First, a word. Everything I write from here on down is going to be based on ESPN's projected stats for each player. You may have different projections, and that's OK. There may be a different site you like, or you may have your own (I recommend a rolling 3-year average if that's the case). But many, if not most, people use ESPN for their fantasy leagues, as I do. So I use their projections. Of course, this also means that my analysis is also only as good as the projections I use; I don't want angry letters because a player failed to live up to ESPN's idea of what they should do this season. Chances are, I'll be just as frustrated as you anyway. But I find the ESPN projections to be quite useable- it certainly helped me a ton last year- and so it will be those projections I use.
The idea behind a roto league- and I'm not sure you need this explained, but indulge me- is to aggregate the best numbers possible in as many of the 10 categories as possible: Runs (R), Home Runs (HR), Runs Batted In (RBI), Stolen Bases (SB), and Batting Average (BA or Avg) for hitters; Wins (W), Strikeouts (K), Earned Run Average (ERA), Saves (Sv), and Walks + Hits per Inning Pitched (WHIP) for pitchers. Insofar as winning your league is concerned, the best thing you can do for any given category is to draft the person who will (or is at least projected to) lead the league in that category. Anyone else you draft is sub-optimal with respect to that goal, and that can be represented by a percent. However, there are positional constraints, and positional expectations. For example, you need at least one first baseman. No first baseman is going to lead the league in steals. As such, low steal totals for a first baseman are not great, but not as bad as a low steal total for an outfielder or middle infielder. So the same SB total for a 1B should be more valuable than an identical total for a middle infielder or outfielder.
To this end, the metric I use is Value Under Maximum, or VUM (though it really should undergo a name change). The formula for VUM for any given statistic is as follows:
VUM= (2*(x-min(l)))/(max(p)+max(l)-(2*min(l)))
where x is the projected value for the stat (Avg, K, SB, etc), min(l) is the MLB minimum projection, max(l) is the maximum MLB projection, and max (p) is the maximum projection for that position. Put another way, I take the projection above minimum performance (say, 0 steals or a .200 BA) as a percent of the maximum league value over the same number:
((x-min(l))/(max(l)-min(l))
do likewise for the projection with respect to the position maximum
((x-min(l))/(max(p)-min(l))
and average the two. All values are between 0 (the absolute worst in all MLB) and 1 (the absolute best). You can do this for every statistic and add them all up, giving you a value between 0 and 5 for all players. You can then rank them by this aggregate value.
This method allows you to compare various types of players- not just pitchers and hitters, but a 20 HR/20 SB shortstop to a 40 HR outfielder. Every little bit of production gets a numeric value, and aggregating them allows you to compare disparate strengths and weaknesses in an objective manner. VUM is also not roster/league dependent. Because it works of maxima and minima, it doesn't matter how many teams are in the league or what the specific roster composition is, as long as the league is 5x5 roto. There are, however, a few caveats.
Caveat #1: As stated above, this value is only as accurate as the projections. Insofar as the projections are off, VUM will not be an accurate way to measure value.
Caveat #2: This measure overvalues relievers. Closers are notoriously hard to predict from one year to the next; this is actually true for relievers in general. VUM does not care about this, and sees winning saves as just as valuable as winning steals (which is true). But it's difficult to predict who will lead the league in saves; it's difficult to even predict who will close for a team in June/July/August. VUM loves Craig Kimbrel, in part because he's projected to lead the league in saves with 100 Ks. If this is true, Kimbrel will indeed be valuable. But VUM loves him as a 2nd round pick. There are two things to note about this. First, saves can be had much later than the 2nd round. Second, no one is going to draft a closer until at least the 6th or 7th round, possibly later. It would be stupid to take Kimbrel at a time when there's no danger of someone else picking him. All VUM is saying is that, if he hits his projections, Kimbrel will return 2nd round value in terms of helping you win your league. I devised this system, and I still wouldn't draft Kimbrel that high; aside from the fact that closers can be had 15 rounds later, there's no telling if he's going to be this year's Soria/Broxton/Franklin/etc. Don't pay for saves.
