Rotographs Top 300 Now Available
On Friday we released our positional rankings with the promise of a Top 300 forthcoming. It is now available on the spreadsheet which you can find in the original post here. Our own Dan Schwartz has embedded the Top 300 for us below:
UPDATE: I’ve adjusted the rankings for instances were one or more rankers didn’t include a guy as pointed out in the comments. Now those guys get a last ranked+1 for that particular ranker. (ex. I ranked 345, so everyone else gets a 346 for the purposes of this rank). It’s not perfect since we didn’t all rank the exact same set, but at least this skews on the side of bringing the guys down to where they belong instead of artificially inflating them within the average.
Rankings based on 5×5 roto, 12 teams, single-catcher with Middle Infield and Corner Infield spots, 14 hitters and 9 pitchers. Each ranker had their own mix of projections and intuition.
Last Update: 2/17/15
I thought Pods had Billy the Kid in his top 20?
I believe that was based on the 15-team format of LABR. This is for 12-teams.
…and that matters because?
His values change based on league settings. And everyone’s should, to be honest, whether you’re using pure math, a mix of math and personal touch, or pure personal touch (aka gut).
How does that matter? If you have a 345 guys to choose from or 276, there shouldn’t be much of a difference if you drop the worst 69 guys from your list of 345.
How do you figure? The number of teams and player pool used is *always* a factor.
Proportionally it’s basically the same. If you drop the worst 69 of the 345 players of a 15 team league, you would have the same player pool than if it was a 12 team league, excluding some shifting in catchers because you’ll easily have more than 6 catchers in that group of 69 players.
It’s not about just dropping the 60 worst players. It’s about the replacement level options, and how many SB are going to be had in a league and the % that Hamilton delivers
I’m quite aware of the replacement level options. So if you add 40 hitters to that pool of replacement players, the valuations are not going to change dramatically.
I’m participating in both the NFBC Primetime (12 team) and NFBC Main Event (15 team) so I’ve done the calculations. Most of the Top 100 doesn’t move that much (0-4 spots) with the exception of catchers. Billy Hamilton goes from #41 in my Primetime rankings – based solely on math – to #37. The variation in the expert lists in this article is greater than that difference.
I’m not saying that you don’t make adjustments because the player pool changes, but the change from 12- to 15-team format has less impact than you think. As I already said the variability in the expert lists above is more variable than the calculations you’ll find in a 12 vs. 15 team format.
If you do the math solely on Hamilton’s SB’s and 23 rounds of hitters, the percentage of SB’s Hamilton gets from the 23 round player pool goes from 2.6% to 3%. Its not much.
mdecav: Replacement level is calculated for each position, not one figure for the entire drafted player pool. Therefore a change in the number of teams will change the replacement level distinctly at each position.
Paul is correct. When I say that values change depending on format, this is what I mean.
The only thing that should change should be catcher rankings between a 15 team and 12 team format. Otherwise there isn’t going to be much difference in the Top 100
mdecav – pretty much any valid value formula takes into account the specific player pool (# of starters at each position) – the actual $ value of the same HRs/R/RBI/SB change depending on who’s in the player pool. Losing the bottom 69 players doesn’t just drop them from the list, it impacts the value of the remaining players’ stats
Is mdecav Russian for “noob”?
It’s not the same. Particularly with someone like Hamilton who is a specialist that excels in a category. Pod can probably expound or you can call up his piece on why he took Hamilton where he did in LABR. But it’s not the same regardless of team count and roster size. Those differences all matter.
The waiver wire in a 10-team and 15-team is VASTLY different and thus how you’d draft in those setups is different. A 12-team is somewhere in between, but still different than both.
In 10-12 team leagues steals are more plentiful on the waiver wire. Guys like Dyson, Fuld etc.
A simple example:
Say there is 4 team league with 3 RP spots. Only 12 closers need to be owned. Closers’ value are greatly devalued because there are plenty of them on the waiver wire. Why would I take Chapman with the 50th pick when I can get a guy who make up 80% of his value freely on the waiver wire?
I don’t want to be this guy… but z-scores, Tulo =75?
Can we throw out the column where Trout is not number 1?
C’mon, its not like he ranked another hitter above him. For a lot of teams last year, having Kershaw basically carried your pitching staff for the season.
Had Trout stolen 30 bases and had that avg 15pts higher I’d be with you, but Kershaw’s 2014 was historic….deserves one waivering vote from the Trout bandwagon.
