July 3-4 Welcome home: Flip the bat vs Flip the Bird

[Schad]

But yeah, it's not just Donaldson, although that was a huge loss. It's that Donaldson along with 3/5 of our rotation (Stroman/Sanchez/Garcia) being well below what was expected of them. Between those 4 players alone, we're 5.3 WAR short of what they were projected to accrue through the first half of the season. Then we've been missing our closer for the better part of the year. And we haven't had anybody play over their head to balance it out, really. Teoscar has been good, sure, but our overall OF production is right about where we should have

Certainly, but it's for guys like Sanchez/Garcia that I really wish they published error bars on those projections. Sanchez's performance to date isn't all that surprising; given that he has been both a Cy Young contender and a complete mess in recent years, he came into the year with a range of reasonable outcomes as wide as Bartolo Colon's belt. Same with Garcia, whose injury/inconsistency issues (with his underperformance dating back to the second half of last year, when his walk rate shot up and he got shelled regularly) are the reason he was available cheaply.

Stroman would be well outside of his 80% range, however, but the pitchers, they break.

[Skin Blues]

Error bars or not, it's still an average projection. Half the time, they'd be better than the projection too. As with Grichuk's luck in April, high volatility can go one way or another to extreme degrees. But that doesn't mean the projection or measure of talent was wrong, even though it's often explained away after the fact with a narrative ("he hits too many fly balls, we should have known he'd suck", "he has no control of his pitches, we should have known he'd get much worse", "he's old so we should have known he'd deteriorate even faster", etc).

So yeah, Sanchez was volatile. He was projected for 2 WAR, but could have been a 4 WAR pitcher like a couple years ago, or a 0 WAR pitcher like last year. Garcia could have been a 3 WAR pitcher or a negative-WAR pitcher. To have all of our luck swing one way is just that - luck. Overall, volatility isn't a bad thing as long as you have enough to field a roster without having a bunch of negative-WAR guys given significant playing time. We probably stuck with Garcia too long, and our pitching depth is obviously really thin to begin with. But no matter how well equipped we were with depth, there's not much chance for you when the WC2 team is sitting at 55-32.

[Schad]

Error bars or not, it's still an average projection. Half the time, they'd be better than the projection too. As with Grichuk's luck in April, high volatility can go one way or another to extreme degrees. But that doesn't mean the projection or measure of talent was wrong, even though it's often explained away after the fact with a narrative ("he hits too many fly balls, we should have known he'd suck", "he has no control of his pitches, we should have known he'd get much worse", "he's old so we should have known he'd deteriorate even faster", etc).

So yeah, Sanchez was volatile. He was projected for 2 WAR, but could have been a 4 WAR pitcher like a couple years ago, or a 0 WAR pitcher like last year. Garcia could have been a 3 WAR pitcher or a negative-WAR pitcher. To have all of our luck swing one way is just that - luck. Overall, volatility isn't a bad thing as long as you have enough to field a roster without having a bunch of negative-WAR guys given significant playing time. We probably stuck with Garcia too long, and our pitching depth is obviously really thin to begin with. But no matter how well equipped we were with depth, there's not much chance for you when the WC2 team is sitting at 55-32.

It's an average projection, but that's what I'm getting at: it's not half the time that he'd be better/worse, necessarily. With Sanchez, the top end of his 80% confidence interval would probably be that 4 WAR pitcher, but it's quite likely that the majority of the simulation results would cluster below the actual projection, in the 1-1.5 WAR range (about where he is now); you'd get a situation where the long tail on the happy side would elevate the average a fair bit above the median. Hence my desire for error bars; they can sometimes tell more of a store with volatile projections.

[BigLeagueChew]

Stroman doing his Jon Gray impression. 6.50era/4.48fip/3.89xfip

Stroman's is a bit more explicable than Gray's. One of the highest percentages of hard-hit balls in baseball, and one of the highest pull percentages in baseball. There's absolutely bad luck in there (his strand rate, mostly) but he was giving up an average exit velo of 91.3 MPH entering today...we're going to need to put helmets on the infielders when he's pitching.

I recall talking about Stroman before his injury, saying something like he should throw outside more and more fastballs. After more research,it doesn't appear that would help very much , after looking at his exit velocity heat maps.

Am also looking for the pitching chart on baseball savant for last game , have started using it more recently but am unfamiliar with the site, do you know how long that takes to show up?

[Skin Blues]

Error bars or not, it's still an average projection. Half the time, they'd be better than the projection too. As with Grichuk's luck in April, high volatility can go one way or another to extreme degrees. But that doesn't mean the projection or measure of talent was wrong, even though it's often explained away after the fact with a narrative ("he hits too many fly balls, we should have known he'd suck", "he has no control of his pitches, we should have known he'd get much worse", "he's old so we should have known he'd deteriorate even faster", etc).

So yeah, Sanchez was volatile. He was projected for 2 WAR, but could have been a 4 WAR pitcher like a couple years ago, or a 0 WAR pitcher like last year. Garcia could have been a 3 WAR pitcher or a negative-WAR pitcher. To have all of our luck swing one way is just that - luck. Overall, volatility isn't a bad thing as long as you have enough to field a roster without having a bunch of negative-WAR guys given significant playing time. We probably stuck with Garcia too long, and our pitching depth is obviously really thin to begin with. But no matter how well equipped we were with depth, there's not much chance for you when the WC2 team is sitting at 55-32.

It's an average projection, but that's what I'm getting at: it's not half the time that he'd be better/worse, necessarily. With Sanchez, the top end of his 80% confidence interval would probably be that 4 WAR pitcher, but it's quite likely that the majority of the simulation results would cluster below the actual projection, in the 1-1.5 WAR range (about where he is now); you'd get a situation where the long tail on the happy side would elevate the average a fair bit above the median. Hence my desire for error bars; they can sometimes tell more of a store with volatile projections.

I would be amazed if it's much different than a normal distribution.

[Schad]

I would think that would be a fairly normal distribution for players with Sanchez's combination of performance and injury history. It's not like he entered the year with rosy projections; Steamer had him at 1.7 fWAR over 139 IP, with a 4.44 FIP, and their revised projection has him at 1 WAR in 130 IP with a 4.65 FIP.