Since I was a little boy (not all that long ago) I have played and loved Strat-o-Matic baseball. As an adult, I have had the opportunity to join a rather intense play-by-mail league called MAX (http://themax.addr.com/). I inherited a franchise that was a borderline contender in our playoff system. Since my team wasn’t stacked, I couldn’t afford to make too many poor lineup decisions. Hence, I decided to take a couple of weekends and tear the Super Advanced Fielding Charts apart to get an idea of exactly how important defense was at each position and how fielding ratings could be directly compared with batting prowess. Below are the results of my research and the charts I developed to help me make logical comparisons.

As most SOM veterans know, the results of a particular AB can be determined in three ways; 50% of the time they come from the batter’s card, 36.1% of the time from the pitcher’s card, and 13.9% of the time from the fielders’ ratings and the fielding charts. Each position on the field determines a certain percentage of the opponent’s batting results as shown in the table below (percentages are rounded to 1 decimal place):

Since batters determine the results of their AB 50% of the time, the average batter in the lineup would determine about 5.56% of his team’s batting results (50/9 = 5.56, rounded to 2 decimal places). It is important to note that a leadoff or #2 hitter will determine more of his teams batting results and a #9 hitter will obviously determine fewer. However, in order to keep things simple, I assume each batter will determine 5.56% of the results. I take lineup position into account only when my methodology does not clearly prefer using one player over another.

Since batters can be effectively compared to one another by using the simple formula of OBP + SLG (OPS), I decided to convert the value of the fielding ratings into OPS so that they could be directly compared to batting OPS. I don’t know that any of the more complex formulae that have been developed correlate with runs scored any better than OPS so I see no reason to use anything else. I calculate a fielder’s OPS by looking at the Super-Advanced fielding charts and calculating the percentage of the time a batter will reach base on an X-chance for that fielder, whether by hit or error (OBP). I then calculate the SLG percentage in the same way, treating all errors as hits, or extra-base hits as appropriate. I then add the OBP to the SLG. I have calculated the OPS for every rating for every player except for pitchers and catchers. I neglected to do anything for pitchers since the effect of the ratings is negligible when compared with the pitcher’s OPS allowed. I call the result of these calculations FOPS.

I neglected to do anything with catchers since their range and errors have little effect on batting results. Throwing arm, PB ratings, etc. have much more impact but can’t be measured using OPS. To take those things into account would require some sort of Linear Weights analysis which is a more ambitious undertaking than I’m willing to pursue (although I know it’s been done by others).

Let me begin this section by assigning variable names to the statistics that I use to create my basis for comparing fielding prowess with hitting ability. Subscripts (1 and 2) distinguish the two batters from one another.

FPct The percentage of time a fielding position will determine the batting results of the opposing team, as listed in column 3 of Table 1.

BPct The percentage of batting results a batter will determine for his team; fixed at 0.0556, or 5.56%.

EF FPct/BPct, or the ratio of importance of a batter's fielding at a given position compared to his hitting.

FOPS The OPS of the fielder’s range and error ratings, as calculated from the SADV Fielding Charts.

OPS The batter’s OPS.

PLAY A single number describing the overall fielding and hitting value of a player, for a particular position. The higher the number, the better the player. This can be a negative number, but generally only if the batter's hitting ability is truly execrable. Generally, only differences in PLAY greater than .050 should be considered significant.

PDiff The difference in value between the two batters being compared. If positive, player 1 is better. If negative, player 2 is more valuable. PDiff is equivalent in magnitude to batting OPS; i.e. if PDiff = 100, batter 1 is better than batter 2 by the equivalent of roughly 100 points of batting OPS.

PPct The percentage of batting results an average batter will determine for his team from the pitcher's card; that is 36.1%, the percentage of all play results that come from the pitcher's card, divided by 9, the number of batters on a team. 36.1 / 9 = 4.012, so PPct is fixed at 4.012.

PF PPct/BPct, or the ratio of importance of the pitcher's card compared to the batter's card, fixed at 0.722.

POPS The pitcher's OPS against opposing batters.

PVP A single number describing the overall fielding and hitting value of a player, for a particular position, against a particular pitcher. The higher the number, the better the player. Like PLAY, PVP is calculated to be analogous to batting OPS.

