Statsbomb and other Data Analytic Services

[Martin_T]

I recently listened to a Podcast that involved a rep from StatsBomb, a football data analytics company who specialise in various metrics. One of the services they offer, is using data to help identify Head Coaches and Managers who are perhaps undervalued relative to their reputation or results and clubs can use this data to aid their recruitment process.

 

There are obviously no guarantees in any recruitment process, but this may be a way that the club could identify and evaluate a candidate for the role, who may not be an obvious choice, but in a MoneyBall sense, could be a good candidate to improve the performance of the team.

 

This Blog  for example on the German centre back options for the European Championships, shows the level of detail in the analysis that they do.

 

https://statsbomb.com/

Edited November 12, 2019 by Martin_T

[redjambo]

I may be becoming an old fuddy duddy, but I yearn for the days when you just felt/knew which players would be good or not - you watched them train and play and understood who would benefit the team (although perhaps with some good hard work to get there first). I think we have an unhealthy infatuation with statistics nowadays when it comes to sport, and these statistics can easily mask all the complex ways in which a player can contribute to the dynamics of a football game. Naismith's contribution on Saturday as one small example - you can't put a statistic on someone who encourages their team members and gets folk up, running and motivated.

Edited November 12, 2019 by redjambo

[jambostuart]

4 minutes ago, redjambo said:

I may be becoming an old fuddy duddy, but I yearn for the days when you just felt/knew which players would be good or not - you watched them train and play and understood who would benefit the team (although perhaps with some good hard work to get there first). I think we have an unhealthy infatuation with statistics nowadays when it comes to sport, and these statistics can easily mask all the complex ways in which a player can contribute to the dynamics of a football game. Naismith's contribution on Saturday as one small example - you can't put a statistic on someone who encourages their team members and gets folk up, running and motivated.

 

 

Was going to post something similar. Stats certainly have a place in the game but in context. Running the most yards is useless if it's headless chicken stuff which no direct contribution.

 

Accurate pass ratio is great but over what difference? Are they all sideways passes with no intention to open up a defence?

 

And as you mention, the human element can't be quantified by stats. Naismith's most telling stats is games won when on the pitch and games lost without him, but this stat on its own doesn't show the real reason, which is the impact he has on those around him as well as his overall goalscoring contribution.

 

We've been very guilty recently I feel of analysing a little too much (Cathro's laptop, and Macphee/Levein playing to stop the opposition instead of setting up our own game). I think we need to be looking at someone who is less analytical and has the unmeasurable factor of encouraging players or sticking a rocket up their arse when needed. Basically a manager who can do what Locke did for derbies but in every match.

[Martin_T]

 

24 minutes ago, jambostuart said:

 

 

Was going to post something similar. Stats certainly have a place in the game but in context. Running the most yards is useless if it's headless chicken stuff which no direct contribution.

 

Accurate pass ratio is great but over what difference? Are they all sideways passes with no intention to open up a defence?

 

And as you mention, the human element can't be quantified by stats. Naismith's most telling stats is games won when on the pitch and games lost without him, but this stat on its own doesn't show the real reason, which is the impact he has on those around him as well as his overall goalscoring contribution.

 

We've been very guilty recently I feel of analysing a little too much (Cathro's laptop, and Macphee/Levein playing to stop the opposition instead of setting up our own game). I think we need to be looking at someone who is less analytical and has the unmeasurable factor of encouraging players or sticking a rocket up their arse when needed. Basically a manager who can do what Locke did for derbies but in every match.

 

I appreciate what you are saying and indeed StatsBomb themselves acknowledge that the human factor often can't be measured in statistics. I still think that this would be a useful tool in addition to a conventional process that may help us identify someone whose ability has perhaps not been recognised and acknowledged.

[Geoff Kilpatrick]

As a statistician and data analyst, the "Moneyball" concept is interesting in sports that lend themselves to data points (and indeed Moneyball is often cited in Data Science courses) but to me football doesn't naturally engender data points that are useful for analysis. Time of possession, for example, is a classic. If you do nothing meaningful with it it is a pointless stat. Similarly, shots on target. If they are 25 yard pings straight at the keeper, they count but they aren't the same as cutting through and just missing the target. The latest daft stat is "expected goals" which is worthy of a facepalm in itself.

 

What's also forgotten about Moneyball is that the Oakland A's didn't actually win anything other than their division. Billy Bean went to the Redsox, got more money, and won!

 

So to me, there is nothing better than the eyeball test and the player's character. Indeed, in a full on team sport like football, the latter is hugely important. For every Meshino you need a Michael Smith. I'm willing to bet that his stats look very different now that he's played more at centre back than full back but he hasn't actually changed as a player.

[Garrincha]

The current mania for “Working the keeper” with shots on target is the one that irritates me. Surely far better to place an effort out of the keepers reach and be a few inches wide rather than hit it straight at him and claim +1 for the shots on target stat.

 

Also the creative player that tries a through ball is regarded as statistically worse than the player that does nothing other than keep possession with 5 yard sideways passes. 

[Martin_T]

19 minutes ago, Geoff Kilpatrick said:

As a statistician and data analyst, the "Moneyball" concept is interesting in sports that lend themselves to data points (and indeed Moneyball is often cited in Data Science courses) but to me football doesn't naturally engender data points that are useful for analysis. Time of possession, for example, is a classic. If you do nothing meaningful with it it is a pointless stat. Similarly, shots on target. If they are 25 yard pings straight at the keeper, they count but they aren't the same as cutting through and just missing the target. The latest daft stat is "expected goals" which is worthy of a facepalm in itself.

