|
Post by psuhistory on Jan 29, 2015 8:26:43 GMT -5
Only 142.3 innings from Bailey in 2015: he can't even stay healthy in a simulation... Projection systems forecast the mean result of a series of outcomes. So I would imagine out of 1 million sims there were quite a lot of zero innings pitched totals for Bailey (and every pitcher for that matter) and quite a few but probably less 220+ innings pitched totals. It is worth considering that Davenport's 2015 projection for Bailey's innings most closely corresponds, among recent years, to the season in which Bailey was injured. This is part of deciding how useful the projection is--and it's not useless--not a criticism of how it's calculated...
|
|
|
Post by kinsm on Jan 29, 2015 8:50:02 GMT -5
Projection systems forecast the mean result of a series of outcomes. So I would imagine out of 1 million sims there were quite a lot of zero innings pitched totals for Bailey (and every pitcher for that matter) and quite a few but probably less 220+ innings pitched totals. It is worth considering that Davenport's 2015 projection for Bailey's innings most closely corresponds, among recent years, to the season in which Bailey was injured. This is part of deciding how useful the projection is--and it's not useless--not a criticism of how it's calculated... No doubt...I would assume his system is weighted with last years' values being worth more than two year's ago.
|
|
|
Post by redsfanman on Jan 29, 2015 8:56:14 GMT -5
In the end projections are just BS, rather than a sign of what is going to happen.
|
|
|
Post by Dustrated on Jan 29, 2015 9:16:45 GMT -5
I don't need a moms basement dwelling nerd to tell me we're gonna suck this season. We're the Reds, damnit. It's a given. He's an educated meteorologist who helped create baseball prospectus, he's probably got more money than all of us put together. My Baseball Prospectus arrives tomorrow!
|
|
|
Post by psuhistory on Jan 29, 2015 9:16:59 GMT -5
In the end projections are just BS, rather than a sign of what is going to happen. They're not meant as predictions though, and it wouldn't be surprising if those who create them aren't authorities on the individual teams. They're just more pieces of information to add to the mix of news about how the team's changing...
|
|
|
Post by redsfanman on Jan 29, 2015 9:37:19 GMT -5
In the end projections are just BS, rather than a sign of what is going to happen. They're not meant as predictions though, and it wouldn't be surprising if those who create them aren't authorities on the individual teams. They're just more pieces of information to add to the mix of news about how the team's changing... Clearly they're not authorities on individual teams, if they calculated Corky Miller's 2015 production. Same with the projection mentioned above that Holmberg and Axelrod will make ~30 starts. It's a mathematical formula calculating potential outcomes... without taking into account obvious decisions, developments (even ones as significant as retirement), and stuff like that. In the past I've found fans citing these sorts of projections like they're facts until they're disproved. Bailey, 142 innings in 2015, guilty until proven innocent! I look for decent, healthy seasons until proven otherwise. Some comparable player getting injured at a similar point in his career is totally irrelevant to anybody else. If anybody's looking for trends, check out Johnny Cueto, who's exceptional every OTHER year... NOT back-to-back years. If that trend holds 24 starts, let alone 24 decisions, seems unlikely for Cueto.
|
|
|
Post by redsfanman on Jan 29, 2015 9:39:06 GMT -5
My Baseball Prospectus arrives in mid-February, around the time players report to spring training. I ordered it along with Game of Thrones Season 4 to get free shipping. Oh well. I figured it'd give me more time to read and do other things before baseball really starts.
|
|
|
Post by kinsm on Jan 29, 2015 10:23:03 GMT -5
In the end projections are just BS, rather than a sign of what is going to happen. They are no different than any projections made about anything. Politics, demand for a new pepsi drink, movie sales, etc. As long as the imputs and formulas are sound they tend to be more accurate than not within confidence intervals. Obviously they aren't crystal balls.
|
|
|
Post by redsfanman on Jan 29, 2015 11:10:37 GMT -5
Don't get me started on political projections. Many seem to view projections of the popular vote as the definitive projection for an upcoming election, although the irrelevance of that measure was shown by Al Gore in 2000. More recently Romney's 'close' race that ignored the reality of his struggles with the electoral college, and netted him only North Carolina (by 2%).
|
|
|
Post by kinsm on Jan 29, 2015 11:23:38 GMT -5
Don't get me started on political projections. Many seem to view projections of the popular vote as the definitive projection for an upcoming election, although the irrelevance of that measure was shown by Al Gore in 2000. More recently Romney's 'close' race that ignored the reality of his struggles with the electoral college, and netted him only North Carolina (by 2%). Listen up Karl Rove, the best projection models are continuously updated with new information. The reason why Bailey's IP projection is so low is because he finished the year injured. If he's healthy going into the regular season and that is imputed into the model it'll increase his IP tally. The reason why Joe Bob and Jimbo are projected to make 30 starts (as of January 2nd) is because no one has yet locked down the 4th and 5th rotation spots - if someone does coming out of ST their GS projections will drop. There is nothing wrong with their being a projection for Corky Miller, just as steamer does projections for Jim Thome, Barry Bonds, and the like. It's a what might they do...not what they are going to do. It's why fangraphs makes continuous team W/L total (RS/RA) projections throughout the year. If Votto breaks his leg tomorrow then his WAR contribution will decline in the next projection and so forth will Reds overall win totals. A better sample (i.e. health come opening day, whether a player has been cut from major league camp, etc.) normally will lead to a better estimate of the population parameter. It's a best guesstimate as of today (actually 1/2/15). When the weather man tells you it'll be 35 degrees a week from now, he's not saying bet on it...he's saying that's a best estimate with the current known variables within a reasonable error (like previous years temps for that day, distant weather patterns, etc.).
