Carey on a free? (joining St Johnstone) | Page 4 | PASOTI
  • Welcome to PASOTI. Sponsored by Lang & Potter

Carey on a free? (joining St Johnstone)

Sep 2, 2008
2,857
478
All this talk about him possibly coming back is pointless anyway. I’m pretty sure that having been in Bulgaria for so long, we won’t have access to any stats to see if he can do a job for us.

Another example of how recruiting using stats restricts us. We only look at players that the stats direct us to which are probably all UK based players. No more Friio’s, Buszaky’s or Halmosi’s.
 
Jun 27, 2021
274
225
All this talk about him possibly coming back is pointless anyway. I’m pretty sure that having been in Bulgaria for so long, we won’t have access to any stats to see if he can do a job for us.

Another example of how recruiting using stats restricts us. We only look at players that the stats direct us to which are probably all UK based players. No more Friio’s, Buszaky’s or Halmosi’s.
Carey I feel is a non starter however I don’t think the recruitment with stats is a closed book, it’s is being used for our recruitment however there will be the odd player that may need a trial or extra scouting/recommendations where like you say the stats are hard to come by.
 
Last edited:

up the line

🚑 Steve Hooper
🌟Sparksy Mural🌟
Mar 7, 2010
7,573
3,801
Manchester
All this talk about him possibly coming back is pointless anyway. I’m pretty sure that having been in Bulgaria for so long, we won’t have access to any stats to see if he can do a job for us.

Another example of how recruiting using stats restricts us. We only look at players that the stats direct us to which are probably all UK based players. No more Friio’s, Buszaky’s or Halmosi’s.
Good luck signing any of those players post Brexit anyway, stats or no stats 🤣
 

Argylegames

Administrator
Staff member
🏆 Callum Wright 23/24
✅ Evergreen
Jade Berrow 23/24
🎫 S.T. Donor 🎫
✨Pasoti Donor✨
🌟Sparksy Mural🌟
Jun 12, 2006
7,742
1,302
69
Hampshire, UK
www.argylegames.org.uk
We only look at players that the stats direct us to which are probably all UK based players. No more Friio’s, Buszaky’s or Halmosi’s.
The stats company we use covers world football.
 

The Doctor

🏆 Callum Wright 23/24
✨Pasoti Donor✨
Sep 15, 2003
8,892
4,296
Plymouth
andapoet.blog
All this talk about him possibly coming back is pointless anyway. I’m pretty sure that having been in Bulgaria for so long, we won’t have access to any stats to see if he can do a job for us.

Another example of how recruiting using stats restricts us. We only look at players that the stats direct us to which are probably all UK based players. No more Friio’s, Buszaky’s or Halmosi’s.
I think this is completely wrong. All of the companies that produce player stats will be cross-calibrating the stats across different leagues and different countries so that their products are useful to the clubs that pay for them. This is done by using games between clubs in different leagues such as domestic cup games, European competitions etc. For example, if team A from Bulgaria plays team B from Italy and B then plays C from England then that ties together the stats for the Bulgarian, Italian and English leagues - obviously this would not be very reliable if you only had the two games to go on but if you take every single cross-league game that is played then you have a huge network of connections that fairly robustly links together all of the different leagues considered (and you would also have a measure of how reliable each linkage is based on how many pieces of information help to form that link). The data companies could also cross-calibrate by looking at the stats for individual players when they move between clubs at different levels or different countries. You don't need to calibrate every individual player in this way, just enough players to enable the relative levels of the leagues within which the players are based to be determined. Otherwise the player stats would be pretty useless - you might know that a striker has the best stats in a given league but if you have no way of cross-calibrating that league against other leagues then the stats are only useful WITHIN that league.

I think that anyone who thinks that Eyeopener's comment is correct cannot really be understanding of just how sophisticated and powerful the world of statistics and data analytics really is (not only in football but in many areas).
 
Mar 16, 2009
1,082
652
London
I think this is completely wrong. All of the companies that produce player stats will be cross-calibrating the stats across different leagues and different countries so that their products are useful to the clubs that pay for them. This is done by using games between clubs in different leagues such as domestic cup games, European competitions etc. For example, if team A from Bulgaria plays team B from Italy and B then plays C from England then that ties together the stats for the Bulgarian, Italian and English leagues - obviously this would not be very reliable if you only had the two games to go on but if you take every single cross-league game that is played then you have a huge network of connections that fairly robustly links together all of the different leagues considered (and you would also have a measure of how reliable each linkage is based on how many pieces of information help to form that link). The data companies could also cross-calibrate by looking at the stats for individual players when they move between clubs at different levels or different countries. You don't need to calibrate every individual player in this way, just enough players to enable the relative levels of the leagues within which the players are based to be determined. Otherwise the player stats would be pretty useless - you might know that a striker has the best stats in a given league but if you have no way of cross-calibrating that league against other leagues then the stats are only useful WITHIN that league.

I think that anyone who thinks that Eyeopener's comment is correct cannot really be understanding of just how sophisticated and powerful the world of statistics and data analytics really is (not only in football but in many areas).

Lots of flaws in this.

Cup games are considerably different from domestic league seasons, as we all know.

