| Analytics 15:39 - Aug 14 with 1556 views | JPC |
Can anyone explain how you get a competitive advantage out of this? I presume that everybody is using some form of analytics for recruitment? I imagine that a few (e.g. Brighton and Brentford through their betting company owners) have more / better data, but for all of the others, is it simply a case of having better analysts to interpret the data? In the old days, a scouting system presumably depended on trusted scouts with a track record of finding players - what drives the advantage today? |  | | |  |
| Analytics on 16:44 - Aug 14 with 1446 views | Dorse |
I was going to explain it but I don't want to lose my competitive advantage.
 |  |
| 'What do we want? We don't know! When do we want it? Now!' |
| |
| Analytics on 17:13 - Aug 14 with 1399 views | NewYorkRanger |
I find this to be a bit of dilemma tbh, not least cos most of my career has been in data, analytics etc and the old adage of more data = better decisions is generally true BUT - it does depend on who is making the decisions with said data and how they are using it.
We are all familiar with Moneyball and how the gave the unfashionable Oakland A's a big advantage but these days I think it's all gone a bit far. If you look at the half time stats they put up on a game shown on Sky its largely meaningless.
All this XG, passes completed nonsense is not for me. And if clubs are buying players based on that and not seeing what they can do in the flesh then they deserve all they get |  |
| |
| Analytics on 17:31 - Aug 14 with 1359 views | dannyblue |
I always had a hunch that the really smart ones, perhaps like Brighton, might learn what metrics their market is looking for, and then play players in such away that they stand out as obvious buys. Gaming the system in a way.
E.g., I know Chelsea are big spenders, and I know they look for <24yo, progressive carries, sprints, passes into final third and pass completion over 80%, but don't really look at interceptions or aerial duels. So I'm going to coach my 22 year old midfielder to play in a way that maximises those stats, and create a team structure that lets him do so. Right, that'll be £100m thanks.
But there's always interpretation. Is it good or bad if goals scored is above XG? Are they clinical or lucky? If goals scored are below XG is that good or bad? Do they lack composure, or have they been unlucky and will revert to mean? |  | | |
| Analytics on 17:39 - Aug 14 with 1328 views | Rsole |
I think Brighton and Brentford have proprietary data which gives them the edge.
Most of the other clubs will be looking at similar data sets and therefore it’s more of an equal footing.
There may well be some better analysts and data scientists using the data. That can also make a difference but not as much as the underlying data. |  |
| Those possessed by devils, try and keep them under control a bit, can't you ?
|
| |
| Analytics on 18:00 - Aug 14 with 1267 views | turnsey |
I'm a data analyst that doesn't work in sport, so treat this with a grain of salt. My instinct is that the quality of model/strategy matters more than the quality of data or analytics personnel at the high professional level.
If you're looking at metrics, the players at the top of your list will vary heavily based on what metrics are most valued, how they are weighed relatively to each other, and how you weigh the context of what the player is doing (e.g. how good is the league? how much do you care about the relative quality of teammates/opposition? etc.)
I would be willing to bet that for us, it's Nourry or a collective group of overseers who establishes the "model"--the metrics/variables/profiles being targeted--and then it's up to subordinates to carry out the grunt work and then present reports to the real decision-makers based on different model variations, different player pools by nation, etc. That's why I say it's more the leadership direction than the quality of analyst/data that really makes a difference.
Old-school scouting must still have a place, though. I'd guess that the data serves as an initial screening process, and then players at the top of lists are then scouted in a more traditional sense? If anyone has genuine knowledge of analysis/scouting these days at this level, I'd be really interested to know. |  | | |
| Analytics on 18:05 - Aug 14 with 1221 views | Tonto |
Think of it like Football Manager, but in real life.
You base your signings on data, not scouting reports |  |
| |
| Analytics on 18:27 - Aug 14 with 1159 views | Boston |
Nobody relies on hunches anymore, unless they're looking for a back. | |
| |
| Analytics on 18:27 - Aug 14 with 1155 views | Rsole |
| Analytics on 18:00 - Aug 14 by turnsey |
I'm a data analyst that doesn't work in sport, so treat this with a grain of salt. My instinct is that the quality of model/strategy matters more than the quality of data or analytics personnel at the high professional level.
If you're looking at metrics, the players at the top of your list will vary heavily based on what metrics are most valued, how they are weighed relatively to each other, and how you weigh the context of what the player is doing (e.g. how good is the league? how much do you care about the relative quality of teammates/opposition? etc.)
