Football Question (5 Viewers)

[Philosoraptor]

As most of you are probably aware by now, my football knowledge is pretty much non-existent. Therefore, I need to throw out stuff on here to get answers to questions.

My question involves Appearances.

When does a player actually get recognized for their Appearances.

Is it when they are in the match day squad or when they touch the grass in a game?

Just to be clear on this, does a substitute make an appearance when they play or does it get added if they make the bench?

I do need some help on this

 

---

[baldy]

Just when they actually get on I’d say (the sub appearances would be classed separate)

 

[Woodingdean_Sky_Blue]

I agree with Baldy, it only counts as an appearance when the substitute player gets onto the pitch. Substitute appearances are generally classed separately to starting line-up appearances.

 

[Magwitch]

Has to be when on the pitch

 

[Oh for an Ian Gibson.]

I was on the pitch twice during the Wolves game when we gained promotion to the old division one, does this count as an appearance?

 

[Philosoraptor]

I was on the pitch twice during the Wolves game when we gained promotion to the old division one, does this count as an appearance?

Yes!

 

[Philosoraptor]

Cheers good people, I think I have the info I need.

 

[Johhny Blue]

I was on the pitch twice during the Wolves game when we gained promotion to the old division one, does this count as an appearance?

Based on that I should qualify for a testimonial

 

[Philosoraptor]

Just when they actually get on I’d say (the sub appearances would be classed separate)

What I am playing around with at the moment gives everyone a rating based upon being on the pitch during a game and the result.

If you involved in the game on the pitch then you get the ratings for that match even if you came off or on, or completed the full game.

I am using the Elo Rating System powered by Glicko-2. I have just caught up to Game 8 and reviewed everything for the first time.

I will have to show it to everyone to see if it is up to scratch.

 

[Woodingdean_Sky_Blue]

What I am playing around with at the moment gives everyone a rating based upon being on the pitch during a game and the result.

If you involved in the game on the pitch then you get the ratings for that match even if you came off or on, or completed the full game.

I am using the Elo Rating System powered by Glicko-2. I have just caught up to Game 8 and reviewed everything for the first time.

I will have to show it to everyone to see if it is up to scratch.

A player's individual minutes on the pitch against team result achieved (goals scored/conceded) during that timeframe might be interesting.

 

[Philosoraptor]

A player's individual minutes on the pitch against team result achieved (goals scored/conceded) during that timeframe might be interesting.

That's an interesting proposition which may be needed to be explored more.

Okay, here's the first few games.

I would like to point out all players are started on 1500 points and there is a sorting action going on until the system finds a stable rating for the players.

Does this chart have any resemblance to the first 8 games played and the individual strengths of players?

 

[Frostie]

What I am playing around with at the moment gives everyone a rating based upon being on the pitch during a game and the result.

If you involved in the game on the pitch then you get the ratings for that match even if you came off or on, or completed the full game.

I am using the Elo Rating System powered by Glicko-2. I have just caught up to Game 8 and reviewed everything for the first time.

I will have to show it to everyone to see if it is up to scratch.

From what's shown so far I have a few queries;

So the player is rated for that match regardless of how many minutes he plays in & regardless of the context of that game during their time on the pitch?

Real world example taking the Nottingham Forest game:

Godden is withdrawn after 56mins, replaced by Walker at which point we are 1-0 down but equalise almost immediately after.

Giles is withdrawn after 90+2mins, replaced by Dacosta with the score still 1-1.

Forest score after 90+7mins to win 2-1.

Godden, Walker, Giles & Dacosta would all score the same for that match?

 

[Philosoraptor]

From what's shown so far I have a few queries;

So the player is rated for that match regardless of how many minutes he plays in & regardless of the context of that game during their time on the pitch?

Real world example taking the Nottingham Forest game:

Godden is withdrawn after 56mins, replaced by Walker at which point we are 1-0 down but equalise almost immediately after.

Giles is withdrawn after 90+2mins, replaced by Dacosta with the score still 1-1.

