Saves and 'Perfect' Saves

[Frayed Knot]

... otherwise known as, answering my own question.

In the 'Stats I'd Like to See' thread I wondered about not only the percentage of games that were saved but also about that fan-comforting feel of the 'perfect save', that 1-2-3, no-mess/no-fuss inning that most fans are convinced happens against them regularly but only rarely for them.

So, not being able to find the actual stats I wanted, I did the next best thing: I made them up.
Well, not quite. I actually took a dive into the land of box scores (via the net Timmy K., not my scrap book) and came up with a representative sample size that should at least approximate what I was looking for.

Here's the criteria:
- I took a month of MLB games all from after this year's AS break. That's the games from Thursday July 15 up to and including yesterday, or around 400+ games in total. Maybe not enough to be a definitive answer on the topic but certainly a sizable sample.
- I looked for all cases of what we might call 'Standard Issue Saves' in this closer's era. IOW, saves where a reliever comes in at the beginning of the 9th inning (or occasionally in extras) with no outs and the bases empty trying to close out a 1, 2, or 3 run lead. So no partial inning saves, no multi-inning saves, no inherited runners.
- I didn't distinguish WHO was saving the game (as in whether or not that pitcher normally has the title 'Closer' tattooed on his ass) only that a team used someone in the above situation
- and I tracked two things: Did that pitcher save the game?, and was that save a perfect 1-2-3 inning?


Results:

- Of the 400-some games, about half (203 to be exact) had a save situation that fit the mold. Actually it was probably a bit fewer than half the games seeing as how some games had more than one save situation (ex: visiting closer blows lead in top 9 giving the home closer a shot to either save or blow the game in the bottom half)

- of those 203, 41.3% were 1-run leads, 29.6% were 2-run leads, 29.1% were 3-run.
I don't think this says anything really except that close games are more common in MLB than not.

- of all the save opps, 166 - or 818% - got saved.
That's about where I thought it would be and maybe even a bit higher (I might have guessed it would be in the 75-80% range). What I probably should have done but didn't was track what pct of the 1-run, 2-run & 3-run saves were converted rather than just taking them as a whole. Maybe next time for that.

- and just under 1/3 of the total save opps (66, or 32.5%) were 'Perfect' saves. Or, looked at another way, of games that are saved, just under 40% are perfect.
I'm betting that a majority of fans would guess that number to be much higher - like well over half.

[Gwreck]

Frayed Knot wrote: - and just under 1/3 of the total save opps (66, or 32.5%) were 'Perfect' saves. Or, looked at another way, of games that are saved, just under 40% are perfect. I'm betting that a majority of fans would guess that number to be much higher - like well over half.

It think it depends on the situation. A "perfect save" when it's a 1 run lead is an accomplishment. Giving up a baserunner (or even two) when it's a 3-run lead is far less consequential. I'm sure there could be a metric for this, something like [(Hits + HBP + Unintentional Walks / "1 run save situations") * x] and we could call is "close save efficiency."

[Frayed Knot]

Gwreck wrote: Frayed Knot wrote: - and just under 1/3 of the total save opps (66, or 32.5%) were 'Perfect' saves. Or, looked at another way, of games that are saved, just under 40% are perfect. I'm betting that a majority of fans would guess that number to be much higher - like well over half. It think it depends on the situation. A "perfect save" when it's a 1 run lead is an accomplishment. Giving up a baserunner (or even two) when it's a 3-run lead is far less consequential. I'm sure there could be a metric for this, something like [(Hits + HBP + Unintentional Walks / "1 run save situations") * x] and we could call is "close save efficiency."

Yeah, differentiating the data between 1, 2, & 3 run saves might have been more informative - although with the way closers are used in modern baseball the managers rarely make a distinction when choosing to go to their closer unless the lead gets to 4 runs or more.
Plus, I was more trying to get a gauge on the usual complaint about how often your team's closer "makes you nervous", or "can't ever make it easy" etc. I suspect most fans think that closers have 1-2-3 innings a lot more than they actually do, so when their closer gives up a 9th inning baserunner under any circumstances the whole stadium/home audience lapses into a kind of 'here we go again' mode no matter what the score.


That proposed metric sounds like something the geeks at 'Baseball Prospectus' would come up with (along with some god-awful accompanying acronym) - and they probably do have one although I doubt they have it broken out to the specific 'closer situations' that I'm using here.

[metsmarathon]

you want an acronym? well, i do work for the army...

Fan - Reliever ANnoyance Concerning (save) [crossout]Opportunities[/crossout] er... Outcomes

[Frayed Knot]

Made a quick check on Mariano Rivera's stats for this season as they pertained to this same set of guidelines: ie. 1 innning, no baserunners, etc.

Remember that the overall stats for MLB was approximately 82%/32% (pct of games saved/"perfect" inning saves)
At age 40 Rivera is at 96%/56% [23 Sv Opps, 22 saved, 13 clean innings]