The only reason I question it is that I find it really really REALLY hard to believe that anyone (including Zach) would draft that way. Even if you make the bare minimum assumption that all leagues, no matter what the format, include some sort of RP category (SV, HD, SVHD, etc). That is one category where Kershaw does nothing for you. Unless the league has an “up high strikeout” category, Trout will contribute to every single one. I understand that the production of one player in four categories can equal the production of one player in five categories, but when you are talking about Trout, it really is not a factor.
But hey, it is only my opinion as Trout at 2 was Zach’s.
Also, if your true intent was to draft Kershaw, but you had the first (or I would even argue second pick if it was McCutchen), you could draft Trout and, more often then not, trade him for Kershaw and a decent outfielder.
I believe that Zach’s rankings are almost entirely based of objective valuations.
zach is presenting his rankings based on his FVARz approach to fantasy valuation (off whatever projections he’s using). He has to present the data objectively. His drafting would be subjective. Yes, i’m talking for Zach. but i know he’d say this because I’ve picked his brain on presenting rankings (using his FVARz approach).
While uncommon, Kershaw has been going before Trout in some drafts.
i wouldnt draft kershaw over trout, but based on z-score approach which i know zach uses, he is #1 based on his zsum relative to other pitchers. he to other pitchers outvalues trout to next few hitters. its when non-zscore approaches such as SGP (70/30 hitter/pitcher valuation) devalues the pitcher. but using z-sums, kershaw is number 1 in my NFBC format valuation as well (although again, i wouldnt draft him in top 3).
My apologies in advance as this question may be asinine, but what is the point of these rankings?
What I am specifically asking is what value does Zach’s ranking provide or any ranking based solely on a calculated score (read: I really do not mean to target Zach. He seems like a decent enough dude and I do enjoy the free content he provides.)? If I wanted such a list, could I not take the same projections and calculate the same scores myself?
My thought was that the point of these rankings were to have that subjective element included. To be able to have another, informed take on how players rank against each other in a draft (or keeper) scenario?
Yeah–what’s up with Zach?
Rotographs crew…thank you as always for putting this together. Your guys’ page is my main resource for Fantasy Baseball.
fyi: I clicked on the link at the top of this article which led me to the positional rankings article from last week and I downloaded that spreadsheet as instructed but don’t see a tab for the top 300 unless I’m missing something. Can you let me know if there’s a different spreadsheet I’m missing or something? Thanks as always.
this should be fixed now J.
Best,
Big D.
Sweet. Thanks, Daniel.
Will there be any info provided on how each writer came up with their rankings?
Pod Projections converted into dollar values and sorted in descending order. No manual adjustments made after.
I project as follows:
I rank by z-sum
I rank by position adjusted z-sum (using FVARz approach)
I rank using best ADP
I then average (weighing z-sum which is actual relative value of the 5 categories for hitters and pitchers that are z-scored) as follows:
(z-sum + z-sum + posadj z-sum + ADP)/4
This incorporates actual value*2; positionally adjusted value based on the league’s format and incorporates ADP.
I will argue for this moving forward.
I used a 12-team SGP model with some personal changes to some players with a consensus of several projections. I found I weighted hitters more than pitchers while adding my ranks.
Ha… I downloaded the positional breakdown probably 5 minutes before the Top 300 tab was added.
Thanks for all the work guys
How much higher would Rendon rank if he was 2B eligible, like he is in Yahoo? Thanks!
He’s 2B eligible everywhere (28 games), so it should be included in his rank.
Wasn’t sure since he was listed as a 3B here, even though he should clearly be drafted as a 2B everywhere this year. Glad to see you’re high on him, though, Paul. Great podcast btw, listening right now at work.
Do you any of you guys have a strategy for a shallow 10 team h2h league? Really simple format (No MI/CI, 3 OF, 1 UTL), so only 9 offensive positions. Normally I don’t mess with leagues this shallow, but its a big money league I’ve been in for almost 10 yars. Im thinking with only 3 OF spots needed, I’m going to focus on the infield positions early on. Thanks.
personally, in h2h (especially shallow) i want absolute value – most fantasy stats. dont care about position. dont care about devaluing pitchers. I want best hitters/pitchers available at their projected value. You can always pick up supplemental guys if they get hurt since its shallow. Pederson, Tomas – risk = reward with them for their potential counting stats also.