PVPDiff The difference in PVP values between two batters. This number is equivalent in magnitude to batting OPS, since we adjust the POPS and FOPS values by coefficient's that are percentages of how much weight they carry in determining play results as compared with batting OPS.

In order to make the system easier to use, I have included a table below which calculates the value of EF (FPct / BPct) for each position in the field. 

To really be useful, the OPS values should be calculated vs. LHP and RHP separately. Since SOM normalizes their cards by adjusting raw numbers based on league norms, the OPS should be calculated from the cards themselves, rather than using the players' actual stats. I use Glen Guzzo's (Strat Fan) ratings disk as the basis for calculating OPS. But one could also "count the cards" by hand or purchase "counted cards" from another vendor.

The formulae I developed would be fairly easy to use if OBP and SLG were readily available. The OBP and SLG averages are available in numerous stats books but these numbers aren’t the ones SOM uses to create player cards. Instead, each batter’s and pitcher’s card is normalized with respect to league averages. In effect, a batter will reproduce his real life stats almost precisely if he faces league average pitching (and fielding) during the course of an entire season. If the real OBP and SLG averages were used, then an above average hitter would reproduce statistical totals that are significantly below what he achieved in real life. For example, if the league batting average is .270, then a .300 hitter would only hit .285 against league average pitching in SOM if his true BA were used to create his SOM card. Since SOM wisely adjusts the cards for league averages, a .300 hitter would have a .330 BA on his card, so that he would hit .300 over the course of a full season against league average pitching.

Consequently, using true OBP and SLG to calculate OPS in the formulae above will favor the weaker hitter and better fielder. This results from the fact that an above average hitter will actually have a better OPS on his card than in real life. And a below average hitter would have a poorer OPS in real life than on his card. I don't believe my formulae have much value using real life batting numbers, but they can be used to help approximate the relative value of two players before the SOM cards come out.

Fortunately, Strat Fan produces a ratings guide that usually comes out 4 to 6 weeks before the cards do. This guide not only gives fielding ratings but actually counts OB chances, H, BB, K, Total Bases, etc. per 108 chances on each side of a SOM batting or pitching card. These ratings can be used to calculate OBP and SLG, which combined produce OPS, for the actual card. For truly precise results, the ballpark factors must be incorporated for the park in question. To obtain a general figure for an average ballpark, I use 1-9 for diamonds and 1-8 for triangles. Using the Strat Fan numbers, OBP and SLG can be calculated in the following manner:

OB The number of On Base chances per 108, as calculated by Strat Fan.

s The number of triangles on the batter’s card, this should be 5 for all batters.

BPS The singles number for the ballpark in question (take the hitter’s batting side into account).

u The number of diamonds on the batter’s card.

BPHR The Homerun number for the ballpark in question (take the hitter’s batting side into account).

TB Total bases on the batter’s card, as calculated by Strat Fan.

HIT The number of hit chances on the batter’s card, as calculated by Strat Fan.

The first example I have chosen comes from my own MAX league team. I have Mark Grace at 1B, with Roberto Petagine as his backup. I also have Chris Hoiles, my number 2 catcher, available to play first for me as well. Among the three, you wouldn't think I'd ever want to use anyone but Grace, especially considering his Gold Glove caliber defense, but the numbers proved otherwise. Below are the fielding ratings (at 1B) and batting stats, taken from the Strat Fan numbers. The OPS numbers were calculated from the cards (using Strat Fan ratings) according to the formulae above. I usually don't calculate them by hand but use a spreadsheet formula to do it for me. But for the purposes of this discussion, I'll do a Mark Grace example. Assume a neutral ballpark (BPS = 8, BPHR = 9 from both sides) unless otherwise indicated.

Grace has 23.1 HIT chances (out of 108), 35.2 OB chances, and 30.1 total bases on his card vs. LHP. He also has 1 diamond. Therefore, Grace's OBP = (35.2 + (5 Ÿ 8 / 20) + (1 Ÿ 9 / 20)) / 108 = 0.349. His slugging percentage is SLG = (30.1 + (5 Ÿ 8 / 20) + (1 Ÿ 9 Ÿ 4 / 20)) / (108 - 35.2 + 23.1) = 0.353. 0.349 + 0.353 = 0.702, Grace's carded OPS vs. LHP in a neutral park. Bear in mind that some players don't have any triangles on their cards. These players have an asterisk following their number of diamonds (BP) in the Strat Fan ratings. Be careful to plug that 0 in for s instead of 5 in such cases.

