| I still don't understand XGf so could someone explain.. 11:04 - Feb 4 with 2982 views | norfsufblue |
How many should we have scored from 63 shots in the last 2 games please? |  | | |  |
| I still don't understand XGf so could someone explain.. on 11:05 - Feb 4 with 2966 views | Simonds92 |
If we're as good as Maidstone, 63. |  | | |
| I still don't understand XGf so could someone explain.. on 11:18 - Feb 4 with 2915 views | DavoIPB |
Very few. We had very few clear cut chances | | | |
| I still don't understand XGf so could someone explain.. on 13:45 - Feb 4 with 2783 views | Sawtrich |
There are different xG figures on different sites from different ways of calculating it. Yesterday footstats gave us 2.43 and in the flashscore stats it's 2.82.
Clearly you can't score 0.43 of a goal and in scoring 2 we slightly underperformed in terms of putting away the chances we created. |  | | |
| I still don't understand XGf so could someone explain.. on 13:46 - Feb 4 with 2775 views | Sawtrich |
| I still don't understand XGf so could someone explain.. on 13:45 - Feb 4 by Sawtrich |
There are different xG figures on different sites from different ways of calculating it. Yesterday footstats gave us 2.43 and in the flashscore stats it's 2.82.
Clearly you can't score 0.43 of a goal and in scoring 2 we slightly underperformed in terms of putting away the chances we created. |
*footystats | | | |
| I still don't understand XGf so could someone explain.. on 13:48 - Feb 4 with 2767 views | Trequartista |
Its a simple formula, Goals = Expected Goals + Unexpected Goals. |  |
| |
| I still don't understand XGf so could someone explain.. on 13:54 - Feb 4 with 2737 views | Sawtrich |
The xG against Maidstone was between 3.3 and 5.
So to answer the OP we would have scored 6 or 7 goals with average finishing over the two games. | | | |
| I still don't understand XGf so could someone explain.. on 14:12 - Feb 4 with 2696 views | norfsufblue |
| I still don't understand XGf so could someone explain.. on 13:54 - Feb 4 by Sawtrich |
The xG against Maidstone was between 3.3 and 5.
So to answer the OP we would have scored 6 or 7 goals with average finishing over the two games. |
So we are 3 or 4 short over the two games.... 4th round and a point possibly 3.....I think the main takeaway is....
I hate losing! | | | |
| I still don't understand XGf so could someone explain.. on 15:03 - Feb 4 with 2610 views | NthQldITFC |
| I still don't understand XGf so could someone explain.. on 13:48 - Feb 4 by Trequartista |
Its a simple formula, Goals = Expected Goals + Unexpected Goals. |
Goals = Expected Goals + Unexpected Goals - Unexpected Misses + A'hole Officiating |  |
| # WE ARE STEALING THE FUTURE FROM OUR CHILDREN --- WE MUST CHANGE COURSE # | | Poll: | It's driving me nuts |
| |
Login to get fewer ads
| I still don't understand XGf so could someone explain.. on 15:47 - Feb 4 with 2531 views | DJR |
I think it's a bit like those calculators you use for online banking.
You enter the number of goals scored, and it comes up with a random number generally not more than one higher, or one lower, than the number of goals scored. But the parameters are slightly higher when you play a non-league team. [Post edited 4 Feb 15:51]
|  | | |
| I still don't understand XGf so could someone explain.. on 15:53 - Feb 4 with 2494 views | NthQldITFC |
| I still don't understand XGf so could someone explain.. on 15:47 - Feb 4 by DJR |
I think it's a bit like those calculators you use for online banking.