Forest score after 90+7mins to win 2-1.

Godden, Walker, Giles & Dacosta would all score the same for that match?

Yes, if they are all involved in the match shouldn't they all be scored equally?

Each game played has been weighted for team strength using Glicko-2 and the score by Elo for that particular game for individual ratings has been adjusted to reflect this.

 

[Sky_Blue_Dreamer]

What I am playing around with at the moment gives everyone a rating based upon being on the pitch during a game and the result.

If you involved in the game on the pitch then you get the ratings for that match even if you came off or on, or completed the full game.

I am using the Elo Rating System powered by Glicko-2. I have just caught up to Game 8 and reviewed everything for the first time.

I will have to show it to everyone to see if it is up to scratch.

Please call your rating system Hey There, Mr Sky Blue

 

[Frostie]

Yes, if they are all involved in the match shouldn't they all be scored equally?

Each game played has been weighted for team strength using Glicko-2 and the score by Elo for that particular game has been adjusted.

Hmmm, not sure about that. Wouldn't it be better to weight it or omit players playing very few minutes?

As we are one of the weaker teams we are likely to lose a lot of games so there is a chance players making short cameos in wins will score unfairly highly.

For example, with the 1500 starting point Marcel Hilßner with 0 minutes played still ranks as one of our better performers as it stands.
If, hypothetically, he came on in stoppage time of 1 or 2 games where we'd already all but secured a win it would make it look as though he's one of our best players?

 

[Philosoraptor]

Hmmm, not sure about that. Wouldn't it be better to weight it or omit players playing very few minutes?

As we are one of the weaker teams we are likely to lose a lot of games so there is a chance players making short cameos in wins will score unfairly highly.

For example, with the 1500 starting point Marcel Hilßner with 0 minutes played still ranks as one of our better performers as it stands.
If, hypothetically, he came on in stoppage time of 1 or 2 games where we'd already all but secured a win it would make it look as though he's one of our best players?

Each player starts ELO on (1500) and as with Glicko-2 there does need to be a certain passage of games for each player before a provisional score becomes an actual score.

 

[Philosoraptor]

I think we need to get more games completed and put up on the chart to see if this way will work or not.

It is all a bit experimental, but I think there may be something to this which we wouldn't normally see usually.

I'm working on a hunch this side.

 

[Johhny Blue]

That's an interesting proposition which may be needed to be explored more.

Okay, here's the first few games.

I would like to point out all players are started on 1500 points and there is a sorting action going on until the system finds a stable rating for the players.

Does this chart have any resemblance to the first 8 games played and the individual strengths of players?

If each player starts on 1500 points and we have more than 20 players surely that’s enough points to stay up?

 

[Philosoraptor]

If each player starts on 1500 points and we have more than 20 players surely that’s enough points to stay up?

I think this is something we need to sit down and watch over the course of season to see if it is going to work or not.

 

[Frostie]

Each player starts ELO on (1500) and as with Glicko-2 there does need to be a certain passage of games for each player before a provisional score becomes an actual score.

Ok. I still don't know how 'fair' or truly reflective it will be but then I don't fully understand Glicko systems either.
Be interested to see how it pans out.

 

[BornSlippySkyBlue]

I think this is something we need to sit down and watch over the course of season to see if it is going to work or not.

(Psst, I think he was jesting )

 

[Philosoraptor]

This one is quite simple folks, anyone can play this game.

Do you think the Individual Statistics displayed are a true reflection of squad strength?

I do not have the necessary expertise to answer that question.

What I can say is the data will get richer over the course of the season and patterns will form.

These patterns might be interesting to view.

 

[JulianDarbyFTW]

This one is quite simple folks, anyone can play this game.

Do you think the Individual Statistics displayed are a true reflection of squad strength?

I do not have the necessary expertise to answer that question.

What I can say is the data will get richer over the course of the season and patterns will form.

These patterns might be interesting to view.

