(*) ROTISSERIE: Adjusting Mayberry Scores can Improve Roster Decisions

Rotisserie

Patrick Davitt

Mar 2 2012 12:01am

(*) ROTISSERIE: Adjusting Mayberry Scores can Improve Roster Decisions

Patrick Davitt

Mar 2 2012

BaseballHQ.com offers readers a lot of tools and concepts for drafting and auctions. We also look for ways to combine these tools to offer still other ways of looking at the player pools, valuations, and roster management.

For example, Ron Shandler introduced the “Mayberry Method” in 2010. As he put it at the time, one impetus was to “embrace imprecision” in forecasting models and allow fantasy owners to slot batters in ranges of likely outcomes for Power (PX), Speed (Spd), Expected Batting Average (xBA) and Playing Time (PT—AB for hitters, IP for pitchers), each on a simple scale of 0-5.

Mayberry also has a score total, arrived at by totaling the four component scores and then multiplying by PT. PT is counted twice because it is used as a “force multiplier.”

Armed with this simplified scoring system, it should be an easy matter to sort the batter universe by MM Score, giving us an easy top-to-bottom list of the most valuable players for our drafts or auctions. (Such a list might seem more useful in drafts, but in fact having a sense of actual player value allows the owner at auction to ignore the projected dollar values in favor of matching player productive value to the “facts on the ground” at the auction table—who else is bidding, how much budget room bidders have, positional needs, and so on.)

In his earlier writing about his "Portfolio 3 Plan,” Shandler was also interested in the importance of reliability, especially in the early rounds of straight drafts or top-dollar auction prices. His idea, which has proved sound, is that owners simply can’t afford to have a top player blow up, because that essentially destroys a team’s competitiveness. We can tolerate a $7 David DeJesus having an off year, but if a $42 Carl Crawford stinks, we’re dead.

So the Mayberry Method scoring also included letter grades (A,B,C,D, and F) for the three key components of reliability: Health, Experience and Consistency.

This really helps make decisions. If you were deciding between, say, two $20-ish projected shortstops like Hanley Ramírez and Starlin Castro, you’d note their Mayberry scores:

Castro   2435 ACA  70
Ramirez  3235 CBC  65

While the two batters are identical for BA and PT, Castro has a good edge in Speed but Ramirez in Power. That might matter to you.

More importantly, perhaps, you’d note that BHQ, at least, considers Ramirez to be quite a bit riskier, with “C” grades for Health and Consistency (compared to Castro’s “A” grades), and only a one-grade countervailing benefit in Experience.

Looking at the Mayberry batter pool

Let’s sort the batter pool and look at the top 45 batters plus ties, which would take us through three-plus rounds of straight draft and the greater share of auction value:

                          |--Mayberry--|
Rk  Batter           Pos   Code    Score
==  ======           ===  ======== =====
 1  Braun,Ryan        7   4355 AAC   85
 2  Reyes,Jose        6   3545 FCC   85
 3  Cabrera,Miguel    3   5155 AAB   80
 4  Cano,Robinson     4   4255 AAB   80
 5  Tulowitzki,Troy   6   4255 BAA   80
 6  Votto,Joey        3   5155 AAB   80
 7  Granderson,Curtis 7   5425 BAB   80
 8  Hamilton,Josh     7   4255 BBF   80
 9  Victorino,Shane   7   3445 BAB   80
10  Hart,Corey        7   4345 BBB   80
11  Pence,Hunter      7   4345 AAB   80
12  Stanton,Mike      7   5335 ABC   80
13  Lawrie,Brett      5   4525 ADF   80
14  Beltran,Carlos    7   4255 FCD   80
15  Fowler,Dexter     7   4525 BBA   80
16  Pujols,Albert     3   4155 AAC   75
17  Kemp,Matt         7   4335 AAF   75
18  Gonzalez,Carlos   7   4245 BBC   75
19  Ellsbury,Jacoby   7   3345 FBF   75
20  Bautista,Jose    75   5235 AAF   75
21  Holliday,Matt     7   4155 AAA   75
22  Upton,Justin      7   4335 BAB   75
23  Sandoval,Pablo    5   4155 BAF   75
24  McCutchen,Andrew  7   4335 AAA   75
25  Kendrick,Howie   47   3345 BBB   75
26  Napoli,Mike      23   5145 BCF   75
27  Bourjos,Peter     7   3525 ABC   75
28  Fielder,Prince    3   4055 AAD   70
29  Gonzalez,Adrian   3   4145 AAB   70
30  Pedroia,Dustin    4   2345 CAA   70
31  Kinsler,Ian       4   4235 CAA   70
32  Longoria,Evan     5   4145 BAA   70
33  Bourn,Michael     7   1535 AAB   70
34  Young,Michael    53   2345 AAC   70
35  Beltre,Adrian     5   4145 CBD   70
36  Crawford,Carl     7   3335 BAC   70
37  Cuddyer,Michael  73   3245 BAB   70
38  Phillips,Brandon  4   3245 AAA   70
39  Castro,Starlin    6   2435 ACA   70
40  Youkilis,Kevin    5   4235 CBB   70
41  Jeter,Derek       6   1445 BAC   70
42  Ortiz,David       0   4145 BAB   70
43  Ramirez,Aramis    5   4145 BBD   70
44  Zobrist,Ben      47   4235 BAF   70
45  Utley,Chase       4   3335 DBB   70
46  Francoeur,Jeff    7   4235 AAC   70
47  Weeks,Rickie      4   4235 FBA   70
48  Molina,Yadier     2   3155 ABC   70
49  Craig,Allen       7   4145 CDC   70
50  Kipnis,Jason      4   4235 AFA   70
51  Morrison,Logan    7   4235 ACA   70
52  Rolen,Scott       5   3245 FCC   70
53  Duda,Lucas       37   4235 ACB   70
54  Aley,Dustin       4   2525 ACB   70
55  Jones,Chipper     5   4145 CBA   70
56  LaHair,Bryan      7   5225 ABB   70