You enter the number of goals scored, and it comes up with a random number generally not more than one higher, or one lower, than the number of goals scored. But the parameters are slightly higher when you play a non-league team. [Post edited 4 Feb 15:51]
|
I didn't read your response properly and thought that you meant the calculators that bankers and economists use, which also seem to throw out random numbers. | |
| # WE ARE STEALING THE FUTURE FROM OUR CHILDREN --- WE MUST CHANGE COURSE # | | Poll: | It's driving me nuts |
| |
| I still don't understand XGf so could someone explain.. on 15:56 - Feb 4 with 2480 views | monty_radio |
| I still don't understand XGf so could someone explain.. on 13:48 - Feb 4 by Trequartista |
Its a simple formula, Goals = Expected Goals + Unexpected Goals. |
For the miserable bloke who sits near me, all Town's goals are unexpected | |
| |
| I still don't understand XGf so could someone explain.. on 16:12 - Feb 4 with 2428 views | DJR |
I wonder whether managers and coaches use these figures, as opposed to their own eyes in judging chances based on the ease of scoring and the potency of the person with the chance.
One weakness it strikes me that we have (at least pre-Moore) is scoring from headers, or even getting headers on target. But XGf wouldn't appear to identify that.
Instead, it is something that I imagine the coaching staff will be aware of, and trying to improve, although some people would just seem better at scoring from headers than others (for example, some of the central defenders we had under McCarthy). And as Chaplin showed recently, it isn't necessarily a height thing. [Post edited 4 Feb 16:16]
| | | |
| I still don't understand XGf so could someone explain.. on 03:36 - Feb 5 with 2156 views | HalifaxBlue |
| I still don't understand XGf so could someone explain.. on 16:12 - Feb 4 by DJR |
I wonder whether managers and coaches use these figures, as opposed to their own eyes in judging chances based on the ease of scoring and the potency of the person with the chance.
One weakness it strikes me that we have (at least pre-Moore) is scoring from headers, or even getting headers on target. But XGf wouldn't appear to identify that.
Instead, it is something that I imagine the coaching staff will be aware of, and trying to improve, although some people would just seem better at scoring from headers than others (for example, some of the central defenders we had under McCarthy). And as Chaplin showed recently, it isn't necessarily a height thing. [Post edited 4 Feb 16:16]
|
It's definitely used by managers and coaches, I would say even the most traditionalist managers in the game have no choice in that regard. It's long been a staple at this point.
Any good xG model would account for headed goals - a headed attempt will almost always score lower than a shot with the foot from the same spot.
xG is very often misunderstood, and some forms are smarter than others. Like all data, it is of no use whatsoever when not paired with a keen eye for the game. I work part-time in data analysis for a professional club and would be happy to answer any questions. |  |
| |
| I still don't understand XGf so could someone explain.. on 08:04 - Feb 5 with 2020 views | DJR |
| I still don't understand XGf so could someone explain.. on 03:36 - Feb 5 by HalifaxBlue |
It's definitely used by managers and coaches, I would say even the most traditionalist managers in the game have no choice in that regard. It's long been a staple at this point.
Any good xG model would account for headed goals - a headed attempt will almost always score lower than a shot with the foot from the same spot.
xG is very often misunderstood, and some forms are smarter than others. Like all data, it is of no use whatsoever when not paired with a keen eye for the game. I work part-time in data analysis for a professional club and would be happy to answer any questions. |
Thanks for your response.
Do the clubs rely on the publicly available information on the sites mentioned in this thread, or is there a bespoke and more detailed version for them, which perhaps analyses every chance? | | | |
| I still don't understand XGf so could someone explain.. on 10:08 - Feb 5 with 1903 views | GeoffSentence |
| I still don't understand XGf so could someone explain.. on 13:48 - Feb 4 by Trequartista |
Its a simple formula, Goals = Expected Goals + Unexpected Goals. |
surely goals = expected goals + unexpected goals - unexpected misses |  |
| |
| I still don't understand XGf so could someone explain.. on 13:55 - Feb 5 with 1718 views | HalifaxBlue |
| I still don't understand XGf so could someone explain.. on 08:04 - Feb 5 by DJR |
Thanks for your response.
Do the clubs rely on the publicly available information on the sites mentioned in this thread, or is there a bespoke and more detailed version for them, which perhaps analyses every chance? |
It depends on the size of the club, must most use models from data providers like Opta or Statsbomb. Clubs like Liverpool however are known for their massive data departments and have probably had their data scientists develop a bespoke model, yes.
