|
Post by tachyon on Dec 13, 2019 20:00:33 GMT
Sure. He collects his own data in house & uses it at Brentford. But the basic premise of xG is the same. IIRC Tippet's recent book explicitly states that there's no skill differential in conversion rates. I do see that as the major flaw though mate. I think the differentials are poorly calibrated. I genuinely know that traders at other firms are rowing away a little bit but that's not saying that the system itself can't evolve. It comes up with too many absurdities in my eyes. Work in the same industry. Models are continually calibrated against out of sample data. It certainly works well in longer term markets. Sell of Reading after the year they nearly got promoted, Atalanta for a top four finish last season, Watford for the drop this year, Avoid Stoke when favs for promotion. No value in tomorrow's game, though for any result :-) It's a tool, but a very useful one.
|
|
|
Post by Little Gary Patel on Dec 14, 2019 0:40:30 GMT
Just got in from Regimes nightclub. I doused myself in Kouros head to toe before I went out and hence my expected snogs (Xs) quota was quite high I stood by the bar in Phase2 all night trying to touch or snog girls but sadly got a Xs quota of zero. This does not mean I will get off with Betty Boo next week, and then Sherilyn Fenn the week after - it simply means I am an ugly annoying little bastard and need to get used to it 1/10 tried to be funny but completely missed the point and the concept. like most on here.
|
|
|
Post by FullerMagic on Dec 26, 2019 17:56:19 GMT
|
|
|
Post by Deleted on Dec 26, 2019 18:19:38 GMT
It’s all pointless drivel. The only stat that ever matters in football is the full time score.
|
|
|
Post by tachyon on Dec 26, 2019 19:05:44 GMT
It’s all pointless drivel. The only stat that ever matters in football is the full time score. Cool, where did the "only stat that matters" predict we'd be sitting in the table on Boxing Day back in September?
|
|
|
Post by Deleted on Dec 26, 2019 19:48:58 GMT
It’s all pointless drivel. The only stat that ever matters in football is the full time score. Cool, where did the "only stat that matters" predict we'd be sitting in the table on Boxing Day back in September? Yeah ok mate, you sit and play with your stats if it makes you feel better. Me, I’ll just watch real football, see what the final score is and then look at the league table.
|
|
|
Post by tachyon on Dec 26, 2019 19:58:21 GMT
Cool, where did the "only stat that matters" predict we'd be sitting in the table on Boxing Day back in September? Yeah ok mate, you sit and play with your stats if it makes you feel better. Me, I’ll just watch real football, see what the final score is and then look at the league table. Simply curious as to why you think the final score is all that matters and why the aversion to an alternative, wider ranging approach :-)
|
|
|
Post by terryconroysmagic on Dec 26, 2019 21:35:47 GMT
Yeah ok mate, you sit and play with your stats if it makes you feel better. Me, I’ll just watch real football, see what the final score is and then look at the league table. Simply curious as to why you think the final score is all that matters and why the aversion to an alternative, wider ranging approach :-) I enjoy this statistical analysis. Have you update the probability of relegation after today’s win?
|
|
|
Post by FullerMagic on Dec 26, 2019 21:39:40 GMT
Unfathomably, this XG table had us sitting 5th going into today's game
|
|
|
Post by Deleted on Dec 26, 2019 22:23:16 GMT
Yeah ok mate, you sit and play with your stats if it makes you feel better. Me, I’ll just watch real football, see what the final score is and then look at the league table. Simply curious as to why you think the final score is all that matters and why the aversion to an alternative, wider ranging approach :-) Is there anything else? Points for good behaviour, or not snotting on the pitch?
|
|
|
Post by apb1 on Dec 26, 2019 22:53:11 GMT
|
|
|
Post by kjpt140v on Dec 26, 2019 23:11:16 GMT
The biggest value expected goals or xG brings to the table is in analysing longer term trends. If you're continually creating more chances than the opposition, even if you're not always converting those chances, it's very likely that over time your outcomes (results) will adhere more closely to your process (xG differential). After six games Stoke had one whole point from a possible 18 (results dire), but a fairly equal share of the xG (process not great, but not terrible) in their matches to go with a similar trend from the previous season. At that season low in September, Infogol simulated the remainder of the season using the xG ratings of all 24 teams to see when it would be more likely than not that Stoke would claw their way out of the bottom three and posted it on Twitter. View AttachmentWe estimated Stoke would start to be slightly more likely than not to escape from the drop zone by game week 22. That's next week and with favourable results we could hit the scheduled recovery at the earliest predicted time.....or we could take until week 30 or we might not manage it at all. That's what projections are, they come with an uncertainty, grouped around the most likely outcome. XG from the start of September has done a pretty good job of projecting where Stoke would be half way through December and currently it thinks our median final points total will be 54, our final median finishing position will be 18th and we currently have a 21% chance of relegation.