Caveat #3: VUM is an unfeeling creature. It doesn't take into account what the more likely outcome is (the basis of TMR's wonderful draft manifesto). VUM loves Brett Lawrie as a 5th/6th round pick, and loves him slightly more than Pablo Sandoval. Lawrie is probably more talented than Sandoval, but I still think Sandoval is more likely to hit his projections than Lawrie. If Lawrie gets figured out by pitchers, he'll be less valuable. If not, he'll be as valuable as predicted. There's less variation with Sandoval, and so while Lawrie may have more potential, Sandoval is safer.
Caveat #4: This VUM only works for 5x5 roto. You can add in more stats as you add more categories, but those numbers will be different from the ones below. I'm in a 7x7 league, and VUM for that league ranges between 0 and 7, with different rankings because additional measures are used. VUM as a general idea still works, but these specific numbers will not.
Caveat #5: VUM works best for filling starting positions: 1B, 2B, SS, 3B, C, and P. It works less well for MI/CI spots, and even less well for Utility/Bench spots. That's not to say it's useless- it allows you to compare having a given SS vs. a given OF on your bench, but it's not quite as good as it is for filling the 14 starting positions.
Caveat #6: Not so much a true caveat, but it's worth noting that the same player with the same statistics has different VUM values for different positions if he has multiple eligibility. This is simply because different positions have different baselines. It also means that there is an optimal position to play someone, and the same player can be a bargain or a reach depending on positions. This is less pronounced at the top of the draft- the person immediately behind Miguel Cabrera the 3B is Miguel Cabrera the 1B- and in general isn't a huge factor. But it is something to keep in mind in the middle to late rounds.
With all that out of the way, I'll get into my positional rankings. I'll go top 15 around the infield, top 50 outfielders, top 50 starters, and top 20 relievers. I'll have VUM values next to each name, as well as a brief discussion.
Catcher
1) Mike Napoli (3.14)
2) Carlos Santana (2.81)
3) Joe Mauer (2.67)
4) Brian McCann (2.66)
5) Matt Wieters (2.57)
6) Miguel Montero (2.54)
7) Buster Posey (2.39)
8) Alex Avila (2.27)
9) Yadier Molina (2.23)
10) Kurt Suzuki (2.19)
11) Russell Martin (2.14)
12) JP Arencibia (2.00)
13) Salvador Perez (1.93)
14) Geovany Soto (1.87)
15) Wilson Ramos (1.86)
Mike Napoli is far and away the #1 catcher. Don't worry about the lack of at-bats- it's comparable to Brian McCann, admittedly one of the most well-rested catchers in the league. After that it's Carlos Santana, then Mauer/McCann, and then you might as well wait. Of all the catchers Posey has the most potential to outperform his ranking, but this still assumes a 16 jacks and a .291 batting average. I like Wieters, but he's not objectively better than Miguel Montero. If you're in a 12 or more team league, Salvador Perez (ADP 321) and Russell Martin (ADP 317) have the potential to be great value picks.
First Base
1) Albert Pujols (4.06)
2) Miguel Cabrera (3.75)
3) Joey Votto (3.68)
4) Adrian Gonzales (3.58)
5) Prince Fielder (3.49)
6) Mark Teixeira (3.22)
7) Michael Cuddyer (2.89)
8) Eric Hosmer (2.81)
9) Paul Konerko (2.78)
10) Mike Napoli (2.72)
11) Michael Young (2.67)
12) Lance Berkman (2.65)
13) Michael Morse (2.65)
14) Mark Reynolds (2.58)
15) Carlos Lee (2.47)
First off- you might have Cabrera at 3rd (not unreasonable), Cuddyer in the outfield (though he's slightly more valuable at 1B), Young at 3B, and you almost certainly have Napoli at catcher. So, for reference, the next five are Freddie Freeman, Adam Lind, Carlos Santana (someone you also have at catcher), Paul Goldschmidt, and Gaby Sanchez. In general, if you don't get one of the first six names you're better of waiting until rounds 8-10 or later to draft a first baseman. First base is sufficiently deep that you should let people reach for talent (Hosmer, Freeman, Konerko) and take what comes to you. However, there are some noticeable talent gaps here. The difference between Pujols and Cabrera is almost as large as the difference between Cuddyer and Lee. If you can snag Pujols with your first round pick, do it. It's a huge value. And there's no reason not to take a 1B with one of your first three picks if you can get Tex or better- they all grade out to top-30 picks. And you need a good 1B to buttress your counting stats.
