This is Better?

image (3)

Advertisements

Bookies-Good/Bad/Lucky/Unlucky

How to read this? 1 example

Leicester City- Got extra ~7.5 points vs individual match odds so far, expected to get another ~7.5 pts in remaining games vs preseason odds

Methodology Problem: Individual game odds adjust and do not reflect what those odds would have been in preseason.

LC gets more Leicesterry and Chelsea gets more Chelseasque.

New one:

image (2)

Beginner ELO +/- Model for Galatasaray (thanks again ClubElo)

Since the state of the analysis is paralysis, I wanted to do my share. Following uses a very basic +/- model with few adjustments(for minimum games played/unplayed) over 3 seasons for Galatasaray. The twist is rather than use goals, it uses ELO points from clubelo.com. Not all goals are equal; so it tries to adjust for that. Other than that it has many failings:

1-Black box- and not Shapley which might be more appropriate

2-3 season of data used to get over noise

3- Starting 11 used as I do not have dynamic Elo numbers(during the game). It is doable but this was easier. So impact subs, etc. are not taken into consideration.

4- I still prefer counting models over agnostic models but if I had a good agnostic one I would try to create a composite one.


On the other hand:

1-Passes the eye/nose test since I probably watched 95% of Galatasaray games that are in this sample.

2-Just food for thought for others with more time/computing power/better statistical skills.


Basic formula: Avg. Elo Points in games player started- Avg. Elo Points player did not start

Adjustment: If player started fewer 10 games avg. Elo point for team is assumed for games he is missing and vice versa for games he didn’t start.

Results:

Player Elo +/-
Sneijder 4.1
Muslera 3.7
Bilal 3.2
Carole 3.0
Sabri 2.7
Selcuk 2.5
Melo 2.3
Yasin 2.0
Emre 1.9
Chedjou 1.8
Denayer 1.6
Semih 1.2
Riera 1.1
Burak 1.1
Telles 1.0
Dany 0.7
Podolski 0.6
Karacan 0.5
Drogba 0.4
Ufuk 0.4
Hamit 0.1
Sinan B. 0.1
Zan 0.0
Umut -0.4
Burdisso -0.4
Hakan -0.4
Jose -0.5
Engin -0.9
Amrabat -0.9
Bruma -0.9
Hajrovic -1.3
Aydin -1.3
Yekta -1.5
Olcan -1.9
Pandev -2.5
Eboue -2.7
Eray -2.9
Berk -3.0
Veysel -3.2
Ceyhun -4.0
Koray -4.3
Dzemaili -5.1
Tarik -5.6

If you have watched any Galatasaray games recently the team has been based on Muslera and Sneijder. Good news is most horrible players are gone or never start except for Olcan.

I know this appeals mostly to Galatasaray fans but EPL ppl can give it a try to create something similar.

dingostats weekly-11/6/15

Lineals

Napoli and Olympiakos are still going very strong. But Bayern is waiting the Greeks in a few games. That won’t be easy.

Helmond Cup dingoB Sandeman
Holder Olympiakos Olympiakos Napoli
Next Opp Iraklis(H) Iraklis(H) Udinese(H)
Date 8-Nov 8-Nov 8-Nov
Lose? 7% 7% 8%

dingoT

Methodology: dingoT uses 6 different inputs ranging from ELO ratings to market values. The weight of inputs change as season progresses to reduce the importance of forward looking metrics such as betting odds in favor of actual achievements.

Benfica and Wolfsburg are back in not because of what they have done but because of no team’s ability to hold on to 15th and 16th spots this year. It was Leverkusen and Valencia’s turn this time around. If Roma  and Liverpool can keep winning then it will be theirs soon.

Rank Team Points
1 Barcelona 140.1
2 Real Madrid 136.1
3 Bayern 134.2
4 Atlético 114.1
5 Paris SG 113.7
6 Man City 112.3
7 Juventus 111.9
8 Chelsea 109.0
9 Dortmund 107.6
10 Arsenal 106.8
11 Man United 103.1
12 Porto 100.1
13 Napoli 98.1
14 Sevilla 95.6
15 Benfica 93.9
16 Wolfsburg 93.3

dingoW

Methodology: Using dingoT to rate games on 2 metrics – quality (ie. 1 vs 2 is higher than 8 vs 9) and competitiveness (ie. 16 vs 17 is higher than 1 vs 32) to combine into one rating. 75+ games are listed.

11/8/15-Sunday

11:00am Arsenal vs Tottenham Hotspur: 75

2:30pm Sevilla vs Real Madrid: 88

Few games missed the mark by few points Roma-Lazio, Dortmund-S04 and Barca- Villarreal. I’m mentioning this in case you want to get your fix before international break.

dingoçãô (or dingoach for Portugal)

If there is a country that is punching above its weight in European club football it is Portugal. The connection to Brazil, smart scouting and even super-agents’ dealings have been credited with Portugal’s success in European competitions. Portugal is smaller than Belgium and Netherlands population wise, and much poorer. Similar to them it exports many of its players but unlike them its clubs are still feared unlike Anderlecht or Ajax who were good in distant 20th century. I am a beginner in Portuguese affairs so I only have one more aspect to point out.

