SoxProspects News
|
|
|
|
Legal
Forum Ground Rules
The views expressed by the members of this Forum do not necessarily reflect the views of SoxProspects, LLC.
© 2003-2024 SoxProspects, LLC
|
|
|
|
|
Forum Home | Search | My Profile | Messages | Members | Help |
Welcome Guest. Please Login or Register.
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,914
|
Post by ericmvan on Apr 6, 2016 12:56:25 GMT -5
This may be the craziest thing I've ever come up with. Those who believe in Statistical Correctness, the belief that small sample sizes are always meaningless regardless of effect size, should pretend that I'm kidding. And I think I halfway am. (The SC have my permission to be appalled and terrified by the half of me that's taking this seriously.)
But this begins with a pure scouting assessment, combined with memory. The two things we love to see David Ortiz do, the two markers that he is on his game, are hit home runs, and drive the ball to the opposite field. My next thought was that the opposite field home-runs should be lumped with the opposite-field doubles as the same phenomenon, just done better.
And then I remembered that when Ortiz was struggling in April, he wasn't doing one or both of these things. I could remember Remy, for instance, pointing out a ball hit hard to the opposite field as a first good sign that Papi may be getting back on track when he'd been struggling for a while.
So this suggested a little study. Each year beginning in 2004 when he started the season as a regular, count how many PA it took Ortiz to do each of these things:
1) Hit a HR to straight-away CF or to RF.
2) Hit a 2B, 3B, or HR to LF or left-center. (There were no triples. Gee.)
My initial hypothesis was that there might be a correlation between how long it took him to accomplish both of these feats, and his wRC+ for the season. But as I started to crunch the numbers, it also occurred to me that I should also look at the correlation for each of the things separately, and for how long it took him to do one or the other.
Here's the raw data (there is no RF / CF / LF data for HR online before 2008, but I had every relevant game from 2004-2007 in my scorecards).
Year Opp R/C-HR wRC+
2004 30 12 147
2005 12 6 157
2006 35 3 157
2007 1 24 175
2008 74 17 124
2009 48 164 100
2010 14 46 134
2011 4 5 154
2012 1 34 170
2013 9 20 151
2014 112 7 134
2015 31 17 138
2016 2 5 ?
And here are the regressions results, correlating each thing to wRC+ for the season. The only comment I will make is that there is no cherry-picking here; this is all the possible tests given my original observation and decision as to what the dependent variables should be (plus logarithmic variants for two where the data fit that better). "CI" is confidence interval.
What r p 95% CI for wRC+
PA to first HR to RF/CF .66 .02 138
PA to first XBH to LF or LCF .60 .04 139
Same, logarithmic .76 .004 142
PA to accomplish both .81 .001 148
PA to accomplish one or the other .89 .00009 149
Same, logarithmic .91 .00005 150
Multiple regression on HR and ln(XBH) .95 .00003 157
The obvious next question is, how much of this is explained by his quick start simply improving the overall numbers? So what I'll do tonight is correlate the data to his wRC+ from July 1 onward.
|
|
|
|
Post by ramireja on Apr 6, 2016 13:04:55 GMT -5
You know, I thought this was a pretty interesting tidbit from Alex Speiers reports today: "Ortiz went 2-for-4 with a ringing double (106 miles per hour exit velocity off the bat) just to the left of center field and a mighty homer (109 m.p.h.) to right. He also had a near-homer (105 m.p.h.), a deep fly ball to right that likely would have found its way into the seats but for the artic 34-degree conditions to open the game. Ortiz also received an intentional walk.
He demolished three pitches – even as he stayed back on the ball and proved capable of staying in the middle of the field on the double. The performance was noteworthy. Last year, after all, Ortiz had just one game in which he scalded three pitches at 105 m.p.h. or more, a two-homer performance in August."
|
|
|
|
Post by brianthetaoist on Apr 6, 2016 13:56:32 GMT -5
This is Peak Eric Van, and I, for one, think it's AWESOME.
|
|
|
|
Post by chrisfromnc on Apr 6, 2016 14:08:01 GMT -5
This is Peak Eric Van, and I, for one, think it's AWESOME. +1.
This is too good, Eric. Meaningless, I guess, but freakin' brilliant. To quote Kenny Banya, "That's Gold!"