I would ignore injury rish and get the best possible player. I could see myself taking Tulo second in this league. Upside is key
I’ll echo these sentiments. With the thick waiver wire, I’m less concerned about high floor. I’d rather have Javy Baez than Erick Aybar in this case. If Baez pops, it’s huge. If not, I can probably pick up Aybar or an Aybar near-clone.
My thoughts exactly. I can see myself going all in on guys like Tulo and Springer. Thanks guys!
12 teams with what lineups?
See orig. ranking post – http://www.fangraphs.com/fantasy/rotographs-positional-rankings/
Thanks!
Does Miggy get to #3 if he still has 3B eligibility in your league?
Not with me in this shallow league type … 3B aren’t getting any bump in value, only catchers.
Santana has C in Yahoo, how much does that alter his value? I imagine significantly. Catchers who dont catch are the only ones worth paying for, IMO.
Harper preceded by Marte and Pence, and followed immediately by Dickerson. No one ranked him higher than 25. No one is betting on a big breakout, I guess. Who would’ve thought in year 4, that’s where he’d be.
The more I dig into Harper, the more I think his upside is limited. What he has been is what he is going to be. I am also worried that his injuries are always going to be an issue.
Way too premature. His 1st 2 season were HUGE on a per game pace for a 19 and 20 yr old. So he had 1 down injury plagued year last year. I dont see how you can say he is what he is. He hit 20hr as a 19 and 20 yr old in limited games.
Zach – What did Mookie Betts do to hurt you so bad????
What are people’s thoughts that one ranker has Kershaw ranked #1 overall?
Doesn’t bother me. It’s like saying “Kershaw is REALLY good” versus “Kershaw is really good.” Decide your level of enthusiasm and draft accordingly.
Agree with Buzzo – I am okay with this. It’s not like someone is pimping Bryce Harper or Tulo or Adam Jones to go first. Perfectly reasonable, even if I would prefer Trout.
Are there any dollar values available for your top $300? An average assigned dollar value for each player?
Thanks,
Easy answer to this ? is no but its formatted for leagues as such: c-1b-2b-3b-ss-ci-mi-5of-2u-9p and used (Steamer+FANS)/2 projections. Maybe for a post, i can run the $ values. I will talk with RG guys about doing said $ values.
I noticed Zach had some guys ranked much lower than others, like Pedro Alvarez at 643 and Brett Lawrie at 566.
Take Alvarez, for example. With Zach’s ranking of 643, Alvarez ends up at the very bottom of the list, with an average of 284.4. Take out Zach’s ranking and Alvarez’ average ranking is 194.75. In a 12-team league, the difference between 284th and 195th is the 24th round versus the 17th round. Zach has 9 guys ranked 350 or higher and I think their overall rankings get skewed because of this.
Just for fun, looking at just the top 250 (average ranking), here is how each person’s rankings correlate to the average of all the rankings:
Paul: 0.94
Dan: 0.93
Mike: 0.93
Jeff: 0.92
Zach: 0.88
zachs rankings are purely objective, and based on his dated and flawed z-score approach (fvarz or some ridiculous moniker he came up with). This issue stirs up comments every year. His rankings are pretty much useless. Given the diligence put in by the other writers, and their established records, not sure why they continue to publish his crap.
Well, last year everyone’s brains blew up when Zach ranked Jose Altuve as a top 10 player. Reading the comments from last year, knowing how it turned out, is pretty hilarious reading.
His rankings are different. That doesn’t make them useless.
More just for fun:
If you remove the MAX and MIN values from each player ranking, there are a few pretty big changes in the ranks. These are the players whose values changed by more than 10 average rank points after removing the “outlier” rankings. (apologies for the crappy copy/paste).
Columns are Original ranking order / Name / Original Ranking Average – Adjusted Ranking Average after removing MAX and MIN from the mean.
