Football Musings

[andy28]

When I say VBA I cheat and do it as a record Macro, but once I have done that I can go into the code and change things if I have to, I find this is the best way to learn. Yeah I have taken early retirement so all the maths and Excel was new to me but I enjoy it so I carry on, will never earn anywhere what most on here earn but I don't have to I just plod along at my own pace. But I will say the more skills I pick up the faster I am learning, definitely a snowball effect.

[wearthefoxhat]

Added to the spreadsheets a Bayesian Thoerem "washing machine" to put through my tissue prices and the live market prices to see what comes out, and then make a value assessment. (probably all pink as some of my ole shirts used to be....)

[wearthefoxhat]

Running a test with the Bayesian value assessment and also added in a League points factor adjustment to "improve" expected goals.

QPR v Sheff Wed

QPR v Sheff Wed Sheet.png

Looks too good to be true (we know what that means). QPR have struck a rich vein of form and are scoring well, whereas, Sheff Wed, although playing okay, are conceding a few goals recently away and won't fancy a long arse trip down to London.

QPR.png

SheffWed.png

Marvellous... FT QPR 0-2 Sheff Wed (Reversion to the mean?)

[firlandsfarm]

If it's assumed Bf markets are reasonably efficient including the Correct Score market then an approximation of goals can be derived from summing the probabilities of the individual scorelines.

[wearthefoxhat]

If it's assumed Bf markets are reasonably efficient including the Correct Score market then an approximation of goals can be derived from summing the probabilities of the individual scorelines.

Yep, the CS score market could be used that way.

From what I've seen, the Betfair market makers tend to follow the Pinnacle prices initially, and Pinnacle aren't too far off the Club Elo prices that are available well in advance.

[firlandsfarm]

If it's assumed Bf markets are reasonably efficient including the Correct Score market then an approximation of goals can be derived from summing the probabilities of the individual scorelines.

Yep, the CS score market could be used that way.

From what I've seen, the Betfair market makers tend to follow the Pinnacle prices initially, and Pinnacle aren't too far off the Club Elo prices that are available well in advance.

and there is the rub ... it's becoming almost impossible to find an edge when betting. I spend most of my time in Forex these days.

[andy28]

In my opinion, while statistics provide valuable insights, it’s also essential to highlight unique strengths that set us apart. Even the big players leave gaps that can be exploited. Currently, team confidence is a significant factor for me. This confidence is subjective and varies by individual, making it hard to quantify.

For example, a team might win 1-0 at home yet gain minimal confidence, the Manchester City v Southampton match, City wouldn’t gain much confidence from that win despite being pleased with the result. Conversely a narrow win against a top team like Liverpool would significantly boost their morale.

It’s all about who you beat and by what margin. Even weaker teams playing away would be content with a 0-0 draw against teams like Arsenal, City, or Liverpool. Recently I noticed Everton's rising confidence they drew with Arsenal and City, and even a narrow 1-0 loss to Bournemouth didn’t dent their confidence too much. They won today against a team experiencing a dip in confidence, so I wasn't surprised with the outcome.

From all the matches played this weekend, only one didn’t favour the team I rated higher on confidence. The Forest match both teams weren't lacking confidence but Forest were rated higher.

It’s still early days so I remain sceptical.

[andy28]

Looking at Everton further they had just one goal from 6 away matches but I had the confidence was building

[firlandsfarm]

Confidence:, the unmeasurable, unseen parameter that has more impact than an injured best player (but such an injury has an affect on confidence!).

[Kai]

Currently, team confidence is a significant factor for me.

Spot on Andy

Psychological momentum is a big factor and you see what happens when top teams lose confidence, or how high others can climb

When you're a fan following your team or a professional specializing one, your confidence in that team will automatically mirror theirs as it ebbs and flows

Both on a collective team level and on an individual player level, so it can be something to compare with prices on offer whether it's a couple days before KO or during play itself

But something to always keep in mind is the more opinion you form the more bias you potentially invite, if your takes are way off base you might fall into some big traps, so some balance there is very much a requirement imo

[andy28]

I don't rely solely on my opinions, I collect data including the odds for matches. For instance, if a team has odds of 1.25 to win and then draws 0-0 against a lower ranked team, that certainly impacts their rating downward. It's crucial not to base ratings purely on personal opinion, as it may introduce bias. However, the challenge lies in translating these ratings into expected goals.

