There is actually something interesting here. It's a very common mistake for people to infer causality from something which is actually driven by randomness based on too small a sample or a misunderstanding of probability. A good example is newspapers publishing the 'hot' numbers in the national lottery - since the lottery has to be random (or a fraud) these have no value whatever. We all believe on here that we can make judgements on how good players and managers are based on qualitative observation and that may be true. But its really much more likely that we are deluding ourselves based on too few observations and or selecting the data to support our prejudice. If its really true that you can't get a significant difference in manager's performance by looking at 100s of matches what chance have we got in being right based on qualitative observations.
Qualitative observation? What as opposed to **** observation? Bin you numbers and use your experience. I have thousands of references in memory from the thousands of football matches that I have played/watched. All these references enable me to reach a very sophisticated conclusion. I can add to that the experiences of others whose observations I respect all based on myriad bits of memorised information. It's called AN OPINION.
These are the only numbers that matter. In 2 months, avb delivered 3 of our most embarrassing PL defeats conceding 14 goals, with out 1 solitary consolation, one team had no recognised striker, and one game offering 0 on target. All this after spending 100 mill. Great points haul though that got us another crack at the .....Europa League! Our most successful manager! ....if I hear someone say that to my face ....
Exactly. Statistics are never conclusive for the questions they're applied to. For example, passing accuracy is hugely affected by the ease of the passes the player took on. So if you look at the players with the greatest passing percentages they tend to be centre backs, full backs and defensive midfielders. It doesn't mean they're the best passers in the league but they have the best conditions for the statistic to come out in their favour. You can dissect it further by looking at the conditions they play in. With regards to the Lenny's post about win percentages, they're not lying. AVB had a better win percentage in his time at Spurs than Redknapp. That's what the statistics say and it is a fact. The trouble is people tend to draw conclusions from the stats when there's no conclusion to be drawn. Like the example with passing percentages, where the numbers don't mean one player is a better passer than another, so AVB having a greater win percentage doesn't mean he's a better manager. It's understandable that people make these conclusions, at a glance it makes sense but in reality it's a huge leap between the statistics and the conclusions. Lets be honest as to what it came down to, people that liked and supported AVB were encouraged by the signs(although I don't remember anyone pointing to win %ages before he was sacked), those that didn't weren't. Had AVB had a lower percentage than Redknapp, I'm sure it would've been used as a stick to beat him with. I'm not one to jump to ott judgements either way with AVB. For me, he proved last season that he was a decent manager, just like at Porto he showed he was no fool, however in his spell at Chelsea and his second season with us he's shown he still has a lot to learn. I hoped he would learn and grow with the squad but that hasn't happened I don't feel silly or like I've been proven wrong by the anti-AVB Spurs fans, far from it. I'm an optimist, I like being optimistic about our club and just like it took me about a day to be fully supportive of AVB(I really didn't want him) I will be the same with Sherwood and whoever comes in, in the future. It's not a case of being pro-AVB, it's a case of being pro-Tottenham. Soldado was almost the exact same as AVB infact, I didn't want him, wasn't sure it'd work but as soon as he signed I fully support him and will continue to until he leaves. I think it's sad that people are trying to come out with the "I told you so" lines now, especially against fellow fans of your own club, it's petty. If you want to be negative, you do you but leave me and the other optimists to live in the clouds, eh COYS
I'm fine with opinions. Its people that know or think they can prove that they are right that get me. I've never suggested using statistics to rate players or managers or anyone else. I'm quite keen on using statistics to guide me as to whether my (or others') preconceived ideas might be wrong. There are a few subjects that I understand very well. I know from those that when people use 'common sense' to draw conclusions about them they are wrong an awful lot of the time.
"Exactly. Statistics are never conclusive for the questions they're applied to." Statistics is merely about discerning patterns in numerical data. From which hypotheses about the reasons for those patterns can be put forward. Those hypotheses can then be supported/refuted with the same and/or different data in isolation/conjunction. lennypops showed not that AVB was better than Arry, merely that that the reason(s) (if any) may be down to nothing more than chance. Similarly I will state that : - The PL winner this season will come from the teams placed 1-3 on Jan 1 - The teams placed 1-3 on Jan 1 will finish in the top 4 come May - Any team that finishes with 76 pts or more this season will exhibit one or more of the following : GD > 30, goals conceded < 40 But I have no hypotheses for why the above is highly likely to be so.
"I'm quite keen on using statistics to guide me as to whether my (or others') preconceived ideas might be wrong." Which was the essence of the "moneyball" system in USA baseball. We touched on football counterpart notions the other day when discussing corners, and the stats thereof.
I know Lenny wasn't saying that, it was others that apparently were. Lenny was attempting disprove it with more stats when really all you had to do to show the error is look at what the stat actually means and what people were trying to say it means. Football statistics are interesting and food for thought but don't people just love going ott. Ultimately with AVB it comes down to the fact that people didn't like the style of football, couldn't see where the improvement was going to come from and weren't prepared to watch other teams overtake us in the meantime. 1 of those things at a time would probably have been acceptable but all 3 at the sametime is tough to take.
"really all you had to do to show the error is look at what the stat actually means and what people were trying to say it means" What it means is that AVB has a higher win percentage than Arry and other PL era Spurs managers. But not significantly different from Arry as to be of any further remark. I am still not particularly clear as to what people are claiming about the stat other than the above.
Here's another one; This season in the PL, Liverpool have scored more goals from open play at WHL than Spurs. Aren't statistics great!
That's precisely what it means. It doesn't mean he's a better manager, or worse manager, but that he a higher win percentage. Apart from Redknapp no one else has had a squad anywhere near as good so he should be up the top but on the other hand it is a good thing that he did have that level of record. Nothing to throw a parade about but still, it's a minor stat that was showing what it should be which is good. Unfortunately there were a lot of negatives too which he has in the end paid the price for.
Does Levy have a sideline employing pilots? I think some of the managers he has employed were dreamers.
