Click here to return to the Education page.
This sabermetric glossary is reprinted in its entirety from the 2008 Baseball Forecaster. Note that some formulas defined here are not part of the content used on this site.
Abbreviations and Beginning Concepts
PQS: Pure Quality Starts
Pw: Linear weighted power
PX:Linear weighted power index
Rand Var: Random Variance Score
RAR: Runs above replacement
RC/G:Runs created per game
REff%: Relief efficiency per cent
RSpd: Roto Speed
Rotisserie Value(R$): The dollar value placed on a player's performance in a Rotisserie league, and designed to measure the impact that player has on the standings. These values are highly variable depending upon a variety of factors:
- the salary cap limit
- the number of teams in the league
- each team's roster size
- the impact of any protected players
- each team's positional demands at the time of bidding
- the statistical category demands at the time of bidding
- external factors, e.g. media inflation or deflation of value
In other words, a $30 player is only a $30 player if someone in your draft pays $30 for him.
There are a variety of methods to calculate value, most involving a delineation of a least valuable performance level (given league size and structure), and then assigning a certain dollar amount for incremental improvement from that base. The method we use is a variation of the Standings Gain Points method described in the book, How to Value Players for Rotisserie Baseball, by Art McGee. (2nd edition available now)
People play Rotisserie in many variations. The most popular game is the 5x5 format. Mixed league participation is soaring; here, player pool penetration falls short of the standard 75%.
Since we currently have no idea who is going to play RF for the Mets, or whether Matt Antonelli is going to break camp with San Deigo, all the projected values are slightly inflated. They are roughly based on a 12-team AL and 13-team NL league. We've attempted to take some contingencies into account, but the values will not total to anywhere near $3120, so don't bother adding them up and save your irate e-mails.
A $25 player in this book might actually be worth $21. Or $28. This level of precision is irrelevant in a process that is going to be driven by market forces anyway. So, don't obsess over it.
How do other writers publish perfect Rotisserie values over the winter? Do they make arbitrary decisions as to where free agents are going to sign and who is going to land jobs in the spring? I'm not about to make those massive leaps of faith. Bottom line... Some things you can predict, to other things you have to react. As roles become more defined over the winter, our online updates will provide better approximations of playing time, and projected Roto values that add up to $3120.
Sabermetrics, Fanalytics and Advanced Concepts
(BB + K) / IP
Measures the level by which a pitcher allows balls to be put into play and helps tie a pitcher's success to his team's level of defensive ability. In general, extreme power pitchers can be successful even with poor defensive teams. Power pitchers tend to have greater longevity in the game. Contact pitchers with poor defenses behind them are high risks to have poor W-L records and ERA. BENCHMARKS: A level of 1.13 or greater describes the pure throwers. A level of .93 or lower describes the high contact pitcher. Tip... if you have to draft a pitcher from a poor defensive team, going with power over contact will usually net you more wins in the long run.
PQS disaster rate (Gene McCaffrey): The percentage of a starting pitcher's outings that rate as a PQS-0 or PQS-1. See the Pitching Logs section for more information on DIS%.
PQS domination rate (Gene McCaffrey): The percentage of a starting pitcher's outings that rate as a PQS-4 or PQS-5. See the Pitching Logs section for more information on DOM%.
Pure Quality Starts: PQS is the next step in following pitching lines. The old Quality Start method — minimum 6 IP, maximum 3 earned runs — is simplistic and does not measure any real skill. Bill James' "game score" methodology is better, but is not feasible for quick calculation.
In PQS, we give a starting pitcher credit for exhibiting certain skills in each of his starts. Then by tracking his "PQS Score" over time, we can follow his progress. A starter earns one point for each of the following criteria...
1. The pitcher must have gone a minimum of 6 innings. This measures stamina. If he goes less than 5 innings, he automatically gets a total PQS score of zero, no matter what other stats he produces.
2. He must have allowed no more than an equal number of hits to the number of IP. This measures hit prevention.
3. His number of strikeouts must be no fewer than two less than his innings pitched. This measures dominance.
4. He must have struck out at least twice as many batters as he walked. This measures command.
5. He must have allowed no more than one home run. This measures his ability to keep the ball in the park.
A perfect PQS score would be 5. Any pitcher who averages 3 or more over the course of the season is probably performing admirably. The nice thing about PQS is it allows you to approach each start as more than an all-or-nothing event.
Note the absence of earned runs. No matter how many runs a pitcher allows, if he scores high on the PQS scale, he has hurled a good game in terms of his base skills. The number of runs allowed — a function of not only the pitcher's ability but that of his bullpen and defense — will even out over time.