Mark Grace 1b-1e9 .702 OPS vs. LHP, 1.110 OPS vs. RHP

Roberto Petagine 1b-3e5 .167 OPS vs. LHP, 1.160 OPS vs. RHP

Chris Hoiles 1b-4e18 .846 OPS vs. LHP, .950 OPS vs. RHP

To start vs. RHP, I narrowed my choices to Grace and Petagine, as Hoiles fields worse than both and hits worse than both vs. RHP. Actually, I was hesitant to even make calculations for this spot, since Petagine hits only slightly better than Grace but has drastically less range around the bag, according to SOM. After crunching the numbers, I was a bit surprised.

According to my fielding spreadsheets included at the end of this article, Grace, a 1b-1e9, has a FOPS of 0.333. Petagine's 1b-3e5 checks in at a FOPS of 0.563. According to Table 1, the firstbaseman determines 0.93% of the opposing team’s batting results. Since each batter in the lineup determines approximately 5.56% of his team's batting results (50% results off the batter's card divided by 9 batters), the equalization factor (EF) turns out to be EF = 0.0093 / 0.0556 = 0.167, as found in Table 2. The PLAY value for Grace then turned out to be PLAY_Grace = 1.110 - (0.167Ÿ 0.333) = 1.055. The PLAY value for Petagine came in at PLAY_Petagine = 1.160 - (0.167Ÿ 0.563) = 1.066. PDiff, which is analogous in magnitude to batting OPS, is therefore . PDiff = PLAY_Grace - PLAY_Petagine = 1.055 - 1.066 = -0.011. Essentially, this means that Petagine's overall value at 1B vs. RHP exceeds Grace's by roughly 11 points of batting OPS. The difference is too small to be significant, but it illustrates well that the difference in value between a 1 and a 3 at 1B is not that great.

The choice vs. LHP was obviously a 2-man race, as Petagine's card is useless against a southpaw. But the choice was interesting because Hoiles hit far better than Grace while fielding far worse. After running my numbers, the choice was none too clear, as you will see below.

As mentioned above, Grace has a FOPS of 0.333 at first base. Hoiles, a 1b-4e18 firstbaseman, has a FOPS of 1.165. As we found above, for first base, EF = 0.0093 / 0.0556 = 0.167, as found in Table 2. Therefore, PLAY_Grace = 0.702 - (0.167Ÿ 0.333) = 0.646. Similarly, PLAY_Hoiles = 0.846 - (0.167 Ÿ 1.165) = 0.652. PDiff = PLAY_Grace - PLAY_Hoiles = 0.646 - 0.652 = -0.006.

So far, all the examples have been in a neutral park, but what happens to our results if we transplant Grace and Hoiles into Coors Field? Coors Field has ballpark singles values of 19 vs. both LH and RH batters, a ballpark HR value of 16 vs. LH batters, and a ballpark HR value of 19 vs. RH batters. In the thin air of Denver, Grace's batting OPS vs. LHP becomes 0.774, up from 0.702. But Hoiles' OPS, due in large measure to the 4 ballpark HR chances on his card (as opposed to Grace's 1 ballpark HR chance), soars from 0.846 to 0.993! Hoiles saw an improvement in batting OPS of nearly 150 points while Grace only got a 70 point boost. How will this affect their overall value? PLAY_Grace = 0.774 - (0.167Ÿ 0.333) = 0.718. PLAY_Hoiles = 1.030 - (0.167 Ÿ 1.165) = 0.835. PDiff = PLAY_Grace - PLAY_Hoiles = 0.718 - 0.835 = -0.117. So while Grace and Hoiles are roughly equal in value in a neutral park, Hoiles is quite a bit more valuable in Coors field vs. LHP.