Additionally it may be better used for in-play betting as well. If a team with 1.25 odds falls behind 1-0 after the first 15 minutes, and they were expected to perform strongly before the match, it might be a good opportunity to back them at longer odds, particularly if the goal came against the run of play , maybe even mean that they will hold on to a lead in a tight finish.

I think it is worth while digging deeper.

[wearthefoxhat]

Some good insights there for sure. I asked GPT (yeah I know) for further consideration, might be of some use going forward.

1. Recent Form

Last 5 Matches: Points accrued (e.g., wins = 3, draws = 1, losses = 0) give an indicator of recent performance trends.
Goal Difference: A positive goal difference reflects dominance in recent matches.
Clean Sheets: The number of games without conceding shows defensive solidity.


2. Home/Away Performance

Home Advantage: Teams often perform better at home. Analyze recent home/away records.
Travel Distance: Long travel times can reduce confidence, especially for away teams.


3. Player Form and Availability

Key Players: Look at recent performance metrics for critical players (e.g., goals scored, assists, saves).
Injuries/Suspensions: Missing key players can significantly impact morale and confidence.
Return of Key Players: A star player's return from injury can boost confidence.


4. Team Chemistry

Passing Accuracy and Possession Stats: High possession and passing accuracy reflect cohesion.
Communication: Interviews and team interactions can provide qualitative insights.


5. Psychological Factors

Head-to-Head Record: A good record against the opponent boosts confidence.
Momentum: Consecutive wins or strong comeback performances improve morale.
Pressure Situations: Teams used to high-pressure games often handle them better.


6. Manager and Tactical Setup

Managerial Stability: Teams with consistent managerial leadership often exude confidence.
Tactical Readiness: Comments from players or managers about being "prepared" tactically.


7. Betting Market Sentiment

Shifts in Odds: A shortening of odds often indicates rising confidence in the market.


8. Training and Preparation

Positive Reports: News about focused training sessions, improved fitness, or morale-boosting activities.
Team Dynamics: Public gestures of unity (e.g., team dinners, social media posts).


9. External Influences

Fan Support: A large, vocal fan presence, particularly at home, boosts morale.
Media Narratives: Favorable media coverage or pundit confidence can indirectly influence mindset.


10. Statistical Ratings

Expected Goals (xG) Trends: An upward trend in xG reflects offensive confidence.
Team Strength Ratings: Use Elo ratings or similar metrics for a more mathematical approach.

By combining these metrics, you can paint a comprehensive picture of how confident a team is heading into a match. Do you want to focus on specific metrics or explore integrating them into your rating system?


------


Could you give each of the 10 a ranking from 1 to 10 (1 being the best). Also produce a likert scale (5 being the highest) for each of the ten in a table


-----

Metrics.png

----

[wearthefoxhat]

If it's assumed Bf markets are reasonably efficient including the Correct Score market then an approximation of goals can be derived from summing the probabilities of the individual scorelines.

Yep, the CS score market could be used that way.

From what I've seen, the Betfair market makers tend to follow the Pinnacle prices initially, and Pinnacle aren't too far off the Club Elo prices that are available well in advance.

and there is the rub ... it's becoming almost impossible to find an edge when betting. I spend most of my time in Forex these days.

I think there's some areas of value from the leagues not covered by Club Elo.

With Forex, for the craic, I put together an excel sheet where you input the price of a Forex pairing, it calculates the Harmean, standard deviation, std error mean, self adjusting Z-scores, and then, predicts a buy or sell signal with a trailing stop-loss,

Can be used for stock prices too.

[firlandsfarm]

With Forex, for the craic, I put together an excel sheet where you input the price of a Forex pairing, it calculates the Harmean, standard deviation, std error mean, self adjusting Z-scores, and then, predicts a buy or sell signal with a trailing stop-loss,

Can be used for stock prices too.

That's way too complicated for me Foxy ... I just see the trend, see the MACD, maybe look at an EMA ribbon for confirmation and have the Parabolic on the chart and go for it if it looks good. Remember, unlike stocks, FX in the main trades within a range. I guess you can reduce my strategy to "see if it looks a good move and use risk/money management to 'encourage' the profit to come"

[andy28]

Todays matches, I had Brentford over CP, Leicester over Spurs, Av over West H & Man U over Fulham. The match with the biggest variation was Spurs match, both teams were below ave but Spurs is amongst the worst.

I asked Chat Gpt as well and got a similar response, however what AI's cant do is mimic the changing rooms. Most of us have played sport and some wins (or losses) boost confidence more than others.

The key will also be how it effects teams going forward, how does this impact Leicesters next match or Spurs for that matter