In the modern game, these are probably sufficiently randomly distributed among middle relievers as to make little difference.

I'm not sure I agree with the definition necessarily. In theory at least you could have the first pitcher come in and move the probability from, say, 18% to 45%. The next guy comes on and does something not too terrible (maybe one walk and one out?) that moves the probability from 45% to 55% then he's replaced and the other guy gives up the lead. In that scenario you could argue the middle guy did the least damage even if he was the guy who pushed it over 50%.

A relief pitcher of the ‘60s who had 23 saves and eight blown saves might very well have been as effective, given the circumstances under which he pitched, as one who saves 47 of 50 today.

Given closers today almost always start an inning, they are always fully responsible. For past relievers, you rarely saw that much situational relieving so you won't be seeing the types of scenarios you laid out. Are you suggesting there were lots of times that the firemen of yesteryear came into innings where the probability of blowing the lead was already well over 50% so their blown saves are only partial?

Anyway, over 54 years, you've identified fewer than 600 games or about 11 per season. Between expansion and bullpen specialization, I'm guessing they are more likely to be in the modern era. It's hard to see how this would shift our perception of the relievers of yore all that dramatically.

Are you suggesting there were lots of times that the firemen of yesteryear came into innings where the probability of blowing the lead was already well over 50% so their blown saves are only partial?

Yes.

-- MWE

Expanding on #2:

Here are some numbers for the first inning of their appearances for three well-known relief pitchers (Retrosheet doesn't have PBP for three of Face's 1959 games):


I'm suggesting that you can use 1-Prob(All) to set an expectation for the held lead percentage for a reliever, and when you do that, while Mo is still fantastic, the conversion percentages for Face and Marshall also look reasonable. (I should note that both Face and Marshall had games where they held a lead during the first inning that they came in, but lost that lead later).

It's not so much that yesterday's closers came in "lots of times" with a percentage over 50% as that today's closers (and today's relievers in general) rarely come in with a percentage over 50%.

I should also note that Jerome Holtzman developed the save statistic in 1960 as a reaction to Face's 1959, comparing Face unfavorably to the Cubs' tandem of Bill Henry and Don Elston, without noting that both of those pitchers lost a fair number of leads as well.

-- MWE

Wish things looked in Live Preview the way they look in the comment.

-- MWE

That's very interesting, Mike. Do you have any "career" marks for the older guys?

Do you have any "career" marks for the older guys?

I do, but I will probably post those in a separate article. I'm missing a fair number of Face's games because Retrosheet doesn't have a lot of PBP for the Pirates in the 50s and 60s (despite the best efforts of several Pittsburgh-based people, it was all but impossible to find Bucco scoresheets for that era).

-- MWE