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.
Update: Red Sox sign Napoli for one year, $5m
|
Post by jmei on Dec 4, 2012 22:23:43 GMT -5
|
|
|
Post by Oregon Norm on Dec 5, 2012 16:34:13 GMT -5
Thanks for posting. I was going to dig it up and this saved me the trouble. Coming to you from the great northwest
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 5, 2012 17:31:39 GMT -5
I am highly skeptical of that study.
At the very least you wouldn't expect there to be a difference for every player. What we really care about is if Mike Napoli will perform better as a 1B than as a catcher as Blue Chip's thesis suggests.
Fortunately the data is all available for download on fangraphs and we can test this thesis. The results? Not what you would expect.
For his career Napoli has performed BETTER as a catcher than on days he played at other positions.
W/OBA as catcher .373 W/OBA as non-catcher .353
Again I really don't buy the idea that Mike Napoli is going to magically turn into a better player just by putting away his tools of ignorance. It hasn't happened throughout his career, and it probably won't happen now just because he signed with our favorite team.
|
|
|
Post by jmei on Dec 5, 2012 17:54:26 GMT -5
Are we looking at the same thing? Here's what I see ( link): As a catcher: 1667 PAs, .265/.362/.516, .377 wOBA (.307 BABIP, 11.9% BB, 25.2% K, .252 ISO) As a first baseman: 414 PAs, .239/.328/.473, .348 wOBA (.265 BABIP, 10.1% BB, 24.6% K, .234 ISO) As a DH: 154 PAs, .286/.390/.578, .414 wOBA (.330 BABIP, 14.8% BB, 24.2% K, .292 ISO) He's hit worse as a first baseman but a lot of that appears to be BABIP bad luck. Given the very limited sample size here, it's hard to make any firm conclusions or projections based on his past performance. I agree with moonstone that I don't see Napoli being appreciably better going forward now that he's moved off 1B. I just happen to think that maintaining his career performance should be more than good enough
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 5, 2012 18:02:30 GMT -5
For forcasting purposes you should pretend that extreme outliers didn't occur unless you have some reasonable basis to assume that the will reoccur. Using a three year sample isn't cherry picking at all. However using a data set that is skewed by one data point implicitly assumes that the data point will reoccur in the next three years. Just because people refer to a year like Napoli's 2011 as an outlier doesn't mean it would actually pass the test for one in a statistical sense. You can't just chuck it out because it looks different to you. That's not just foolish, but intellectually dishonest. Since you called me a "foolish" and "intellectually dishonest" I decided to actually test how often we should expect a player with the talent level of a .371 W/OBA player to perform as Napoli did in 2011. The answer is very rarely. Mike Napoli's season represented a performance slightly more 2 standard deviations above the mean. For some perspective in every other season, Napoli's performance was within 1 standard deviation of the mean. You would expect Napoli to have such a season by random chance once every 50 years. Unless there is some reasonable basis as to why this will reoccur, any responsible forecast of future performance should exclude this performance or at least give it significantly less weight. If you are using the average of the the last three years as a forecast for the future as jmei did. You are implicitly assuming that a player of Napoli's talent level will reach that level 31% of the time. If you use his career averages you are assuming that he'll reach the level 16% of the time. The evidence suggests that in fact, Napoli is a .350s W/OBA type hitter as opposed to a .380s W/OBA type hitter. That makes him roughly average. And finally I would suggest if you are going to use the type of language that you used here that you actually do the work necessary to test
|
|
|
Post by fenwaythehardway on Dec 5, 2012 18:20:57 GMT -5
Again I really don't buy the idea that Mike Napoli is going to magically turn into a better player just by putting away his tools of ignorance. It hasn't happened throughout his career, and it probably won't happen now just because he signed with our favorite team. It's not magic at all. Catcher is the most physically taxing position on the diamond by a wide margin; it's particularly hard on the lower body. Why wouldn't that carry over to hitting?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 5, 2012 18:21:12 GMT -5
My numbers might be slightly different because I gave him credit for the entire game at the position he played the most that day. For instance if he played eight innings at 1B and moved to C in the 9th. I'd give him credit for all of his ABs as a 1B. Implicitly that makes sense. If you buy the thesis that he would hit better when he didn't catch, a hit he after playing one inning at catcher would be the result of his rest that day.
In case I wasn't clear I am not implying that his bat plays better as a catcher for some reason. Given the sample sizes you are correct, the difference could easilly be explained by random chance.
This is where we get into how to treat the 2011 season which is a two standard deviation event and I explained in more detail in the next post.
Yes in theory a below average fielding 1B who's talent level was in the .350s W/OBA wise could be worth close to $40M over three years even if he only gets 500 PAs a year. However then we get into the risks involved which I think you stated quite well in your earlier post.
