2B Rankings Using ZiPS ROS

The folks at Baseball Think Factory are kind enough to supply Fangraphs with projected ZiPS values for each player covering the rest of the season. These values take the player’s preseason projection and how they have produced so far in 2012 and then estimate the player’s stats for the rest of the season.

I used Zach Sander’s formula to rank only 2B for the rest of the season.using z-scores. Here are the rest of season values for 2B order by their current z-score and some thoughts on the rankings:

Name AVG Runs HRs RBIs SB Z-Score
Robinson Cano 0.299 72 18 76 5 7.3
Ian Kinsler 0.273 77 18 56 19 4.9
Michael Cuddyer 0.287 62 16 66 8 4.6
Dustin Pedroia 0.296 69 14 56 15 4.5
Dan Uggla 0.252 67 22 65 1 4.4
Brandon Phillips 0.276 66 13 60 12 3.4
Ben Zobrist 0.258 65 15 61 12 2.8
Michael Young 0.290 55 11 60 5 2.6
Jason Kipnis 0.260 66 14 56 12 2.4
Jose Altuve 0.295 63 8 51 25 2.3
Neil Walker 0.268 57 11 65 7 2.1
Howie Kendrick 0.274 58 12 52 11 1.6
Aaron Hill 0.262 53 14 54 9 1.1
Allen Craig 0.286 42 13 54 3 1.1
Kelly Johnson 0.244 59 17 48 10 0.9
Danny Espinosa 0.226 61 15 51 13 -0.1
Dustin Ackley 0.259 61 10 41 8 -0.2
Rickie Weeks 0.242 56 15 40 8 -0.6
Johnny Giavotella 0.270 57 5 44 9 -0.7
Omar Infante 0.290 44 7 41 4 -0.7
Mike Aviles 0.271 44 9 43 10 -1.1
Daniel Murphy 0.297 34 5 42 4 -1.6
Gordon Beckham 0.247 52 9 41 4 -1.8
Darwin Barney 0.271 55 2 36 7 -2.0
Ryan Roberts 0.245 46 10 41 11 -2.3
Steve Lombardozzi 0.269 49 4 35 13 -2.4
Trevor Plouffe 0.236 48 11 38 3 -2.6
Jemile Weeks 0.257 53 4 34 19 -2.7
Sean Rodriguez 0.232 46 10 38 8 -3.1
Mark Ellis 0.257 44 5 33 9 -3.4
Ruben Tejada 0.264 44 2 35 6 -3.5
Orlando Hudson 0.243 48 6 34 9 -3.5
Maicer Izturis 0.271 35 3 30 9 -4.2
Ryan Raburn 0.238 35 9 34 2 -4.4
Alexi Casilla 0.260 39 2 23 12 -5.2


– Cano is the top rated 2B by a bit. Most of his extra value, compared to the 2B ranked after him, is from the extra RBIs he looks to produce.

– Kinsler, Cuddyer, Pedroia and Uggla are grouped closely together. A team could look at moving one of these 4 for one of the others depending on a team’s need. An example would be trading Pedroia for Uggla if a team is in need of a little extra power (HR and RBIs) and can take a hit in AVG and SBs.

Jose Altuve looks to be the 2B that has moved up the most since the beginning of the season.

Kelly Johnson and Rickie Weeks see their value depressed due to their low AVG. If a team needs power and can handle the low AVG, they are nice options.

Ryan Roberts looks like he remembered who he was and is now back to being irrelevant in fantasy baseball.

Jemile Weeks, even though ranked low, will be an option for a team desperate for SBs.

– I am a little surprised to see Mark Ellis so low. He has been a nice source of Runs this season.

Print This Post

Jeff writes for FanGraphs, The Hardball Times and Royals Review, as well as his own website, Baseball Heat Maps with his brother Darrell. In tandem with Bill Petti, he won the 2013 SABR Analytics Research Award for Contemporary Analysis. Follow him on Twitter @jeffwzimmerman.

27 Responses to “2B Rankings Using ZiPS ROS”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Mike M says:

    Where’s Robert Andino?

    +7 Vote -1 Vote +1

  2. Ben G. says:

    Wow, ZiPs doesn’t really care for Jason Kipnis, especially since it had him almost going 20-20 in the preseason. I feel he’s at least kept that pace.

    Vote -1 Vote +1

    • Scott Clarkson says:

      This is ZiPS “ROS” : rest of season (i.e. these numbers do NOT include #’s already accumulated this year). Add his 6 HR and 6 SB already in the bank to these totals for final projected season line which would be 20HR/18SB….I guess he is who we thought he was.

      Vote -1 Vote +1

  3. Matt says:

    This is great! Could you post for other positions as well?

    Vote -1 Vote +1

  4. Michael says:

    Allen Craig?

    Vote -1 Vote +1

  5. PillsburyFlowboy says:

    marco scutaro?

    Vote -1 Vote +1

  6. Daniel says:

    How does Kinsler only have 5 more runs scored projected than Cano? Utterly ridiculous. He already leads the majors with 33, and he isn’t doing all that great. A leadoff hitter for the best offense who gets on base more than Cano will score more runs. by far.

    Vote -1 Vote +1

    • jcxy says:

      This is a reasonable criticism. Someone with a .400 obp is likely to be more valuable on a team that will score 900 vs a team that will score 600. I suspect–although I admit, actually have no clue–that ZiPS projection system is more limited (could that be a good thing?) in projecting R/RBI values vs HR/SB values because the former are context dependent and the latter less so.

