Hitter Evaluations: 2B 2012 Talent Projections
With the core background work done on my hitter evaluations, I have decided to apply the process to second basemen for 2012. This ranking is based off of all player’s stats given the same number of PA. This a look to see which players are the most talented. You can follow the process with this spreadsheet.
Background and and Procedures (skip if desired)
To get a 2012 projection (columns R to X), I used a one-part weighting of 2011 stats (Column J to P) and a two-part weighting of the 2011 ZIPS projections (column B to H). I hoped to include most 2012 2B-eligible players (10 games at 2B in 2011). If there is a second baseman you want added, let me know.
Note: There was not a ZIPS projection for Tsuyoshi Nishioka so I created one. It was the average values of the other second basemen.
The standard values were changed to rates (columns AA to AI) to be used in further calculations.
Next, I created a projection using the five standard stats (AVG, HR, RBI, Runs, SB) for each of the second basemen based on 600 PA (columns AK to AQ). Besides these five stats, Hits, Ks, BBs and OBP can be determined using the information at hand. The values were then weighted using Zach Sander’s formula for finding fantasy above replacement values (columns AS to AW).
Note: We understand that the process is flawed when comparing different positions, and are working on updating it, but the calculator seems to work just fine for one position.
Finally, the individual overall rankings are calculated (columns AY and AZ).
2012 2B Rankings (based off of 600 PA):
| Rank | Name | HR | Runs | RBI | AVG | SB | Ranking Value |
| 1 | Ian Kinsler | 23 | 78 | 78 | 0.270 | 26 | 7.3 |
| 2 | Dustin Pedroia | 16 | 80 | 80 | 0.301 | 18 | 7.2 |
| 3 | Robinson Cano | 23 | 79 | 79 | 0.299 | 5 | 5.7 |
| 4 | Chase Utley | 20 | 76 | 76 | 0.270 | 16 | 4.5 |
| 5 | Brandon Phillips | 17 | 72 | 72 | 0.281 | 17 | 3.5 |
| 6 | Michael Young | 14 | 75 | 75 | 0.300 | 5 | 3.3 |
| 7 | Ben Zobrist | 18 | 74 | 74 | 0.252 | 18 | 2.9 |
| 8 | Dan Uggla | 29 | 76 | 76 | 0.244 | 3 | 2.5 |
| 9 | Rickie Weeks | 22 | 73 | 73 | 0.248 | 13 | 2.1 |
| 10 | Brian Roberts | 11 | 69 | 69 | 0.262 | 29 | 2.1 |
| 11 | Kelly Johnson | 20 | 72 | 72 | 0.251 | 13 | 1.6 |
| 12 | Michael Cuddyer | 17 | 73 | 73 | 0.269 | 8 | 1.6 |
| 13 | Daniel Murphy | 11 | 72 | 72 | 0.292 | 8 | 1.5 |
| 14 | Ryan Raburn | 23 | 72 | 72 | 0.262 | 5 | 1.5 |
| 15 | Howie Kendrick | 14 | 69 | 69 | 0.282 | 15 | 1.5 |
| 16 | Martin Prado | 13 | 70 | 70 | 0.283 | 4 | 0.3 |
| 17 | Neil Walker | 15 | 70 | 70 | 0.268 | 8 | 0.2 |
| 18 | Ryan Roberts | 15 | 69 | 69 | 0.244 | 17 | -0.1 |
| 19 | Aaron Hill | 19 | 68 | 68 | 0.253 | 10 | -0.3 |
| 20 | Omar Infante | 7 | 68 | 68 | 0.289 | 5 | -1.0 |
| 21 | Tsuyoshi Nishioka | 13 | 67 | 67 | 0.258 | 12 | -1.1 |
| 22 | Maicer Izturis | 7 | 67 | 67 | 0.273 | 13 | -1.1 |
| 23 | Chone Figgins | 2 | 62 | 62 | 0.244 | 32 | -2.3 |
| 24 | Danny Espinosa | 18 | 64 | 64 | 0.223 | 17 | -2.6 |
| 25 | Gordon Beckham | 13 | 65 | 65 | 0.250 | 7 | -2.7 |
| 26 | Johnny Giavotella | 6 | 64 | 64 | 0.264 | 13 | -2.7 |
| 27 | Orlando Hudson | 8 | 65 | 65 | 0.249 | 12 | -2.8 |
| 28 | Jason Kipnis | 11 | 65 | 65 | 0.258 | 7 | -2.9 |
| 29 | Ryan Theriot | 2 | 63 | 63 | 0.271 | 15 | -3.0 |
| 30 | Mark Ellis | 10 | 63 | 63 | 0.254 | 12 | -3.0 |
| 31 | Justin Turner | 8 | 64 | 64 | 0.264 | 7 | -3.1 |
| 32 | Sean Rodriguez | 17 | 63 | 63 | 0.225 | 13 | -3.4 |
| 33 | Robert Andino | 11 | 62 | 62 | 0.248 | 13 | -3.5 |
| 34 | Dustin Ackley | 7 | 64 | 64 | 0.246 | 8 | -4.1 |
| 35 | Darwin Barney | 2 | 61 | 61 | 0.273 | 9 | -4.6 |
| 36 | Jemile Weeks | 5 | 58 | 58 | 0.244 | 19 | -5.2 |
I will not go over the list today because I will begin looking at it in detail over the next few weeks. In the mean time, let me know if you have any suggestions or questions.
