DJBOB Posted April 12, 2023 Share Posted April 12, 2023 9 hours ago, nick76 said: Completely pointless percentages…the data is so small and one result or two makes such a different that 8th position has changed by 27% in a month. This is exactly right - too few data points (8 games) and the difference between a win/draw/loss in points can skew the percentages wildly from week to week. If you re-ran the simulator with a win against Newcastle, the results would be wildly different. 1 Link to comment Share on other sites More sharing options...
allani Posted April 12, 2023 Share Posted April 12, 2023 4 hours ago, tomsky_11 said: Just noticed the order on the Brighton games above is slightly off and they'll have only played Chelsea away in the league in the time we play Newcastle, Brentford and Fulham. Their FA Cup semi comes while we are playing Brentford and then Forest away is a day after we play Fulham. Obviously this is going to add to their games in hand, but it does give us a big opportunity to really open the gap and put the pressure on them, especially if they also drop points at Chelsea. The semi between the Chelsea and Forest games hopefully draws some focus from those games as well. If we can go the next 3 unbeaten and get at least one win, we could be looking at a 6-8 point gap to Brighton, them needing to get something from a Forest side who have been much better at home than away to start clawing that back, with three games in hand that few would expect them to win any of. It also means that they will have at least 3 extra matches to fit into the last few weeks of the season so they are almost certainly going to be weekend - midweek - weekend - midweek - weekend. 5 matches in 14 or 15 days will be tough at this point of the season. 2 Link to comment Share on other sites More sharing options...
DJBOB Posted April 12, 2023 Share Posted April 12, 2023 (edited) 5 minutes ago, allani said: It also means that they will have at least 3 extra matches to fit into the last few weeks of the season so they are almost certainly going to be weekend - midweek - weekend - midweek - weekend. 5 matches in 14 or 15 days will be tough at this point of the season. Ye - and while it doesn't necessarily mean they will struggle - that it is a big mental task to try to play a match every 3 days. I don't even think De Zerbi and all the analysts at Brighton can keep up with every opponent tweak. Edited April 12, 2023 by DJBOB 2 Link to comment Share on other sites More sharing options...
Sulberto21 Posted April 12, 2023 Share Posted April 12, 2023 On 10/04/2023 at 13:17, Delphinho123 said: Let’s be honest, all of us on here are going to be a little disappointed if we don’t make Europe now. The upside of making Europe with THIS manager in charge could be a game changer both in terms of the players we can attract and the extra revenue it will bring in. Not only that, we have a serial winner of that competition in charge. It just feels right. But, we shouldn’t be disappointed. We could have been looking at a relegation scrap had Gerrard been here any longer. Like others have said, we get greedy, but it’s just human nature. I actually won’t see it as a failure because for me it’s about preparation for next season. I might have said it when Don Unai arrived that us finishing on 19 wins will be a challenging but realistic challenge and something that Don Unai can help us to achieve. We can become top 4 challengers under Don Unai. Link to comment Share on other sites More sharing options...
allani Posted April 12, 2023 Share Posted April 12, 2023 8 hours ago, tomsky_11 said: Newcastle and Man Utd too far ahead at this point, Fulham and Chelsea too far behind. Outside chance of us catching Spurs with their remaining fixtures looking tough and the management situation there. 3 wins probably enough for them to stay ahead. Brentford probably need 6 wins from 8 to have any chance of 7th. And/or two of us, Brighton and Liverpool tanking. Liverpool should be able to get at least 5 wins from that run. Means be probably need to at least equal it to stay ahead of them. Brighton have 3 should win home games. Maybe then only need the game against us. Looking really likely that game of the last day is going to be crucial. Yeah I almost excluded those 4 clubs from the breakdown but decided that as we still have to play 3 of them to leave them in. I think Arsenal, City, Utd and Newcastle will finish top 4. I still think that Spurs could collapse - so much is wrong with the club that it could easily impact the players. Without Son and Kane they really aren't that good. I don't think anyone has an "easy" run in - ours is probably tougher than most but we're in better form and playing with more confidence than most of the others. Newcastle is a massive game. Link to comment Share on other sites More sharing options...
allani Posted April 12, 2023 Share Posted April 12, 2023 11 minutes ago, DJBOB said: Ye - and while it doesn't necessarily mean they will struggle - that it is a big mental task to try to play a match every 3 days. I don't even think De Zerbi and all the analysts at Brighton can keep up with every opponent tweak. Mentally and physically it is a big ask. Link to comment Share on other sites More sharing options...