      If that’s the case, would we be better off just using hr/sb/avg as inputs when computing the zscore?

      Vote -1 Vote +1

      • phoenix2042 says:

        so the yankees are only a 600 run team this year? laughably bad projection.

        Vote -1 Vote +1

      • jcxy says:

        was commenting generally on how ZiPS projects rather than on the kinsler/cano question specifically. i mean…come on…use your head.

        Vote -1 Vote +1

  7. Nick M says:

    If only “Dat Dude” B. Phillips can put up those numbers the rest of the year. He has 1 home run so far…

    Enjoy the stats though.

    Vote -1 Vote +1

  8. Brent Crossman says:

    Did you use the Stdev of the 2B position or for the entire player pool. This was discussed heavily in the other post and it has to be the entire player pool, otherwise when you get to Catcher’s you’ll over value any catchers that steal bases (as an example). http://www.insidethebook.com/ee/index.php/site/comments/economics_of_fantasy_valuation/ (post 58-61 for Tango’s thoughts)

    Vote -1 Vote +1

  9. Utley? says:

    No data so far this season, but where’d he fit in

    Vote -1 Vote +1

  10. wjylaw says:

    Again, another projection system totally ignoring reality. Cano over Kinsler , especially by such a large margin is farcical. Nice to see that actual production accounts for almost nothing.

    Vote -1 Vote +1

    • Ralph says:

      We it IS a projection system, it takes previous seasons heavily into account. If you just want to see a ranking where this seasons stats are projected over 162 games just put it into a spreadsheet and assign a ranking and voila! A ranking system built for you.

      (I don’t necessarily think Cano will outperform Kinsler ROS, but evidence does suggest its at least likely).

      Vote -1 Vote +1

  11. Sean O says:

    So, it looks like Craig may have been underrated in the off-season. I know we have only a small sample size with him, but he’s been crushing it since last season.

    If Craig stays healthy for the ROS, here are my projections for him: (# of StL games remaining = 126) * (5/6 appearances) = ~105 game started, remaining.
    (Beltran and Berkman will have regular days off, not to mention a DL visit or two, + other positions).

    Multiplied this by the average of his rates last year and this year… (yes, that overly weights the outrageous rates he’s put up to date, but let’s call that the “crazy ceiling” = .5R/g; .25HR/g; 1.0RBI/g; .1SB/g; 1.10 OPS; .340 AVE. I’d say of these, AVE and OPS are probably a little high, as is RBI because the whole team is killing it; SB and HR rates seem fairly accurate, and R is probably a bit low.

    This gives us, ROS: 52 R, 26 HR, 105 RBI, 10 SB, 1.10 OPS and .340 AVE.
    My guess, ROS is: 60 R, 24 HR, 85 RBI 10 SB, .910 OPS and .315 AVE and when all is said and done, will challenge Holliday for the Cards offensive MVP.

    Hey, even if he puts up 80% of these numbers, he’s a top 5 second baseman.

    Vote -1 Vote +1

  12. mcbrown says:

    This definitely shows the perils of calculating z-scores within positions, as opposed to using all players and then just calculating replacement level within positions. Does anyone really think Cano’s projected edge in BA and RBI is enough to make him 50% more valuable than Kinsler?

    Vote -1 Vote +1

  13. Marc Ross says:

    Why isn’t Freddie Galvis on the list?

    Vote -1 Vote +1

    • cable fixer says:

      most leagues don’t play with uzr as a category.

      Vote -1 Vote +1

      • shaso says:

        Well, he’s 8th in RBIs; do most leagues play with RBIs as a category?

        Vote -1 Vote +1

      • jcxy says:

        I’m not sure his RBI total paints the correct picture of Galvis.

        ZiPS (R) is: 47/5/36/12 in 480 ABs.

        The two problems I see are:
        1. Although there is a nonzero chance (Utley or Rollins injury/injuries) he actually gets to 480, smart money most certainly is on the way, way under.
        2. Does that stat-line, turned into rates, play well for the next month or two? A: Probably not…unless ZiPS is way low on speed, he’s roughly ODawg territory…(-)3.5 zscore. Not great.

        Vote -1 Vote +1

  14. LD Smith says:

    “These values take the player’s preseason projection and how they have produced so far in 2012 and then estimate the player’s stats for the rest of the season.”

    So, the preseason projections are unchanged? So, if a player is off to a solid start, like Mark Ellis, they already have provided a good portion of the preseason projection and RoS will need to be below average to come back to projected total.

    This would seem to favor a slow starter and penalize a hot starter?

    Vote -1 Vote +1

    • mcbrown says:

      I’m quite sure that is not how ZiPS works. If a player is projected preseason to hit 10 HRs, and hits 5 in the first 20% of the season, ZiPS doesn’t subtract the 5 hit so far and assume the “remaining” 5 will be spread over 80% of the season. What I assume it does is take its prior (the player will hit at a rate of X HRs/PA, such that the player was projected for 10 over a season, or 8 over the remaining 80% of the season) adjusted for his performance to date (5 HRs in a shorter time frame might make you think you had perhaps underestimated the player’s power, and now you think his true talent is maybe Y HRs/PA which gives a rate of 11 over a season, or 8.8 over the remaining 80%). Thus hot starters receive some benefit and slow starters receive some penalty. You can this see from checking out the difference between ZiPS preseason and updated projections for various hitters. ZiPS ROS projects Pujols to have slightly lower rates of production than ZiPS preseason. On the other hand ZiPS projects Adam Dunn to have slightly higher rates of production than the preseason projection.

      Vote -1 Vote +1