SWC75
Bored Historian
- Joined
- Aug 26, 2011
- Messages
- 33,966
- Like
- 65,506
Last year I introduced the concept of “bases produced” and wedded it to the old concept of “runs produced”. There’s a lot of debate about which baseball statistics are the best ones. My points in a somewhat lengthy introduction were that:
1) It’s sports, not a moon shot: let’s keep things simple. Stats should be easy to compute and what the resulting number represents should be clear. You should be able to see a play in a game and know the impact of that play on the stat, perhaps even compute it in your head.
2) I like the concept of “OBPS”: adding on base percentage to slugging percentage to wed the two most important stats and supplant the over-rated batting average, home runs and stolen bases in evaluating players. But I don’t like the stat itself. You adding together two percentages, each with a different divisor, (total plate appearances vs. official at bats). You are including hits on both sides of the equation and thus counting them twice. You are including something the batter is not actually, (or normally), trying to do, being hit by a pitch, which is part of on base percentage. You are excluding stolen bases, which is something the player is trying to do once he gets on base. Finally you are producing a number that, while it can be used to rank players, doesn’t represent how often or how frequently he produced something. Albert Pujol’s OBPS for his career is 1.013. For this season so far it’s 0.723. That tells me he’s off to a bad start but what else does it tell me? For his career he’s produced 1.013 what? He’s producing 0.723 what so far this year? Suppose I see him hit a home run. What’s his OBPS now? I know he’s got 8 home runs, 32 RBIs and 30 runs scored. Two guys were on base so I know he’s now got 9 home runs, 35 RBIs and 31 runs scored. He was 53 for 218 on the season, (.244). He’s now 54 for 219. Even if I don’t have a calculator handy, I could, with some effort, figure out that he’s now batting. .247. What’s his OBPS? I don’t know.
3) Let’s simplify this by computing it this way: Add total bases, which is the hit total broken out to 1 base for a single, 2 for a double, three for a triple and four for a home run, the dividend in a player “slugging percentage”. Let’s add to that the major additional component in “on base percentage: the walks, and then the other thing a batter can do to advance himself when he gets on base: the stolen base. That’s “bases produced”. You could add in some other stuff: the number of bases he got by tagging up on outs, the number of bases he took on hits that were more than the batter took, (Enos Slaughter gets an extra base for scoring off Harry Walker’s double to win the 1946 series), even bases taken on the batter’s ground outs. You could add in the bases the bases other players get to take because of the batter’s hits. But those things aren’t kept track off so we will “keep it simple”: Total Bases + Walks + Steals = Base Production, (BP). Everything is counted once and everything is something the batter is actually trying to do to obtain bases for himself and his team. Pujols in hitting the three run homer I surmised, produced four bases, increasing his season total to 116: 93 total bases + 23 walks + 0 steals.
4) The “sister stat” to BP would be “Runs Produced”, (RP), which has been around for years: Runs scored + runs driven in – home runs so you don’t count them twice. Pujols, after hitting the 3 run homer, (with the help of my imagination), went from 54 runs produced (30 scored + 32 driven in – 8 homers) to 57, (31 + 35 -9).
5) There’s a lot to be said for gross totals: you don’t know if a rate of production would have been kept up and the impact of a player playing in a game is always going to be greater than the impact of a player who doesn’t. If you must have a percentage or an average the instinct would be to divide BP and RP by plate appearances, giving Albert, after his three run homer, a BPA of .466, (116/249), and an RPA of .229, (57/249). But we will tend to be comparing players who start and who will play entire games unless injured, so I would propose we use an average obtained by simply dividing by the number of games played. You’d have to wait until the game was over to redo this stat. But if Albert produced no more runs and bases by the end of today’s game, he’s have a BPA of 2.04, (116/57) and a RPA of 1.18, (67/57). His career numbers are BPA 2.84 and RPA 1.25. Albert normally produces 3 bases and one run per game. This year it’s 2 and 1. If 9 guys produced 2.04 based and 1.18 runs per game, that would be 18 bases and 11 runs per game. So he’s still doing pretty good. But using his career averages, X 9 would be 26 bases and 11 runs. That’s pretty easy to comprehend.
I think BP and RP are a much easier to compute and comprehend way of evaluating player’s offensive skills than the “super numbers” the sabermaticians have come up in recent years, like Total Average, (an oxymoron), Offensive Winning Percentage, Runs Created and Win Shares, which Bill James used 86 pages to explain in a book. The numbers would be more relevant than batting average, homers and steals, yet nearly as easy to compute and comprehend. Maybe BP and RP would be the stats everybody would know about their favorite player or the guy they think should be MVP, (although that call should be based on more than just a stat).