All and all I'd worry about giving a bad bodied, average hitting 1B, on the wrong side of 30 a three year deal especially when he's coming off a season where he struck out 30% of the time.
|
|
|
Post by Oregon Norm on Dec 5, 2012 18:30:24 GMT -5
Am I the only one who doesn't like the Napoli signing? Can't stay healthy, awful 1st baseman who makes a very good infield D suspect, and a skill set that looks like it is in decline. I would've rather had Youk on a 1 yr, Swisher on a 3 yr, or even a Youk/Pena platoon at first. And I still think Cherrington is in waaaay over his head. Is he injury prone, or did he just get a lot of time off because he was a catcher and Scioscia had a thing for Mathis? Also, according to every defensive metric he's essentially average at 1B. There were injuries with the Angels but he was not Scioscia's favorite player and that's an understatement. He has had a history of nagging injuries. Whether that's helped now that he has a regular gig at first we'll have to wait to see. If the guy stays healthy - that's a big if - there's every chance that he duplicates his career numbers given his BABIP last year with it's very low outlier against lefties. At that level of production, his .850+ OPS makes him quite valuable - the equal of the guy he replaced. He's far from being washed-up as some of the posts on the board would have it.
|
|
|
Post by jmei on Dec 5, 2012 19:41:52 GMT -5
Just because people refer to a year like Napoli's 2011 as an outlier doesn't mean it would actually pass the test for one in a statistical sense. You can't just chuck it out because it looks different to you. That's not just foolish, but intellectually dishonest. Since you called me a "foolish" and "intellectually dishonest" I decided to actually test how often we should expect a player with the talent level of a .371 W/OBA player to perform as Napoli did in 2011. The answer is very rarely. Mike Napoli's season represented a performance slightly more 2 standard deviations above the mean. For some perspective in every other season, Napoli's performance was within 1 standard deviation of the mean. You would expect Napoli to have such a season by random chance once every 50 years. Unless there is some reasonable basis as to why this will reoccur, any responsible forecast of future performance should exclude this performance or at least give it significantly less weight. If you are using the average of the the last three years as a forecast for the future as jmei did. You are implicitly assuming that a player of Napoli's talent level will reach that level 31% of the time. If you use his career averages you are assuming that he'll reach the level 16% of the time. The evidence suggests that in fact, Napoli is a .350s W/OBA type hitter as opposed to a .380s W/OBA type hitter. That makes him roughly average. And finally I would suggest if you are going to use the type of language that you used here that you actually do the work necessary to test Some quick things: -Player performance does not necessarily correspond to a normal distribution. -I'm not sure exactly how many standard deviations Napoli's 2011 was relative to his career, but if I'm remembering my statistics right, two standard deviations above the mean should occur naturally by chance roughly one in twenty times, not one in fifty. -Other variables changed in 2011-- namely, he moved from a park which suppressed RH power to one which inflated it and he was surrounded by much better teammates. This was not a strictly controlled experiment and thus we cannot consider all variation to have been random.
|
|
|
Post by chavopepe2 on Dec 5, 2012 21:19:25 GMT -5
Bottom line: You can't just randomly remove a year from a sample because it is an outlier. That is simply not how statistics work. The goal should be to increase the sample size while being inclusive of all possible data points. Avoid drawing lines and removing data because it looks to be an outlier. It may not be "intellectually dishonest", but it is absolutely a poor way to analyze a data set.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 5, 2012 23:52:39 GMT -5
Then you can't use past mean performance to forecast future results.
You are remembering your statistics incorrectly. Events in a normal distribution should occur within two standard deviations of the mean 95% of the time. Half of those will be above the mean and half below. That means that 2.5% of events should occur two standard deviations above the mean.
You might have an argument if not for his 2012 season in which he performed right in line with his career average W/OBA since 2011.
Every single forecasting class and every single text book on the subject recommends that outliers not be treated as normal values. Paper after paper shows that removing outliers or at least underweighting them SIGNIFICANTLY improves the accuracy of forecast.
Unless there is a basis as to why an outlier is a credible predictor of the future the outlier should be removed. Calling a basic tenant of forecasting "poor data analysis" because it doesn't favor your position is baloney to say the least.
No it's not..... far more likely to occur than a .445 W/OBA in 2011.
