What exactly do our Prem. statistics show ?

[david in sweden]

Having watched Rasmus Ankersen's very interesting Tedx  presentation (available on You Tube) ... I began to wonder how do Saints stats. look - just  now ?

(Those of us who either believe in, or utterly reject  Stats.  go along with;  AntiVaxers, UFO-ers  and stolen elections.   You either believe in them ...or not.

 

Currently 14th  and halfway into the season, we have played 19 games and accrued 21 points.  Quite enough to survive in a normal season  IF ..we continued in the same form. 

We've scored  20 goals but conceded 29.   Bad news you say,  yet only two sides in the bottom half have conceded less.  (Burnley is one, but have played  2 games less).

Our biggest team weakness is not in defence at all,  but in attack -  just 20 goals scored - which is one of the lowest goal tallies in the entire Prem.

 

Our worst defeats have all been AWAY ( to 3 of the Top 4 sides;  Chelsea, Liverpool and Arsenal),  yet we drew at Man. City and even won v. WHam ( now 5th place)

We've lost ONLY one home game (0-1 v. Wolves),  (which we should  have won.....with  57% possession and 18 shots against their 5 ).

However, we have (sadly) drawn 6  of the other 8  games.  One goal more in each of our 9 home fixtures would have given  13 points more  (and a place in the top 6 ).

Our shots to goals ratio in the last 8 games is;    100 shots scoring just 10 times,  whilst opposition teams have  scored 17 times from their (cumulative) 114 shots. 

Positive outcome is;  AT HOME ...we have already played  6 of the top 10 clubs,   and against ALL of the top 5 teams in AWAY games.

 

Statistically speaking  ...the second half of the season should be "smoother" , and despite a few disappointing results we played better in the first half of season

than the results actually show, but without scoring that one extra goal at home that would have made a real difference. 

Sound simple?......perhaps.... but the real problem is the lack of scoring goals at home, and NOT the overall number of goals conceded, but as the majority of our

remaining fixtures (both home and away) are against mid-table and bottom-end sides, we can hopefully look forward to a decent top half finish.

 

IF.. Rasmus' theories are correct, then it should be an interesting question to return to  at the end of the season.   

 

 

 

 

 

[cloggy saint]

What exactly do our Prem. statistics show ? 

That we're 14th with a GD of -9

•  

[Turkish]

it's not a case of believing them or rejecting them. It's that some people simply believe they are all that matter, whilst some of us believe they are relevant but also only part of the picture. They are important and can be a good way of assessing players, opposition etc but they are not the only way and should not be the only way, which is where some are over reliant. For example people were raving about Vestergaard this time last year because his stats said he was one of the best defenders in the league, anyone who knew anything about football knows that wasn't the case.

[Diabolus Ex Machina]

The thing with stats is that they are simply data - a record of what has actually happened though without context. They can be used as an indicator of possible future performance but they are certainly not a crystal ball that will tell you what will happen, there are simply too many variables for that. It's the interpretation of them that seems to be the problem most of the time.

For example - you say an extra goal a game would have given us x more points, which while true is speculative based on the fact that the extra goal would be the only thing that changed in the game - as the goals never happened we have no way of definitively knowing if they would have affected the points we would actually get. On the other hand I could point out that out of the 5 times we've scored 2+ goals this season we've only won once (and still conceded twice in that game - total points for those 5 games = 7pts) whereas in the 5 games where we've kept clean sheets we've managed to get 3 wins (11 pts total over the 5 games). These were results that actually happened and show that a better points return has come this season when we've prevented goals rather than scoring multiple goals ourselves and so the data would show that improving our defense would lead to a greater points return.

But even that could be deemed too simplistic really - games are rarely super one-sided and a single mistake or bad ref call or injury/illness at the wrong time can turn a result around in a moment and could make a big difference on the stats for a season when dealing with a smallish set of data. If you want a more accurate picture you'd have to take a look at the data over a longer period of time to see if any actual pattern emerges and even then things like managers and playing personnel can change season to season and so a direct comparison will always be difficult.

The more stats you have to back up a position the better but that still doesn't make that position definitive, though i'd take the opinion of someone who tries and back it up with some figures and reasoning more seriously every time over someone who just pulled stuff out their backside because it fits the narrative they are trying to push.