Anyway, we are embarked on a new season and I am again going to post the monthly top ten rankings in BP and RP, along with the per- game averages, although the ranking will be based on the total.
1) It’s sports, not a moon shot: let’s keep things simple. Stats should be easy to compute and what the resulting number represents should be clear. You should be able to see a play in a game and know the impact of that play on the stat, perhaps even compute it in your head.
2) I like the concept of “OBPS”: adding on base percentage to slugging percentage to wed the two most important stats and supplant the over-rated batting average, home runs and stolen bases in evaluating players. But I don’t like the stat itself. You adding together two percentages, each with a different divisor, (total plate appearances vs. official at bats). You are including hits on both sides of the equation and thus counting them twice. You are including something the batter is not actually, (or normally), trying to do, being hit by a pitch, which is part of on base percentage. You are excluding stolen bases, which is something the player is trying to do once he gets on base. Finally you are producing a number that, while it can be used to rank players, doesn’t represent how often or how frequently he produced something. Albert Pujol’s OBPS for his career is 1.013. For this season so far it’s 0.723. That tells me he’s off to a bad start but what else does it tell me? For his career he’s produced 1.013 what? He’s producing 0.723 what so far this year? Suppose I see him hit a home run. What’s his OBPS now? I know he’s got 8 home runs, 32 RBIs and 30 runs scored. Two guys were on base so I know he’s now got 9 home runs, 35 RBIs and 31 runs scored. He was 53 for 218 on the season, (.244). He’s now 54 for 219. Even if I don’t have a calculator handy, I could, with some effort, figure out that he’s now batting. .247. What’s his OBPS? I don’t know.
3) Let’s simplify this by computing it this way: Add total bases, which is the hit total broken out to 1 base for a single, 2 for a double, three for a triple and four for a home run, the dividend in a player “slugging percentage”. Let’s add to that the major additional component in “on base percentage: the walks, and then the other thing a batter can do to advance himself when he gets on base: the stolen base. That’s “bases produced”. You could add in some other stuff: the number of bases he got by tagging up on outs, the number of bases he took on hits that were more than the batter took, (Enos Slaughter gets an extra base for scoring off Harry Walker’s double to win the 1946 series), even bases taken on the batter’s ground outs. You could add in the bases the bases other players get to take because of the batter’s hits. But those things aren’t kept track off so we will “keep it simple”: Total Bases + Walks + Steals = Base Production, (BP). Everything is counted once and everything is something the batter is actually trying to do to obtain bases for himself and his team. Pujols in hitting the three run homer I surmised, produced four bases, increasing his season total to 116: 93 total bases + 23 walks + 0 steals.
4) The “sister stat” to BP would be “Runs Produced”, (RP), which has been around for years: Runs scored + runs driven in – home runs so you don’t count them twice. Pujols, after hitting the 3 run homer, (with the help of my imagination), went from 54 runs produced (30 scored + 32 driven in – 8 homers) to 57, (31 + 35 -9).
5) There’s a lot to be said for gross totals: you don’t know if a rate of production would have been kept up and the impact of a player playing in a game is always going to be greater than the impact of a player who doesn’t. If you must have a percentage or an average the instinct would be to divide BP and RP by plate appearances, giving Albert, after his three run homer, a BPA of .466, (116/249), and an RPA of .229, (57/249). But we will tend to be comparing players who start and who will play entire games unless injured, so I would propose we use an average obtained by simply dividing by the number of games played. You’d have to wait until the game was over to redo this stat. But if Albert produced no more runs and bases by the end of today’s game, he’s have a BPA of 2.04, (116/57) and a RPA of 1.18, (67/57). His career numbers are BPA 2.84 and RPA 1.25. Albert normally produces 3 bases and one run per game. This year it’s 2 and 1. If 9 guys produced 2.04 based and 1.18 runs per game, that would be 18 bases and 11 runs per game. So he’s still doing pretty good. But using his career averages, X 9 would be 26 bases and 11 runs. That’s pretty easy to comprehend.
I think BP and RP are a much easier to compute and comprehend way of evaluating player’s offensive skills than the “super numbers” the sabermaticians have come up in recent years, like Total Average, (an oxymoron), Offensive Winning Percentage, Runs Created and Win Shares, which Bill James used 86 pages to explain in a book. The numbers would be more relevant than batting average, homers and steals, yet nearly as easy to compute and comprehend. Maybe BP and RP would be the stats everybody would know about their favorite player or the guy they think should be MVP, (although that call should be based on more than just a stat).
Anyway, we are embarked on a new season and I am again going to post the monthly top ten rankings in BP and RP, along with the per- game averages, although the ranking will be based on the total.
