flamengo RJ   away win 2.41
26-Aug-18 | 23:00 | Vasco | Chapecoense-SC    draw   3.32 
 

[keef75]

Still watching this with interest.

[Data]

I've been looking further into various options for this experiment in artificial intellingence football predictions. Using goal differences instead of simply the three options of Home/Away/Draw that I began with didn't appear to get me any closer to a solution. However, using a classification method with correct scores did show some promise . . .

Again, nothing but current league results to date (where the engine assumes each team's home & away results are two distinctly separate individual teams). Results used are all those where the home and away sides scored 2 goals or fewer, and other scores are all lumped together as "other". This approach I figured was working with more than 50% of all results, and I wasn't really bothered about attempting to predict EVERY game, rather than to concentrate on those games where scores were more 'reasonable' and as such perhaps allowing a little more confidence in the outcomes.

We shall see, can this possibly make a level stake profit?

The 3 most common football outcomes are 1-0, 1-1 & 0-1. So, I'll use these scores as control variables, logging the percentage hit rate and profit/loss for each and comparing progress.
 

Japan	26-Aug-18	11:00	Shonan		FC Tokyo	2~0
Japan	26-Aug-18	11:00	Sagan Tosu		G-Osaka	1~0
China	26-Aug-18	12:35	Jiangsu Suning		Guangzhou R&F	2~2
Belarus	26-Aug-18	13:45	Smolevichi STI		Brest	2~0
Sweden	26-Aug-18	14:00	Malmo FF		Sirius	1~1
Sweden	26-Aug-18	14:00	Ostersunds		Sundsvall	1~1
Iceland	26-Aug-18	15:00	KR Reykjavik		Vestmannaeyjar	0~1
Brazil	26-Aug-18	15:00	Sao Paulo		Ceara	1~0
Chile	26-Aug-18	16:00	Palestino		San Luis	1~1
Sweden	26-Aug-18	16:30	Djurgarden		Elfsborg	1~0
Norway	26-Aug-18	17:00	Rosenborg		Stromsgodset	2~2
Iceland	26-Aug-18	19:00	Keflavik		Hafnarfjordur	1~1
Norway	26-Aug-18	19:00	Valerenga		Brann	2~2
Brazil	26-Aug-18	20:00	Internacional		Palmeiras	1~0
Brazil	26-Aug-18	20:00	Vitoria		Atletico-MG	1~0
Chile	26-Aug-18	21:30	Union La Calera		Deportes Iquique	0~0
Brazil	26-Aug-18	23:00	Vasco		Chapecoense-SC	1~0
Chile	27-Aug-18	00:00	A. Italiano		U. Catolica	2~2
Belarus	27-Aug-18	17:00	Slutsk		Luch Minsk	1~1
Sweden	27-Aug-18	18:00	Norrkoping		Goteborg	1~0
Iceland	27-Aug-18	19:00	Fylkir		Grindavik	1~1

[Data]

... and one more correct score prediction for today that I missed ...

Norway

26-Aug-18

17:00

Haugesund

Stabaek

1~0

[Data]

On 25/08/2018 at 10:29 AM, Data said:

A couple of Brazil Serie A best-bet picks from this method to add to the above;
(one result in so far, win at 1.92)

26-Aug-18 | 20:00 | America MG | Flamengo RJ   away win 2.41
26-Aug-18 | 23:00 | Vasco | Chapecoense-SC    draw   3.32 
 

Well, the above two additions both lost leaving a poor return for the weekend's 7pt outlay. Just a single winner from the seven proposed and an overall loss of 5.08pts.

I'm pinning greater hopes on my revised correct scores engine to rescue this thread.

[Data]

Correct scored round-up. No profit from these first ones, but very early days yet. A combined summary is printed here which also includes a handy comparison of my ML picks against backing every game to finish 1-0, every game 1-1 and every game 0-1. Finally the predictions from a web site that puts its correct score on line for free (Bettingclosed.com)
 

[Data]

The full W/L record for the previous correct score predictions as reflected in the previous post were as follows;

My machine learning (ML pred) came second behind backing all to finish 1-0. Early days yet, I'll keep going with this whilst there's still hope.

[Data]

A couple of predictions for midweek games;

[pronoskiller]

End of your experimentation,

Which plateform do you use to run your predictif model ?

I'm interest to process data in this way.

[Sir Puntalot]

 to PL @pronoskiller   

[pronoskiller]

thxs

[Data]

On 12/09/2018 at 9:02 AM, pronoskiller said:

End of your experimentation,

Which plateform do you use to run your predictif model ?

I'm interest to process data in this way.

I'm busy in my darkened room in an attempt to refine this process. The software I'm using is Neuroshell Classifier from the AI Trilogy package. (I've had this for years now ~ back when it was less than a quarter of today's price!)

Their classifier is unique, so far as I know, inasmuch the predicted output is not restricted to two state binary, but can be as many variants as the user chooses (I've been restricting possible outputs/scores to every variant between 0-0 & 3-2)

. . . but still work in progress!

[pronoskiller]

thks for your information about your software.

After reflection, I think this approach of setting up a classification algorithm to highlight variables, impacting the results to find values, it's the best and short way because bookmakers spend a lot of time and money to work in this way and they have better sources more sharp and high quality of data. So i thinking to find another way to explore and find spot value. Good luck for your research.

[Data]

My deliberations have moved on, time for a Mk II version m'thinks, on a new thread.