 

What's also forgotten about Moneyball is that the Oakland A's didn't actually win anything other than their division. Billy Bean went to the Redsox, got more money, and won!

 

So to me, there is nothing better than the eyeball test and the player's character. Indeed, in a full on team sport like football, the latter is hugely important. For every Meshino you need a Michael Smith. I'm willing to bet that his stats look very different now that he's played more at centre back than full back but he hasn't actually changed as a player.

 

Billy Beane turned down the Red Sox and still works for the A's to this day. It was his method that the Red Sox adopted to then go on to win the World Series.

[Geoff Kilpatrick]

5 minutes ago, Martin_T said:

 

Billy Beane turned down the Red Sox and still works for the A's to this day. It was his method that the Red Sox adopted to then go on to win the World Series.

 

Apologies. You are correct on Beane (who has still won nothing). I meant the claim that the RedSox adopted Moneyball. I think the evidence they did is a little spurious.

[Toxteth O'Grady]

Apologies for Quoting an ex Aberdeen Manager but “Statistics are like miniskirts: They give you good ideas but hide the important things.”

Edited November 12, 2019 by Toxteth O'Grady

[graygo]

2 hours ago, Garrincha said:

The current mania for “Working the keeper” with shots on target is the one that irritates me. Surely far better to place an effort out of the keepers reach and be a few inches wide rather than hit it straight at him and claim +1 for the shots on target stat.

 

Also the creative player that tries a through ball is regarded as statistically worse than the player that does nothing other than keep possession with 5 yard sideways passes. 

 

Not always the case, Templeton at Anfield for instance.

[SpruceBringsteen]

I watch as much hockey as I do football (probably more tbh) and analytics has become HUGE in the sport over the last 4/5 years. There's some folk live and die by them, there's as many others think it's the devil's work and that you should rely on scouts.

 

I think there's room for both tbh.

[spacerjoe]

2 hours ago, Geoff Kilpatrick said:

 

Apologies. You are correct on Beane (who has still won nothing). I meant the claim that the RedSox adopted Moneyball. I think the evidence they did is a little spurious.

Didnt they hire the guy that wrote the book on it? Somebody James or whatever.

But for sure, it's easier with baseball as roles are much stricter. Stats mean what they mean.

 

[OTT]

3 hours ago, jambostuart said:

 

 

Was going to post something similar. Stats certainly have a place in the game but in context. Running the most yards is useless if it's headless chicken stuff which no direct contribution.

 

Accurate pass ratio is great but over what difference? Are they all sideways passes with no intention to open up a defence?

 

And as you mention, the human element can't be quantified by stats. Naismith's most telling stats is games won when on the pitch and games lost without him, but this stat on its own doesn't show the real reason, which is the impact he has on those around him as well as his overall goalscoring contribution.

 

We've been very guilty recently I feel of analysing a little too much (Cathro's laptop, and Macphee/Levein playing to stop the opposition instead of setting up our own game). I think we need to be looking at someone who is less analytical and has the unmeasurable factor of encouraging players or sticking a rocket up their arse when needed. Basically a manager who can do what Locke did for derbies but in every match.

 

Yeah, I think stats used in isolation is stupid. Clare -for example, runs the furthest most games yet is woefully ineffective. Its an utter non-entity of a stat. I dare say (and I hope i'm not doing him a disservice) that Glen Whelan probably does less running than most of our team but is so effective in his positioning and vision that he doesn't need to run about. 

 

I think stats can tell you a lot about the player in terms of how they play but shouldn't be used as some sort of budget scouting processing where you don't actually go and watch the player. 

[jambostuart]

16 minutes ago, OTT said:

 

Yeah, I think stats used in isolation is stupid. Clare -for example, runs the furthest most games yet is woefully ineffective. Its an utter non-entity of a stat. I dare say (and I hope i'm not doing him a disservice) that Glen Whelan probably does less running than most of our team but is so effective in his positioning and vision that he doesn't need to run about. 

 

I think stats can tell you a lot about the player in terms of how they play but shouldn't be used as some sort of budget scouting processing where you don't actually go and watch the player. 

 

Exactly this. Hearts have definitely been guilty of over analysing and not utilising stats and players properly. I would hope we don't do that with our manager. Clare is the best example, as you say, he runs a lot but he's a headless chicken. I really hope our next manager can watch a game with his own two eyes and set a team up to our own strengths based on what he can see and not who runs fastest or covers most ground...

[i8hibsh]

It is all interesting but I just don't think it makes the game any better, it just inreases cash values. I also hate the sports science stuff. A player straps a device on his chest and back to make him look like the hunchback of Notre Dam only to tell some computer what 60,000 fans in a stadium witnessed anyway. He played shite, never made many tackles and is a lazy *******.

[Tommy Brown]

10 hours ago, Geoff Kilpatrick said:

 The latest daft stat is "expected goals" which is worthy of a facepalm in itself.

 

 

I see this at the emd of every game on MOTD.

 

Anybody undderstand it

 

[Martin_T]

17 hours ago, Garrincha said:

The current mania for “Working the keeper” with shots on target is the one that irritates me. Surely far better to place an effort out of the keepers reach and be a few inches wide rather than hit it straight at him and claim +1 for the shots on target stat.

 

Also the creative player that tries a through ball is regarded as statistically worse than the player that does nothing other than keep possession with 5 yard sideways passes. 

 

Your point about creative players I would say is inaccurate. Different positions on the pitch are gauged by different measures. For example for defender you are looking at headers won, duels, tackles made, bookings etc. A creative midfielder would be measured by assists, goals, dribbles, time in possession etc.

 

If anyone is interested in this topic, the video below goes into it in greater detail.