As many of you will be aware I've been working with spreadsheets of varying kinds of sophistication for quite some time but only for TFF. I only started working on this in any detail in 2005 and even then I wasn't really putting too much effort in.
Up until the end of the 2005-6 season I was only recording my then personal favourite statistic which was the average PPG per player minus 0.5*the standard deviation of their weekly scores. The purpose of PPG is obvious but the idea of the SD element was to try and find out the streaky players who turn out to be something of a long term gamble. I guess I used this to reasonable effect with a finish in the top 120 in TFF but sadly I still lost my works SL to a guy who came 40th! (But I'm not bitter
)
At the start of the 06-07 season I tried to compare players expected PPG-SD to their values and this, along with my personal opinion of him, led me to believe that Stuart Downing would be the value pick of the season. And I was happy to tell anyone who would listen. (Sadly I didn't stick with him and sold him before he really kicked off around Xmas).
My work is made enormously easier due to the kind generosity of Bramernic posting the weekly scores on his page - as I feed this data directly into Excel using a Web Query. I can then do the statistical gubbins very easily.
In 06-07 I devoted probably more time than is absolutely healthy in working on my sheet, expanding my range of stats to cover pure PPG, pts to date and my new personal favourite - PPG multiplied by games remaining - a good one for use at the end of the season. I also now have a page where my first XI players in each category and presented to me without having to re-sort the data first. Very handy as a quick reference tool and the eagle eyed among you may have seen the outcome of some of this work in my FISO XI thread.
Perhaps even more importantly, I've connected my sheet to some additional sources of data. The FISO Injury Table is now fed in via a web query and I've set my sheet up to exclude injured players from my calculations of the best players.
Similarly, I've connected up the FA Suspensions lists and again excluded suspended players from consideration.
The harsh reality of this all though is that I do this all because I'm a wee bit sad and enjoy both TFF and working in Excel, so combining these two activities seemed wholly natural to me.
Only on Tuesday night I was working on a version of this that transfers the statistical data into Access and then go through the player list with the aim of spotting the highest scoring valid team (by whichever stat you choose). Fortunately this agreed with the papers assessment of the highest scoring team so I guess it proved it worked, although it did take about 20 minutes to run through all the reasonable combinations and that was after fixing Lampard and Ronaldo into the team. So could be a pretty unwieldy tool to use, however I'm sure I'll use it in some way in deciding my starting XI for 07-08.
DixieDean wrote:but I know that some people such as Mystery use much more sophisticated tools in their spreadsheets. He's also posted that he's willing to share his work.
Very true. I have two available at present. The sheet from last season described above and a version of the EPL 07-08 fixture grid. Please PM with an e-mail address if interested.
)
