The reason why I don't use plate appearances straight up is simple. Clifford Van Beek would have had no access to that number.
I think we could argue that he might have used a combination of at bats and walks. The more I dig in, however, the more I believe that Van Beek did something a lot more simple than that...
AB + BB is much better than AB. AB + BB is a great approximation of PA.
We also still have the issue that you say the cards are based on 1929 stats but 1930 stats are consistently used in the analyses. I know "it is complicated" but if Van Beek didn't have access to 1930 stats when the cards were calculated/issued then shouldn't we be looking at 1929 stats?
First of all — I'm not interested in arguing a position here. As I've looked into the evidence, I've discovered strong evidence that National Pastime is indeed based on 1930. This is obvious for two reasons that I've already written about: the rosters, which include players who were not in Major League Baseball in 1929 (and, in some cases, players who weren't even in Minor League Baseball), and the lineups, which fit extremely well with commonly used lineups in 1930.
Second — there is actually really good evidence that Van Beek didn't base his cards on ANY season. That's why it is significant that every single hitter's card has at least one extra base hit number, regardless of the number of extra base hits the player had in real life.
I'm interested in discovering what Clifford Van Beek actually did to create his cards. As we'll see in future posts, there is a lot of evidence that he swapped out play result numbers between dice rolls for whatever reason. As we've already seen, there is also a lot of strong evidence that he didn't make strict use of walk and strikeout numbers to come up with the 13s and 14s on each card.
Somebody asked me in a private message a few months ago about double plays in National Pastime. It turns out that Van Beek couldn't have known exact numbers of double plays grounded into, since those statistics simply weren't kept. Play result numbers 24 and 25 appear to have been assigned because of preset patterns, and not because of some mathematical algorithm.
If we're going to be honest about what the data shows us, we really need to move away from the idea of which season's statistics the cards were strictly based on. That should be obvious given the number of players with no extra base hits in real life who have extra base hit numbers on their cards.
I'd give you a direct answer to the question about which year he used as a base if I had it. I don't have it. Think of this blog as a work in progress — an exploration into what actually happened, as opposed to a polemic argument about why I think it was this or why it was that.
Right now, I'm leaning towards a theory that Van Beek created the cards using end of 1930 season at bat, batting average, and home run numbers for certain key players. My guess is that even the walk and strikeout numbers were likely based on reputation more than anything else. It's going to take some time, though, to dig in and figure out what actually happened — and I intended to continue blogging about it each step of the way.
I still stand by my viewpoint that NP cards are based on PA (AB + BB) not AB.
You thought it was important enough to point it out in the blog, "If you look at Sweetland’s 1930 performance alone, he clearly had a higher percentage of his at bats go for extra base hits than Ruffing. Now, it’s not as much of an advantage when you factor in his 9 walks"
The impression I'm getting is that Van Beek didn't have pitchers' batting stats, at least at the time he calculated the cards. If he was using in-season stat listings, Idon't know of any that included pitcher batting, and if he already had to base some of the lesser-used nonpitchers -- or weaker hitters who got lopped off the bottom of the Sporting News lists -- on compilations from box scores, it's not unlikely that he didn't feel it was worth his time and effort to do that for the majority of pitchers.
(There's also the point that the five pitchers on each NP roster were standing in, so to speak, for all the pitchers used by each team. He may have felt -- or may have convinced himself -- that a literal rendering of those pitchers' batting stats didn't really serve his purpose.)
He'd have known who the more notable good-hitting pitchers were, especially pitchers like Ruffing, Lucas, Wes Ferrell and Uhle who were used extensively as pinch-hitters, and he carded them accordingly, if not necessarily literally. (And Smith of Boston had been a middle infielder when he came up; I actually wound up playing him there a couple of games in my 1930 APBA solo league when things got really bleak.)
I would feel better if you used plate appearances throughout.
Thanks, Doug!
The reason why I don't use plate appearances straight up is simple. Clifford Van Beek would have had no access to that number.
I think we could argue that he might have used a combination of at bats and walks. The more I dig in, however, the more I believe that Van Beek did something a lot more simple than that...
AB + BB is much better than AB. AB + BB is a great approximation of PA.
We also still have the issue that you say the cards are based on 1929 stats but 1930 stats are consistently used in the analyses. I know "it is complicated" but if Van Beek didn't have access to 1930 stats when the cards were calculated/issued then shouldn't we be looking at 1929 stats?
I think you're misunderstanding things.
First of all — I'm not interested in arguing a position here. As I've looked into the evidence, I've discovered strong evidence that National Pastime is indeed based on 1930. This is obvious for two reasons that I've already written about: the rosters, which include players who were not in Major League Baseball in 1929 (and, in some cases, players who weren't even in Minor League Baseball), and the lineups, which fit extremely well with commonly used lineups in 1930.
Second — there is actually really good evidence that Van Beek didn't base his cards on ANY season. That's why it is significant that every single hitter's card has at least one extra base hit number, regardless of the number of extra base hits the player had in real life.
I'm interested in discovering what Clifford Van Beek actually did to create his cards. As we'll see in future posts, there is a lot of evidence that he swapped out play result numbers between dice rolls for whatever reason. As we've already seen, there is also a lot of strong evidence that he didn't make strict use of walk and strikeout numbers to come up with the 13s and 14s on each card.
Somebody asked me in a private message a few months ago about double plays in National Pastime. It turns out that Van Beek couldn't have known exact numbers of double plays grounded into, since those statistics simply weren't kept. Play result numbers 24 and 25 appear to have been assigned because of preset patterns, and not because of some mathematical algorithm.
If we're going to be honest about what the data shows us, we really need to move away from the idea of which season's statistics the cards were strictly based on. That should be obvious given the number of players with no extra base hits in real life who have extra base hit numbers on their cards.
I'd give you a direct answer to the question about which year he used as a base if I had it. I don't have it. Think of this blog as a work in progress — an exploration into what actually happened, as opposed to a polemic argument about why I think it was this or why it was that.
Right now, I'm leaning towards a theory that Van Beek created the cards using end of 1930 season at bat, batting average, and home run numbers for certain key players. My guess is that even the walk and strikeout numbers were likely based on reputation more than anything else. It's going to take some time, though, to dig in and figure out what actually happened — and I intended to continue blogging about it each step of the way.
Thank you for the further explanation.
I still stand by my viewpoint that NP cards are based on PA (AB + BB) not AB.
You thought it was important enough to point it out in the blog, "If you look at Sweetland’s 1930 performance alone, he clearly had a higher percentage of his at bats go for extra base hits than Ruffing. Now, it’s not as much of an advantage when you factor in his 9 walks"
The impression I'm getting is that Van Beek didn't have pitchers' batting stats, at least at the time he calculated the cards. If he was using in-season stat listings, Idon't know of any that included pitcher batting, and if he already had to base some of the lesser-used nonpitchers -- or weaker hitters who got lopped off the bottom of the Sporting News lists -- on compilations from box scores, it's not unlikely that he didn't feel it was worth his time and effort to do that for the majority of pitchers.
(There's also the point that the five pitchers on each NP roster were standing in, so to speak, for all the pitchers used by each team. He may have felt -- or may have convinced himself -- that a literal rendering of those pitchers' batting stats didn't really serve his purpose.)
He'd have known who the more notable good-hitting pitchers were, especially pitchers like Ruffing, Lucas, Wes Ferrell and Uhle who were used extensively as pinch-hitters, and he carded them accordingly, if not necessarily literally. (And Smith of Boston had been a middle infielder when he came up; I actually wound up playing him there a couple of games in my 1930 APBA solo league when things got really bleak.)