Hi all,
Thanks for this thread -- it's good to see how people are responding to the transfer list.
I wanted to make a couple clarifications, though, based on what I've read here.
1) Players are binned on the information that is listed (age, potential, expected position, and top relevant skill). We have found that these four factors have the biggest effect on the transfer price, so they were used.
2) Secondary skills are taken into account: That is, within a bin, your player is compared to the most similar players with the same age, potential, position, and top skill. So, a bin might have 100 transfers in the range of $1k to $1000k, but the most similar few might be in the range of $500k - $700k. So, it's more than just the top skill.
3) The data that is given is purely based on previous transfers. So, by definition it isn't going to undervalue allstar potential players, or condemn over-30 players -- it only shows what the market thinks of them.
4) Because of the minimal data we have available, some players will have very wide price ranges. Our model is intended to give as much information as is possible -- for some players, it is just impossible to give too much info. So, if the price estimator works well for 50% of players, and doesn't work well for the other 50%, then it's still very useful -- just ignore it for the latter 50%.
5) The tool will get better over time, as more and more transfers are made. We only look at the most recent few weeks, but as the game grows, the tool will get better and better.
I hope this helps explain the most common concerns I've seen in this thread. Keep the discussion going, though, and I'll check back every so often.
Cheers!