|
|
|
Post by kjpt140v on Dec 26, 2019 23:11:28 GMT
The biggest value expected goals or xG brings to the table is in analysing longer term trends. If you're continually creating more chances than the opposition, even if you're not always converting those chances, it's very likely that over time your outcomes (results) will adhere more closely to your process (xG differential). After six games Stoke had one whole point from a possible 18 (results dire), but a fairly equal share of the xG (process not great, but not terrible) in their matches to go with a similar trend from the previous season. At that season low in September, Infogol simulated the remainder of the season using the xG ratings of all 24 teams to see when it would be more likely than not that Stoke would claw their way out of the bottom three and posted it on Twitter. View AttachmentWe estimated Stoke would start to be slightly more likely than not to escape from the drop zone by game week 22. That's next week and with favourable results we could hit the scheduled recovery at the earliest predicted time.....or we could take until week 30 or we might not manage it at all. That's what projections are, they come with an uncertainty, grouped around the most likely outcome. XG from the start of September has done a pretty good job of projecting where Stoke would be half way through December and currently it thinks our median final points total will be 54, our final median finishing position will be 18th and we currently have a 21% chance of relegation.
|
|
|
Post by tachyon on Dec 27, 2019 7:58:53 GMT
Simply curious as to why you think the final score is all that matters and why the aversion to an alternative, wider ranging approach :-) I enjoy this statistical analysis. Have you update the probability of relegation after today’s win? cheers, around 20% chance we go down. Median final points projection is back up to 54, median final position 19th
|
|
|
Post by tachyon on Dec 27, 2019 8:15:09 GMT
Simply curious as to why you think the final score is all that matters and why the aversion to an alternative, wider ranging approach :-) Is there anything else? Points for good behaviour, or not snotting on the pitch? No you get nowt extra, but the goals you score and allow tend to come from the chances you create and allow. That's where xG takes a deeper dive into a side's performance levels instead of bouncing around like demented tiggers every time Stoke win or lose. "Get out of my club" / "A corner's been turned" I do however add 0.01 of an xG to a team's xG rating for every incident of ear cupping.
|
|
|
Post by rawli on Dec 27, 2019 8:38:50 GMT
Is there anything else? Points for good behaviour, or not snotting on the pitch? No you get nowt extra, but the goals you score and allow tend to come from the chances you create and allow. That's where xG takes a deeper dive into a side's performance levels instead of bouncing around like demented tiggers every time Stoke win or lose. "Get out of my club" / "A corner's been turned" I do however add 0.01 of an xG to a team's xG rating for every incident of ear cupping. I find that a glass of water gets rid of ear cups
|
|
|
Post by FullerMagic on Dec 27, 2019 10:56:12 GMT
|
|
|
Post by bayernoatcake on Dec 27, 2019 10:59:22 GMT
There’s no way we should be that high. We’re lucky to be 4th bottom.
|
|
|
Post by shipshape on Dec 27, 2019 12:01:36 GMT
No predictive model on the planet will ever be perfect. However, if they are good they should get it right within a given tolerance level most of the time. They are used all the time. Weather forecasting, ads you see on Facebook, exit polls etc. The science behind them is getting more advanced all the time, with machine learning taking more of a front seat all the time. Sometimes though, they will get it wrong. I can't have that we should be fourth either, it is surely an outlier. Our bookies odds for relegation have never dropped below around 3-1, which I found amazing, but maybe they are using these stats in their odds compiling at the moment? As someone who works in data, I do find these numbers encouraging, if somewhat contradictory to what my eyes tell me.
|
|
|
Post by Vadiation_Ribe on Dec 27, 2019 12:02:14 GMT
I don't find it that mystifying given the number of goals conceded due to ridiculous mistakes or unluckiness. I wouldn't have thought playoff-high though, more mid-table.
