Bloody bastards.

Isn't this just a fancy way of saying that OBP is the most important stat?

Kind of interesting. A standard game has 27 outs of course, so this suggests that an average game has 11 PAs that don't end in an out. Which seems ... rather low, actually. That's an OBP of .289, while the actual rate in the NL was .318 last year.

Kind of interesting. A standard game has 27 outs of course, so this suggests that an average game has 11 PAs that don't end in an out. Which seems ... rather low, actually. That's an OBP of .289, while the actual rate in the NL was .318 last year.

Some of the outs wouldn't appear in that calculation of OBP because they would have come from double plays, caught stealing, pick-offs and outs on the bases immediately after a hit.

Some of the outs wouldn't appear in that calculation of OBP because they would have come from double plays, caught stealing, pick-offs and outs on the bases immediately after a hit.

Plus sacrifice bunts which don't count against OBP

edit. And not only that, but outs on the bases mean another PA ends up on base. Of course, that can also be balanced by reaching on an error.

Reading the article, it's another way for someone to put up the goal of having starting pitchers go only three innings(one time through the lineup) This is something I used to like, but over the years I have grown to not like the concept, and to think it wouldn't work.

Of course I think this is another of those causation = correlation messes that people try to interpret as having meaning. Factor in that you can probably get similar results by looking at runs allowed and win percentage and I'm not sure this article really offers much that is new or isn't plainly obvious.

Of course, the WBC and spring training both give examples of limiting starting pitchers to a far shorter leash. Are runs up or down in those situations?

Of course, the WBC and spring training both give examples of limiting starting pitchers to a far shorter leash. Are runs up or down in those situations?

Not really comparable though, in spring training they aren't trying to win and often times are looking at just working one or two pitches. And the WBC doesn't really have the same level of competition that MLB has.

The big thing y'all missed on 38 PA per game and OBP is that often the home team only makes 24 outs. If for no other reason than that I missed it and probably should have tried to calculate it as PA per 9 innings which is probably closer to 39-40.

In fact I'd wager that teams have outstanding winning percentages in games in which they make fewer than 27 outs while they almost always lose when they give up fewer than 27 outs.

cfb #8 ... I think that was the point of #1 and #2. If you have allowed fewer than 38 PA per game, you have probably allowed fewer runs than average so of course the win % goes up. It would make no sense at all to use relievers more often to reduce the number of batters faced -- you'd do it to reduce the number of runs scored.

Batters faced per game (or 9 innings) is of course very closely related to WHIP. AL average WHIP*9 + 27 gives 38.77 BF per 9 unadjusted for DP, CS, HBP (I think), etc. It does include SH though.

On BF, it would be interesting to see whether (after controlling for pitcher quality) RA goes up after 38 batters (or wherever). I don't see any particular reason for that to be the case -- it's one of the reasons to have relievers -- but at least then it might serve as a magic number of some sort.