Unfortunately, I'd say it doesn't fully reflect the squad strength at present.

Hilsner hasn't kicked a ball in anger, yet appears to be rated near the top. Biamou has most definitely not performed better than Godden. Pask was poor, and played a grand total of something like 2 games. McCallum has not been the worst player this season (his form was shit for 2 or 3 games recently, but that aside he's been solid).

Also, I can only make out a handful of player names on there, which kinda defeats the purpose of a chart like this.

Rating players is iffy. I love InfoGol's player ratings, but they don't always accurately represent what we see on the pitch either. I'm not sure it's easy to quantify a player's performance numerically, other than through more traditional stats like goals, assists, key chances created, pass completion and the like, and even then you have a weighting issue - is one tackle worth the same as one completed pass?

 

[Philosoraptor]

So out of curiosity I have completed up to Game 16. Still 7 games to go.

This is what popped up.

And yes, I may need to change visualizations to a different chart. I'm just looking if there are any patterns or information which may come off this.

 

[SlowerThanPlatt]

So out of curiosity I have completed up to Game 16. Still 7 games to go.

This is what popped up.

And yes, I may need to change visualizations to a different chart. I'm just looking if there are any patterns or information which may come off this.

Forgive my ignorance but what is it supposed to be showing? If it’s player performance then Hilssner couldn’t possibly be above McFadzean or Giles as he hasn’t played a game yet.

The formula needs some work?...

 

[Philosoraptor]

Forgive my ignorance but what is it supposed to be showing? If it’s player performance then Hilssner couldn’t possibly be above McFadzean or Giles as he hasn’t played a game yet.

The formula needs some work?...

Imagine you are looking at a balance that is situated on 1500 points

This is not a chart that goes from 0 - 1000+ chronologically.

1500 is the axis.

 

[BornSlippySkyBlue]

Forgive my ignorance but what is it supposed to be showing? If it’s player performance then Hilssner couldn’t possibly be above McFadzean or Giles as he hasn’t played a game yet.

The formula needs some work?...

Think it’s explained above that the ‘default’ score of 1500 can go down as well as up, so if a player doesn’t play his score can be higher than someone that has played 10 games and lost points (so to speak).

 

[Philosoraptor]

Football not a zero-sum game. Discuss!

 

[shmmeee]

That's an interesting proposition which may be needed to be explored more.

Okay, here's the first few games.

I would like to point out all players are started on 1500 points and there is a sorting action going on until the system finds a stable rating for the players.

Does this chart have any resemblance to the first 8 games played and the individual strengths of players?

So you’re saying we should play Hilßner as he can’t be any worse that what we’ve got?

How hard would it be to account for changes to the result when they’re on the pitch?

 

[shmmeee]

Surely you’ve got to apply a penalty for not getting picked (ideally allowing for injury)?

 

[Philosoraptor]

Surely you’ve got to apply a penalty for not getting picked (ideally allowing for injury)?

I think if any player gets to have a continuously high Elo then the Premiership is beckoning for them.

 

[speedie87]

A rating must surely take into account.

If started game
If injured so not available
Minutes played
Result of game whilst on pitch
Opposition rating

Someone Who cant even get in the team basically shouldn’t score higher than those that do. Ie pask

 

[Philosoraptor]

Okay, a few things to post.

Does anyone remember last season?

I don't think I have ever shown this properly.

 

[Philosoraptor]

This season so far.

I just like to point out that I never had any intention of doing this.

Also, this is not a measure of individual ability.

Also, I believe the 1500 individual ratings (Elo) ceteris paribus is tied directly to the centre point of Glicko-2 version of the Championship table. I am not too sure though if the measurements of the range are going to be the same between Glicko-2 and Elo. This is purely experimental. Both rating systems were never designed to be set up to be used this way.

If it works then it will be a bonus. If it doesn't, then we would have learned something else.

----

Goalkeepers




Defenders

 

[Philosoraptor]

Part 2

--

Midfielders




Forwards