Okay, in my main project of analyzing football data I have sheets broken down by position--one sheet per position.
Each position has different attributes and stats to be analyzed and I used a bonus/malus approach to determine value. If you're over the average for that stat then you take the difference over the average and multiply it by a small bonus but if you're below the average you multiple the difference by a much harsher penalty. The catch with this approach is that it works great but with one issue:
Virtually everyone ends up with a negative number and it's a different "severity" depending on the position. A simple position like kicker only considers a few attributes where a position like quarterback takes into account two dozen. Thus kickers might range in valuation from say, -120 for the best kicker to -1200 for the worst. When you look at QB's the best would be lucky to be -1200 and the worst could easily be -10,000.
Again, this works great for evaluating talent within each position but now I need to evaluate talent between positions. So how can I compare the best kicker with the best quarterback?
An idea hit me on this but I'm sure there's a better way. That stats that get evaluated (for the most part) range from 1-99. Thus, there is a worst case scenario (someone with all 1's). And, of course, on the other end is the best case scenario (someone with all 99's). If you then calculated the range from worst to best you'd end up with a range that could--in some way that's escaping me--be compared to the numbers actually calculated and then multiple that into a positive integer (say between 1 and 1,000). Then if you repeat per position you'd end up being able to compare a QB to a kicker and know who to draft in what order.
Am I over-complicating all this? Is there a better way to go about this comparison? Either way, what does a formula look like to make this work?
Thanks.
Each position has different attributes and stats to be analyzed and I used a bonus/malus approach to determine value. If you're over the average for that stat then you take the difference over the average and multiply it by a small bonus but if you're below the average you multiple the difference by a much harsher penalty. The catch with this approach is that it works great but with one issue:
Virtually everyone ends up with a negative number and it's a different "severity" depending on the position. A simple position like kicker only considers a few attributes where a position like quarterback takes into account two dozen. Thus kickers might range in valuation from say, -120 for the best kicker to -1200 for the worst. When you look at QB's the best would be lucky to be -1200 and the worst could easily be -10,000.
Again, this works great for evaluating talent within each position but now I need to evaluate talent between positions. So how can I compare the best kicker with the best quarterback?
An idea hit me on this but I'm sure there's a better way. That stats that get evaluated (for the most part) range from 1-99. Thus, there is a worst case scenario (someone with all 1's). And, of course, on the other end is the best case scenario (someone with all 99's). If you then calculated the range from worst to best you'd end up with a range that could--in some way that's escaping me--be compared to the numbers actually calculated and then multiple that into a positive integer (say between 1 and 1,000). Then if you repeat per position you'd end up being able to compare a QB to a kicker and know who to draft in what order.
Am I over-complicating all this? Is there a better way to go about this comparison? Either way, what does a formula look like to make this work?
Thanks.