viivvaa66 Posted April 12, 2023 Share Posted April 12, 2023 11 hours ago, nick76 said: Completely pointless percentages…the data is so small and one result or two makes such a different that 8th position has changed by 27% in a month. The percentages are not pointless, it just gives the most likely outcomes of the league position at the end of the season. However, as you said, if a team is over performing compared to what the model predicted, that will change the percentages. Remember this is not the final answer, it is just a prediction, that changes as results get inputted in to the model. Let’s take the match against Newcastle as an example. If the model predicts that a Newcastle win is 55%, draw is 25%, and us winning is 20%, and if we actually win the match it will change our percentages because it was an unlikely outcome, but if we don’t win, the percentages would mostly stay the same. The last month we have taken 13 out of 15 points on offer, that is more than what most fans in here were predicting, and I assume way more than the model was predicting, meaning that the model will now predict us higher up the table, as you showed. I do believe we will win the remaining matches and qualify for Champions League, the model predict that the chance for that is 0.1%, but in my view it just show what a miracle Emery have performed when we actually do it. 1 Link to comment Share on other sites More sharing options...
allani Posted April 12, 2023 Share Posted April 12, 2023 12 minutes ago, viivvaa66 said: The percentages are not pointless, it just gives the most likely outcomes of the league position at the end of the season. However, as you said, if a team is over performing compared to what the model predicted, that will change the percentages. Remember this is not the final answer, it is just a prediction, that changes as results get inputted in to the model. Let’s take the match against Newcastle as an example. If the model predicts that a Newcastle win is 55%, draw is 25%, and us winning is 20%, and if we actually win the match it will change our percentages because it was an unlikely outcome, but if we don’t win, the percentages would mostly stay the same. The last month we have taken 13 out of 15 points on offer, that is more than what most fans in here were predicting, and I assume way more than the model was predicting, meaning that the model will now predict us higher up the table, as you showed. I do believe we will win the remaining matches and qualify for Champions League, the model predict that the chance for that is 0.1%, but in my view it just show what a miracle Emery have performed when we actually do it. It that happens Wills needs to be having a word with his old man (seconded by George) and making it Sir Unai Emery! Link to comment Share on other sites More sharing options...
nick76 Posted April 12, 2023 Share Posted April 12, 2023 4 minutes ago, viivvaa66 said: The percentages are not pointless, it just gives the most likely outcomes of the league position at the end of the season. However, as you said, if a team is over performing compared to what the model predicted, that will change the percentages. Remember this is not the final answer, it is just a prediction, that changes as results get inputted in to the model. Let’s take the match against Newcastle as an example. If the model predicts that a Newcastle win is 55%, draw is 25%, and us winning is 20%, and if we actually win the match it will change our percentages because it was an unlikely outcome, but if we don’t win, the percentages would mostly stay the same. The last month we have taken 13 out of 15 points on offer, that is more than what most fans in here were predicting, and I assume way more than the model was predicting, meaning that the model will now predict us higher up the table, as you showed. I do believe we will win the remaining matches and qualify for Champions League, the model predict that the chance for that is 0.1%, but in my view it just show what a miracle Emery have performed when we actually do it. I’m sorry it is pointless because you then go on to point out one reason of a number of reasons why it has a much reliability as somebody guessing. The sample data is too small, it doesn’t take in other factors like form, injuries, fixture congestion just to name a few of many, the rewards per game ie 3 points can make a massive difference in such a small data sample. This falls down dramatically as a statistical model and as I said it’s pointless. Sorry! Link to comment Share on other sites More sharing options...
MrBlack Posted April 12, 2023 Share Posted April 12, 2023 19 minutes ago, nick76 said: I’m sorry it is pointless because you then go on to point out one reason of a number of reasons why it has a much reliability as somebody guessing. The sample data is too small, it doesn’t take in other factors like form, injuries, fixture congestion just to name a few of many, the rewards per game ie 3 points can make a massive difference in such a small data sample. This falls down dramatically as a statistical model and as I said it’s pointless. Sorry! How do you know the data behind the percentage calculation doesn't factor in likelihood of injuries and their relevant impact on each team? Opta have stats for everything so likely factored in more than just opposition strength and isolated result likelihood. It's only pointless if they didn't use much data to form the view. Otherwise, it's not pointless, and as mentioned, a win when one is not expected would (and should) affect the likely finishing position. That's a factual reality. Are we expected to beat Newcastle? If not, then a win against them improved our chances of finishing higher than any model predicts at this point in time. Its just logic. 1 Link to comment Share on other sites More sharing options...
nick76 Posted April 12, 2023 Share Posted April 12, 2023 10 minutes ago, MrBlack said: How do you know the data behind the percentage calculation doesn't factor in likelihood of injuries and their relevant impact on each team? Opta have stats for everything so likely factored in more than just opposition strength and isolated result likelihood. It's only pointless if they didn't use much data to form the view. Otherwise, it's not pointless, and as mentioned, a win when one is not expected would (and should) affect the likely finishing position. That's a factual reality. Are we expected to beat Newcastle? If not, then a win against them improved our chances of finishing higher than any model predicts at this point in time. It’s just logic. I don’t know whether you are pulling my leg or not with your comments! Anyway I have some professional background in Statistical Modelling, so based on that experience I think it’s pointless model. If you think it is a good model (you may have more or less knowledge on the subject than me) then fair enough, you enjoy it. Link to comment Share on other sites More sharing options...
