Football Musings

[Kai]

Good stuff Mr Fox but isn't saying "Xg" instead of the standardized term "xG" a big faux pas for the analytically minded?

I'm amazed it has never bothered you one bit

Fallen at the first hurdle....

FoxG would be better the FoXg that's fir sure.

Hehe, I quite like the sound of foxG to be fair. And wtfxG too if you wanted to use your full username

[wearthefoxhat]

interpretation of Paul Steele's power ratings from Page 125 to Page 129 into excel.

steele1.ot png

All I need to do is carry forward the new power rating for each team's match and then see how they match up to create a difference. (+ for home superiority - for away).

The next part talks about penetration plus ratings to create a scoreline based on each games stats for last 5 games. I'll have that covered with FoxG.

The game form section for the last 5 games, awards, Home Win = 3pts Home Draw = 1pt Away Win = 5 pts Away Draw = 2 pts.

The book then goes on to use all 3 elements for comparison where all 3 elements are at their strongest and suggests to check on injuries, suspensions, new managers...etc

[wearthefoxhat]

Using the 23/24 premier league final table stats, it looks like Man Utd are a fair bet at 1.57 with pinnacle. Betfair tends to follow their lead so the price is the same on the exchanges.

My view going forward, is that I'd prefer to use recent form, ie: last 8 games, and use xG (FoxG) over normal score-lines.

ManUtd v Fulham.png

[jtrader]

If (FoxG) is changed for xG,we could lay Man Utd.

Screenshot (940).png

[wearthefoxhat]

If (FoxG) is changed for xG,we could lay Man Utd.



Screenshot (940).png

I agree, reckon xG, or a version of xG is a better metric overall.

[wearthefoxhat]

The FoxG still seems to have a better balance than the normal xG

Cov v Oxf xG.png

---

Coventry were clearly dominant and it wasn't a surprise to see a late winner,

Cov v Oxf.png

[wearthefoxhat]

Championship:

Burnley 5 Cardiff 0

Burnley must have battered Cardiff....not necessarily. Both the real xG and FoxG look similar.

BurnleyvCardiff.png

Defensive mistakes and worldy goals won't always be factored in. The actual result has to be, so 3-1 maybe a fairer score-line overall.

[avs_kz]

The market was against Northampton as the price ended up around 3.10 just before the off. They weren't far wrong as Stockport came out fast and furious and deservedly took an early lead, and maybe should have scored again.

It turned out to be a game of 2 halves, as key subs were made at HT, making all the difference, with Northampton winning 2-1.


HT.png

hi! where i can find this spreadsheet?

[wearthefoxhat]

The market was against Northampton as the price ended up around 3.10 just before the off. They weren't far wrong as Stockport came out fast and furious and deservedly took an early lead, and maybe should have scored again.

It turned out to be a game of 2 halves, as key subs were made at HT, making all the difference, with Northampton winning 2-1.


HT.png

hi! where i can find this spreadsheet?

That one was/is my own concoction of stats trying to bring a predictive element into the mix.

I'm currently collecting data for my own xG to use for predictive purposes. The excel sheet I'm using for this, I found on the internet a couple of years ago. It's someone else's work, so I shouldn't share it on an open forum.

If you've got any specific questions on footy excel stuff, fire away, I'll do my best to answer.

[wearthefoxhat]

An update on how things are progressing and my latest musings.

The Championship league have now had 5 games completed. For comparison purposes, I wanted to compare the FoxG with the normal league table. As the season develops, the L5 games will become a rolling form table.

ChampionshipTable.png

GamceScoreL5Form.png

The Attack/Defence figures can be then input into the poisson prediction model for each game.

For illustration purposes, I'm using the Stoke v Hull upcoming game. (20th Sept 24)

Input in cells C4 and C5 are for the League Home & Away average
Input in cells E4 and G4 are for the Home team (Stoke) last 5 games
Input in cells E5 and G5 are for the Away team (Hull) last 5 games

Then Hey Presto!