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,914
|
Post by ericmvan on Apr 6, 2016 14:30:56 GMT -5
The Speier stuff is golden. And does anyone doubt that a) wanting to go out on top, and knowing that he'll never have to work out all winter again, Papi would get himself into you-know-what for his swan song, and that b) he wouldn't talk about it?
It gets better on my end, too, maybe. Can we predict his wRC+ from July 1 on, from something as simple as, how long in the season did it take him to either hit a homer, or hit a 2B to the opposite field?
r = .80, p < .002.
Year PA Pred Actual
2004 12 147 150
2005 6 160 165
2006 3 173 184
2007 1 194 187
2008 17 141 133
2009 48 121 117
2010 14 144 128
2011 4 168 153
2012 1 194 204
2013 9 152 138
2014 7 157 151
2015 17 141 182
2016 2 181 ?
And far and away the outlier here is 2015. Exclude that, and you get r = .94, p = .000015.
Year PA Pred Actual Diff
2004 12 141 150 9
2005 6 157 165 8
2006 3 172 184 12
2007 1 195 187 -8
2008 17 134 133 -1
2009 48 111 117 6
2010 14 138 128 -10
2011 4 165 153 -12
2012 1 195 204 9
2013 9 148 138 -10
2014 7 153 151 -2
2015 17 134 182 48
2016 2 180
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,914
|
Post by ericmvan on May 11, 2016 0:28:53 GMT -5
Well, after precisely 1/5 of the season, Papi is rocking a 198 wRC+, and needs just a 147 the rest of the way to satisfy the initial 95% confidence that he would post a 157 or better ... based on his Opening Day performance.
Meanwhile, PECOTA projected him at .299 TAv, and his 20% of the season at .340 has boosted their rest-of-season projection all the way to ... .302.
Some other original, to date, and ROS projections from PECOTA (excluding last night, of course)
.292, .242, .288 Betts
.274, .264, .275 Pedroia
.266, .292, .268 Bogaerts
.280, .264, .279 Ramirez
.257, .291, .261 Shaw
.252, .273, .255 Holt
.254, .297, .257 Bradley
Something tells me their ROS system is completely non-player-specific. What Bradley and Shaw have done in their SSS is much more meaningful than what Ortiz has done, but they've all gotten the same degree of re-assessment.
|
|
|
|
Post by costpet on May 11, 2016 5:56:15 GMT -5
I think this is the reason for Ortiz's hot bat this year. He's relaxed. Before, he was playing for a contract. Now, he's decided to just have fun. No pressure, just go up and hit and laugh. Baseball be best played relaxed. It's not like football where you have to get all charged up. If you get too intense, you won't play well. He's having more fun this year than he ever did. That's why every at bat is just a pleasure.
On the other hand, maybe sometime in October Dom. might saddle up to him and whisper "$30 next year?" Could happen.
|
|
|
|
Post by voiceofreason on May 12, 2016 7:18:30 GMT -5
I think this is the reason for Ortiz's hot bat this year. He's relaxed. Before, he was playing for a contract. Now, he's decided to just have fun. No pressure, just go up and hit and laugh. Baseball be best played relaxed. It's not like football where you have to get all charged up. If you get too intense, you won't play well. He's having more fun this year than he ever did. That's why every at bat is just a pleasure. On the other hand, maybe sometime in October Dom. might saddle up to him and whisper "$30 next year?" Could happen. I think their is a lot of merit to this thought, he is off to an incredible start. I really like the idea of him coming back. |
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,914
|
Post by ericmvan on May 12, 2016 9:39:11 GMT -5
I think this is the reason for Ortiz's hot bat this year. He's relaxed. Before, he was playing for a contract. Now, he's decided to just have fun. No pressure, just go up and hit and laugh. Baseball be best played relaxed. It's not like football where you have to get all charged up. If you get too intense, you won't play well. He's having more fun this year than he ever did. That's why every at bat is just a pleasure. On the other hand, maybe sometime in October Dom. might saddle up to him and whisper "$30 next year?" Could happen. Makes no sense to me ... how could a guy with his track record of clutch performance have any problem with relaxation or excess intensity? The simpler explanation is that he worked out harder than ever last winter because he knew he'd never have to do it again, and because he really wanted to go out on top. He has said that his longevity is the product of a tremendous off-season work regimen, and that the only reason he is retiring is that he wants to finally stop doing it. But one last time tackling something hard, that you have mastered? If it's just a matter of effort, and you're proud, you do it better than ever. |
|
|
|
Post by okin15 on May 12, 2016 11:20:46 GMT -5
I can certainly see him coming back for an abbreviated season a-la Clemens. No reason not to make him the richest contract offer in the game, but then pay a pro-rated version that starts on May 10.