Jeff, one of the authors of the fantasy baseball guide,The Process, writes for RotoGraphs, The Hardball Times, Rotowire, Baseball America, and BaseballHQ. He has been nominated for two SABR Analytics Research Award for Contemporary Analysis and won it in 2013 in tandem with Bill Petti. He has won four FSWA Awards including on for his Mining the News series. He's won Tout Wars three times, LABR twice, and got his first NFBC Main Event win in 2021. Follow him on Twitter @jeffwzimmerman.
That chart really seems to punish 2nd year players. Ackley, Kipnis and Jemile Weeks seem low.
sophomore slumps much? i think ackley can beat a .248 BA…
The ZIPs projections hated them.
Ackley was projected with a 0.244 average and 0.280 BABIP
Just out of curiosity, what’s the reason for giving the prior year ZIPS projections twice the weight of the actual 2011 stats? I think that’s probably what’s driving down the value of 2nd year players so much.
Most weighting systems use a 5,4,3 (or 8,5,4,3) weighting for the last 3 years. It should actually be a 1.4 to 1 weighting. The projection needs more weight. I added a little more weight to projection because it adds in some regression. Another problem is the projections are near 600 PA and the players are only at 300 PA for 2011. The weighting is closer to almost 4:1 for the rookies.
Uggla with under 30 HRs? He is the Adam Dunn of 2bman, no?
I also don’t like the Cano projection….he could go 30-100 and possibly approach 35 HRs next year. His power is trending up.
That’s a ridiculously pessimistic projection for Cano and Pedroia. I understand that it’s based on 600 PAs (Cano’s had at least 670 for the past three seasons), but still, if either of them only put up around 80 runs/rbi in that many PAs I would be shocked. Does it factor in the quality of the players hitting around them at all? The problem with this chart is that not enough variables are accounted for, therefore we see very little variation between the very best and very worst second basemen – everyone ends up with around 68 runs and RBI. It’s also tough to see 36 second basemen getting 600 PA. I don’t fully understand the methodology, but an eye test will reveal that it seems very flawed.
The R/RBI totals smell extremely hinky. I’d be less critical if we weren’t holding PA constant for this exercise. The HR, SB, and AVG number are probably of some use, but the R/RBI should just be discarded. I’ll set the line on R^2 to actual performance for R/RBI at .175. Any takers?
Clearly the first problem is that there’s no attempt to project lineup slot. Ian Kinsler is not going to go 78/78 in R/RBI. He’s going to score way more R than RBI. Similarly, Uggla will drive in a lot more runs than he’ll score.
Yes. Please read:
http://www.fangraphs.com/fantasy/index.php/hitter-evaluation-runs-and-rbis-part-2/
So Cano, the 3 hole hitter in the Yankees lineup is only gonna have 79 RBI’s next year? Kinsler’s power seems low too. I don’t personally care for this list for a whole lotta reasons, too many of which to ramble on in this post…
Also:
http://www.fangraphs.com/fantasy/index.php/hitter-evaluation-runs-and-rbis-part-2/
Jeff, I recommend you either combine R and RBI into 1category, or apply the batting order adjustments based on expected order. You can also apply the team runs adjustment. I think it’s silly to present these “preliminary” numbers at all, since they are obviously not meant to be indicative of anything.
Also, why predict R+RBI together? The problem is that factors that can contribute to one (Walks contributing to Runs) might have adverse effect on the other (putting balls in play probably increases RBI). The only way that BB would likely contribute positively to RBI is probably caused by interactions with other elements in your model.
Sorry I didnt notice this stuff earlier, as I just now looked at the google spreadsheet. Seeing the positive coefficient on BB in the RBI predictor gave me pause.
Lastly, if you are giving projections for fantasy advice, it makes no sense to normalize to 600 PA. I can only understand this decision if you have zero methods for projecting PA, but you’ve apparently developed methods based on batting order. I recommend using them. Maybe present both the normalized version and the one based on projected PA.
I am looking at the player’s talent and this all I want to look at in October. 1/3 of these players will probably be on other teams or lose their job by the time ST comes around. The top 10 may have secure jobs, but the rest of them could change their roles and teams easily.
This is how the players would perform on an average team with an average spot in the lineup.
So maybe Runs and RBI should be withheld from this presentation, since they are so context dependent. Including them “as is” undermines your work, since everyone projects to a narrow range and R=RBI.
what does “an average spot in the lineup” entail exactly?
Chone Figgins, leading all MLB 2bmen in SBs next year…
you heard it here first.
some GM would have to be drunk enough to give him the 600 PA, but Juan Pierre still plays everyday.
