Tiebreaker function

[luisgomezesp]

Hi, I'm trying to make a Spanish league table and I have found a little problem. I start with a database with all the matches played with their results,each match has two entries, one for each team, there you can find the result and the points won by each team. It´s not dificcult to make a program for calculating the table at each week but my problem is with the tiebreaker function. In Spain when 2 or more teams are tied, the first team at the table is the one who have won more points at the head to head matches, so I have programmed that:

tiebreaker=function(data,tied){
data2=filter(data,Team %in% tied,
Rival %in% tied)
data2=summarise_at(group_by(data,Team),vars(Points,goalaverage),funs(sum))
data2=arrange(m, -Points,-goalaverage)
return(data2)
}
Where tied are the teams with the same amount of points.

So, my question is how can I use this function after having making the table with the arrange function.

[kaijabean]

Hi @luisgomezesp,
Thanks for describing your problem and providing the code you've already tried! Could you also provide us with some sample data so that we can run your code and help you more easily? You will get an answer much quicker if we can see some of your data to use as an example.

This article gives some help on how to create a minimal reproducible example. If you can post a small sample of your data, or of some fake data that reproduces your problem (in case your actual data are sensitive etc), that would be great!

Thanks

[luisgomezesp]

Thanks for your message, I can give you the data and the code, it isn't sensitive at all hahahaha.
The database I have done is like this:

Equipo Rival Jornada LoV Marcados Recibidos Resultado Puntos

1 Villarrubia Villanovense 1 L 0 3 P 0
2 Villanovense Villarrubia 1 V 3 0 G 3
3 Socu Badajoz 1 L 1 1 E 1
4 Badajoz Socu 1 V 1 1 E 1
5 Melilla Villarrobledo 1 L 3 0 V 3
6 Villarrobledo Melilla 1 V 0 3 D 0
7 Merida Talavera 1 L 0 1 D 0
8 Talavera Merida 1 V 1 0 V 3
9 Donbe Extremadura 1 L 1 2 D 0
10 Extremadura Donbe 1 V 2 1 V 3

Where the important variables are Equipo (team), Rival and Puntos (Points).
The table is ordered in first place by the points won by each team, but if there is a tie between two or more teams we see the head to head matches.
For example, now Donbe and Talavera have 20 points each one but Donbe is in a better position than Talavera because in the matches between then the results were:
Donbe 2 Talavera 0
Talavera 1 Donbe 1
Of course we can have 2, 3 or more teams with the same amount of points.

So, I can order the teams by the points won and I can make the tiebreaker function (the one I posted yesterday) but I dont know how to put both together.
For ordering the teams we can use:

l=summarise_at(group_by(datos,Equipo),vars(Puntos),funs(sum))
l=arrange(l, -Puntos)

The real tiebreaker function I have is:

desempate=function(datos2,empatados){
  datos3=filter(datos2,Equipo %in% empatados,
                Rival %in% empatados)
  datos3$goalaverage=datos3$Marcados-datos3$Recibidos
  if(dim(datos3)[1]==
     4*factorial(length(empatados))/(2*factorial(length(empatados)-2))){
    datos3=filter(datos2,Equipo %in% empatados,
                  Rival %in% empatados)
    datos3$goalaverage=datos3$Marcados-datos3$Recibidos
    m=summarise_at(group_by(datos3,Equipo),vars(Puntos,goalaverage),funs(sum))
    m=arrange(m, -Puntos,-goalaverage)
    
  }else{
    
    datos2$goalaverage=datos2$Marcados-datos2$Recibidos
    datos2=transform(datos2,averagecum= ave(goalaverage,Equipo, FUN =function(x) cumsum(x)))   
    m=summarise_at(group_by(datos2,Equipo),vars(averagecum),funs(last))
    datos3=filter(m,Equipo %in% empatados)
    m=arrange(datos3,-averagecum)
  }
  
  return(m)
}

Its not very important but if the teams tied have not played all the matches between the other ones the tiebreaker function would be the difference between scored and received goals.

Sorry for bad English and for bad programming I hope you have understood me and you could help me.
Thank you

[nirgrahamuk]

@luisgomezesp

Equipo Rival Jornada LoV Marcados Recibidos Resultado Puntos

1 Villarrubia Villanovense 1 L 0 3 P 0
2 ....

Once we copy and paste this into our editors how do we make an R object out of it to use in the code for your example?

(Hint : read the linked guide)