Alright guys, so after some analysis in Excel and some discussion as well, this is the alternative proposal for arbitration salaries going forward. This approach is centered around the ESPN Player Rater, which provides a player rating based on their performance, relative to the categories we're looking for. Some advantages of using the player rater in lieu of the current bounds:
The current bands do not differentiate between .400 OBP over 200 PA or 600 PA. The nature of the player rater evaluating a whole season should account for this. Same goes for a 1.54 ERA over 100 or 250 innings.
Simply put, it would be easier. Instead of adding up several categories to calculate arbitration values, you only use one number into a linear formula.
This system is going to be based directly on current Contracts, similar to MLB arbitration.
The plan is to split players into their separate positional categories. SP, RP, OF, 2B, etc when performing this calculation. We'll pull all players with a positive player rating from ESPN for each category, and then isolate the players who in our league are under contract. Then, plotting just these players versus their contract values, we can get a sliding scale of player rater performance versus salary.
This will generate a formula that looks something like:
Scales won't be able to be determined until the end of the regular season, as we'll be waiting on final rater values. Only contracts active the previous season will be used. As a result, the formulae will have minor changes on a year to year basis.
The biggest advantage here is having a fully functional salary range in lieu of using a ton of stepped data ranges to determine salaries. And being based off the relationship between performance and rating for contracted players, there will always be an understandable relationship between the two.
**Whatever discounts(50%/70%/90%, etc) and arbitration maximums/minimums are going to be independent of this. This is simply for the calculation of the raw arbitration salaries.
I'm not really familiar on how the ESPN player rater is calculated so I have a question: If a player has an amazing 30 day run before he gets hurt for the season what would his player rating be like for that season? Or does the player rater amortize the stats?
The rater should evenly account for an entire season, as there's separate sorting options for ratings from the last 7/14/30 days. It provides a weighting for contribution to each valid category we use, and sums them for the desired time period.
Turb, do you have representative bandings for this season if this method had been in action?
Not at the moment, and I'm not sure when I'll get a chance to do that. It would be doing everything an extra time over considering the player pool will likely change between now and the end of the season to some extent.
I don't have concerns, I just think that examples would help me to understand how the values will be calculated and how the system would work in reality. If there are any changes to arb salary calculations we need to have clear instructions on how they are calculated, and at the moment they're either not clear enough or I'm not bright enough to follow how it will be done.