Not a bad list, though probably a little controversial with Albert Pujols buried below Brett Lawrie and Carlos Beltrán because of their higher Spd scores, and other anomalies sure to catch the eye of the discerning reader.

We instinctively know that Pujols is a first-rounder, and that Lawrie and Beltran (and José Reyes, for that matter) are not. Why? Because we perceive Pujols as a much surer bet. In other words, we want to avoid risk in the early rounds, and Lawrie, a rookie, and Beltran, a longtime injury bug, are very risky batters. So is Reyes, with his checkered health. And note that all three of these batters have “F” reliability ratings, Lawrie for Consistency (and a “D” for Experience, which could have been an “F”), and Beltran and Reyes for Health.

Adjusting for Reliability

So what if we adjusted the overall Mayberry score to take those Reliability scores into account? We did just that. Using a spreadsheet, we broke the Mayberry scores out into separate digits, and then after some trial-and-error, assigned bonuses or penalties for the letter grades:

Grade  Bonus
=====  =====
  A     +10% 
  B     + 5%
  C       0%
  D     -10%
  F     -20%

The bonuses were multiplied together, not applied cumulatively. So an “AAA” reliability score would give the batter a +33% bonus (1.1 x 1.1 x 1.1), a “BBB” would be +16%; all the way down to an “FFF,” which would mean a -49% penalty. We then recalculated the batter pool:

