Originally Posted by bucksfan2
If you want to find out how good or how bad an offense is, you need to take into consideration outliers. Every team in baseball is going to have games where they bust out offensively. Every team with professional ball players is going to find themselves in a game where everyone his hitting. When the opposing team goes from trying to win the game to trying to end the game as soon as possible.
The question becomes if the Reds are struggling to score runs are 5 games of an offensive explosion more indicative of their performance that 16 games of ineptitude?
First, let me concede something. Given a certain level of overall run production
, a team is better off scoring it's average number of runs each day than having the occasional huge day and a lot of below average days.
Here's where I think you're misguided. You're essentially suggesting that those high scoring games are a fluke. That a team's having a handful of huge scoring games create an artificially high perception of the team's actual, predictable ability to score runs.
Put another way, you're saying that there is no relationship between having big spikes in run scoring and being a good run scoring team the rest of the time.
So, a few thoughts
1. There is a lower bound of run scoring -- zero. There is no upper bound. Every single team in MLB will have an average R/G that is higher than their median R/G. Put differently, every team will have more than half of their games include a run scoring total that is below their average.
2. Teams that are generally better at scoring runs are more likely to have games where they score a ton of 'em.
3. Therefore, teams with generally strong offenses will have a higher percentage of their games in which their run total is below average.
It's sort of like the runners left on base "problem". Guess which teams lead MLB every year with runners left on base? It's the ones that score the most runs. It seems counter-intuitive, because we want to imagine a scenario where they convert more of those LOB guys to runs, but that's not how it works in practice. The more runs you score, the less efficiently you tend to do it.
So, I'd suggest that if you aren't comfortable with using the average R/G to judge the Reds offensive, look at run scoring this way: How often do the Reds score at least X runs per game compared to the average team? (or use X runs or fewer if you prefer) Use whatever number you want for X, but you have to do the same for whomever you want to compare the Reds to. I think you'll be pleasantly surprised.
While I do think your basic instinct is correct in terms of big nights inflating the average, a few things:
- What you're seeing with the Reds is not a function of something special with the Reds offense -- it's something that happens to good offenses, period. Teams that score 4 runs or 5 runs more than other teams are the same ones that score 10 runs more than other teams.
- Those 10 run nights aren't coming at the expense of 2 and 3 run nights. Rather, they are a result of an offense than turns 2 run nights in to 3 run nights, 3 run nights in to 5 run nights, and 5 run nights in to 10 run nights. In other words, you're thinking about those nights backwards.
- So instead of thinking, "If only the Reds could take some of those runs and use them on lower scoring nights" think "those big nights are a sign of an offense that turns close wins in to blowouts and which turns close losses in to close wins"
- Having an offense that DOESN'T have big scoring nights isn't a sign of an efficient offense; it's a sign of a poor one.
- And of course, small sample size. Our run scoring distribution in a subset of our games is less predictive our ability to score runs in all of our games.