I would be willing to bet that for us, it's Nourry or a collective group of overseers who establishes the "model"--the metrics/variables/profiles being targeted--and then it's up to subordinates to carry out the grunt work and then present reports to the real decision-makers based on different model variations, different player pools by nation, etc. That's why I say it's more the leadership direction than the quality of analyst/data that really makes a difference.
Old-school scouting must still have a place, though. I'd guess that the data serves as an initial screening process, and then players at the top of lists are then scouted in a more traditional sense? If anyone has genuine knowledge of analysis/scouting these days at this level, I'd be really interested to know. |
Agreed on the strategy - data and analytics can be used and misused to fit objectives and how you present results also.
I think what the 2Bs have is data from markets that others probably don’t have - South America for example = Caicedo. It’s how they make a lot of money as Bookies with their odds.
They are also pretty shrewd in using the data to target scarcity of skills. For example, they know Chelsea will need that type of player and they had money to spend on that role, that season when ready to sell.
Not much point in stockpiling ‘keepers unless they are:
English
Good with their feet
On Man City’s shopping list
Aged under 22
Etc
Meantime, I guess everyone else looks at shared data from similar sources and packaged analytics that give you an idea but not much more than that. Next step up would be good analysts, better insights, models and algorithms that use the same data.
You would also hope that a scout has seen the player in multiple games and their feedback is another variable that should be considered but it’s somewhat subject to opinion, as you saw in Moneyball. | |
| Those possessed by devils, try and keep them under control a bit, can't you ?
|
| |
Login to get fewer ads
| Analytics on 19:07 - Aug 14 with 1091 views | Rsole |
| Analytics on 18:27 - Aug 14 by Boston |
Nobody relies on hunches anymore, unless they're looking for a back. |
FC Notre Dame | |
| Those possessed by devils, try and keep them under control a bit, can't you ?
|
| |
| Analytics on 00:14 - Aug 15 with 790 views | queensparker |
| Analytics on 17:31 - Aug 14 by dannyblue |
I always had a hunch that the really smart ones, perhaps like Brighton, might learn what metrics their market is looking for, and then play players in such away that they stand out as obvious buys. Gaming the system in a way.
E.g., I know Chelsea are big spenders, and I know they look for <24yo, progressive carries, sprints, passes into final third and pass completion over 80%, but don't really look at interceptions or aerial duels. So I'm going to coach my 22 year old midfielder to play in a way that maximises those stats, and create a team structure that lets him do so. Right, that'll be £100m thanks.
But there's always interpretation. Is it good or bad if goals scored is above XG? Are they clinical or lucky? If goals scored are below XG is that good or bad? Do they lack composure, or have they been unlucky and will revert to mean? |
The trouble with that is you have to win games doing it.
I think that's closer to what we were doing last season - buying players who fit a data model who don't actually work in real life.
Brighton's astonishing rise up the ranks would suggest they're better at it than we are. |  | | |
| Analytics on 00:37 - Aug 15 with 758 views | Trom |
Whilst data is important you can’t dismiss the human element.
You have to identify players and then convince them that the opportunity and salary makes you viable. Given we can’t really compete on salary the career enhancement has to be the key to get deals over the line.
So you both need to identify players with the right characteristics and then offer them a pathway to advance their career.
So structure of the club, manager, past success and notable coaches will all be important. Identifying is one thing and attracting is another. Some cautious signs of improvement from the club given our summer signings. As ever we will see as the season plays out. |  | | |
| Analytics on 01:36 - Aug 15 with 683 views | NW5Hoop |
| Analytics on 17:31 - Aug 14 by dannyblue |
I always had a hunch that the really smart ones, perhaps like Brighton, might learn what metrics their market is looking for, and then play players in such away that they stand out as obvious buys. Gaming the system in a way.
E.g., I know Chelsea are big spenders, and I know they look for <24yo, progressive carries, sprints, passes into final third and pass completion over 80%, but don't really look at interceptions or aerial duels. So I'm going to coach my 22 year old midfielder to play in a way that maximises those stats, and create a team structure that lets him do so. Right, that'll be £100m thanks.
But there's always interpretation. Is it good or bad if goals scored is above XG? Are they clinical or lucky? If goals scored are below XG is that good or bad? Do they lack composure, or have they been unlucky and will revert to mean? |
I don't think that can be right. Take Caicedo. Allow six months for scouting and all that. Then he was at Brighton for two and a half years. The notion of targeting specfic players to sell to specific clubs would then depend on Brighton at some point in 2021 knowing exactly what kind of player Chelsea would need in 2024.
Even Chelsea didn't know what kind of player they would need in 2024. In 2024. | | | |
| Analytics on 06:08 - Aug 15 with 552 views | PlanetHonneywood |
Footy data analytics is a bit like IT and cars for me. Yeah, fascinating stuff, but for other people to get excited about.
