Fun with Run Distribution
Run distribution over a given season is an amusing thing. Most offenses are judged off their seasonal runs scored average, but that’s often a bit misleading, as Studes showed us a while back. So far this season, despite being nearly a month engaged, the San Francisco Giants are yet to score 70 runs. In fact, the Astros, Diamondbacks, Athletics, and Reds have yet to score even 75 runs. On the other hand the only team to score more than 120 runs is the Blue Jays of all teams.
125 runs for the Jays, 65 for the Giants, that’s not a perfect 2:1 ratio, but it’ll work. What you see below is a run distribution chart. Basically, I took the amount of runs scored, plugged in how many games Team X scored Y amount of runs, divided that amount of games by the total amount of games played, and bam, we have liftoff. What’s the difference between the best and worse run producing lineups so far It’s important to note that while the Jays have played three additional games, the Giants are not being punished for a less-hectic schedule.
The Jays offense has scored 4 or 5 runs in nearly a combined 40% of their games. Compare that to the Giants ability to score 4+ runs, and you end up seeing that the Jays are scoring 4 or 5 runs in nearly more games than the Giants are scoring at least four. Interestingly, the two are close to equal on the amount of one run games while the Giants are blowing the Jays out of the water in two run games.
The big key to the difference: the Jays have yet to end a game with zero on the scoreboard, the Giants are about 5% of the time.
25
A small note on presentation — you might want to use a bar chart instead of a scatter plot so you get a histogram.
I second that — also, it’s a bit misleading to say that the Giants have been shut out “about 5% of the time,” when you’re talking about one game out of 18. With such a tiny sample, you don’t really have any reason to expect that the Giants will continue to score zero runs in 5% of games the rest of the season.
Right. Really small sample size.
To me the “big difference” is that the Jays have score 6 runs or more eight times, compared to the Giants’ three times. Combining that with the shut out, one could probably expect four or five more wins by the Jays thus far this season. Looks about right.
This would be interesting to see at the end of the season when sample size is not only larger, but wouldn’t matter so much because there’d be nothing to project.
I do wonder at what point we start to think that the offensive performance we expected from Toronto was a bit pessimistic.
Of course it’s a small sample size guys but I don’t think he was analyzing anything. He was just comparing the two teams who have scored the most and least amount of runs. Nothing wrong with that.
Also, a histogram would have definietly been easier to see visually.
One thing I will say is that when comparing these two teams you need to take in account for the league differences in the AL (DH) and NL (pitcher batting 9th). Not sure it would make that much of a difference with only a small number of games played, though.
Everyone expected the Giants to be low scoring. I think the more interesting question is what happened to the A’s high-potential offense powered by Matt Holliday? It’s been MIA. Funny how my local newspaper, the SJ Mercury ignores their problems but trumps up the Giants (i.e. kissing up to a potentially new and significant advertiser if the A’s move to San Jose).
I agree that run distribution is interesting, I think any baseball website with team data should be presenting that.
And it is not only run distribution that is interesting in baseball. I think the distribution in the draft is interesting in that way as well. People focus too much on the average value there when all we really care about is getting a good starter there, a star. Using averages, you get what you already know, that the value does down with every pick, and also learn about HS vs. College and other nuances.
Using distribution, I found that the odds were less than 50% of finding a good player with the Top 5 picks, and just plummets from there. People think having a first round pick is suppose to be good, but the odds I found there was roughly 9 to 1 against finding a good player when you pick in the 21-30 range, overall, that is, where you draft if you are a division/pennant contender. That means roughly, on average, if you are contending for 10 years, you will find, on average, ONE good player with that first round pick. A loser, say, like the Rays, getting a top 5 pick almost every year, over 10 years would pick up 4-5 good players. It’s tough to stay up on top with those odds against.