Adj   Un                           |-Mayberry-| |--Adjust-|
 Rk   Rk  Chg  Batter         Pos  Code   Score Bonus Score
===  ===  ===  ======         ===  ============ ===== =====
 1     1    0  Braun,Ryan        7  4355 AAC  85  +21%  103
 2     3   +1  Cabrera,Miguel    3  5155 AAB  80  +27%  102
 3     4   +1  Cano,Robinson     4  4255 AAB  80  +27%  102
 4     5   +1  Tulowitzki,Troy   6  4255 BAA  80  +27%  102
 5     6   +1  Votto,Joey        3  5155 AAB  80  +27%  102
 6    11   +5  Pence,Hunter      7  4345 AAB  80  +27%  102
 7    21  +14  Holliday,Matt     7  4155 AAA  75  +33%  100
 8    24  +16  McCutchen,Andrew  7  4335 AAA  75  +33%  100
 9     7   -2  Granderson,Curtis 7  5425 BAB  80  +21%   97
10     9   -1  Victorino,Shane   7  3445 BAB  80  +21%   97
11    15   +4  Fowler,Dexter     7  4525 BBA  80  +21%   97
12    10   -2  Hart,Corey        7  4345 BBB  80  +16%   93
13    38  +25  Phillips,Brandon  4  3245 AAA  70  +33%   93
14    12   -2  Stanton,Mike      7  5335 ABC  80  +16%   92
15    16   +1  Pujols,Albert     3  4155 AAC  75  +21%   91
16    22   +6  Upton,Justin      7  4335 BAB  75  +21%   91
17    29  +12  Gonzalez,Adrian   3  4145 AAB  70  +27%   89
18    32  +14  Longoria,Evan     5  4145 BAA  70  +27%   89
19    33  +14  Bourn,Michael     7  1535 AAB  70  +27%   89
20    25   +5  Kendrick,Howie   47  3345 BBB  75  +16%   87
21    27   +6  Bourjos,Peter     7  3525 ABC  75  +16%   87
22    70  +48  Ramirez,Alexei    6  2335 AAA  65  +33%   87
23    30   +7  Pedroia,Dustin    4  2345 CAA  70  +21%   85
24    31   +7  Kinsler,Ian       4  4235 CAA  70  +21%   85
25    34   +9  Young,Michael    53  2345 AAC  70  +21%   85
26    37  +11  Cuddyer,Michael  73  3245 BAB  70  +21%   85
27    39  +12  Castro,Starlin    6  2435 ACA  70  +21%   85
28    42  +14  Ortiz,David       0  4145 BAB  70  +21%   85
29    46  +17  Francoeur,Jeff    7  4235 AAC  70  +21%   85
30    51  +21  Morrison,Logan    7  4235 ACA  70  +21%   85
31    56  +25  LaHair,Bryan      7  5225 ABB  70  +21%   85
32    18  -14  Gonzalez,Carlos   7  4245 BBC  75  +10%   83
33    58  +25  Teixeira,Mark     3  4135 AAB  65  +27%   83
34    59  +25  Jones,Adam        7  3235 BAA  65  +27%   83
35    81  +46  Young,Chris       7  4315 AAB  65  +27%   83
36    91  +55  Escobar,Alcides   6  1525 ABA  65  +27%   83
37    36   -1  Crawford,Carl     7  3335 BAC  70  +16%   81
38    41   +3  Jeter,Derek       6  1445 BAC  70  +16%   81
39    48   +9  Molina,Yadier     2  3155 ABC  70  +16%   81
40    53  +13  Duda,Lucas       37  4235 ACB  70  +16%   81
41    54  +13  Ackley,Dustin     4  2525 ACB  70  +16%   81
42    55  +13  Jones,Chipper     5  4145 CBA  70  +16%   81
43   102  +59  Ethier,Andre      7  3135 AAA  60  +33%   80
44   104  +60  Upton,B J         7  3405 AAA  60  +33%   80
45   107  +62  Markakis,Nick     7  2235 AAA  60  +33%   80
46   114  +68  Swisher,Nick      7  4125 AAA  60  +33%   80
47    64  +17  Gardner,Brett     7  1525 ABB  65  +21%   79
48    76  +28  Howard,Ryan       3  5125 BAB  65  +21%   79
49    79  +30  Wieters,Matt      2  4135 ABB  65  +21%   79
50    87  +37  Stubbs,Drew       7  3505 AAC  65  +21%   79
51    88  +37  Bonifacio,Emil  675  1525 ABB  65  +21%   79
52    92  +40  Ibanez,Raul       7  4135 AAC  65  +21%   79
53    96  +43  Espinosa,Danny    4  4315 ACA  65  +21%   79
54    98  +44  Barney,Darwin     4  1435 ABB  65  +21%   79
55    40  -15  Youkilis,Kevin    5  4235 CBB  70  +10%   77
56    28  -28  Fielder,Prince    3  4055 AAD  70  + 9%   76
57   101  +44  Andrus,Elvis      6  1335 AAB  60  +27%   76
58   112  +54  Santana,Carlos   23  4125 ABA  60  +27%   76
59   121  +62  Jackson,Austin    7  2505 AAB  60  +27%   76
60   122  +62  McCann,Brian      2  4125 ABA  60  +27%   76

Adjusting for risk, Reyes falls off the list, and Miguel Cabrera, Robinson Canó, Troy Tulowitzki and Joey Votto step up a place each. Dexter Fowler, already a tremendous bargain if we believe current ADPs, looks even better, climbing to 11th overall. And several players make huge leaps—Matt Holliday and Andrew McCutchen climb from the 20s into the top 10, and Brandon Phillips jumps 25 full places on the basis of solid skills plus “AAA” reliability.

Of course, we wouldn’t necessarily take Holliday, McCutchen or Phillips that high, if we suspected from ADPs and/or knowledge of competitors that any of them would be available two rounds later or for $10 less. But we might well want to go the extra buck or nab a guy a round early, knowing that he stands a top chance of delivering the goods.

We might also want to look at other batters who made the Adjusted Mayberry list who weren’t on the unadjusted list:

Adj   Un                           |-Mayberry-| |--Adjust-|
 Rk   Rk  Chg  Batter         Pos  Code   Score Bonus Score
===  ===  ===  ======         ===  ============ ===== =====
22    70  +48  Ramirez,Alexei   6  2335 AAA  65  +33%   87
36    91  +55  Escobar,Alcides  6  1525 ABA  65  +27%   83
35    81  +46  Young,Chris      7  4315 AAB  65  +27%   83
33    58  +25  Teixeira,Mark    3  4135 AAB  65  +27%   83
34    59  +25  Jones,Adam       7  3235 BAA  65  +27%   83
46   114  +68  Swisher,Nick     7  4125 AAA  60  +33%   80
45   107  +62  Markakis,Nick    7  2235 AAA  60  +33%   80
44   104  +60  Upton,B J        7  3405 AAA  60  +33%   80
43   102  +59  Ethier,Andre     7  3135 AAA  60  +33%   80
54    98  +44  Barney,Darwin    4  1435 ABB  65  +21%   79
53    96  +43  Espinosa,Danny   4  4315 ACA  65  +21%   79
52    92  +40  Ibanez,Raul      7  4135 AAC  65  +21%   79
50    87  +37  Stubbs,Drew      7  3505 AAC  65  +21%   79
51    88  +37  Bonifacio,Emil 567  1525 ABB  65  +21%   79
49    79  +30  Wieters,Matt     2  4135 ABB  65  +21%   79
48    76  +28  Howard,Ryan      3  5125 BAB  65  +21%   79
47    64  +17  Gardner,Brett    7  1525 ABB  65  +21%   79

While many of these batters are not top-tier, they are nonetheless much less risky than batters with similar skills, and therefore inherently more valuable. Among catchers, for instance, BAL backstop Matt Wieters is the 79th-ranked batter overall on the unadjusted list. But giving him points for his ABB reliability boosts him nearly 30 places, and takes him past such seeming prize catchers as Miguel Montero and Mike Napoli.

Other big gainers and decliners

Other batters had huge reliability-based jumps in score but did not make the list of top batters: Kurt Suzuki (C, OAK +80 places), James Loney (1B, LA, +73), Danny Valencia (3B, MIN, +69), Billy Butler (DH, KC, +62) and Garret Jones, (1B/OF, PIT, +60) all jumped 60 spots or more.

And we would be remiss if we didn’t show the big decliners based on reliability adjustments:

Adj   Un                           |-Mayberry-| |--Adjust-|
 Rk   Rk  Chg  Batter         Pos  Code   Score Bonus Score
===  ===  ===  ======         ===  ============ ===== =====
184   19 -165  Ellsbury,Jacoby  7  3345 FBF  75  -33%   50
149   14 -135  Beltran,Carlos   7  4255 FCD  80  -28%   58
202   86 -116  Murphy,Daniel  354  2245 FDD  65  -35%   42
190   75 -115  Davis,Ike        3  4225 FDC  65  -28%   47
192   80 -112  Cespedes,Yoenis  7  2245 AFF  65  -30%   46
123   13 -110  Lawrie,Brett     5  4525 ADF  80  -21%   63
158   52 -106  Rolen,Scott      5  3245 FCC  70  -20%   56
124   26  -98  Napoli,Mike     23  5145 BCF  75  -16%   63
204  109  -95  Goldschmidt,Paul 3  4125 AFF  60  -30%   42
261  167  -94  Lowrie,Jed      65  2405 FFF  55  -49%   28
 95    2  -93  Reyes,Jose       6  3545 FCC  85  -20%   68
143   65  -78  Morse,Michael   37  4135 DCC  65  -10%   59
205  127  -78  Altuve,Jose      4  1335 AFF  60  -30%   42
144   68  -76  Cruz,Nelson      7  4135 DCC  65  -10%   59
 83    8  -75  Hamilton,Josh    7  4255 BBF  80  -12%   71
120   49  -71  Craig,Allen      7  4145 CDC  70  -10%   63
104   35  -69  Beltre,Adrian    5  4145 CBD  70  - 5%   66
112   44  -68  Zobrist,Ben     47  4235 BAF  70  - 8%   65
125   57  -68  Hosmer,Eric      3  2245 AFA  65  - 3%   63
 90   23  -67  Sandoval,Pablo   5  4155 BAF  75  - 8%   69

Some big names here, and some relatively well-known batters who will be sought after at draft. It might be worth letting someone else find them.

Conclusion

The simplicity of the coding and the overall score make adjusting Mayberry easy—another advantage of the Method. For instance, we could also award bonuses and penalties based on positions, to adjust for scarcity. We could add 10% to catchers, subtract 5% from OFs, do something between for scarce IFs. Of course, no formulas (tweaked or otherwise), projections, or applied expertise can or should substitute for your own thinking. Depending on your own risk tolerance, for instance, you might be much more forgiving of low “Consistency” grades and harsher about “Health” grades. You might be feeling more positive about an untried import like Yoenis Céspedes, whose legitimate “AFF” reliability score sinks his Mayberry by 30%.

In his recent essay on 2012 Mayberry target players, Shandler wrote:

“The reliability grades of both (Brett) Lawrie (lack of experience) and (Carlos) Beltran (health history) would make me reluctant to draft them too high. (Corey) Hart and (Hunter) Pence (both surprisingly high Mayberry players) are much more interesting...”

What Shandler was doing was informally adjusting his list by looking at batters’ reliability scores. It’s no great leap to formally adjust those scores, to quantify batter reliability. In fact, it seems like a natural next step.