VillaJ100 Posted April 12, 2023 Share Posted April 12, 2023 5 minutes ago, nick76 said: I don’t know whether you are pulling my leg or not with your comments! Anyway I have some professional background in Statistical Modelling, so based on that experience I think it’s pointless model. If you think it is a good model (you may have more or less knowledge on the subject than me) then fair enough, you enjoy it. Talking of stats I recently found out that xG doesn't take into account who it is... So Kane with a 1 on 1 is the same xG as if some clogger of a centre half had the same chance 1 1 Link to comment Share on other sites More sharing options...
briny_ear Posted April 12, 2023 Share Posted April 12, 2023 On 10/04/2023 at 13:17, Delphinho123 said: Let’s be honest, all of us on here are going to be a little disappointed if we don’t make Europe now. The upside of making Europe with THIS manager in charge could be a game changer both in terms of the players we can attract and the extra revenue it will bring in. Not only that, we have a serial winner of that competition in charge. It just feels right. But, we shouldn’t be disappointed. We could have been looking at a relegation scrap had Gerrard been here any longer. Like others have said, we get greedy, but it’s just human nature. Not me. Back on 5th November I was hoping for an 11-17 finish so the fact that we can even credibly talk of qualifying for Europe now is evidence of an amazing turnaround. I'm just enjoying that. Sadly, I think Unai came on the scene just a little too late to see us home and our remaining fixtures a pretty tough. I just hope we pit up a tough fight and create a few scares. 2 Link to comment Share on other sites More sharing options...
nick76 Posted April 12, 2023 Share Posted April 12, 2023 2 minutes ago, VillaJ100 said: Talking of stats I recently found out that xG doesn't take into account who it is... So Kane with a 1 on 1 is the same xG as if some clogger of a centre half had the same chance Yep I know Link to comment Share on other sites More sharing options...
duke313 Posted April 12, 2023 Share Posted April 12, 2023 36 minutes ago, briny_ear said: Not me. Back on 5th November I was hoping for an 11-17 finish so the fact that we can even credibly talk of qualifying for Europe now is evidence of an amazing turnaround. I'm just enjoying that. Sadly, I think Unai came on the scene just a little too late to see us home and our remaining fixtures a pretty tough. I just hope we pit up a tough fight and create a few scares. 17th? You sure you’re not setting the bar too high? Link to comment Share on other sites More sharing options...
MWARLEY2 Posted April 12, 2023 Share Posted April 12, 2023 Emery has 15 seasons experience of this chase for Europe to draw upon for these next 8 games. He will be taking the pressure off while doubling down on his expectation of every one of our players . We will not be underprepared. We might lose but it will not be through lack of preparation. 1 Link to comment Share on other sites More sharing options...
Kiwivillan Posted April 12, 2023 Share Posted April 12, 2023 1 hour ago, VillaJ100 said: Talking of stats I recently found out that xG doesn't take into account who it is... So Kane with a 1 on 1 is the same xG as if some clogger of a centre half had the same chance Also hitting post or cross bar doesn't count as shot on target IIRC when it's usually closer to scoring than on target straight at the keeper 1 Link to comment Share on other sites More sharing options...
Enda Posted April 12, 2023 Share Posted April 12, 2023 5 hours ago, DJBOB said: If you re-ran the simulator with a win against Newcastle, the results would be wildly different. That’s just how probability works, though? For example, the probability of a rolling 6 twice in a row from two rolls of a die is about 3%. But if your first one come up as a 6, that probability jumps to about 17%. Similarly if we get three points from Newcastle when most of us would take a draw, the probability of hitting Europe should jump up. 1 Link to comment Share on other sites More sharing options...
Enda Posted April 12, 2023 Share Posted April 12, 2023 1 hour ago, Kiwivillan said: Also hitting post or cross bar doesn't count as shot on target IIRC when it's usually closer to scoring than on target straight at the keeper And a whipped in ball across the six yard box that nobody gets a touch to counts as zero xG. Link to comment Share on other sites More sharing options...
Zhan_Zhuang Posted April 12, 2023 Share Posted April 12, 2023 Listening to the radio today and they spoke at length at how feasible it was for Brighton to achieve Europe (possibly CL) and how wonderful they were. I'm sitting there thinking WTF, Villa are a point better off and one place ahead of them and not a mention of us at all. Weird. Link to comment Share on other sites More sharing options...
Recommended Posts