Second Base
1) Robinson Cano (3.68)
2) Ian Kinsler (3.63)
3) Dustin Pedroia (3.62)
4) Brandon Phillips (3.13)
5) Ben Zobrist (3.06)
6) Dan Uggla (2.78)
7) Chase Utley (2.71)
8) Howard Kendrick (2.69)
9) Ryan Roberts (2.66)
10) Rickie Weeks (2.62)
11) Neil Walker (2.57)
12) Jason Kipnis (2.42)
13) Danny Espinosa (2.40)
14) Jemile Weeks (2.37)
15) Dustin Ackley (2.36)
Cano, Kinsler, and Pedroia are pretty tightly clustered, even with Kinsler not getting as many at-bats as the other two. That's what 30 steal/30 homer potential gets you. If you want to drop Kinsler based on injury potential, though, I won't argue with you placing him 3rd. There's also major value/reach potential here. Mock Draft Central has Uggla going 4th among second-sackers (53 overall), Rickie Weeks 7th (77th overall), and Ackley going 10th (134); meanwhile Zobrist (8th 2B/81 overall), Kendrick (9/108) and especially Roberts (15/196) have the potential for great value. Those last three all have positional flexibility too, making them great adds generally. Roberts seems especially puzzling for such a high ranking, especially with the names below him, but in spite of a .260 batting average he's got 15/15 potential with decent run/RBI numbers. Ackley and J Weeks seem to suffer for the offenses around them, and though Ackley has great potential I still wouldn't place him in the top 10.
Third Base
1) Miguel Cabrera (3.85)
2) Jose Bautista (3.71)
3) Evan Longoria (3.50)
4) Hanley Ramirez (3.31)
5) Adrian Beltre (3.19)
6) David Wright (3.10)
7) Brett Lawrie (2.93)
8) Pablo Sandoval (2.84)
9) Michael Young (2.74)
10) Aramis Ramirez (2.72)
11) Kevin Youkilis (2.71)
12) Ryan Zimmerman (2.70)
13) Mark Reynolds (2.63)
14) Ryan Roberts (2.58)
15) Alex Rodriguez (2.56)
Cabrera has slightly more value at 3B (if/when he gets there) than 1B, but it's a toss-up. Hanley Ramirez is a slightly better play at SS, but I guess it's a matter of how you draft if you decide to slide him over after 10 or 20 games. Numbers 9 to 12 are a pretty close cluster- a HR there, a few runs there makes the difference- so I wouldn't worry about who you prefer over whom in that range. Young and his boring stats excepted, it's injury risks that cluster them down there anyway. A-Rod is the big stay-away here. Yes, he's A-Rod. On the other hand, he's projected to get fewer than 500 at-bats, which sounds right, and he's been trading contact for power, which means a low BA. Given that he doesn't steal anymore either, the fact that he's next to a 15/15 Ryan Roberts actually sounds about right. Draft A-Rod at your own risk.
Shortstop
1) Troy Tulowitzki (3.56)
2) Hanley Ramirez (3.36)
3) Jose Reyes (3.11)
4) Jimmy Rollins (2.86)
5) Elvis Andrus (2.69)
6) Starlin Castro (2.68)
7) Alexei Ramirez (2.61)
8) Marco Scutaro (2.54)
9) JJ Hardy (2.50)
10) Derek Jeter (2.45)
11) Asdrubal Cabrera (2.45)
12) Dee Gordon (2.31)
13) Jhonny Peralta (2.31)
14) Erick Aybar (2.24)
15) Emilio Bonifacio (2.19)
Hanley is slightly more valuable at SS than 3B, and given the relative strengths of the positions I'd probably use him at short. In general, though, if you don't get one of the top 4 there's not much difference in numbers five to eleven. It's also worth noting the names not on this list- Ian Desmond, Stephen Drew, Alex Gonzales, Rafael Furcal, and Jed Lowrie. Those are middle infield/bench options at best.
