Repricing Players Cards v2

Sewell's 2.12 or 1.86 card is popular as an offensive value. So, before long, Joe Sewell will have a salary of 6m?

Perhaps an exageration for effect. But for discussion how long would it take for a usage based model to drive a price that high? How about some math? For simplicity sake just going to average those two and start with a 2 mil card.

So by whatever definition used the card is over used and gets a 5% boost. I use 5% because that is what I remember someone suggesting and it makes for easy math. 5% of $2 mil is 0.1 mil. For purposes of this math exercise we will assume repricing every 6 months. At that rate it would take 11 years for that card to get priced at 6 mil.

It would take 4 years for the card to reach $3 mil. It would take 7 years to pass $4 mil.

Now I assume those who think he is a bargain at $2 mil and use him would not consider him a bargain at 3 or 4 mil. They would quit using him and his price would level off or go down. It would likely go up and down until it reached an equilibrium where it would hover at the community's perceived value going up one year and back down the next.

Basically the logic is this. Start with a formula that at present is relatively accurate. While it is better than that used in ATG1 it may still miss on some players. Usage based tweaking would serve to over time correct those minor pricing errors. There is no way though that it is likely to ever result in doubling or cutting in half a salary. Usage patterns would prevent that. But if there is concern it would that problem is easy to solve. Put a cap on how much a player can vary from his formula based value. For discussion let's just say 50%. That means Sewell would never go higher than 3 mil or lower than 1 mil.

One could also only reprice the top 10% overused or top 10% underused. That would leave 80% unchanged only correcting for the most mispriced cards.

As to the cap levels it is going to be true that 200 mil leagues would make one group of cards look under used while 60 mil leagues would make a different set under used. But they would balance each other out and would only skew the results to the extent those caps are popular. Including all caps would mean the more popular caps would have the most influence on the results with an overall balance of being the most equal. But for the purpose of using usage rates to identify under or over priced cards Ithink only using 80 to 140 mil leagues would be just as efficient.

Of course all of the above is just for fun discussion as none of this has a chance of ever being done. Remember Strat does not even read these.

Just think same things are true of adding new cards for example- Bill does most of the work (he does a great job imo) in gathering who is requested and compiling a list and choosing hopefully.
Strats only has to input the card data, but that doesn't seem to be happening right now.

A very accurate statement which underscores that this discussion is really just for something interesting to talk about.