|
|
|
Post by redsfanman on Jan 29, 2015 11:28:15 GMT -5
Don't get me started on political projections. Many seem to view projections of the popular vote as the definitive projection for an upcoming election, although the irrelevance of that measure was shown by Al Gore in 2000. More recently Romney's 'close' race that ignored the reality of his struggles with the electoral college, and netted him only North Carolina (by 2%). Listen up Karl Rove, the best projection models are continuously updated with new information. The reason why Bailey's IP projection is so low is because he finished the year injured. If he's healthy going into the regular season and that is imputed into the model it'll increase his IP tally. The reason why Joe Bob and Jimbo are projected to make 30 starts (as of January 2nd) is because no one has yet locked down the 4th and 5th rotation spots - if someone does coming out of ST their GS projections will drop. There is nothing wrong with their being a projection for Corky Miller, just as steamer does projections for Jim Thome, Barry Bonds, and the like. It's a what might they do...not what they are going to do. It's why fangraphs makes continuous team W/L total (RS/RA) projections throughout the year. If Votto breaks his leg tomorrow then his WAR contribution will decline in the next projection and so forth will Reds overall win totals. A better sample (i.e. health come opening day, whether a player has been cut from major league camp, etc.) normally will lead to a better estimate of the population parameter. It's a best guesstimate as of today (actually 1/2/15). When the weather man tells you it'll be 35 degrees a week from now, he's not saying bet on it...he's saying that's a best estimate with the current known variables within a reasonable error (like previous years temps for that day, distant weather patterns, etc.). Karl Rove? OUCH! Talk about an insult!!!!
|
|
|
Post by Dustrated on Jan 29, 2015 12:22:38 GMT -5
Don't get me started on political projections. Many seem to view projections of the popular vote as the definitive projection for an upcoming election, although the irrelevance of that measure was shown by Al Gore in 2000. More recently Romney's 'close' race that ignored the reality of his struggles with the electoral college, and netted him only North Carolina (by 2%). Listen up Karl Rove, the best projection models are continuously updated with new information. The reason why Bailey's IP projection is so low is because he finished the year injured. If he's healthy going into the regular season and that is imputed into the model it'll increase his IP tally. The reason why Joe Bob and Jimbo are projected to make 30 starts (as of January 2nd) is because no one has yet locked down the 4th and 5th rotation spots - if someone does coming out of ST their GS projections will drop. There is nothing wrong with their being a projection for Corky Miller, just as steamer does projections for Jim Thome, Barry Bonds, and the like. It's a what might they do...not what they are going to do. It's why fangraphs makes continuous team W/L total (RS/RA) projections throughout the year. If Votto breaks his leg tomorrow then his WAR contribution will decline in the next projection and so forth will Reds overall win totals. A better sample (i.e. health come opening day, whether a player has been cut from major league camp, etc.) normally will lead to a better estimate of the population parameter. It's a best guesstimate as of today (actually 1/2/15). When the weather man tells you it'll be 35 degrees a week from now, he's not saying bet on it...he's saying that's a best estimate with the current known variables within a reasonable error (like previous years temps for that day, distant weather patterns, etc.). Very well explained but do you honestly think he will comprehend this? His main attacks on these are only because he can't wrap his mind around such things. He doesn't understand them. His brain isn't and will never be wired that way.
|
|
|
Post by psuhistory on Jan 29, 2015 12:37:33 GMT -5
Don't get me started on political projections. Many seem to view projections of the popular vote as the definitive projection for an upcoming election, although the irrelevance of that measure was shown by Al Gore in 2000. More recently Romney's 'close' race that ignored the reality of his struggles with the electoral college, and netted him only North Carolina (by 2%). The best projection models are continuously updated with new information. The reason why Bailey's IP projection is so low is because he finished the year injured. If he's healthy going into the regular season and that is imputed into the model it'll increase his IP tally. The reason why Joe Bob and Jimbo are projected to make 30 starts (as of January 2nd) is because no one has yet locked down the 4th and 5th rotation spots - if someone does coming out of ST their GS projections will drop. It's a best guesstimate as of today (actually 1/2/15). Very clear explanation of why the limitations of these projections are often the most interesting discussion points for those who actually follow these teams. The less you know about the local climate, the greater your dependence on the weather report...
|
|
|
Post by redsfanman on Jan 29, 2015 13:36:11 GMT -5
Very well explained but do you honestly think he will comprehend this? His main attacks on these are only because he can't wrap his mind around such things. He doesn't understand them. His brain isn't and will never be wired that way. Me's Dustrated, me's smart. Ho ho ho.
|
|
|
Post by general on Jan 29, 2015 18:40:06 GMT -5
I don't need a moms basement dwelling nerd to tell me we're gonna suck this season. We're the Reds, damnit. It's a given. He's an educated meteorologist who helped create baseball prospectus, he's probably got more money than all of us put together. Still, to kramer's point, anyone can tell us the Reds outlook for 2014 isn't bright.
|
|