Outfield
1) Ryan Braun (4.08)
2) Matt Kemp (3.96)
3) Jacoby Ellsbury (3.71)
4) Jose Bautista (3.63)
5) Justin Upton (3.38)
6) Curtis Granderson (3.32)
7) Carlos Gonzales (3.32)
8) Andrew McCutchen (3.17)
9) Mike Stanton (3.12)
10) Matt Holliday (3.10)
11) Josh Hamilton (2.99)
12) Jay Bruce (2.97)
13) Hunter Pence (2.94)
14) Corey Hart (2.90)
15) Michael Cuddyer (2.88)
16) Ben Zobrist (2.87)
17) Michael Bourn (2.86)
18) Nelson Cruz (2.86)
19) BJ Upton (2.82)
20) Michael Morse (2.73)
21) Alex Gordon (2.72)
22) Lance Berkman (2.71)
23) Shane Victorino (2.66)
24) Adam Jones (2.64)
25) Chris Young (2.64)
26) Melky Cabrera (2.63)
27) Shin-Soo Choo (2.62)
28) Carl Crawford (2.61)
29) Jayson Werth (2.60)
30) Carlos Beltran (2.58)
31) Jeff Francoeur (2.56)
32) Carlos Lee (2.55)
33) Desmond Jennings (2.52)
34) Torii Hunter (2.52)
35) Howard Kendrick (2.51)
36) Nick Swisher (2.50)
37) Ichiro Suzuki (2.49)
38) Angel Pagan (2.49)
39) Brett Gardner (2.44)
40) Jason Kubel (2.44)
41) Josh Willingham (2.43)
42) Nick Markakis (2.42)
43) Coco Crisp (2.40)
44) Cameron Maybin (2.38)
45) Drew Stubbs (2.37)
46) Colby Rasmus (2.35)
47) Andre Ethier (2.34)
48) Delmon Young (2.33)
49) Allen Craig (2.32)
50) Alex Rios (2.31)
Outside the top 15, you're stuck with players who have measurable flaws and/or a large downside. There are a few players outside the top 50 I like- Heyward and Logan Morrison, to name two. Given the upside, I would take them if the draft falls in such a way that I don't have anyone else to take when they come up.
Starting Pitcher
1) Justin Verlander (3.65)
2) Clayton Kershaw (3.63)
3) Roy Halladay (3.53)
4) Cliff Lee (3.35)
5) Felix Hernandez (3.24)
6) Jered Weaver (3.23)
7) Dan Haren (3.20)
8) Cole Hamels (3.10)
9) Tim Lincecum (3.10)
10) CC Sabathia(3.04)
11) Zack Grienke (3.02)
12) Yovanni Gallardo (3.01)
13) Matt Moore (3.01)
14) David Price (2.98)
15) Jon Lester (2.96)
16) James Shields (2.94)
17) Stephen Strasburg (2.92)
18) Matt Cain (2.89)
19) CJ Wilson (2.87)
20) Ian Kennedy (2.84)
21) Mat Latos (2.79)
22) Tommy Hanson (2.76)
23) Michael Pineda (2.75)
24) Chris Carpenter (2.69)
25) Ubaldo Jimenez (2.68)
26) Jeremy Hellickson (2.66)
27) Madison Bumgarner (2.66)
28) Adam Wainwright (2.65)
29) Ricky Romero (2.63)
30) Daniel Hudson (2.63)
31) Josh Beckett (2.60)
32) Josh Johnson (2.51)
33) Matt Garza (2.50)
34) Jordan Zimmerman (2.49)
35) Hiroki Kuroda (2.44)
36) Brandon Beachy (2.42)
37) Shaun Marcum (2.40)
38) Jaime Garcia (2.40)
39) Max Scherzer (2.39)
40) Ervin Santana (2.34)
41) Cory Luebke (2.33)
42) Tim Hudson (2.33)
43) Anibal Sanchez (2.32)
44) Brandon McCarthy (2.32)
45) Ted Lilly (2.31)
46) Justin Masterson (2.30)
47) Doug Fister (2.30)
48) Tim Stauffer (2.29)
49) Scott Baker (2.29)
50) Colby Lewis (2.29)
The usual words of caution about young pitchers, hurlers coming off injury, etc., all apply here. That said, I'm bullish on most of the pitchers listed here, and would avoid most pitchers not on this list. I would, however, keep an eye on Jhoulys Chacin, Wandy Rodriguez and Brandon Morrow (for the Ks), John Danks, Mike Minor, and Bud Norris. The thing about pitchers is that only one of their stats is completely under their control- Ks. The H in WHIP, ERA, and especially Wins are influenced by outside factors, and Wins especially are entirely unpredictable. Given that, I'd look at these ranks as a general guide more so than with positional players.
