Can’t believe Boston is at the bottom.

Team P C 1B 2B 3B SS LF CF RF Total
Baltimore Orioles 2 -4 -8 2 -3 9 -8 3 -4 -11
Minnesota Twins 2 -2 -5 -15 11 -7 6 3 -7 -14
KC Royals 4 -2 1 -11 -7 -5 8 9 -13 -16
Florida Marlins -15 -2 -2 -4 1 1 -4 3 5 -17
Boston Red Sox -7 -4 7 1 -20 -10 -1 -2 4 -36

If the Rangers were like last year and ranked near the bottom, say where Florida is, the difference in runs allowed would be 53, which would put them at 432 runs in 82 games. At that rate, they would have given up 506 runs through 96 games, which is the number of games before the ASB in 2008 when they gave up 515.

With a team +/- ranking similar to 2008, the improvement in the starting rotation would’ve been basically nullified, and it speaks to the large randomness of baseball statistics that clear improvements in BB/9 by the starters but not in other stats (HR/9 and K/9 or defensive efficiency) could have had a very small impact on the bottom line number of runs allowed.

One thing that isn’t quantified well by this number parsing is the sustainability of the pitching staff’s performance. I think the big improvement of the pitching staff may not be in the bottom line so much as in the staying power of their performance.

and I’m wondering if the pitchers aren’t more comfortable throwing strikes and not nibbling so much with the solid defense behind them. The accountability stance from Ryan might help in this area as well.

But things like this are why I don’t really like to hang my hat on PECOTA or the like. Not that it doesn’t have a place and isn’t interesting to look at, but I don’t know why some people look at it as the preseason gospel.

I think you blame the application of PECOTA in a situation with so many unknowns. Not to say that you can’t trust projection systems, but that you should also recognize the variability of their accuracies when you’re talking about a brand new 20-year-old shortstop and a new third baseman as well.

I also find it hard to discount Kinsler’s ridiculous improvement between last year (-7.3) and this year (+5.4). Of all the things Evan noted, I was surprised to see him not mention that in the realm of “unexpected improvements”.

His reasoning was because they pitch to contact, just like our philosophy.

and am pleasantly surprised that we are seeing the level of improvement we are primarily based on fielding differences. Realistically the pitching really just isn’t any different from last year.

I’m still waiting for AJM to explain what the RISP line should look like for Millwood or anyone else.

if it is sustainable over many years.

Most things attributed to luck just don’t get sustained very long, so if a player had control over it, they’ve forgotton how to control it.

We apply it to baseball quite heavily because of the amount of skill refinement at the major league level.

I often use this comparison with my friends: how many minor leagues does the NFL have? No official ones. How many does basketball have? One. Hockey? One. Baseball? Six. Also, think about the repeatable actions taken in baseball (that is, pitching and hitting mechanics) and the length of the season. Baseball seasons are roughly twice as long as basketball/hockey seasons, and ten times the length of football seasons.

We aren’t saying that no variability exists (we are still talking about humans, even given what I just said), just that it isn’t statistically significant outside of luck/small sample size.

don’t postulate that athlete are robots with no varience.

If a person is always trying to do something, and is unusually good at it for 3 months, and average at it over his whole career, he is having a lucky 3 months. Or you can say it is a small sample size anomaly if you don’t like luck. Its all the same.