StokevHull.png

It would appear a tight game, with an edge towards the away team.

If I had complete trust in the model, I would then hunt down games that likely offer value, in various markets, when comparing it with the live markets.

[wearthefoxhat]

12/5 (3.40) about a "predicted" 2.29 poke, a good result.

Stokev Hull Odds.png

The FoxG not too far off either.

FoxGFT.png

[wearthefoxhat]

I found some time on my hands, so I revisited the good ole Poisson Distribution goal prediction strategy.

Shout out to Kie Millett on you-tube for his video, as he goes through step by step on how to set one up in excel.

https://www.youtube.com/watch?v=24e_Z4WHR48&t=545s

It gets a bit little full on, especially if you are weak on excel, but worth the effort if you pause and work through it. I bolted on Archerys' old spreadsheet to calculate the Attack/Defence stats. I split his sheet into Home/Away and works well with the data validation used for the drop down menus.

Testing my own version (based on Kie's) on tonights Scots Premiership, it aligns very well when compared to all the live markets.

DundeeFC v Rangers.png

Match Info.png

Correct Scores.png

The outlier results could skew the stats sometimes. For example the recent 3-6 Spurs v Liverpool game will influence their Att/Def stats in the short-term. So, using xG instead might be advisable. Kie says this in the video and shows how to copy and paste from the understats site.

https://understat.com/

The understats xG site has all the main european leagues, so for the other leagues, soccer stats normal form tables home/away can be used.

I also found the Fishy site a good resource too.

https://thefishy.co.uk/formtable.php?ta ... tab-last10

[andy28]

I liked that video and have spent the day sorting it out, the difference I made was I use a web site that I can scrape scores and Xg from so all I have to do now is refresh every week, I use offset in my table to give me the last amount of games. I have just got it for EPL but it is just a matter of copying now and adding in data from each league

[wearthefoxhat]

I liked that video and have spent the day sorting it out, the difference I made was I use a web site that I can scrape scores and Xg from so all I have to do now is refresh every week, I use offset in my table to give me the last amount of games. I have just got it for EPL but it is just a matter of copying now and adding in data from each league

Nice one.

Haven't been able to get the web scrape going yet. Similar to you, I've semi-automated it by highlighting a range of data on a website and using one of the paste options to dump in directly into the home table. then repeat for the away table. Once done for each league, the drop downs automatically bring the data across the markets for the each team. Any league with a Home/Away form table or Date range setting can to used at any time.

It's then a question of cross checking with the market for value opportunities.

Peter Webb and one of the forum posters, Derek, have posted up some good stuff on the use of Poisson. I think we all find our own way that works, and with all the readily available data (free) on various websites, there's no reason we can't make a few sovs as well.

My plan is, to specialise in one league, create my on xG (FoxG) for each game based on stats in sofa score, and use Poisson for predictive purposes for identifying value in all the markets. That way, you get to know all the teams and players well, and will likely gain an ongoing edge over the market.

There was no xG on sofa-score's match preview for the DundeeFC v Rangers game last night, so by creating your own, at least there's something to measure the game on.

Dundee v Rangers.png

[andy28]

I use this site https://fbref.com/en/. When I say scrape I mean I just load the pages I want via Data, From Web into excel. Then just refresh each time you want it updated.

I am still not sure about using Xg tho as I looked at Arsenal Spurs match and it's saying the XG will be around 1.6, I am not sure that will be the case. I know you have to look at other data for the full picture. I am thinking of trying a binary method into Xg as that will factor out outlier 6-3 score lines

I think I will concentrate on 1 league as well, I think confidence is important, I see teams that are level with a few minutes to go but manage to pull off a win. I watched the A league match last night and Western Utd scored 2 goals in extra time to win 3-4, They did they same thing at Perth down 2-1 they got up 2-3, 2 goals in extra time. You could see the momentum shifting and at 3-3 I knew they would score again, but more importantly so did they players.

A few years back I had a confidence meter for the NRL that worked quite well. So might have to dust it off and adjust for football