Further I think that a guy on the wrong side of 30 who just struck out 30% of the time last year is quite possibly washed up.
|
|
|
Post by jmei on Dec 6, 2012 0:55:13 GMT -5
Then you can't use past mean performance to forecast future results. No, it means you cannot use a normal distribution's standard deviation odds to argue that a given statistical performance is an outlier. This analysis shows that enough players have "outlier" seasons across MLB history that it is hard to regard Napoli's 2011 as particularly rare. You are remembering your statistics incorrectly. Events in a normal distribution should occur within two standard deviations of the mean 95% of the time. Half of those will be above the mean and half below. That means that 2.5% of events should occur two standard deviations above the mean. Thanks for pointing this out-- I just remembered 95% confidence intervals and such. Still, based on just the eye test, I see no reason why one in fifty odds is enough to toss out a season's worth of data. Every single forecasting class and every single text book on the subject recommends that outliers not be treated as normal values. Paper after paper shows that removing outliers or at least underweighting them SIGNIFICANTLY improves the accuracy of forecast. Unless there is a basis as to why an outlier is a credible predictor of the future the outlier should be removed. Calling a basic tenant of forecasting "poor data analysis" because it doesn't favor your position is baloney to say the least. (a) Excluding outliers is actually fairly controversial, especially in cases where the data point is less than three standard deviations from the mean. See, for instance, here and here. (b) You're applying outlier theory incorrectly. By grouping together a season's worth of Napoli's PAs into one bucket and labeling it an outlier rather than seeing it as 432 distinct data points, you're relying unnecessarily on arbitrary endpoints. Yes, his seasonal performance may be significantly different from his career performance, but we can break the data farther down that that, and on a per-PA basis, there are obviously no outliers in the data. You cannot label every hot streak or cold streak as an "outlier" and summarily exclude it since in every case, it only looks like an outlier because you're applying arbitrary endpoints.
|
|
|
Post by jmei on Dec 6, 2012 1:06:48 GMT -5
More on why we can't just throw out "outlier" seasons ( link): The deeper and more important point is that by looking at one-year deviations as establishing a new level of performance that thus takes on a greater weight (breakout!) or as being irrelevant and thus in need of exclusion (outlier), both positions implicitly assume they already know what we’re trying to find out when projecting a player: his “true talent.” Recall the “general formula for player performance” from Monday’s post: performance = true talent + luck. The various methods that projection systems use (regression, weighted averages, age adjustments, etc.) are meant to take the (limited) data we have for a player and filter out luck in order to estimate his current true talent. These methods are predicated on the fact that we can’t pinpoint the player’s true talent given the limited performance samples we have, so we make our best estimate based on probabilities.
Labeling a single season as irrelevant or supremely relevant to estimating a player’s true talent implicitly assumes that one already knows that player’s true talent. One can certainly cite examples of each kind to support the case for a “breakout” or “outlier.” One could just as easily come up with (many more) examples of the opposite — where a perceived “breakout” or “outlier” turned out not to have the (in)significance assigned to it. But to do either obscures the important point. It is true that individual players age differently and deviate from expectations. However, projection systems only obtain the overall accuracy they have by projecting players as a whole based on the data on hand. An apparent “outlier” season from two years ago may weigh less heavily because time passing and/or, say, BABIP being regressed more heavily than other skills. An apparent “breakout” by a young player may have more impact on the projection because of age adjustment, greater playing time, etc. But projection systems do not and should not take these into account beyond their standard adjustments.
A famous German sabermetrician once wrote, “the owl of Minerva begins its flight only with the onset of dusk.” Although in retrospect we can look back on the careers of particular players and identify certain seasons as “outliers” or “breakouts,” this can only be done years later when we have a perspicuous overview of a period of a player’s career as a whole. Projection systems work in the midst of player performance without the benefit of historical perspective, and have to do the best they can based on the information at hand. Doing anything more would revoke the humble presuppositions upon which player projection rests.
|
|
|
Post by Oregon Norm on Dec 6, 2012 1:09:13 GMT -5
..., Further I think that a guy on the wrong side of 30 who just struck out 30% of the time last year is quite possibly washed up. Napoli's career strikeout rate is just that, 30%. Some hitters have lower contact rates but lots of power. It also goes along with the patience thing. The guy sees a lot of pitches.
|
|
|
Post by redsoxprospects on Dec 6, 2012 4:26:31 GMT -5
I barely clepped out of statistics after 6 weeks of study in HS so I can't compare my understanding to some of you dudes but the fact that Napoli's outlier season is fairly recent should make it somewhat more relevant and the fact that he is suited to Fenway should help him. Bill James is projecting him at 1 HR per every 15.17 AB. And he should benefit some as a hitter from additional rest as more of a DH/1st base option for us ( Ortiz playing 1st in NL parks ). His worse BAPIP ever was last year so some improvement with that is very likely. He fits the team philosophy of seeing a lot of pitches. Overall I like this signing, given the roster flexibility it affords, the dirth of many other quality options and the length of the deal. He's a good flexible fit who isn't costing us a fortune while helping to keep us competitive. He's not Adrian Gonzalez but he's not costing us $155 mil either. He is a business move. He keeps us competitive until the cavalry arrives and he didn't cost a pick. We should be sending Henry a thank you note for this signing.
