A-League Spreadsheet

It has all the usual functions contained in one of my league sheets:

-  League tables and fixtures
- Cumulating points graph
- Changing positions graph
- Collective or individual results entry interface
- Strength of opposition & remaining opponents calculations
- Results grid with user definable highlighting
- Dynamic prediction engine*
- User predictions facility
- Dynamic team and division stats for season
- Form tables with user defined parameters
- All-time league head-to-head meetings

- Previous season league tables

 

As always, it's free and available to anyone who is interested.

 

*The historic data component for the prediction engine was of course empty for the Gold Coast and North Queensland teams, so I had to extrapolate a starting point from the A-League odds for those teams. Once we get to the end of August, the current form component should start to out-weigh the simplistic inputs of the odds. I may also have to tweak the weight of the previous fixture where a repeat occurs - not something my existing algorithm had to be concerned with.

 

 

This is the first time I've managed to convert the template to cope with leagues that use multiple home and/or multiple away meetings between teams in the same season. I've kept its functionality and look/feel consistent with the existing sheets. Now that I have such a system, there is a possibility of expanding coverage to some additional leagues that were previously deemed too hard. Not sure if I will find time to do that for 2009/10 - I will most likely see if the new template encounters any bugs during the course of this first application before doing more.

 

Any problems give me a shout (see www.michaelwray.co.nz).

 

It should work fine with Excel 2007 or prior. (Note if you have Excel 2000 or 97, you should be fine only if you are not on IE7 or above - IE7 breaks all but the simplest Office 2000 documents.)

Cheers,
Michael.

 

I take it we update every result into the results sheet and it does the rest

 

Results have to be entered - either using the individual entry or the collective enter result for date-range functions. The predictions and stats will automatically be generated from the results.

 

Importing results and date changes direct from an on-line source is on the development list, but I haven't got round to it yet - when I do, I'll want to do it for all of the league sheets (Top 5 English divisions, top flights of Italy, Spain, Germany) and possibly the Euro Champ League and World Cup sheets.

Yep. 442 have me a mention a few years ago too. It was only one paragraph, but it was cool.

 

according to your 'oracle' will the Tigers stay up (you do realise I'll only believe you if the answer is "yes"

http://www.michaelwray.co.nz/Premiership-0910.zip

 

It has you finishing 17th this year. Of course, that's the starting point so is purely based on meetings between the teams over the last five years and has no current form. I have thought about trying to carry over form from the end of a previous season, but it doesn't seem appropriate to do so.

 

Thanks Tony.

 

Yep, you can enter results and your own predictions. You will need to have enabled macros to do so.

 

You can enter results using the sheet "Enter Result" for a single match or "Enter Results for Date-Range" collectively. Similarly, your own predictions can be loaded using "User Preds by Team" or "User Preds for Date-Range".

 

The sheet is predicting 9th, not me! (I'm guessing 5th-ish pre-season). I've not incorporated any bias or personal feeling into the calculations.... which is why Brighton are tipped for another relegation scrape in English League One.

 

The predictions are using past meetings between the teams, with weightings and score distributions according to a weighting of history and current season distributions. Obviously, current season info is blank at the start though, so it is purely history.

 

There is no history for the two new teams, so I had to resort to using the betting odds. That penalised us against them, as we're second favourites to finish bottom.

 

Once we get a month into the season, the current form component of the prediction engine will kick-in properly.

 

I entered the sheets into some prediction comps a few years ago, where the picks were submitted each week (as opposed to all being submitted pre-season). Relative to the human competitors, it performed best for the German Bundesliga and English Championship, worst was the English Conference. The rest did okay.

 

Very much appreciated.

indeed not

 

p.s. Tony P seems well impressed

 

Us Brighton boys are just impressing the Nix management all the time now....