|
|
|
|
Post by ancientsoxfogey on May 12, 2016 12:00:46 GMT -5
I can certainly see him coming back for an abbreviated season a-la Clemens. No reason not to make him the richest contract offer in the game, but then pay a pro-rated version that starts on May 10. I see two things problematic about this: (1) If his real issue is working out to prepare for the season rather than simply playing the games, why should he be extra motivated to return because of an offer to start the season late? Simply sit him against more LH and get him more consistent rest that way. (2) If he were to play a short season -- without spring training -- who's to say he wouldn't slack off in preparation, especially if, as postulated, he may have geared up for this season thinking it would be his last go-round? If he did slack off, that is precisely when, for a 40+ player, he'd lose the edge that has made him so great, and he'd be just another player, or worse, take that sudden plummet to oblivion. I'd prefer not to even think about this. If Ortiz wants to retire after this season and he ends it with a top-flight year (as it certainly appears he might), say thanks for the memories, let him go with nothing but best wishes, and move on as an organization. All things end, an an icon like Ortiz has earned the right to end on his terms. |
|
|
|
Post by okin15 on May 12, 2016 13:07:16 GMT -5
I hear you on those ASF. But I'm clinging to him for as long as I can. I don't think Papi would agree to show up for the season and not show up in shape.
|
|
|
|
Post by Coreno on May 12, 2016 18:40:15 GMT -5
I just don't want him to Favre it
|
|
|
|
Post by soxfan06 on May 12, 2016 20:41:02 GMT -5
I hear you on those ASF. But I'm clinging to him for as long as I can. I don't think Papi would agree to show up for the season and not show up in shape. I don't think Papi would agree to show up or the season and not show up for 162 games. Don't see him as a half year guy. He seems like the type whose in or out. I am also clinging to the hope that he decides to not retire. |
|
|
|
Post by larrycook on May 12, 2016 21:19:08 GMT -5
Julio Franco played this game a billion years because his bat speed and mechanics never declined.
As long as ortiz's bat speed and mechanics remain, he will be a dangerous hitter.
Watching Ortiz coming out of the batters box, you have to wonder if his Achilles heels don't one day pop.
|
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,914
|
Post by ericmvan on Jul 13, 2016 23:52:22 GMT -5
And here are the regressions results, correlating each thing to wRC+ for the season. The only comment I will make is that there is no cherry-picking here; this is all the possible tests given my original observation and decision as to what the dependent variables should be (plus logarithmic variants for two where the data fit that better). "CI" is confidence interval. What r p 95% CI for wRC+
Multiple regression on HR and ln(XBH) .95 .00003 157I didn't report the actual projection here because it seemed so wildly optimistic. But it was in fact 170. He's at 182 at present. What I should have done is report that, and the upper 95% confidence interval. Which is .... 182. It gets better on my end, too, maybe. Can we predict his wRC+ from July 1 on, from something as simple as, how long in the season did it take him to either hit a homer, or hit a 2B to the opposite field? r = .80, p < .002. Year PA Pred Actual
2016 2 181 ?And far and away the outlier here is 2015. Exclude that, and you get r = .94, p = .000015. Year PA Pred Actual Diff
2016 2 180 That prediction of 180 from 7/1 on seemed even crazier than 170 for the season, but all he'd have to do to meet it is keep hitting precisely as he has. The 95% confidence interval is 164 to 197. |
|
ericmvan
Veteran
Supposed to be working on something more important
Posts: 8,914
|
Post by ericmvan on Oct 5, 2016 15:01:10 GMT -5
Well, he did tail off from July 1 on .. to only a 143 wRC+. (In fact, he was 200 on June 3 and was 143 afterwards. That's the wear and tear.)
The 95% confidence interval for the season, based on his first two games in the context of his career, was 157 to 183. He ended up with 163.
Projections based on his entire career previous to those 2 games: 126 ZiPS, 144 Steamer, 128 fans at FG.
Feel free to use this in the future as an example of a how adding 10 properly evaluated PA to a sample of 9465 can hugely improve a projection. It's a microscopic sample, but it meant something, and the meaning started with a scouting observation based on 13 years of observation. This is why the entire scouts vs. analysts "opposition" is nonsense.
|
|
|