|
|
|
Post by redsoxprospects on Dec 6, 2012 4:50:12 GMT -5
It will be very interesting where these new guys fit in the lineup. Is Napoli our #4?
|
|
|
Post by welovewally on Dec 6, 2012 7:24:35 GMT -5
I'm not on board with this signing, he is not a 3/4/5 hitter or a 1B & he can't Pitch.
|
|
|
Post by The Town Sports Cards on Dec 6, 2012 7:26:08 GMT -5
It will be very interesting where these new guys fit in the lineup. Is Napoli our #4? Depends on the #1-3. Does Victorino hit in the top 3 bumping down Ellsbury or Pedroia? Or does he go down the bottom at #6-#7? Most likely, Napoli is #4 or #5 though, here's what I see for possible lineups: 1) Ellsbury-Pedroia-Ortiz-Napoli 2) Ellsbury-Victorino-Pedroia-Ortiz-Napoli 3) Victorino-Pedroia-Ellsbury-Ortiz-Napoli
|
|
|
Post by jioh on Dec 6, 2012 7:31:06 GMT -5
Are we looking at the same thing? Here's what I see ( link): As a catcher: 1667 PAs, .265/.362/.516, .377 wOBA (.307 BABIP, 11.9% BB, 25.2% K, .252 ISO) As a first baseman: 414 PAs, .239/.328/.473, .348 wOBA (.265 BABIP, 10.1% BB, 24.6% K, .234 ISO) As a DH: 154 PAs, .286/.390/.578, .414 wOBA (.330 BABIP, 14.8% BB, 24.2% K, .292 ISO) He's hit worse as a first baseman but a lot of that appears to be BABIP bad luck. Given the very limited sample size here, it's hard to make any firm conclusions or projections based on his past performance. I agree with moonstone that I don't see Napoli being appreciably better going forward now that he's moved off 1B. I just happen to think that maintaining his career performance should be more than good enough Can you say anything about how Napoli would hit as a full-time 1b from how he hit at each position while going back and forth from C to 1b? If he caught on Mon and Wed and played first base on Tues, I would think that he would feel great on Wed after a day without squatting on Tues.
|
|
|
Post by polarbear91 on Dec 6, 2012 7:57:25 GMT -5
I'm not on board with this signing, he is not a 3/4/5 hitter or a 1B & he can't Pitch. This makes sense only if the Napoli signing precludes them from not obtaining another pitcher. Does anyone believe that is the case?
|
|
|
Post by welovewally on Dec 6, 2012 8:02:18 GMT -5
I'm not on board with this signing, he is not a 3/4/5 hitter or a 1B & he can't Pitch. This makes sense only if the Napoli signing precludes them from not obtaining another pitcher. Does anyone believe that is the case? True. But I still don't like the Napoli signing. And I haven't heard anything about the Sox pursuing any Pitching of consequence. Dempster- pass. Trade for Dickey - pass
|
|
|
Post by buffs4444 on Dec 6, 2012 8:13:56 GMT -5
I agree patching the rotation with a Dempster is a bad sign, this team needs something more. But as long as Napoli is a 5/6 hitter in this lineup, I don't see a problem. I expect him to rake at Fenway and be decent away. Definitely expecting the home W/L record to improve next year....
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Dec 6, 2012 11:17:05 GMT -5
Again I really don't buy the idea that Mike Napoli is going to magically turn into a better player just by putting away his tools of ignorance. It hasn't happened throughout his career, and it probably won't happen now just because he signed with our favorite team. It's not magic at all. Catcher is the most physically taxing position on the diamond by a wide margin; it's particularly hard on the lower body. Why wouldn't that carry over to hitting? Because in Napoli's, case it hasn't. He's performed better on days that he caught than on days he hasn't. Yes the hypothesis does make logical sense, but there is no statistical backing for it in the case of this particular player.
|
|
|
Post by honkbal on Dec 6, 2012 11:32:30 GMT -5
SSS
|
|
|
Post by beasleyrockah on Dec 6, 2012 11:43:26 GMT -5
Napoli has never had a season where he played the large majority of his at bats at 1b. The one season where he got more at bats at 1b than catcher was still essentially a 50-50 time share. You don't get a clean slate on the random days you get at 1b, catching is a cumulative grind over the course of the season. The fact is he's never had the opportunity to be a real first baseman. I think it's silly to ignore the logical benefits of avoiding catching just because he didn't hit better at 1b while he was splitting time between the positions. Even if his performance isn't significantly effected, it should allow him to play more and create less stress for his body. If you catch three days in a row and then get placed at 1b for a game, that random game at 1b isn't predictive of how you'd play if you avoided catching entirely (or almost entirely).
|
|
|