SS-Notes, Hitting
We always get some questions about how we juggle your batting order when bench players have to start. (We don't just put the sub in the same spot as the starter because too often the sub is a far weaker hitter than your starter. For instance, you may have Barry Larkin batting third, but if Rafael Belliard is your backup SS you certainly don't want him hitting third if he starts a game!)
Before we do any lineup juggling we figure out who your starters are; the juggling procedures occur after your starters are determined. Here are the rules we follow when juggling lineups.
A) A starter never moves down in the order.
B) A starter batting leadoff, 2nd, or 3rd never moves at all.
C) A starter who is in your 3, 4, 5 or 6 slot never moves higher than the third slot.
D) Since we never move a 3 thru 6 hitter above the 3rd slot, your new leadoff and 2nd batter (if either is subbed for) comes from either a sub or your 7th or 8th hitter (and/or 9th hitter in the AL.) If your leadoff or second batter is out we first look for a batter who has a Y in his steal column to fill that spot. If more than one batter has a Y we put the batters with the better PH rank (the lower the number the better - #1 is best) at the top of the order. Also, a catcher will not be moved to the leadoff spot.
E) In the 3 thru 8 (or 9) spots, we put the guy with the best pinch hit number highest in the batting order. (We never move a starter down in the order. But when a sub comes in the decision of whether to move other guys up, or to bat the sub higher than a starter, is based on PH rank.) Lineup juggling is another reason that ranking your hitters in the PH column is important, as we do put the hitters with the better rank higher up in the order.
F) If there is no one left with pinch hit numbers and the pitcher is up, Scoresheet will still pinch-hit -- with players (without numbers in the PH column) in the order that they are listed on the lineup card. Once those players have been used, the pinch-hitters will come from the taxi squad.
G) When looking for a sub for your DH (or for your 1B if you have no one on the bench with 1B listed as a position or who qualifies at 1B), we take players based on their PH rank. Unlike all of the other positions, if you need a sub DH to start we do NOT look for a guy on the bench you have listed at DH, instead we simply take the guy with the best PH rank.
Platoon Splits
This from Scoresheet's Jeff Barton in January, 2004
I've been analyzing individual batter differences between facing LHPs and RHPs,
i.e. platoon splits, and have come up with a couple of corrections to the
scoresheet-talk platoon.txt files for 2004. First of all, the league average
platoon splits for 2002 plus 2003 are:
BAvL SLvL OBvL BAvR SLvR OBvR
Average LHB -18 -41 -28 6 13 9
Average RHB 9 21 19 -3 -8 -7
Average BHB -1 -6 -6 0 2 2
These are the numbers that will be used for rookies in the 2004 summer season.
(Our old-timers game uses somewhat different numbers, given in the old-timers
newsletter.
Secondly, individual players often deviate from the above numbers, especially if
they don't have very many AB. For instance, here are Doug Mirabelli's numbers:
PA AB BAvg SlgA OBA
2001 LHP 60 53 .283 .717 .367
RHP 164 137 .204 .350 .319
2002 LHP 50 44 .364 .750 .440
RHP 123 107 .168 .271 .260
This certainly makes it look like Mirabelli's the best platoon player in the
history of baseball. Yet here are his 2003 numbers:
2003 LHP 57 52 .250 .288 .298
RHP 119 111 .261 .523 .311
He actually hit RHPs better than LHPs in 2003! I think Bill James first did a
study pointing out that this is typical, big platoon splits are more luck than
anything else, and tend to vary almost completely from year to year. While
platoon splits do matter some (for instance some switch hitters like Roberto
Alomar seem to consistently hit RHPs better than LHPs, presumably because they
get a lot more practice batting from the left side), there's definitely also a
lot of luck.