Random Variance score: A gauge that measures the probability that a player's performance in the subsequent year will exceed or fall short of the immediate past year's numbers. Expressed on a scale of -5 to +5, this only measures variables for which outliers tend to regress. Positive scores indicate rebounds; negative scores indicate corrections. The further the score is from zero, the higher the likelihood of a performance swing. The variables tracked for batters and pitchers:
Batting: Outlyinglevels for h%, hr/f, and xBA.
Pitching: Outlyinglevels for H%, s%, hr/f, and xERA.
Reliability Grades (REL)
Health: "A" level players would have accumulated fewer than 30 days on the Major League DL over the past five years. "F" grades go to those who’ve spent more than 120 days on the DL. Recent DL stays are given a heavier weight in the calculation.
Playing Time and Experience: For batters, we simply track plate appearances. Major league PAs have greater weight than minor league PAs. "A" level players would have averaged at least 550 major league PAs per year over the past three years. "F" graded players averaged fewer than 250 major league PAs per year.
For pitchers, workload can be a double-edged sword. On one hand, small IP samples are deceptive in providing a read on a pitcher’s true potential. Even a consistent 65-inning reliever can be considered higher risk; just one bad outing can skew an entire year’s work.
On the flipside, high workload levels also need to be monitored, especially in the formative years of a pitcher’s career. Exceeding those levels elevates the risk of injury, burnout, or breakdown. So, tracking workload must be done within a range of innings. The grades capture this.
Consistency: "A" level players are those whose runs created per game level (xERA for pitchers) has fluctuated by less than half a run during each of the past three years. "F" grades go to those whose RC/G has fluctuated by two runs or more.
Remember that these grades have nothing to do with quality of performance; they strictly refer to confidence in our expectations. So a grade of "AAA" for Jason Marquis, for instance, only means that there is a high probability he will perform as poorly as we’ve projected.
Reliever efficiency per cent (REff%)
(Wins + Saves + Holds) / (Wins + Losses + SaveOpps + Holds)
This is a measure of how often a reliever contributes positively to the outcome of a game. A record of consistent, positive impact on game outcomes breeds managerial confidence, and that confidence could pave the way to save opportunities. For those pitchers suddenly thrust into a closer's role, this formula helps gauge their potential to succeed based on past successes in similar roles. BENCHMARK: Minimum of 80%.
Roto Speed (RSpd)
Spd x (SBO + SB%)
This is the metric used to reflect speed in the Mayberry Method (third revision). An adjustment to the measure for raw speed (Spd) that takes into account a runner’s opportunities to steal and his success rate. This stat is intended to provide a more accurate predictive measure of stolen bases for the Mayberry Method.
Runs above replacement(RAR): An estimate of the number of runs a player contributes above a "replacement level" player. "Replacement" is defined as the level of performance at which another player can easily be found at little or no cost to a team. What constitutes replacement level is a topic that is hotly debated. There are a variety of formulas and rules of thumb used to determine this level for each position (replacement level for a shortstop will be very different from replacement level for an outfielder). Our estimates appear below.
One of the major values of RAR for fantasy applications is that it can be used to assemble an integrated ranking of batters and pitchers for drafting purposes.
Batters create runs; pitchers save runs. But are batters and pitchers who have comparable RAR levels truly equal in value? Pitchers might be considered to have higher value. Saving an additional run is more important than producing an additional run. A pitcher who throws a shutout is guaranteed to win that game, whereas no matter how many runs a batter produces, his team can still lose given poor pitching support.
To calculate RAR for batters:
Start with a batter's runs created per game (RC/G).
Subtract his position's replacement level RC/G.
Multiply by number of games played: (AB - H + CS) / 25.5.
Replacement levels used in this book, for 2007:
To calculate RAR for pitchers:
Start with the replacement level league ERA.
Subtract the pitcher's ERA. (To calculate projected RAR, use the pitcher's xERA.)
Multiply by number of games played, calculated as plate appearances (IP x 4.34) divided by 38.
Multiply the resulting RAR level by 1.08 to account for the variance between earned runs and total runs.
RAR can also be used to calculate rough projected team won-loss records. (Roger Miller) Total the RAR levels for all the players on a team, divide by 10 and add to 53 wins.
Runs created(Bill James)
(H + BB - CS) x (Total bases + (.55 x SB)) / (AB + BB)
A formula that converts all offensive events into a total of runs scored. As calculated for individual teams, the result approximates a club's actual run total with great accuracy.
Runs created per game(Bill James)
Runs Created / ((AB - H + CS) / 25.5)
RC expressed on a per-game basis might be considered the hypothetical ERA compiled against a particular batter. Another way to look at it... a batter with a RC/G of 7.00 would be expected to score 7 runs per game if he were cloned nine times and faced an average pitcher in every at bat. However, cloning batters is not a practice we recommend.
BENCHMARKS: Few players surpass the level of a 10.00 RC/G in any given season, but any level over 7.50 can still be considered very good. At the bottom are levels below 3.00.
This is the version that is currently used in this book.