I’m going to make these posts tidier, and give only the tables and a list of teams that could be lost from each race in the coming week. To check out the methodology see here (link), and to see how these compare to last week tables see here (link).
Race to 38 (To avoid relegation)
Arsenal will be safe with a point (home v Liverpool, Wednesday, and home v Stoke, Wednesday)
Liverpool will be safe with a combination of 4 points (away v Arsenal, Wednesday, and away v Man City, Sunday)
West Brom will be safe with a combination of 4 points (away v Everton, Wednesday, and home v Spurs, Sunday)
Swansea will be safe with a pair of wins (away v Sunderland, Tuesday, and away v Reading, Saturday)
Race to 68 (To finish in the top four)
Southampton will be eliminated unless they win both games (away v Man United, Wednesday, away v Wigan, Saturday)
Fulham will be eliminated unless they get a combination of 4 points (home v West Ham, Wednesday, and home v Man United, Saturday)
Norwich will be eliminated unless they get a combination of 3 points (home v Spurs, Wednesday, and away v QPR, Saturday)
West Ham will be eliminated unless they get a combination of 2 points (away v Fulham, Wednesday, and home v Swansea, Saturday)
Sunderland will be eliminated unless they pick up 1 point (home v Swansea, Tuesday, and away v Reading, Saturday)
Race to 87 (To win the league)
Chelsea will be eliminated unless they get at least one win (away v Reading, Wednesday, away v Newcastle, Saturday)
two questions … one a personal favour and the second about this post … first the post.
- Would it be interesting to add a “quality” element to the race situation?
What I mean is this … QPR need 23 points from their remaining games to reach 38 points … the two percentages (65% and 35.8%) give a good indication of how often such a run has occurred … now … what is their chance of stringing such results together?
I’m sure that for example one team within those percentages is a title winning team (e.g. last season ManC) … but how many teams of the quality of QPR have put a string of results together that they need?
i.e. if we take QPR’s TSR do they have a chance? How does it relate to the average TSR of teams that put such a run together?
An answer might plummet QPR fans into a further depression … or it might give them hope (i.e. QPR’s TSR is equal or better than the average).
I know the percentages already give a good indication but was wondering if it could be spiced up a bit.
other question will come later … out of time.
cheers.
That’s a great idea. The best I’ve heard in a while. I’ll have a think about the best way to do it before I put the next one up.
Hi James,
Cheers for the compliment …
Just a touch more explanation … what I thus mean is that QPR’s TSR could put them at equal (or better) chances in avoiding relegation with 23 points needed as Reading’s TSR and the 19 points they need … or Villa’ TSR and the 18 points they need … as it could be that Reading’s PDO is what has them in that position. etc etc.
Anyway, I think you understand.
Speaking of PDO … I had a discussion on the Counter Attack blog the other day after I answered someone’s question on PDO (in a Liverpool post) … I still think it is a “green” stat and in essence not that useful … yet … has more novelty value at the moment.
Anyway, threw up some ideas that I’ve been mulling over, might be interesting to look into (maybe you already have) … I don’t have the data to do so.
Speaking of data I have …
… here’s my second question … I’ve got 2 excel files of this season’s data and have been working on some new stuff but I don’t know how to combine it into something interesting … could I mail the files to you and if you have time, could you glance at them? Maybe give me tips on where to go? It is mostly on individual player data.
I’ve started writing the e-mail (long one) already and it has an explanation of what I’m trying to do …
Haven’t become a member of the forum yet, don’t feel I could contribute too much at this time (aren’t that smart stats wise and don’t have the time to invest into it).
I’ve sent you an e-mail before with files but wanted to run it by you first and not send you an unsolicited e-mail.
cheers.
Absolutely – send it along. I can’t promise to be overly timely with a response (I still owe a guy some stuff from January) but I’ll take a look when I get a chance.