I have analyzed all this a lot more, both empirically and theoretically, to find
the best amount of league average platoon splits to include when predicting an
individual's future performance. I believe the answer is roughly 1500 plate
appearances, so this is what we'll now add to each players' splits from the last
two years. This still gives some 2004 results such as:
BAvL SLvL OBvL BAvR SLvR OBvR
Olivo,Mi 18 54 22 -7 -20 -8
Delgado,Ca -22 -88 -38 8 34 14
Giambi,Ja -26 -75 -38 9 26 13
Harvey,Ke 26 60 32 -10 -24 -13
Ortiz,Da -40 -83 -59 13 28 19
Colbrunn,Gr 20 52 26 -10 -26 -13
Catalanott,Fr -39 -70 -39 10 18 10
Chavez,Er -42 -83 -56 15 30 20
Blalock,Ha -46 -105 -56 14 33 17
Valentin,Jo -29 -88 -35 8 25 10
Gerut,Jo -33 -80 -43 11 27 14
Nixon,Tr -32 -88 -46 10 26 13
Monroe,Cr 19 67 26 -8 -28 -11
Thomas,Fr 9 40 24 -3 -14 -9
Martinez,Ed 11 49 36 -4 -18 -14
Ausmus,Br 17 51 35 -5 -16 -11
Thome,Ji -25 -80 -43 9 30 15
Karros,Er 29 49 46 -10 -18 -17
Franco,Ju 31 52 42 -12 -21 -17
Kent,Je 25 40 39 -8 -13 -13
Spivey,Ju 17 63 36 -7 -26 -16
Ramirez,Ar 13 58 18 -4 -20 -6
Nevin,Ph 24 77 33 -9 -28 -13
Abreu,Bo -16 -69 -32 6 25 11
Berkman,La -14 -69 -20 4 22 6
Grissom,Ma 22 78 33 -8 -31 -13
Stairs,Ma -27 -77 -42 6 18 10
Kinkade,Mi 30 66 35 -12 -27 -15
(We'll publish a complete file in mid-January when we produce the 2004 player
lists.) As you can see, there are still many players with large platoon splits,
but perhaps not quite as large as in the past. I think this makes the game much
more accurate. In fact, I believe this new method gives the most accurate
predicted platoon splits available, and would be very interested if anyone
thinks there are better ones elsewhere.
Cheers, and Play Ball!
- Dave Barton
Also, this exchange on SS-talk (March, 1999) on the net (on a little trick to save roster spots AND achieve a platoon): (Basically you put only one catcher -- Stinnett -- on your roster and hope the other -- Mayne -- comes in when you're up against a rhp.)
From: Bernard Luttbeg :
Suppose I have two catchers that are a righty and a lefty: Stinnett and Mayne. Even with the platoon I would rather have Stinnett starting against a righty. So, I don't want a strict platoon. However, is there any trick to increase my chances that when Mayne plays he plays against righties?
Douglas Dolbear wrote:
A year or two back, someone asked the same question and someone answered saying that if you start Stinnett every day and he runs short of playing time, the program will attempt to put Mayne in against the RHP. That gives you exactly the result you are looking for. i.e. play Stinnett as much as possible, but never have Mayne against LHP. I've never really tested the theory, but I recall it was from a reliable source.
From Jeff Barton at Scoresheet:
This is somewhat true, though does not work perfectly. When the computer benches a starter because he does not have enough at-bats to start every game, it 'tries' to bench that player against pitchers who pitch from the same side the batter hits from. But, 'trying' is a complicated thing, and as said above, it will not work perfectly every time. But over the long haul, doing what is described below will have your right handed batters sitting more against right handed pitchers than against left handed pitchers. - Jeff
Baserunning
1) RBI and RUNS used to determine how far runners advance, and SF chances.
2) Fast players (guys with a lot of major league steals, triples, and runs scored) do score in Scoresheet more often from second on a single, or from first on a double. (Major league runs scored are used in our game to determine how far a guy advances on a base hit, and if the batter has a lot of RBIs in the majors then your runner also will have a better chance of scoring.) However, in both Scoresheet and the majors, even the fastest guys do NOT always score from second on a single. There are infield singles, and sometimes the ball is just hit too hard for even a Kenny Lofton to score. And scoring from first on a double is definitely not a sure thing, though it does happen more often in Scoresheet when there are 2 outs. Over the course of a season your 'faster' players will take more extra bases - and overall we have players go from second to home and first to home at the same rate as in the majors.
GIDP Actual results modified by batting order. Also pitchers who have wins or saves are assisted to increase their chances of a DP.
Rank or PH For
This column always seems to cause a lot of confusion.
Very Simply: If you put a dash in a player's rank or PH for column he gets pinch hit for, if you put a number in his rank or PH for column then he does NOT get pinch hit for, versus that type of pitcher. This does apply to hitters listed in both the starter and in the bench sections - if you don't put a number in the PH or rank column for a starter, then he will get pinch hit for if there is an available player on the bench who has a number in that column. Probably the easiest way to think of how to number your players for that column is to think of the starters and bench players as all being listed in the same section, and then number them accordingly. Simply give the best hitter on your team versus that type of pitcher (RHP or LHP) a 1, the second best a 2, and so on. The weaker players that you would want to pinch hit for if you are behind late in a game should get a dash. We do look at the pitcher at that time to decide which column to use (vs. LHP or vs. RHP). For instance, you can have a dash for a player vs. LHP's and a number for that same player vs. RHP's, and then if that hitter faces a LHP late in the game he'd be pinch hit for (since he has a dash versus LHPs), and if he faces a RHP he'd hit for himself (since he has a number versus RHPs.) Pinch hit numbers are also used to juggle your lineup when subs have to start a game.