Could be worse, we could be like Everton. (P.S I haven't read the thread so I'm sorry if this has been posted 127 times already)
Mathematically this means that our defence ability to defend is 5 times worse than our attack's ability to score
Interesting stats. In comparison with two other clubs, using shots on target only rather than all shots: 1. LFC. Shots for on target 46 Scored 12 Shots/goal: 3.8 Shots conceded 23 Scored 11 S/g: 2.1 2 Man City SoT for 44 scored 21 Shots/goal: 2.2 Shots conceded 11 Scored 2. S/g: 5.5 3 Man United SoT For 40 Scored 17 Shots/goal: 2.3 Shots conceded 18 Scored 2. Shots/goal: 9.0 Conclusions: Our attack is profligate and needing more shots on target to score. Our defence allows not only more shots on target but our GK is also not good as others to stop the goal once the shot is struck. Every other shot on target on our goal gets scored. But City only allows one out of 5 or 6 shots to be scored against them. The most miserly are the red mancs: they only allow one out of 9 shots on target to be scored. Mind you, these stats only confirm what we already knew.
Just a quick point on stats. The sample size is too small to be reliable. The degree on confidence in their accuracy would be quite low. However a tutor of mine used to say that after about 8 points of data you could go into a dark room and plot a run chart as long as nobody saw it. These stats are really a confirmation bias. We see we are **** therefore see stats that prove it. It is patently obvious that we are throwing goals in but when you have the lions share of possession then over commit bodies you are wide open yo a break. Add that to set piece incompetance and the chances we are hiving up are indeed guilt edged. Then you must wonder just who have Utd and city played yet etc etc. Lfc are poor enough thus far but 8 of the goals come from 2 games. Watford away and city disaster away. If you took the city game out of those stats what would they become? I would call the city game a special cause of variation and remove.
I agree in statistical terms these differences may not have reached significance I.e p < 0.05. But they are nevertheless interesting. Yes they may be recall bias but these are actual and objective data not inferred or subject to interpretation. After more matches we'll be more certain of the differences in attacking and defence/goalkeeping abilities of the 3 teams.
Sure but I would still eliminate the city game from them as an assignable cause. We all know We are ****, but how **** are we?
I would only remove the City game if the game did not represent a premiership match. Yes, there was a sending off and we had to play with 10 men. But that's not so unusual as to be totally exceptional. It is not a once in a lifetime occurrence (e.g 3 of our players, after substitutions have been done, collapsing together from illness leading to playing with 8 for 30 minutes) so that that match can be excluded as an outlier. This match was part and parcel of a league as much as our 4-0 thrashing of Arsenal was. The loss of Mane could well represent the usual characteristic recklessness or aggression of the player rather than an unusual exceptional event. We can hardly say that playing with 10 men is a very rare event can we? otherwise we could remove the 4-0 thrashing of Arsenal as Wenger having a brainstorm and not playing lacazette his best striker and Oxley who was being transferred to us instead.
No it's a clear case of a special cause variation caused by a red card. We would not have lost 5 goals in one game if not for that so if you plot all the goals conceded stats for 16 games i will bet anything that we see this being an out of control signal
Red cards are part and parcel of the game. It is also obvious that any red card will have serious consequences on the outcome. As much as extreme scorelines like 4-0, 5-0 or 7-1 etc. If we did remove any matches with a red card then where do we stop? One where Coutinho and Mane have not played as it has affected our scoring ability? Or another that the ref made mistakes by awarding "wrong" penalties against us? I would prefer to just use the data raw without any subjective filtering. But it is up to the analyst manipulating the data. That is why stats can be made to say anything.
No it's not if it's shown to be an out of control signal. Give me say oh....10 data points and plot them. Have we played 10 games? I think so. Then see if 5-0 is outside the limits. I'm saying the result is freak not the red card. Quite simply if the data point is outside the co tell limits then it is removed from the stats otherwise no. That is not the analyst doctoring data it is analyst making data more accurate. Say you did a deplhi analysis on our results 4-0 would be one extreme, the 0-5 would be the other extreme. You could get feedback on those extreme from each person and the zone in on subsequent rounds on a best estimate of just how **** or good we are. So extreme do affect data, that's all I am saying.
You don't need stats to know that City's strikers are better than Firmino or that De Gea is a much better keeper than Ming. They just back up the bleeding obvious.
i could allow taking out matches where red cards apply if you really wanted to doctor them games. However to say you guys rarely get stuffed 5-0 also doesn't take into account city have been stufing teams that scoreline pretty consistently this season so actually is a perfectly normal result and not an anomaly
I agree with MITO, we should take the Stoke game out of Uniteds stats too as it was an anomaly on two counts, we don't normally concede or drop points
