An analytical assessment of league one this season


Status
Not open for further replies.
Don't really need all those stats to tell us this. It's looked clear all season. There's been loads of games where our extra quality at the other end has made the difference with the opposition missing loads.
 
It's just a terrible model that treats a team with Camp and Altidore in it as statistically likely to have the same outcomes as one with say Pickford and Defoe (or even McLaughlin and Maja).

The quality of a player at the most critical moments of the game is sort of important in football.

There is simply no way to tell if this 'model' shows if SAFC is just 'running good' (above EV if you ever seriously played poker) or is merely performing in accordance with its players' quality. Or indeed is currently underperforming.

I think those of us who have watched every second of SAFC's games this season would say we are running a little 'warm' but not outrageously so, and are likely to improve from now on and so sustain results rather than 'regress towards the mean'.
 
Last edited:
OK, after a long career in performance measurement (in the real world, not the FIFA manager type) I was willing to give this a chance.

My conclusion is it is trying to do something very sophisticated with the minimum of facts, in my experience this has never worked in any context. The idea that you can take a few simple facts and use it to accurately predict or explain reality is attractive, of course, but bullshit,

Firstly, half the time the data sets are more or less dependent upon each other, which is a problem with 'prediction' rather than 'analysis' and the author does do this for prediction, doesn't stop at (vaguely interesting albeit) analysis. Its a bit like 'can we use goal difference to more or less predict points' - yes! what a shocker!! then we display those that tally as 'proof' and give subjective 'explanations' for the (whopping big) exceptions - that subjectivity defeating the whole point of believing the tarot card calculations in the first place.

Secondly though, and far more importantly, key factors are simply ignored (or isolated to individual exceptions). We know better than most that the goalkeeper is not just there to retrieve the ball, we tried that and were relegated. It follows that the conversion quality of shots (and at the other end defending) may be affected a bit by the keeper? A good percentage of shots never reach the keeper, do you think maybe better defenders block or disrupt more shots than poorer ones?

If quality of finishing is such a big factor for us, as claimed, why is it then irrelevant for everyone else?? The analysis is not 'wrong', its just its only a 'level one trawl' that if enough effort was put in would then add other factors like these until you explain nearly everything - then you raise really interesting questions about what is left. It might be that a team's injury crisis jumps out for example or a wide range of other things.

If we are after prediction, just take results to date, look at who you've played so far and pay a lot of attention to current form - that will yield a far more accurate prediction IMO than this 'analysis' which interestingly seems to accept it is only short term (see 'Barnsley') and needs re-run regularly which is disappointing if we wanted to rush to Betfair to bet against Sunderland based on the analysis.

It seems, in summary, that a teams overall performance is critically influenced by its future performance (in a longer sentence it says that). Maybe it's not a simple game after all.
 
The way I understand it is saying Portsmouth are the dogs whatever and will stroll to promotion yet we are lucky to be where we are.

However if we win our game in hand we go above Portsmouth.

I think this report stinks. :)
 
Played 15, scored 30, that’s 2 goals scored per game. Conceded 13 that’s less than one goal per game

Points 32

Continue at this reality then we’ll achieve 96 points, whilst scoring 92 and conceding 40

Champions
 
This section is the icing on the cake

Sunderland are again noticeable in the context that in the “real life” table they sit 3rd and with the best goal difference in the division. In this, they’re slightly above-average defensively and slightly below-average offensively. Now, individual quality isn’t accounted for which could certainly be a contributing factor as to why they have scored as many as they have – simply having better finishers. To what degree that is true and that it isn’t just dumb luck is debatable though and it also wouldn’t be the full cause for a gap this large.

Anyone who has actually watched us regularly would know we are much better offensively than defensively (even taking into account the recent clean sheets which particularly at Donny we have McLaughlin to thank).

Yes we have players who will generally take their chances but to say it’s ‘debatable’ that it’s ‘dumb luck’ is laughable.
 
For me it just proves stats mean fuck all really we are doing well because of the quality and we rarely miss a chance. If look at Doncaster the other night to glorious chances before half time missed both we went on to win because the chance we got we put away similar to Shrewsbury
 
There can't be any dispute our finishing has been of a higher standard than the teams we've played. I'd take a punt on Maja's goals to shots ratio this season being up there with the best in the country.

I disagree with article's prediction that our stats can't be maintained though. Imo, the forwards have genuine finishing ability and it would take a crisis in confidence for it all to tits up.
 
So the actual data disproves his analytical model, time to abandon or rethink the analytical models methinks, clearly he hasn't or can't model all the data successfully
 
There can't be any dispute our finishing has been of a higher standard than the teams we've played. I'd take a punt on Maja's goals to shots ratio this season being up there with the best in the country.

I disagree with article's prediction that our stats can't be maintained though. Imo, the forwards have genuine finishing ability and it would take a crisis in confidence for it all to tits up.
Maja is something like 70% of shots on target and 50% scoring - way ahead of the rest of League 1.
 
I find with all this statistics palaver, that the statation has something to prove then invents a statistic formula to prove his point
 
I haven't done any hard scientific analysis on this, but in the first part of a season, I personally have always looked at the goal difference of teams as a good indicator of how good that team is. When there haven't been a huge number of results it appears to be a better indicator of how they will perform over the rest of the season than the actual points gained.

The fact that as of now we have the best goal difference on League One can't be a bad thing.

I think all the other statisticians disagree with his outcome:

League 1 Promotion Betting Odds | Football

They are punters, not statisticians. That just reflects where the money is going.
 
Status
Not open for further replies.

Back
Top