|
|
|
Post by lordb on Dec 27, 2019 12:42:42 GMT
I don't find it that mystifying given the number of goals conceded due to ridiculous mistakes or unluckiness. I wouldn't have thought playoff-high though, more mid-table. Correct We haven't been hammered, time and time again the opposition have scored with their first or second chance and more often than not as a result of an error. The good news in that front is Ryan is back.
|
|
|
Post by FullerMagic on Jan 1, 2020 22:21:24 GMT
|
|
|
Post by FullerMagic on Jan 1, 2020 22:27:00 GMT
|
|
|
Post by Pugsley on Jan 1, 2020 22:32:14 GMT
March was right, goals are irrelevant. Expected goals on the other hand...
|
|
|
Post by mrcoke on Jan 2, 2020 0:50:30 GMT
I don't know about expected goals, but we watch football to see actual goals and today's 5 take us to 33.
At the actual rate so far this season, we will score the most goals since the last promotion season, with the advantage of the longer season over the Prem.
Our actual goal score rate since MON arrived is on a par with a promotion position, over a whole season. Unfortunately it is negated by our actual goals conceded.
|
|
|
Post by bayernoatcake on Jan 2, 2020 0:51:38 GMT
Shows what a load of old bollocks it is
|
|
|
Post by tachyon on Jan 2, 2020 8:46:58 GMT
Shows what a load of old bollocks it is No, not really. It shows what a powerful predictive and descriptive tool it is compared to mere outcome based conjecture. xG projected we'd be out of the bottom three in time for the New Year when we had just one point from 21. We were. It has been projecting we'd get a final points total in the low to mid 50's and we're on course for that as well. xG chances from yesterday gave us a 75% chance of winning the game and a 95% chance of taking at least a point. Can't really argue with that. A fifth of the time we'd beat Huddersfield by three or more goals having created and allowed those quality and quantity of chances. We created two big chances to huddersfield's one and that's another promising, process based performance indicator. A lot of low probability efforts are prone to a wider varience in outcome. Occasionally, lots will go in, like yesterday, more typically they won't, no matter how "clinical" we want to believe our players are.
|
|
|
Post by Little Gary Patel on Jan 2, 2020 13:18:00 GMT
Shows what a load of old bollocks it is Back to exposing yourself as an idiot in this thread now you've had to give up on Tom Ince. How's your stats pHD going out of interest? Every time tachyon comes on here and explains you refuse to listen and stick to your idiotic guns about it being "nonsense". You complain like fuck about our recruitment, yet admire Brentfords. Are they just plucking it out of thin air? Think they might have some predictors about future player performance or players who are undervalued? Nah, they're nonsense.
|
|
|
Post by bayernoatcake on Jan 2, 2020 13:22:39 GMT
Shows what a load of old bollocks it is Back to exposing yourself as an idiot in this thread now you've had to give up on Tom Ince. How's your stats pHD going out of interest? Every time tachyon comes on here and explains you refuse to listen and stick to your idiotic guns about it being "nonsense". You complain like fuck about our recruitment, yet admire Brentfords. Are they just plucking it out of thin air? Think they might have some predictors about future player performance or players who are undervalued? Nah, they're nonsense. Of course it’s bollocks. That was about 3-0 game yesterday or should have been. It just doesn’t relate to reality.
|
|
|
Post by bayernoatcake on Jan 2, 2020 13:23:57 GMT
Shows what a load of old bollocks it is No, not really. It shows what a powerful predictive and descriptive tool it is compared to mere outcome based conjecture. xG projected we'd be out of the bottom three in time for the New Year when we had just one point from 21. We were. It has been projecting we'd get a final points total in the low to mid 50's and we're on course for that as well. View AttachmentxG chances from yesterday gave us a 75% chance of winning the game and a 95% chance of taking at least a point. Can't really argue with that. A fifth of the time we'd beat Huddersfield by three or more goals having created and allowed those quality and quantity of chances. We created two big chances to huddersfield's one and that's another promising, process based performance indicator. A lot of low probability efforts are prone to a wider varience in outcome. Occasionally, lots will go in, like yesterday, more typically they won't, no matter how "clinical" we want to believe our players are. It was saying that when we had Jones in charge. That was never going to happen.
|
|