Frankfurt would not finish anywhere near 7th in the Premier League and yet they beat West Ham over 2 legs in the Europa League recently. Loads of reasons for this, so you can't extrapolate reliable stats from this type of match up.

Similarly with players crossing borders to play for new clubs in different countries. Often players take at least a season to acclimatise. Especially top leagues like the EPL. What happens to first season stats in those instances?
 

The Doctor

🏆 Callum Wright 23/24
✨Pasoti Donor✨
Sep 15, 2003
8,892
4,296
Plymouth
andapoet.blog
Lots of flaws in this.

Cup games are considerably different from domestic league seasons, as we all know.

Frankfurt would not finish anywhere near 7th in the Premier League and yet they beat West Ham over 2 legs in the Europa League recently. Loads of reasons for this, so you can't extrapolate reliable stats from this type of match up.

There are lots of examples like that, and you wouldn't trust stats that

Similarly with players crossing borders to play for new clubs in different countries. Often players take at least a season to acclimatise. Especially top leagues like the EPL. What happens to first season stats in those examples?
That's why you wouldn't base the calibration on just a few games/cases - it would be completely unreliable. But if you base it on 100s or 1000s of examples you can certainly calibrate across leagues and levels to a known, useful level (albeit obviously not perfect).

Even in a simple analysis you could easily exclude first season stats from an analysis if you wanted to. Or you could use a taper to give more weight to more games as time goes on (or to more recent games) etc.

I think a lot of people have almost no idea how much data, how much computational effort and how sophisticated the data analysis/modelling methods used will be. This is high level mathematical and statistical modelling using methods designed by really knowledgeable and intelligent people who have studied such things for years. It's not just counting or calculating a few averages (the kind of statistics that most people encounter in school). It's not like saying that if we'd beaten Chelsea in the FA Cup this season we'd have been the best team in the world because they won the World Club Championship (like conkers).
 
Mar 16, 2009
1,082
652
London
That's why you wouldn't base the calibration on just a few games/cases - it would be completely unreliable. But if you base it on 100s or 1000s of examples you can certainly calibrate across leagues and levels to a known, useful level (albeit obviously not perfect).

Even in a simple analysis you could easily exclude first season stats from an analysis if you wanted to. Or you could use a taper to give more weight to more games as time goes on (or to more recent games) etc.

I think a lot of people have almost no idea how much data, how much computational effort and how sophisticated the data analysis/modelling methods used will be. This is high level mathematical and statistical modelling using methods designed by really knowledgeable and intelligent people who have studied such things for years. It's not just counting or calculating a few averages (the kind of statistics that most people encounter in school). It's not like saying that if we'd beaten Chelsea in the FA Cup this season we'd have been the best team in the world because they won the World Club Championship (like conkers).

I do have a good understanding of the amount of data available and the analysis that takes place - so agree with you there.

There are a variety of pro platforms that can be subscribed to, such as Wyscout, that are used by clubs, media etc. It's not difficult to get hold go this data. It's just that much of it is questionable. I guess that's where a good data analyst comes into their own.
 
Last edited:

jespafc

✨Pasoti Donor✨
🌟Sparksy Mural🌟
Oct 23, 2005
2,094
736
www.johnstanlake.com
Lots of flaws in this.

Cup games are considerably different from domestic league seasons, as we all know.

Frankfurt would not finish anywhere near 7th in the Premier League and yet they beat West Ham over 2 legs in the Europa League recently. Loads of reasons for this, so you can't extrapolate reliable stats from this type of match up.

Similarly with players crossing borders to play for new clubs in different countries. Often players take at least a season to acclimatise. Especially top leagues like the EPL. What happens to first season stats in those instances?
I don't see why Frankfurt wouldn't finish 7th in the Premier League. Are they that far behind a team like West Ham or Wolves? They finished 5th in the Bundesliga last season (2021) and 7th in 2019 and have just won the Europa League, so they must have something about them.
 
Mar 16, 2009
1,082
652
London
I don't see why Frankfurt wouldn't finish 7th in the Premier League. Are they that far behind a team like West Ham or Wolves? They finished 5th in the Bundesliga last season (2021) and 7th in 2019 and have just won the Europa League, so they must have something about them.

The Bundesliga is nowhere near the level of the Premier League.

It's an entertaining league full of goals for sure, but way behind in terms of quality of defences, competitiveness and quantity of elite players at the top of their game. It's a good place to nurture and develop players of promise. Plenty of EPL clubs send players on loan to Bundesliga clubs to give them game time. Plenty never make it in the EPL on return.

Conversely, lots of Bundesliga stars struggle in the Prem. Take Timo Werner. Scored for fun at Leipzig.
 
Sep 6, 2006
16,422
3,960
The Bundesliga is nowhere near the level of the Premier League.

It's an entertaining league full of goals for sure, but way behind in terms of quality of defences, competitiveness and quantity of elite players at the top of their game. It's a good place to nurture and develop players of promise. Plenty of EPL clubs send players on loan to Bundesliga clubs to give them game time. Plenty never make it in the EPL on return.

Conversely, lots of Bundesliga stars struggle in the Prem. Take Timo Werner. Scored for fun at Leipzig.
Frankfurt also beat Barcelona Justin.