Eric Hosmer San Diego Padres

Age: 31 (October 24, 1989) | 6' 4" | 225lbs. | Bats: Left 1B-32 DH-5 PH-2
Tm Lg YEAR G AB R H BB SO 2B 3B HR RBI SB CS BA OBP SLG BB% SO% BABIP G/L/F % $4x4 $5x5
KC AL 2017 162 603 98 192 66 104 31 1 25 94 6 1 .318 .385 .498 10 15 .351 56/22/22 31 30
SD NL 2018 157 613 72 155 62 142 31 2 18 69 7 4 .253 .322 .398 9 21 .302 60/20/20 17 17
SD NL 2019 160 619 72 164 40 163 29 2 22 99 0 3 .265 .310 .425 6 24 .323 56/21/23 17 17
SD NL 2020 38 143 23 41 9 28 6 0 9 36 4 0 .287 .333 .517 6 18 .296 46/20/34 28 24
SD NL 2021 34 127 12 39 13 19 6 0 4 20 3 1 .307 .373 .449 9 13 .333 61/17/22 7 6
Career 11yrs 1437 5493 726 1531 484 1066 278 20 180 790 74 27 .279 .337 .435 8 18 .315 n/a
Welcome! You are invited to wander around and read all of the comments that have been posted here at Patton & Co., but as soon as you register you can see the bid limits that Alex, Peter and Mike propose for each player, and you can post your own comments. Registering is free, so please join us!
Hmmm. My pan of him in The Guide is looking pretty good so far. Helps that it's sitting next to about 8 picks or so.
Mike Gianella MikeG
Apr 19 '12
We know for a fact that a system that tries to guess the outliers will be less accurate. I know it because I've done tons of work on this, modeling systems and then measuring the error.

The reason I tweak my projections by hand is to incorporate info that no standard set of calculations could include, but it has to be judicious because while it means I get some more more right, the ones you're wrong on will mean getting two more wrong. Especially so if you're making everything add up.

I just came across another article at Fangraphs that explains what projections do (or should do) pretty clearly.

http://www.fangraphs.com/library/index.php/understanding-projections-true-talent-level-and-variability/

Peter Kreutzer Rotoman
Mar 10 '12
This in the end is a kind of silly debate. Because ours is a bidding contest, not a projecting contest. Bids are how we acquire players, and we wait for actual stats to tell us whether they were good bids or bad bids.

But, to continue the debate (because it's fun!), is it absolutely known for sure that wild ass guesses for projections are less accurate? As long as they aren't totally crazy, and add up to a realistic target for each league, and the totals for hitters and pitchers are consistent with each other, why can't your more hunchy projections have the same or better success rate? Even if it's just, you know, luck.
Alex Patton Alex
Mar 10 '12
The problem with using projections for pricing is that good projections don't have outliers. Performance is regressed for the sake of accuracy.

You can then make wild ass guesses to bring your projections into align with your bid prices, but you will then be making your projections less accurate.

You can have one (accurate projections) or the other (projections that look like bids). You can't have both.
Peter Kreutzer Rotoman
Mar 10 '12
The problem I have with ZIPS and these young players isn't ZIPS saying that they'll each earn $28 but rather with the fact that ZIPS is saying that Lawrie will be the 4th best player in the American League and Hosmer the 7th. Could they be? Well, sure, they could. But based on historical data the odds are very, very poor. And one thing ZIPS is "wrong" about - season after season - is on what it is the best players will earn. The 4th best hitter in the A.L. last year earned $36. The 7th best hitter earned $32. So for ZIPS to be "right", either Lawrie and Hosmer have to earn something in the neighborhood of $32-36 OR the 4th/7th best hitters have to be $28 players this year. If Vegas would let me wager on such things, I'd feel pretty safe wagering that neither one of these things is going to happen.

Since you bring up Nellie Cruz, do you happen to have the ZIPS projections from that year please? Is Cruz the exception or the rule to ZIPS getting rookies right.
Mike Gianella MikeG
Mar 10 '12
In a Moyer keeper league, I first drafted Hosmer before his senior year. Then, he was viewed as a future Triple Crown threat. The stroke is so sweet that mammoth production will ensue with just normal progress in being a little more selective, though the contact rate given his age and power is quite impressive.
michael salfino Emjay
Mar 10 '12
Of course ZiPS projected Nelson Cruz right. Minor league stats are projectable to the major league level. They might not be as robust inputs as major league numbers, but even those are only 70% correct. I would venture to say that Lawrie and Hosmer have as good a chance to earn the $26 and $25 or so ZiPS projects them to earn as Granderson and Gardner do of earning $25 (their ZiPS projection). The biggest risks I see in the 2nd year players is the high SB totals. We don't know how their managers will use their potential basestealing abilities right now.

Alex is right about what I am saying. Take ZiPS perfect full-time projection and then change the ABs/IP in the software based upon your own playing time estimate to get a true value.
Eugene Freedman EugeneFreed
Mar 10 '12
They didn't "rank" there, but that's where the fall out if you take their numbers and apply Alex's formulas to them (well, my variations to Alex's formulas but the baseline values are very close). I can buy what you're saying about Trout, but Eric Hosmer and Brett Lawrie are likely to play the entirety of 2012 in the Major Leagues. Yes, you can and should make adjustments to formulaic projections in your bids, but any system that has players ranked that high using raw data is, again, problematic.
Mike Gianella MikeG
Mar 10 '12
Marcel gives everyone who hasn't played in the big leagues the exact same "projection," which is about 200 AB at .265, with 28 runs and RBI and four homer and steals. Everyone. (Not these exact numbers, I made those up, but something similar). This is maximum regression and puts up good R scores, even though on a player by player basis it is meaningless.

But it tells you a lot about how much variance there is in performance of young players.
Peter Kreutzer Rotoman
Mar 10 '12
Using the software, customers ultimately are responsibel for their own bids. I think Eugene is saying the same thing about ABs if you use ZiPS.
Alex Patton Alex
Mar 10 '12