Power as element of Contact

[jcard]

I would rethink the removal of Power from the Contact formula. Specifically, replace BABIP with BACON. BABIP is useful in the absence of Statcast data (and associated xN metrics) for isolating hit luck (or, with a big enough sample, as a proxy for quality of contact). In OOTP, however, BABIP is an attribute of the batter; a measure of underlying skill. From this perspective, BACON is a better complement to Avoid K for representing the batter’s Contact attribute / skill.

[Rain King]

I think they probably pulled it out because they wanted the "base" ratings (i.e. Contact, Power, Eye) to work independently of each other.

I don't remember for sure, but I believe this change was made in part to alleviate some issues with cross-era play (i.e. Random Debut, Perfect Team, etc.).

I also believe the current Contact rating is kind of confusing, but I just take out the middle man and remove it from most of my views and instead show both BABIP and Avoid K...which tells a better story about the player anyway.

[jcard]

Ideally, the scouting model would be brought into the 21st century by giving you the option of projected batting statistics (basically, like the editor but with scouting error / incompleteness. As for the ratings: Certainly, there are trade-offs in the different methods of representing the attributes. I personally find the current model less efficient. For example, if Arraez and Judge are scouted to hit for similar averages, it would be easier to look at their Contact rating without having to look also at their Power rating and then guesstimate how many hits (HR’s) have been left out of Judge’s Contact rating. Perhaps Power could be used like Gap, representing a proportion of the Contact hits that will be HRs (directly or indirectly through HR opportunities, however the engine works. Again, though, my strong preference would be the option to avoid all of this runic mystery with ratings and use analytic-style projections.

[kidd_05_u2]

I've always pushed for the complete opposite approach to what jcard wants. I don't get how OOTP can be any fun if the player ratings are basically straight stat projections.

To me at least, a big chunk of the GM-ing fun is that I have these players with these "skills" ratings and I don't know exactly what to expect of their performance. Of course, I hide the Babip ratings, use a 20-80 scale, and I incorporate stats into scouting reports, to make the ratings even less of a straight stat projection.

[jcard]

Originally Posted by kidd_05_u2

I've always pushed for the complete opposite approach to what jcard wants. I don't get how OOTP can be any fun if the player ratings are basically straight stat projections.

To me at least, a big chunk of the GM-ing fun is that I have these players with these "skills" ratings and I don't know exactly what to expect of their performance. Of course, I hide the Babip ratings, use a 20-80 scale, and I incorporate stats into scouting reports, to make the ratings even less of a straight stat projection.

Well, I intentionally spoke of it as an option.

I actually appreciate the desire to avoid a direct statistical translation from ratings. However:

1. Certain ratings (Eye, for example) have a correlated 1:1 rating:stat relationship.

2. Modern front offices use projected stats, at least once they have enough reliable data (stats, environment).

3. Even when using scouting grades, these have some understood translation: a GM knows what his scout means by “60”, beyond just “better than 50, worse than 70”.

4. Generally, I wonder how predictive the translations would really be. If you play the same season multiple times, you will see a player with no meaningful ratings changes from initial nevertheless put up a wide range of performances. I sometimes wonder whether the ratings (or a projection thus defined) are meant to represent a scouted underlying talent level, while short and media-term factors affecting performance (league adjustments, mechanics issues, psychological components) are abstracted into adjustments unrepresented in the scouted ratings / projection. Perhaps including stats in the scouting report would then provide nearer term information supplementing underlying skill ratings. This may have been what the “show ratings > max” option that preceded the current “include stats in scouting reports” tool was trying to capture.

Or not‚…