Depending on the site, the publicly available models generally tend to be quite robust as they are also provided by companies like Opta. These sites would subscribe to an API just like clubs would.
Is analyzing every chance even without a shot at the end what you mean? There are similar metrics to xG that do things like this. Expected Threat (xT) is an example which attempts to measure the degree to which each pass makes a goal more or less likely. It's less common in the public sphere but many clubs use this or similar metrics. Packing is another example which attempts to measure how many opponents each pass "eliminates" from that phase of play (ie. a pass from a CB to a 6 that splits two strikers might score 2 as a very basic example)
It's a real rabbit hole. Again - all useless if not backed up by a keen eye for the game that can contextualize the data. [Post edited 5 Feb 13:57]
| |
| |
| I still don't understand XGf so could someone explain.. on 14:27 - Feb 5 with 1640 views | SmithersJones |
| I still don't understand XGf so could someone explain.. on 03:36 - Feb 5 by HalifaxBlue |
It's definitely used by managers and coaches, I would say even the most traditionalist managers in the game have no choice in that regard. It's long been a staple at this point.
Any good xG model would account for headed goals - a headed attempt will almost always score lower than a shot with the foot from the same spot.
xG is very often misunderstood, and some forms are smarter than others. Like all data, it is of no use whatsoever when not paired with a keen eye for the game. I work part-time in data analysis for a professional club and would be happy to answer any questions. |
Am I right in thinking that every shot has an xG attached to it regardless of context? For example if I look on FotMob it looks like Al Hamedi’s shot at the end had an xG of 0.74 and Moore’s follow up (second goal) had an xG of 0.51. But of course if Al Hamedi’s shot had gone in there would have been no follow up. So our xG of 2.6 for the game is somewhat overstated. | | | |
| I still don't understand XGf so could someone explain.. on 14:51 - Feb 5 with 1603 views | Brandons_Socks |
Slight aside but are there any stats/records on the number of blocked shots per game, per team?
I might well be biased as I pretty much only watch ITFC games these days, but we do seem to have an incredible number of blocked shots, which I largely put down to the style of our build up play. It would be good to back up these observations with data though. | | | |
| I still don't understand XGf so could someone explain.. on 15:15 - Feb 5 with 1547 views | DJR |
| I still don't understand XGf so could someone explain.. on 13:55 - Feb 5 by HalifaxBlue |
It depends on the size of the club, must most use models from data providers like Opta or Statsbomb. Clubs like Liverpool however are known for their massive data departments and have probably had their data scientists develop a bespoke model, yes.
Depending on the site, the publicly available models generally tend to be quite robust as they are also provided by companies like Opta. These sites would subscribe to an API just like clubs would.
Is analyzing every chance even without a shot at the end what you mean? There are similar metrics to xG that do things like this. Expected Threat (xT) is an example which attempts to measure the degree to which each pass makes a goal more or less likely. It's less common in the public sphere but many clubs use this or similar metrics. Packing is another example which attempts to measure how many opponents each pass "eliminates" from that phase of play (ie. a pass from a CB to a 6 that splits two strikers might score 2 as a very basic example)
It's a real rabbit hole. Again - all useless if not backed up by a keen eye for the game that can contextualize the data. [Post edited 5 Feb 13:57]
|
I probably was meaning both in the sense that, say, a blocked shot might involve other considerations such as whether the pass or cross was in the right place or the attacker took up the right position.
All very interesting, and thanks for explaining. [Post edited 5 Feb 15:16]
| | | |
| I still don't understand XGf so could someone explain.. on 15:28 - Feb 5 with 1512 views | hoppy |
| I still don't understand XGf so could someone explain.. on 15:15 - Feb 5 by DJR |
I probably was meaning both in the sense that, say, a blocked shot might involve other considerations such as whether the pass or cross was in the right place or the attacker took up the right position.
