So we took a look at Trevor in non-saves vs saves. I thought it would be interesting to see how the other great closer of our time handles himself on non-save situations and with some prompting from sdsuaztec4, I went ahead and did it.. Keep in mind, these are just tip of the iceberg types of numbers, but they're interesting to look at anyways.
I think the big reason I wanted to take a look at these is because, as good as Geoff's original article was, he makes reference to "small samples". The one critique I have of the article is the fact that when looking at career numbers for guys like Trevor of Mariano Rivera, we're not dealing with small samples, but the actual population of their innings pitched, so we actually do have some statistical say in whether or not there's a difference in how they performed according to their splits.
More (boring?) baseball talk after the jump. I CAN'T WAIT.
Here we are again. And here's another set of numbers:
| PA | AB | OBP | |
|---|---|---|---|
| save situation | 2358 | 2174 | .256 |
| non-save situation | 1255 | 1145 | .279 |
These are Mariano's numbers and his OBP against looks a little better than Trevor's. You know what I realize I did wrong in Trevors' calculation was adding back in the HBP. Silly me. Bean balls already count in OBP. Luckily, it didn't actually change the results of the numbers, so we move forward like nothing happened. Looking up the HBP actually did give me something interesting to look at with Mariano though, who plunked 26 batters in save situations and 7 in non-save situations, which is, in itself, a statistically significant difference. If a save is on the line, we're 95% confident that Mariano Rivera will hit batters at a greater rate than in a non-save situation.
How weird is that? Isn't that weird? Does Mariano tend to pitch inside more during a save? We know that Trevor likes to use the fastball to set up the change and vice versa, while Mariano likes to jam hitters with the cutter, so maybe he's letting it fly just a bit more when the save's on the line and when he misses his spot, he hits somebody. In a non-save situation, Mariano, through his career, wouldn't shy away from pitching more than one inning, so he may have saved the cutter as an out pitch. In any case, having a tendency to hit batters isn't a bad strategy per se, because the next guy up may not be so quick to take away the inside.
Anyways, doing what we did for Trevor, we're 85% confident that Mariano allows more runners at a greater rate in non-save situations. And again, taking into acout IBB, we are not very confident that the difference is significant. So, both Trevor and Mariano allow more runners, but that seems to primarily be due to the greater number of intentional walks they're asked to perform during non-save situations.
Here's how people scored against Mariano:
| PA | Runs Allowed | Earned Runs |
|
|---|---|---|---|
| save situation | 2358 | 141 | 131 |
| non-save situation | 1255 | 94 | 85 |
Running these numbers through the t-test and we become 95% confident that Mariano allows runs at a greater rate during a non-save situation than a save situation. So we're even more confident that there's a difference here than we were with Trevor. With earned runs, we're 90% confident that Mariano allows more earned runs during non-save situations than save situations. Score one for Trevor. In Hoffy's case, we were not very confident that the difference was significant for earned runs, but reasonably confident that overall runs allowed showed a difference. We are much more confident that Mariano gives up runs at a greater rate during non-save situations.
Home runs and strikeouts:
| PA | HR | Ks | |
|---|---|---|---|
| save situation | 2358 | 24 | 531 |
| non-save situation | 1255 | 17 | 292 |
Looking at the home runs, we find that the difference isn't statistically significant, even though I thought off hand it looked like it might have been. Remember, this is the same thing that we found with Trevor. The difference here is that we're even less confident that Mariano's strikeout rate changes in a save vs non-save situation, which is commendable. The rate at which Mariano Rivera strikes out batters or gives up home runs is not significantly different if it's a save situation or a non-save situation.
So did we learn anything?
- We're pretty confident that Mariano gives up earned and unearned runs at a greater rate during non-save situations
- Mariano doesn't appear to give up home runs or strike people out at a rate that's statistically significant
- We're pretty sure that Mariano hits people at a greater rate when he's in the middle of a save situation
So there you have it. We looked at the two premier closers of our time and they both appear to be doing something a little bit different when they're saving games versus not saving games. Does this make us feel better about the fact that Trevor sucks when we put him in tie games and the like? I dunno. At least with Trevor, we can't be sure he's giving up earned runs at a greater clip than with Mariano, but we can't avoid the fact that Trevor doesn't strike people out at the same rate, which is where a decent amount of his effectiveness lies.
All in all, is this enough to keep Trevor (and Mariano) out of non-save situations? I'd think yes. It doesn't make much sense when many of the key factors show a statistically significant difference. Unfortunately, we're kinda at the twilights of both their careers, so maybe it's just future warning to young closers.
All this calculatoring make my head hurt. Somebody check my numbers.