skmsw wrote:And statistically, SLG correlates more closely to runscoring than OBP does (of course, OPS correlates more closely than either alone).
I'm glad you mentioned this because I meant to ask you before. I have read that 1.8*OBP+SLG correlates most highly with run scoring. Which would make it seem like OBP is the more important component.
I feel like I am missing something here, what is it?
What this means is the scales for the two measures are different. Although on average they add the same number of runs, adding 10 points of OBP isn't as 'reliable' as adding 18 points of SLG.
I don't like this question because it's kinda rare that a team has a choice between, say, an average defender with a 325/500 and another average defender at the same position with a 375/450.
you need both really. ops is a pretty good stat only because it is convenient from existing stats. if slg and obp didn't already exist, we'd probably look at something more like (TB+walks) / outs, lets call it ERP for expected run production.
the key to OBP is that its as much about not making outs as it is getting on base. "out %" = 1-OBP. if you are getting on base 40% of the time, you are only making outs 60% of the time. but getting on base isn't valuable unless your team-mates can drive you in.
consider the classic case of Dwarves vs. Trolls. The Dwarves never swing but are so small that they draw a lot of walks. The Trolls swing so hard that they always either strike out or hit a HR.
The Trolls typical performance:
1 HR and 6 Ks per every 7 PA = 4.5 Runs per game.
average = 1/7 = 0.143
obp = 1/7 = 0.143
slugging = 4/7 = 0.571
ops = 5/7 = 0.714
ERP = 4 / 6 = 0.667
The Dwarves typical performance
3 BB and 4 Ks per 7 PA = 4.5 Runs per game
average = 0.000
obp = ~3/7 = 0.429
slg = 0.000
ops = 0.429
ERP = 3 / 4 = 0.750
you can see for these extreme cases, both produce the same scoring but ops is drastically different ... 0.714 vs 0.429. ERP is similar considering the extreme nature of the offenses ... 0.667 vs 0.750
So before integration, there was great parity between the two types of teams. Dwarf teams and Troll teams had ferocious battles for the pennant each year.
Then an innovator came along. Branch Rickey was the first general manager to put a Troll on a team of Dwarves, thinking that it would be more effective. Now some of the trolls home runs would not be solo shots but would be 2-run, 3-run or grand slam homers.
Surprisingly, in baseball terms, the mixing turned out to make very little difference, no matter how he shook up the line-up and mixed Trolls and Dwarves. The Troll personally had more RBIs, but in trade he made so many outs that the Dwarves were no longer producing as many runs as before. The net effect was surprisingly--- almost zero.
In terms of the relationships between Dwarves and Trolls, however, a great change was seen. The races become friends, introduced their sisters to eligible bachelors of the opposite races, and the frequent interbreeding led to the creation of the human race as we now know it.
StatmanCrothers wrote:you need both really. ops is a pretty good stat only because it is convenient from existing stats. if slg and obp didn't already exist, we'd probably look at something more like (TB+walks) / outs, lets call it ERP for expected run production.
the key to OBP is that its as much about not making outs as it is getting on base. "out %" = 1-OBP. if you are getting on base 40% of the time, you are only making outs 60% of the time. but getting on base isn't valuable unless your team-mates can drive you in.
consider the classic case of Dwarves vs. Trolls. The Dwarves never swing but are so small that they draw a lot of walks. The Trolls swing so hard that they always either strike out or hit a HR.
The Trolls typical performance:
1 HR and 6 Ks per every 7 PA = 4.5 Runs per game.
average = 1/7 = 0.143
obp = 1/7 = 0.143
slugging = 4/7 = 0.571
ops = 5/7 = 0.714
ERP = 4 / 6 = 0.667
The Dwarves typical performance
3 BB and 4 Ks per 7 PA = 4.5 Runs per game
average = 0.000
obp = ~3/7 = 0.429
slg = 0.000
ops = 0.429
ERP = 3 / 4 = 0.750
you can see for these extreme cases, both produce the same scoring but ops is drastically different ... 0.714 vs 0.429. ERP is similar considering the extreme nature of the offenses ... 0.667 vs 0.750
So before integration, there was great parity between the two types of teams. Dwarf teams and Troll teams had ferocious battles for the pennant each year.
Then an innovator came along. Branch Rickey was the first general manager to put a Troll on a team of Dwarves, thinking that it would be more effective. Now some of the trolls home runs would not be solo shots but would be 2-run, 3-run or grand slam homers.
Surprisingly, in baseball terms, the mixing turned out to make very little difference, no matter how he shook up the line-up and mixed Trolls and Dwarves. The Troll personally had more RBIs, but in trade he made so many outs that the Dwarves were no longer producing as many runs as before. The net effect was surprisingly--- almost zero.
In terms of the relationships between Dwarves and Trolls, however, a great change was seen. The races become friends, introduced their sisters to eligible bachelors of the opposite races, and the frequent interbreeding led to the creation of the human race as we now know it.
That's great, Statman! If I had kids, I would read that to them for their bedtime story as it is both educational and endearing.