“I am looking at the player’s talent and this all I want to look at in October. 1/3 of these players will probably be on other teams or lose their job by the time ST comes around. The top 10 may have secure jobs, but the rest of them could change their roles and teams easily.”
Jeff, this comment begs the question of why even do this analysis now.
To give people as much information as possible for now.
In one league I have Kelly Johnson, Cuddyer and Neil Walker. 2 will become FAs at the end of the season. It will give an idea of their potential.
Also as information become available, the adjustment to the numbers can be made.
I appreciate the efforts, Jeff. I’d already compiled my initial 2012 rankings at every position. It’s never too early!
When using the average and standard deviations of the whole group…will you be using the same amount for other 1 starter positions also(C,1B,SS,3B)? Differences in the total amount in the group can artificially give players more or less value by mistake yes?
How do you project Rickie Weeks to only hit 22 hr with a .248 BA?
ZIPs doesn’t like him.
0.300 BABIP with a 20% K%
Freddy Sanchez?
Will add.
Nishi may not hit 3 career homers let alone 13 in 1 year.
The extra power is because I used an average player as his ZIPs projection.
Out of curiousity, If Allen Craig somehow manages to get a full time job out of spring training, and considering he qualifies for 2nd, where would you put him in your rankings?
I will add him and see.
Figgins won’t even get 100 PA next year.
I’d take Kipnis and Ackley top 10.
I am not sure why ZIPs hates them so much, I will see what the projections are this up coming year.
So Cano will have is worse year since 2008? By far? And pedroia will regress a ton as well?
This is the number for 600PA, hitting 5th for an average team, like the Indians.
Let’s take Cano for a second.
Take these numbers and assume all the surrounding circumstances stay the same (to early to tell how this will work out with many of the players)
He gets 671 PA
Yankees score 5.35 R/G
Cano bats fifth all season
(All of these he has no control over)
Using the adjustments from here, he ends up with:
26 HRs, 107 Runs, 114 RBIs, 0.299 AVG and 5 SB
for some bizarre reason he was traded to Minnesota who averaged 3.85 R/G and had a team 0.306 OBP vs 0.340 (less PA). He would get on average 660 PA and his numbers would change to:
25 HR, 75 Runs, 80 RBIs, 0.299 and 5 SB
The top players are going to look worse, but if they were stuck at Houston or Minnesota last year, their numbers would be below the numbers above.
Anyone knows the top few 2B are going into next year will be good. I am looking for the diamonds in the rough and how changing teams will effect the player’s numbers.
Finally, with 2B the top 3 all play for high power offenses, so their numbers take huge drop off when compared to the rest of the league.
A couple things:
1) Cano’s not going to be traded to Minnesota, neither are any of the top 10 SB’s, so why are these fictious numbers relevant?
2) Assuming these numbers are relevant, do your calculations take into account the increased R/G and OBP that the “average team” would have if Cano/Pedroia, etc. was batting 5th for them?
I misunderstood the point of your article, and now I want you to admit your numbers are wrong!
(just trying to join in with the general theme in comments section here)
Thanks, it had been a couple of hours since someone reminded me.
Howdy would you mind letting me know which webhost you’re working with? I’ve loaded your blog in 3 completely different internet browsers and I must say this blog loads a lot faster then most. Can you recommend a good web hosting provider at a fair price? Thanks a lot, I appreciate it!
this is idiotic.
Please let me know if you’re looking for a writer for your weblog. You have some really great posts and I feel I would be a good asset. If you ever want to take some of the load off, I’d love to write some material for your blog in exchange for a link back to mine. Please shoot me an email if interested. Regards!
Hey I am so glad I found your blog page, I really found you by mistake, while I was searching on Bing for something else, Nonetheless I am here now and would just like to say many thanks for a tremendous post and a all round enjoyable blog (I also love the theme/design), I don’t have time to read through it all at the moment but I have book-marked it and also added your RSS feeds, so when I have time I will be back to read more, Please do keep up the superb work.
Contrary to the spam just above me and the whiners too lazy to read the introduction about team, lineup et al adjustments, this is great work.
I’ve been toying around with projection systems recently, seeing if I could get increased accuracy on rate figures by regressing Marcel, CAIRO, ZiPs, etc., but haven’t been able to get r^2 north of ~0.55 and standard error of about 1.1 hr per 100 PA. Seeing as this translates to a +/- of 6 HR over an average season, the difference between Ryan Howard and Freddie Freeman, I didn’t find my combinations particularly helpful.
I’m curious if you could(/would) retroactively run your projections using 2010 stats and 2011 ZiPs on 2011 actuals (rates, as to eliminate injuries etc.) and report the r^2 and stdevs…?
You are probably about as close as you are going to get. The numbers move that much (see Adam Dunn).
One key I have found it to combine 3 projection systems together (I am not sure if 3 is the right number). Tom Tango’s battle of the projection systems has shown that the composite is better than just 1 system.
I may get back to it later, but we are about to start the 2012 fantasy writeups so time is short. If I haven’t gotten back to this question by Xmas, email me at wydiyd ~ hotmail and I will look into it. -Jeff
Thanks so much. Looking forward to the upcoming content.