The AL had a .531 winning % this year vs the NL. In past years, it’s been worse.

No one would argue the AL isn’t better than the NL, that’s well accepted fact. The only question is how much better.

I would also note that few of the NL teams have as high of straight winning percentages as the better AL teams… And that’s against weaker competition.

there are 4 teams from the east league.
very competitive imo
very good and good wont get you a playoff spot in the east.

But that’s just nitpicking. Great as usual.

But the Giants are a much better team than you give them credit for. xFIP does not measure well with Giants. Zito and Cain for instance almost always beat their xFIP.

But two questions, why include slg and obp when woba is already included? Doesn’t woba include both aspects in its calculation and whats the difference between xfip and xfiprns?

I’m having some issues with BtB right now, between it not letting me sign in (had to go to another SBN site to sign in and come back) and now not being able to reply to comments, so I’ll just put this as a stand-alone comment….

I also think this is not a very accurate evaluation of the Giants, either in our chances going forward or an evaluation of our past performance. I believe the problem is almost entirely because of xFIP. I’ve seen various metrics on how our park suppresses HR’s, ranging from the one provided above (.95) to ones far below that (something along the lines of .95 to RF, .7 to LF off the top of my head), so of course our park is a significant factor to an arguable degree. That said, I’m not sure what the appropriate step would be – you’re only looking at one type of batted ball in xFIP, which normally is fine (since variation between pitchers is so small ignoring it doesn’t cause problems), but in this case is a problem. Just park adjusting one factor (HR’s) and assuming all other balls in play are equal at each park just isn’t accurate in this situation, as evidenced by AT&T being tough on HR’s but still pretty close to neutral overall.

Second, the skill aspect of batted balls. I’m not sure I could statistically conclude that guys like Cain and Zito are harder to hit in a meaningful way than most pitchers, but their statistics are least suggest we should be open to the possibility. Both have career BABIP’s under .280. Both have high career IFFB%‘s (it appears to me league average is somewhere in the high single digits, they’re both at 12.7%+) in a park that isn’t particularly big, though I will note Zito spent 6 years in the huge foul territory across the Bay. Not trying to get into too many details here, just pointing out that the Giants as a team may have an advantage in balls put in play due to pitcher skill. I’m not sure they do, but this is the second year in a row you guys have been off in evaluating their performance, and at least two of their big inning guys seem to indicate that…..it could be some sort of systematic way the Giants do things.

I think you made a good point, JinAZ, that it might not be worth making any changes just for the Giants, since it could make the system less accurate on other teams, so I’m not trying to advocate you do anything (and as I said, I’m not completely sure it’s wrong on the Giants nor am I necessarily sure, if it is, what any appropriate fixes would be)….I just suspect there’s some systematic bias against them right now. I do believe an accurate evaluation of the Giants should have them much higher than they are.

On an unrelated note, just because this is a huge pet peeve of mine….I hate the AL/NL distinction. I don’t see any reason to use it, nor do I understand why we think it’s a good idea. Do it at the divisional level. Teams play half their games within their division, it’s very meaningful. The talent gap between divisions is significant. We shouldn’t be giving the AL West credit for how good the AL East is. Sure, it might be harder to do, but it makes a lot more sense to look at the difference between divisions than the difference between leagues.

Bizarre – still unable to use the reply feature.

For point 1, I’ll start by saying some of it is likely skill. Just for a quick example, Cain has a career 48% FB% at home compared to 45.4% on the road. I’m guessing this is significant, as we’re looking at 551 and 464 IP, respectively, but I honestly have no idea how variable FB% is so I could be wrong. Again, not trying to make any definite conclusions here, just pointing out the evidence may suggest that there’s some skill involved (in terms of “pitching to the park”), in addition to skill indicated by his career HR/FB ’s of 7.2 on the road and 6.6% at home (and Cain is just the quick and easy example, here.

My next point is I’m having a hard time thinking through the park effects logically. You’re park adjusting the offense, which in reality is just adjusting for the run environment of the park (as opposed to a theoretical “what would he have done in a neutral environment”), the defense is somehow park adjusted (not having ever seen the details of what adjustments they make I don’t know details to say anything more), but then for the batted balls, we’re ignoring the park, and just adjusting one aspect of batted ball profiles by regressing to a league mean while completely ignoring the rest. I guess it doesn’t strike me as consistent or telling us exactly what we’re looking for? My thoughts are still….a bit unclear on this matter, I’d enjoy a conversation going into more depth into what end result you’re going for, precisely, and how park adjustments play a role in getting to that, as well as how to apply them consistently across every input.

Maybe an easier way to put it is this:

Yes, pitchers at SFG will probably allow fewer actual runs than expected because they allow fewer homers than expected…but that’s not due to pitcher skill, that’s due to park.

Numbers all hypothetical:

Say the Giants allow 500 runs (with their FIP saying they should have allowed 500 runs), but after you regress to a league average HR/FB you determine they should have allowed 520 by xFIP (since we expect their HR/FB to be favorable solely due to the park), so that’s what you would say is their “true talent level”, or whatever you want to call it. On the other hand, offensively, they score 500 runs (and were expected to based on your pre-park adjusted calculations). If you’re using an exactly neutral park factor (the Giants are close to neutral, not sure it’s exact, but this is just for illustration purposes), you determine offensively, they should have scored 500 runs. So your power rankings for offense/defense would have them as a team that should score 500 runs but allow 520….yet the difference is entirely because you’re essentially saying the park helped them allow only 500 (instead of 520), but then didn’t aid them at all offensively. So that’s the main thing putting me at a loss.

A couple of things I’ll say – again, we’re only discussing my team, I’ll acknowledge that it still might, league wide, be the best way to do it. Just because there might be an exception that it doesn’t handle well doesn’t invalidate the whole system, of course. I’ll also note that things that make sense at the individual level might not always make as much sense at the team level, and vice versa.

As for the league adjustment, I understand why it’s important. My point isn’t to say we shouldn’t have an adjustment, my point is that doing it at the league level, rather than the division level, just isn’t the way we should be doing it. I do not think the errors this adds are insignificant, and especially when you start talking about the effects on individual players, I’m not sure making a general league adjustment removes any more error than it creates, which is why it bothers me to see. I guess at the team level it probably is a net positive, though – thinking about it it strikes me as one of those things that works a lot better at the team level than the individual level.