Portugal has some great managers. Not just Mourinho. Portuguese clubs are not afraid of giving very young coaches a chance. AVB is still 38, and so is Marco Silva of Olympiakos. It is rare for coaches below 40 in top leagues even rarer to export them. This year 6 of 32 CL teams are managed by Portuguese managers.  In my adjusted dingoach ratings(for opponent strength) 6 managers come in top 50 in career ELO points among all European active club managers(anyone who coached within last 4 years):

Rank Career ELO Games
5 Mourinho 1261 629
16 Jorge Jesus 730 426
34 AVB 399 246
39 Jardim 342 178
45 Marco Silva 298 120
50 Paulo Sousa 270 241

Good but not great you are thinking. That’s where youth comes in; most of these managers coached relatively few games. If we look at per game ELO Points:

Rank Career ELO Games ELO/G
1 Marco Silva 298 120 2.48
6 Mourinho 1261 629 2.00
7 Jardim 342 178 1.92
9 Jorge Jesus 730 426 1.71
14 AVB 399 246 1.62
35 Paulo Sousa 270 241 1.12
37 Paulo Fonseca 109 102 1.07

Much much better.

Well of course not all Portugese managers have been that good. Here is a full list of anyone over 0. Anyone over per game average 2 is unsustainable but exceptional. Between 1 and 2 you have top coaches. Below 1, just average guys.  Fenerbahce and Valencia got the wrong-ish dudes. (Lopetegui is not in the above ranking as he is yet to get 100 games).

Career ELO Games ELO/G
Marco Silva 298 120 2.48
Mourinho 1261 629 2.00
Jardim 342 178 1.92
Jorge Jesus 730 426 1.71
AVB 399 246 1.62
Lopetegui 98 65 1.51
Paulo Sousa 270 241 1.12
Paulo Fonseca 109 102 1.07
Sa Pinto 63 87 0.72
Vitor Pereira 68 108 0.63
Nuno 66 109 0.61
Sergio Conceicao 56 102 0.55
Martins 71 178 0.40
Rui Vitoria 64 169 0.38
Carvahal 66 196 0.34
Manuel Machado 15 403 0.04

Apologies for many spelling errors with Portuguese names.

dingostats weekly-10/30/15

Lineals

All quiet on the southern front.

Helmond Cup dingoB Sandeman
Holder Olympiakos Olympiakos Napoli
Next Opp Veroia(A) Veroia(A) Genoa(A)
Date 31-Oct 31-Oct 1-Nov
Lose? 7% 7% 20%

dingoT

Methodology: dingoT uses 6 different inputs ranging from ELO ratings to market values. The weight of inputs change as season progresses to reduce the importance of forward looking metrics such as betting odds in favor of actual achievements.

Napoli is doing really well; Juve not so much. In top 32 now we have 6 Italian teams (Inter is in). Sporting and Jorge Jesus  are rising higher at the expense of Benfica.  Soon to come dingoach of Portuguese managers; my new pet project.

Rank Team Points
1 Barcelona 140.5
2 Real Madrid 135.3
3 Bayern 134.6
4 Atlético 114.8
5 Paris SG 113.8
6 Man City 111.4
7 Juventus 111.4
8 Chelsea 109.1
9 Dortmund 107.2
10 Arsenal 106.8
11 Man United 102.1
12 Porto 100.9
13 Napoli 97.2
14 Sevilla 97.1
15 Valencia 95.4
16 Leverkusen 94.9

dingoW

Methodology: Using dingoT to rate games on 2 metrics – quality (ie. 1 vs 2 is higher than 8 vs 9) and competitiveness (ie. 16 vs 17 is higher than 1 vs 32) to combine into one rating. 75+ games are listed.

10/31/15-Saturday

8:45am Chelsea vs Liverpool: 78

1:30pm Wolfsburg vs Bayer Leverkusen: 77

11/3/15-Tuesday

2:45pm Real Madrid vs PSG: 97

2:45pm Sevilla vs Manchester City: 88

11/4/15-Wednesday

2:45pm Roma vs Bayer Leverkusen: 77

2:45pm Bayern München vs Arsenal: 91

Bookies- Good/Bad/Lucky/Unlucky

image (1)

How to read this? 2 examples

Leicester City- Got extra ~5 points vs individual match odds so far, expected to get another ~6 pts in remaining games vs preseason odds

Stoke vs Swansea- Both got as expected so far. But going forward Swansea maintains its preseason number for remaining games while Stoke is expected to get 6 less. Bookies not impressed with Stoke.

Methodology Problem: Individual game odds adjust and do not reflect what those odds would have been in preseason.