The Grace vs. Hoiles example is incomplete as listed above, because Grace is a left-handed hitter whereas Hoiles bats from the right side. The numbers above are good for formulating a general opinion about a player but are useless for deciding who should start a particular game against anything but a balanced LHP. In the example below, we'll compare the two again in a neutral ballpark versus two different LH starters, Randy Johnson and Chuck Finley. Below find Johnson's (full-season) and Finley's OBP/SLG/OPS numbers as calculated from Strat Fan ratings:

Randy Johnson vs. LHB: 0.139 0.011 0.150 vs. RHB: 0.213 0.259 0.472

Chuck Finley vs. LHB: 0.350 0.325 0.675 vs. RHB: 0.244 0.217 0.460

As mentioned above, Grace and Hoiles have batting OPS's of 0.702 and 0.846, respectively, vs. LHP. Their repective FOPS scores are 0.333 for Grace and 1.165 for Hoiles. So, versus the Big Unit, PVP_Grace = 0.702 + (0.722 Ÿ 0.150) - (0.167Ÿ 0.333) = 0.754. Similarly, PVP_Hoiles = 0.846 + (0.722 Ÿ 0.472) - (0.167 Ÿ 1.165) = 0.992. PVPDiff = PVP_Grace - PVP_Hoiles = 0. 754- 0.992= -0.238. Hoiles is light years ahead of Grace vs. Johnson, the equivalent of nearly 240 points of OPS ahead!

Against Chuck Finley, though, the story is quite different, since he doesn't torment the lefties like the Big Unit does. PVP_Grace = 0.702 + (0.722 Ÿ 0.675) - (0.167Ÿ 0.333) = 1.133. Similarly, PVP_Hoiles = 0.846 + (0.722 Ÿ 0.460) - (0.167 Ÿ 1.165) = 0.984. PVPDiff = PVP_Grace - PVP_Hoiles = 1.133 - 0. 984 = 0.150. Against a backward lefty like Finley (or at least a backward LHP card), Grace is significantly better than Hoiles, all things considered.

The above examples illustrate quite well that using strict lefty/righty platoons in Strat-o-Matic baseball is not always a good way to maximize the talent on your roster. The Compare feature in the computer game helps, but only so far as batting prowess in concerned. Compare will usually indicate the best pinch-hitter for a given situation, but falls short of giving you a clear idea of what player should start against a particular pitcher.

I must admit that I don’t follow my system to the letter. In fact, I shy away from playing the glove man over the slugger unless his PLAY value is at least .050 better than the slugger's. I leave that margin due to the effects in rounding out the X-chance percentages to one decimal place and the fact that the position in the batting order does make a difference, just not a terribly significant difference. In essence, if the difference in PLAY (or PVP) values is less than 50 (analogous to 50 points of OPS), there really is no clear advantage to playing one player over the other. In such circumstances, it makes sense to bat the better hitter high in the order and then replace him in the field later in the game, preferably with the poorer hitter not coming up to bat for a while.

Further, my system does not take into account the increased number of gbA’s that will come off of the fielding charts for an infielder with a higher range rating. I’m not sure exactly how to incorporate this knowledge into my current system except to say that for infielders, my system slightly underrates fielding skill as opposed to batting skill. By how much I’m not certain.

Bear in mind that if the two players you are comparing bat from different sides of the plate, as in the Mark Grace vs. Chris Hoiles example, the pitcher’s card will have a very large effect and must be taken into account before each game. However, if you are trying to decide whether or not to trade for somebody in particular, the basic numbers will generally suffice. If you have an extreme park (like Coors Field or Dodger Stadium) or a "lop-sided" park (like Shea Stadium), be sure to compare the players using your home park numbers as well, since half of their PA's will come in that non-neutral environment.

The basic idea for my method of analyzing hitting and fielding in SOM is my own. However, many people have helped me to refine my system and correct some bad assumptions I made in my early research. Special thanks go to Adam Wodon who reviewed the original version of this document and provided me with methods and data for calculating OPS directly off of the cards. Thanks go as well to the many active members of the "strat-bb" internet mailing list (circa 1995) who provided lots of comments, suggestions, and help. If the reader has any comments, suggestions, or especially corrections to my work, feel free to email me at dcmadsen@inconnect.com. I will correct or update my research as I have time and try to find a place for it on the web. Finally, thanks again to Adam for putting together an awesome web site which has become an invaluable resource for the serious SOM gamer. Hopefully he can one day find the time to actively maintain it again.

In the pages that follow, you will find EXCEL spreadsheets (printed as tables) that contain the OPS values for every strat fielding rating except those assigned to pitchers and catchers. I left out the intermediate values for the sake of brevity.