With this first ever foray into match day player ratings the focus will be on just five players. One that keeps it a bit simpler from the data gathering perspective. Two, it is a smaller form. And it is new, a test does not need the full 14 players that took the pitch. Some ratings systems also include the coach, but for now we will ignore that.
The five players chosen are a distribution of seasoned veterans and youngsters and offer a strong view of how these ratings will differ from the conventional, overused and under-calibrated 1 to 10 scales that lack definitions. The main reason that Sounder at Heart has avoided that system has to do with calibration and definition. Is a 6 in MLS play from the same outlet that rates USMNT play reflective of the same performance? If multiple people write for an outlet and do the ratings how does that outlet prevent drift?
There are challenges with objectifying subjective data. As someone who did that for over four years in the flavor environment it is both difficult and necessary to calibrate. Our question here is simple - if the player performed at the level of play from the judged match for a three month period of time what quality player would they be?
- All Star - When counting the full All Star roster for MLS it is usually the top 2 to 4 players at the position. Thirty-two players in total are named All Stars per season.
- Regular Starter - At that level of performance would the player start 22 or more matches for 12 or more MLS sides barring injuries?
- Rotational Player - Would they get a mix of starts and substitution appearances on the season contributing but not a regular, not a difference maker.
- Reservist - Should they feature mainly in Reserve matches in order develop a tool or two that would help them succeed in the league while being available for an MLS side in non-League play.
- Lower League - Should they be sent down to a lower level team where they can advance more fundamental skills and grow their game.
The 1-10 scale normally just sees players in the 4-8 range with half points given out. The scale given here allows for similar differences in rating while eliminating the unused numbers that serve little purpose.