Relief Pitchers
1) Craig Kimbrel (3.37)
2) Mariano Rivera (3.19)
3) Jonathan Papelbon (2.85)
4) Rafael Betancourt (2.83)
5) Huston Street (2.8)
6) Tyler Clippard (2.73)
7) Drew Storen (2.68)
8) JJ Putz (2.65)
9) Joel Hanrahan (2.64)
10) Kyle Farnsworth (2.58)
11) Ryan Madson (2.57)
12) Jason Motte (2.53)
13) Mike Adams (2.50)
14) Andrew Bailey (2.41)
15) Kenley Jansen (2.38)
16) Jose Valverde (2.37)
17) Brian Wilson (2.36)
18) John Axford (2.32)
19) Joakim Soria (2.22)
20) Grant Balfour (2.18)
I eliminated RP-eligible starters like Chris Sale and Cory Luebke. After that, we see two non-closers on the list: Clippard and Adams. Both are high-strikeout, low-ERA, low-WHIP guys who should get the odd win/save to boost their value. Essentially, you're targeting skills and not roles. Wilson and Axford have potential WHIP issues, which knocks their value down a bit, but they do have a good handle on the job. Overall, these rankings and numbers should be taken with a grain of salt; closers are unpredictable and will be there in later rounds.
Strict Mock Draft
I did a mock draft following these rankings strictly; that is, I took the highest-rated player by VUM on my board regardless of need, likelihood of being available later, etc. I picked with the 8th spot in the 1st round, and here are the results:
Round 1: Jacoby Ellsbury, OF
Round 2: Justin Verlander, SP
Round 3: Craig Kimbrel, RP
Round 4: Jered Weaver, SP
Round 5: Mariano Rivera, RP
Round 6: Yovanni Gallardo, SP
Round 7: Matt Moore, SP
Round 8: Matt Cain, SP
Round 9: Michael Cuddyer, 1B/OF
Round 10: Jonathan Papelbon, RP
Round 11: Rafael Betancourt, RP
Round 12: Huston Street, RP
Round 13: Tyler Clippard, RP
Round 14: Ryan Roberts, 2B/3B
Round 15: Mark Reynolds, 1B/3B
Round 16: Melky Cabrera, OF
Round 17: Kyle Farnsworth, RP
Round 18: Jeff Francouer, OF
Round 19: Danny Espinosa, 2B
Round 20: Colby Rasmus, OF
Round 21: Delmon Young, OF
Round 22: Jemile Weeks, 2B
Round 23: Daniel Murphy, 1B/2B/3B
Round 24: Kurt Suzuki, C
Round 25: Jason Bartlett, SS
Obviously, that's one hell of a pitching staff that should lead across the board. The position players leave something to be desired, though, and that's a problem. It looks like the system overvalues pitchers relative to their actual ADP, and I could likely get a similar pitching staff while taking better hitters in early rounds. The massive and systemic overvaluation of closers is also apparent, with me taking 7 overall, and 6 of the 30 available closers. It makes for good trade ammunition, but that's a risky route to go. As constituted, this is probably a 3rd or 4th place team, but a solid trade or two of pitching for hitting (say, Verlander and Kimbrel for Votto), even if I overpay, leaves me in good shape. This highlights the risk of following the rankings strictly. However, I'd still say this is a solid foundation to build from.
One more thing- it can be a little difficult dealing with fine distinctions on a 0-5 number scale. If you multiply the numbers by 20, making the scale 0-100, it's a little easier to deal with.
No comments:
Post a Comment