Schoenfield's ESPN Sweet Spot featuring the O's today

[Lucky Jim]

Because its also being assumed that the replacements will be no better than who they are replacing, when its all said and done.

Yes. And this is a limit to the predictive power of the formula - to apply it predictively, you have to assume that over however many games, a team's inputs will remain roughly the same (i.e., that they're indicative of some true talent). This is a relatively safe assumption: the addition of individual players usually results in a fairly small bump in run production (there aren't many Mike Trouts); it is also (relatively) safe to assume that a team will send its best MLB-level pitchers to the mound to start the season, or those individuals will rise up quickly (though there is also room for a Mike Trout pitcher, I would guess).

What we're asking by judging fairness is that pythag take into account that we're replacing something like 60% of our rotation with better MLB pitchers. If this is true, then pythag will be of limited predictive value - of course, pythag generally recognizes that because better pitching would result in a change in differential. But no one has said anything else - they've just said that it's a low-probability bet that we'll find the kind of substantial internal improvement needed to contend. By making the argument that it's unfair, we're essentially saying that pythag should optimistically take into account that the franchise couldn't field more than two decent starting pitchers for something like 65% of the season.

What's amazing to me - and this has been stated elsewhere - is that we're arguing that pythag can't account for the awesomeness of Chris Tillman and Miguel Gonzalez when one was a pariah not two months ago and the other a complete unknown.

[osfan83]

Fairness in terms of judging the outlook for the rest of the season.

Essentially what is happening is that people are "predicting" how the rest of the season will go based on the performances of players that are not on the team anymore. Doesn't add up to me.

This is what I said earlier. If the O's signed 5 new starters for 2013, would we base our expection of performance based the the 2012 rotation that is no longer on the team? Why are the Pirates a good team now...they sucked 5 years ago! Runs allowed by pitchers no longer on the team should not be factored in to what we expect these pitchers to do.

[Lucky Jim]

I don't think we really need to worry about other teams right now because the Orioles are statistical anomaly. There really is no sense in comparing.

This makes no sense. Literally, none.

[Dark Helmet]

Well .... sure. I think everyone understands that. I think everyone understands the historical correlation between RS and RA.

I see about a third of the posts here explaining the pythag, a third mis-interpreting them, and a third saying outrunning the pythag by a wide margin is highly unusual.

I see little attempt to explain why we are winning at a detailed level of analysis despite being outscored by a comfortable margin other than "good bullpen, lucky in close games".

Really the only explanation is in the blowout games win or lose, we've had it taken to us in the losses, and haven't given it out as much in the wins. What would our odds be if we even up in the run differential department? Would it greatly change our chances either way that much?

[Mondo Trasho]

This is what I said earlier. If the O's signed 5 new starters for 2013, would we base our expection of performance based the the 2012 rotation that is no longer on the team? Why are the Pirates a good team now...they sucked 5 years ago! Runs allowed by pitchers no longer on the team should not be factored in to what we expect these pitchers to do.

As far as I can tell, yes. That's exactly how you use the predictive models.

[Pedro Cerrano]

I'm sorry but Britton Matusz and Arrieta have pitched to a 6.07 ERA over 209 innings this year. You'll have a hard time convincing me that replacing those guys, even with unprovens like Tillman and Gonzalez, won't do wonders for our overall run differential.

A 6.07 ERA over 209 innings is positively awful.

[hoosiers]

I agree it is unsatisfactory. What you're actually looking for is the model that explains how outliers to the Pythag happen. This model may or may not exist, but once again it wouldn't prove the Pythag to be wrong, it would just better explain how one might over perform their Pythagorean.

I am not looking for a new model to explain outliers to the Pythag. I am not advocating that DD and BS have cooked up a secret sauce to goose the Pythag. I am not mocking what Lucky Jim and Drungo (one of the very best posters on the OH as I have posted before) bring to the board or this particular discussion.

I am looking for more effort into an analysis (one admittedly that I am not performing) to explain our outperformance of the pythag other than "good bullpen, luck".

Yesterday, I asked what our record would be if we had scored one more run in every game (plus one more run in half of the games we lost by one run to create a 50/50 winning percentage in our actual one-run losses). If we were granted an additional 110-115+ runs scored, what does that do to our pythag? What does it do to our actual record? No response. My guess is that we would still be outperforming our pythag by a margin that would considered unsustainable as well. Go figure - 110 free runs and we still might be considered lucky - which might hint to just how much of an outlier we really are and just how little real explanation we have for our record so far this year.

[Pedro Cerrano]

Plus we're leaving the human element and other intangibles out of the equation. When a team is down 10-0, do they keep fighting? Do they say "hey guys, let's try to score a handful of runs here so our run differential stays in line with our record!"

Nah, what usually happens is the team quits, and this is illustrated by the fact that mangers often empty out their benches and the guys who have been tasked as "mop-up duty" men (Gregg) get to get battered for a few more innings.

[Lucky Jim]

Plus we're leaving the human element and other intangibles out of the equation. When a team is down 10-0, do they keep fighting? Do they say "hey guys, let's try to score a handful of runs here so our run differential stays in line with our record!"

Nah, what usually happens is the team quits, and this is illustrated by the fact that mangers often empty out their benches and the guys who have been tasked as "mop-up duty" men (Gregg) get to get battered for a few more innings.

And this is different than other teams how?

[Pedro Cerrano]

And this is different than other teams how?

A good point.

My only counter would be that perhaps the Orioles are more prone to blow outs when they lose (partially because 3/5 of their rotation sucked) and thus their RD is further exacerbated, but now we're getting too far into speculation than even I care for.

[DrungoHazewood]

This makes no sense. Literally, none.

That's the point. The 2012 Orioles have transcended the normal rules of the universe. We're arguing logic based on physics and other properties as science understands them. Silent James is telling us to stop. Because those rules don't apply any more.

[Mondo Trasho]

Perhaps the injuries to other teams in our division (and the AL as a whole) has helped us beat our pythagorean number. It has been a strange season for injuries too, IIRC.

[Lucky Jim]

A good point.

My only counter would be that perhaps the Orioles are more prone to blow outs when they lose (partially because 3/5 of their rotation sucked) and thus their RD is further exacerbated.

Right. Once again, we're arguing that, because the Orioles perform horribly more often than other teams that they're actually a pretty good team.

[Lucky Jim]

Perhaps the injuries to other teams in our division (and the AL as a whole) has helped us beat our pythagorean number. It has been a strange season for injuries too, IIRC.

Why would that effect wins and not runs?

[Mondo Trasho]

Right. Once again, we're arguing that, because the Orioles perform horribly more often than other teams that they're actually a pretty good team.

Or conversely, they pick and choose their battles more carefully than other teams. They know when to hold 'em and when to fold 'em.