Someone mentioned above xG (no idea what it is and feel free NOT to enlighten me) and the plethora of data that TV companies throw at you before, during, at the halfway point, and at the end of the gane as being excessive. I couldn't agree with you more.
In fact, in my experience of knowing everything I need to know about a game, I find the score does an excellent job of letting me know what's what! It has an uncanny reliability of predicting what the end result will be, too.
I'd also add, the league table is also a pretty reliable indicator of how well you're doing and the form guide over the last six games at how well you might do in your next game.
The rest: just unnecessary noise. [Post edited 15 Aug 7:23]
|  |
| |
| Analytics on 07:49 - Aug 15 with 451 views | BrizR |
| Analytics on 01:36 - Aug 15 by NW5Hoop |
I don't think that can be right. Take Caicedo. Allow six months for scouting and all that. Then he was at Brighton for two and a half years. The notion of targeting specfic players to sell to specific clubs would then depend on Brighton at some point in 2021 knowing exactly what kind of player Chelsea would need in 2024.
Even Chelsea didn't know what kind of player they would need in 2024. In 2024. |
It's explicitly part of the strategy. Nourry has mentioned it a few times. It's about identifying which positions are currently underpriced in the market and picking those players up now expecting the value to increase.
As a simplistic example, imagine all the top teams have 29 year old defensive midfielders. Right now that type of player is probably cheaper than normal, because none of the big teams need to buy one, but in 2 years' time they'll all need to replace them. So you buy a 25 year old now who fits that position and then he ought to jump in value massively.
It's a lot more complicated than that; they'll also be trying to predict trends, like everyone wanting a keeper who can play out from the back, and they'll be looking at much more detailed profiles than just vague positional ones, but you get the idea. [Post edited 15 Aug 7:50]
| | | |
| Analytics on 12:01 - Aug 15 with 304 views | Trom |
Whilst data is important you can’t dismiss the human element.
You have to identify players and then convince them that the opportunity and salary makes you viable. Given we can’t really compete on salary the career enhancement has to be the key to get deals over the line.
So you both need to identify players with the right characteristics and then offer them a pathway to advance their career.
So structure of the club, manager, past success and notable coaches will all be important. Identifying is one thing and attracting is another. Some cautious signs of improvement from the club given our summer signings. As ever we will see as the season plays out. | | | |
| Analytics on 12:17 - Aug 15 with 274 views | KensalT |
| Analytics on 17:13 - Aug 14 by NewYorkRanger |
I find this to be a bit of dilemma tbh, not least cos most of my career has been in data, analytics etc and the old adage of more data = better decisions is generally true BUT - it does depend on who is making the decisions with said data and how they are using it.
We are all familiar with Moneyball and how the gave the unfashionable Oakland A's a big advantage but these days I think it's all gone a bit far. If you look at the half time stats they put up on a game shown on Sky its largely meaningless.
All this XG, passes completed nonsense is not for me. And if clubs are buying players based on that and not seeing what they can do in the flesh then they deserve all they get |
Pretty much agree with all of that.
Although I would say that better data is more important than more data. | | | |
| Analytics on 12:23 - Aug 15 with 259 views | KensalT |
| Analytics on 18:00 - Aug 14 by turnsey |
I'm a data analyst that doesn't work in sport, so treat this with a grain of salt. My instinct is that the quality of model/strategy matters more than the quality of data or analytics personnel at the high professional level.
If you're looking at metrics, the players at the top of your list will vary heavily based on what metrics are most valued, how they are weighed relatively to each other, and how you weigh the context of what the player is doing (e.g. how good is the league? how much do you care about the relative quality of teammates/opposition? etc.)
I would be willing to bet that for us, it's Nourry or a collective group of overseers who establishes the "model"--the metrics/variables/profiles being targeted--and then it's up to subordinates to carry out the grunt work and then present reports to the real decision-makers based on different model variations, different player pools by nation, etc. That's why I say it's more the leadership direction than the quality of analyst/data that really makes a difference.
Old-school scouting must still have a place, though. I'd guess that the data serves as an initial screening process, and then players at the top of lists are then scouted in a more traditional sense? If anyone has genuine knowledge of analysis/scouting these days at this level, I'd be really interested to know. |
Sorry mate but I can't agree with your first paragraph.
If you're building your models on data then you have to get the data right first, otherwise everything that follows is going to be iffy.
It's like the foundations of a house. If you don't get the foundations right then it doesn't matter how pretty or expensive your house is, it's going to start subsiding sooner or later.
Totally agree with your last paragraph. Old-school scouting and knowledge of the game are still essential for me. | | | |
| |
By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions.