All very interesting, and thanks for explaining. [Post edited 5 Feb 15:16]
|
It also confuses me, as I've never really understood it either. What would the 3rd Preston goal go down as, for example, where Hladky made a very uncharacteristic error which presented an unexpected opportunity to Preston? |  |
| |
| I still don't understand XGf so could someone explain.. on 16:02 - Feb 5 with 1419 views | DJR |
Talk of data brings to mind Don Revie's dossiers, and unless my mind is playing a trick, I remember a co-commentator joking, when England weren't doing well in a particular game and there were bits of paper floating about, that Revie had torn up his dossiers. [Post edited 5 Feb 16:53]
| | | |
| I still don't understand XGf so could someone explain.. on 19:15 - Feb 5 with 1265 views | HalifaxBlue |
| I still don't understand XGf so could someone explain.. on 14:27 - Feb 5 by SmithersJones |
Am I right in thinking that every shot has an xG attached to it regardless of context? For example if I look on FotMob it looks like Al Hamedi’s shot at the end had an xG of 0.74 and Moore’s follow up (second goal) had an xG of 0.51. But of course if Al Hamedi’s shot had gone in there would have been no follow up. So our xG of 2.6 for the game is somewhat overstated. |
Spot on - you point out a very common misconception of xG. There are models that would attribute both instances to the players individual tallies, but only the higher one to the team total, but I'm not sure how Fotmob functions.
I wouldn't say it is overstated, I'd just say it is misunderstood. You can't really directly correlate xG to goals. If you think about it, a chance with an xG of .51 is a chance that should be scored more often than not by a good finisher. So an xG total of 2 should actually result in more than two goals if finishing is up to snuff. This is all theoretical of course, and not really how it always works out in reality. | |
| |
| I still don't understand XGf so could someone explain.. on 19:19 - Feb 5 with 1249 views | HalifaxBlue |
| I still don't understand XGf so could someone explain.. on 15:28 - Feb 5 by hoppy |
It also confuses me, as I've never really understood it either. What would the 3rd Preston goal go down as, for example, where Hladky made a very uncharacteristic error which presented an unexpected opportunity to Preston? |
It really depends on how sophisticated the model you are looking at is. I just checked Fotmob and it had the 3rd goal with an xG of .17. I would have expected it to be higher. I don't think many models would account for the fact that the ball has just been given away, but good models would take into account the positioning of our defenders and goalkeeper at the moment the shot is taken. | |
| |
| I still don't understand XGf so could someone explain.. on 19:23 - Feb 5 with 1239 views | HalifaxBlue |
| I still don't understand XGf so could someone explain.. on 15:15 - Feb 5 by DJR |
I probably was meaning both in the sense that, say, a blocked shot might involve other considerations such as whether the pass or cross was in the right place or the attacker took up the right position.
All very interesting, and thanks for explaining. [Post edited 5 Feb 15:16]
|
You have the right idea. Blocked shots are accounted for in xG models though. xG does not account for anything that happens after the shot is taken - at it's most basic, it is a pure measure of how likely a shot from a specific location is to result in a goal, based on a massive database of similar shots in past matches. xG does not care if the shot is blocked, saved, scored, or missed.
There is a 2nd metric, post-shot xG (or xGOT on fotmob) that does account for what happens after the shot. In this metric, any shot that is blocked/missed would register 0. | |
| |
| I still don't understand XGf so could someone explain.. on 19:24 - Feb 5 with 1233 views | HalifaxBlue |
| I still don't understand XGf so could someone explain.. on 14:51 - Feb 5 by Brandons_Socks |
Slight aside but are there any stats/records on the number of blocked shots per game, per team?
I might well be biased as I pretty much only watch ITFC games these days, but we do seem to have an incredible number of blocked shots, which I largely put down to the style of our build up play. It would be good to back up these observations with data though. |
FBRef is an excellent free resource for stats and would almost definitely track this data. Check here:
https://fbref.com/en/comps/10/Championship-Stats | |
| |
| |
By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions.