ORIG. NAME AVG-ADJ
286 Pedro Alvarez 71.7
271 Brett Lawrie 56.3
211 Justin Verlander 29.5
107 Mookie Betts 27.2
234 Aramis Ramirez 26.7
75 Aroldis Chapman 21.5
298 A.J. Burnett 20.1
258 Javier Baez 19.1
103 Carlos Santana 18.5
188 Mike Fiers 18.1
192 Huston Street 17.8
82 Joey Votto 17.3
149 Steve Pearce 17.2
219 Xander Bogaerts 15.9
202 Martin Prado 15.5
94 Pablo Sandoval 15.4
117 Elvis Andrus 15.3
122 Hyun-Jin Ryu 15.2
127 Ryan Zimmerman 15.1
165 Anibal Sanchez 14.7
67 Craig Kimbrel 14.1
101 Devin Mesoraco 13.1
259 Wilin Rosario 12.7
214 John Lackey 12.2
111 Eric Hosmer 12.1
55 Carlos Gonzalez 11.4
130 Dellin Betances 11.1
78 Greg Holland 10.9
147 Salvador Perez 10.6
89 Alex Cobb 10.5
175 Joaquin Benoit 10.3
264 Mike Zunino -11.0
247 Tyler Clippard -11.7
241 Wilson Ramos -11.9
288 Danny Duffy -12.0
227 Kennys Vargas -12.1
171 Garrett Richards -12.3
133 Jayson Werth -13.3
285 Dustin Ackley -14.7
268 Chris Owings -14.8
235 Joc Pederson -15.7
242 Jake McGee -16.9
170 Matt Shoemaker -17.9
191 Yasmany Tomas -20.0
229 R.A. Dickey -22.7
Where is position eligibility coming from? Yahoo seems to be much more liberal with dual eligibility than this site. Miggy and Davis have 3b. Santana has C. Changes their value alot.
Thank you so much for making this available for download in spreadhseet form. I’ve been working on a big composite ranking spreadsheet and this makes my life sooo much easier.
13 team 6×6 (OPS and QS) Roto – 2 C, 1B, 2B, 3B, SS, IF, 4 OF, 3 U, 4 SP, 3 RP, 2 P, 6 Bench, 3 DL
I can keep up to 4 of these players for just this year while forfeiting the provided round’s pick:
Jose Abreu – 6
Jose Altuve – 7
Corey Kluber – 16
George Springer – 19
Corey Dickerson – 26
I think I gotta dump Altuve. That’s a ridiculous group, nicely done.
Sorry did not mean to down vote. I thought I was out of the running last season, so I ended up trading a boatload for Abreu, Altuve, and Kluber for this year. Then I ended up finishing in the money anyway.
I have to agree with paul based on the round association. Or I’d actually trade altuve and dixkerson/springer if possible for something even better round association-wise. Nice group.
I noticed that the SP rankings are different in the overall 300 tab than they are in the SP tab – is there a reason for this? Were updates made to the top 300 but not on the positional tabs? Just want to make sure of which one is the most up to date.
To some extent, I assume the answer to this is that the overall 300 can be more impacted by extreme opinions. For example, Scherzer vs. Strasburg. On the SP positional rankings, you see that Jeff (2 vs 7), Dan (4 vs 5), and Mike (4 vs 5) all like Scherzer better. Paul (3 vs 5) and Zach (4 vs 7) like Strasburg. On the top 300 ranking, Dan (22 vs 25) and Mike (26 vs 27) still like Scherzer better, but the gaps between Strasburg and Scherzer are bigger for Paul (25 vs 32) and Zach (27 vs 33), plus Jeff flipped (having Strasburg ranked higher now, 17 vs 21). Even without Jeff flipping, though, the bigger gaps from the guys who like Strasburg better would push him higher in the overall rankings. I don’t know if I explained that well, but basically because there can be more variance in a ranking across positions, the math can come out different, even if the rankings are internally consistent with each other.
I’m not sure what’s going on with Jeff’s rankings, though. The Strasburg/Scherzer reversal isn’t the only time his rankings are different in the top 300 than in the positional rankings. One that stood out to me (possibly not the most extreme) is that Jeff has Joc Pederson ranked 85th among OFs, but 103rd overall (coming in 40th if you filter to OF). Is this a result of something in Jeff’s methodology (comparing within position to across positions), is one a more recent opinion, or is this just an error?
have with Tulo on your DL, that guy always gets hurt.
All your grammar are belong to us.
Thanks for this great work on the rankings. You had mentioned that you would separate the AL and NL players for each position and the Top 300. I see this has been done for SP,RP and C positions but not the others. Do you plan on completing the rest
On the unranked player problem, you might consider logging that average of other rankers if more than one expert listed the player, then using “346” or possibly a lower number if only one of the experts ranked the player. That way you push down outlier rankings while keeping players ranked more than once at the mean of their rank. Of course this may not be an issue for you, but happens often when I mash together multiple lists of experts from different sites.
No Matt Holliday? I have to believe that is an oversight is it not?
do you have this as a google spreadsheet instead of .xls?
Thanks for all the good work you guys do. Will there be an update to the Top 300 list?
Are you guys going to update the ranks in the coming weeks to reflect injuries and recent developments in terms of playing time ect?