Points/game and points/IP


A great addition. Thanks, @nivshah


P/G and P/IP have been added to the watchlist as well.


I think adding points per PA would more accurately depict how many points you could expect from a player. Currently, if someone pinch hits it counts as an entire game played. It makes judging fringe players tougher, because they may have a lower P/G due to occasionally being used as a pinch hitter or being subbed out for a defensive replacement. P/PA would more accurately represent these players, especially when they get to start.


Up thread here, you can see I considered P/PA but reversed course because I wanted to match the denominator with the playing limits, which are G and IP.

I do see your point about P/PA, but I would like to hear from some more folks who use these numbers.


I use P/PA a lot when comparing guys at the end of my bench. A lot of platoon guys don’t play everyday and get a lot of pinch hit at bats (which almost always lowers P/G), but when they play, they do pretty well. Mark Reynolds is a good example from last year – very low 3.52 P/G, but solid 1.29 P/PA. Springer is just one example of a guy with a lower P/PA than Reynolds (1.21), but a much higher P/G (5.37).

I do understand the point about using P/G because of the player limits, but don’t think it could hurt to show both.


Space is at a premium already, so showing both isn’t a great option. I’d like to pick one if possible. Both might make the most sense though…


Yeah, I understand. If I had a vote, I’d say to go with P/PA, but also see why it makes sense to have P/G.


I think the ideal would be P-in-GS/GS (solves both needs in one column), but P/G with P/PA is an OK compromise if we can’t have the other. If that is too much real estate, then P/G solves the primary need and leaves platoon guys for the owners who do extra research.


The data might be harder to aggregate for P/GS, but I agree, that would be the best way to do it.